NaOH Alkali

ECOLE DES MINES
DE PARIS

Ecole Doctorale 364 : Sciences Fondamentales et Appliques

Nattribu par la bibliothque

|__|__|__|__|__|__|__|__|__|__|

T H E S E

pour obtenir le grade de
Docteur de lEcole des Mines de Paris
Spcialit Science et gnie des matriaux

prsente et soutenue publiquement par

Mlle Magali EGAL

le 8 Dcembre 2006

STRUCTURE AND PROPERTIES OF CELLULOSE / NAOH
AQUEOUS SOLUTIONS, GELS
AND REGENERATED OBJECTS

Directeur de thse : Tatiana Budtova et Patrick Navard

Jury :

M. Jacques Desbrires Rapporteur
M. Volker Ribitsch Rapporteur
M. Jonathan Marshall Examinateur
M. Michel Pierre Examinateur
Mme Tatiana Budtova Examinateur
M. Patrick Navard Examinateur

Table of contents
_____________________________________________________________________________________________________
- 1 -
T TA AB BL LE E O OF F C CO ON NT TE EN NT TS S

RESUME DU TRAVAIL DE THESE ..

INTRODUCTION ...

Rsum du Chapitre I
Etat de lart : La cellulose et sa dissolution dans des solutions aqueuses dhydroxyde de sodium

CHAPTER 1.
STATE OF THE ART: CELLULOSE AND ITS DISSOLUTION IN ALKALINE HYDROXIDE AQUEOUS SOLUTIONS ...

I.1- Cellulose: structure, morphology and treatment ...
I.1.1- Generalities ...
I.1.2- Chemical structure of cellulose ..
I.1.3- Crystalline structures of cellulose ..
I.1.3.1. Cellulose I or native cellulose
I.1.3.2. Cellulose II
I.1.3.3. Cellulose III ...
I.1.3.4. Cellulose IV ..
I.1.4- Supra-molecular structure of cellulose .
I.1.4.1. Micro-fibrils
I.1.4.2. Wood and cotton cellulose fibres ..
I.1.5- Extraction of cellulose and activation
I.1.5.1. Manufacture of cellulosic pulp ...
I.1.5.2. Activation of cellulose .

I.2- Cellulose dissolution in alkaline hydroxide aqueous solutions
I.2.1- Xanthation in viscose process
I.2.2- Action of alkaline hydroxide solutions on cellulose fibres ..
I.2.2.1. Introduction on mercerisation process and cellulose dissolution .
I.2.2.2. Formation of alkali-celluloses
I.2.2.3. Hydration of alkali ions
I.2.2.4. Mechanism of mercerisation ..
I.2.3- Influence of additives on cellulose dissolution .
I.2.3.1. Influence of zinc oxide (ZnO) .
I.2.3.2. Influence of urea and thiourea ...
I.2.3.3. Other additives .

I.3- Conclusions .

REFERENCES

Rsum du chapitre II
Matriaux et mthodes exprimentales .

CHAPTER II
MATERIALS AND METHODS

II.1- Materials ..
II.1.1- Cellulose samples ...
II.1.1.1. Avicel cellulose
II.1.1.2. Borregaard cellulose ..
II.1.1.3. Other cellulose samples
II.1.2- Solvents and additives ...
II.1.2.1. Alkali hydroxides .
II.1.2.2. Additives ..
II.1.2.3. Surfactants ..
II.1.3- Other components used in sponge manufacture ..
5

17

21

23

25
25
26
27
27
28
29
29
29
29
30
32
32
33

35
36
37
37
40
41
43
45
45
46
49

50

51

57

59

61
61
61
61
62
62
62
62
62
62
Table of contents
_____________________________________________________________________________________________________
- 2 -
II.1.3.1. Reinforcing fibres
II.1.3.2. Adhesion promoters ...
II.1.3.3. Na2SO4,10H2O crystals .
II.1.3.4. Carrageenan

II.2- Solutions preparation ...
II.2.1- Drying of steam exploded cellulose .
II.2.2- Dissolution procedure .
II.2.3- Storage .
II.2.4- Filtration

II.3- Methods
II.3.1- Differential Scanning Calorimetry (DSC) .
II.3.1.1. Main principles
II.3.1.2. Temperature measurement ..
II.3.1.3. Melting temperature measurement ..
II.3.1.4. Melting enthalpy measurement
II.3.1.5. Calibration and experimental conditions .
II.3.1.6. Software for peak deconvolution: PeakFit 4.12 ..
II.3.2- X-ray diffraction
II.3.3- Rheology ..
II.3.3.1. Semi-dilute cellulose solution ...
II.3.3.2. Dilute cellulose solution .
II.3.4- Microscopy ...
II.3.4.1. Optical Microscopy (OM) ...
II.3.4.2. Scanning Electron Microscopy (SEM) .
II.3.5- Mechanical property: tensile tests
II.3.5.1. Tensile tests on films of regenerated cellulose ..
II.3.5.2. Tensile tests on regenerated cellulose + reinforcing fibres .
II.3.5.3. Tensile tests on sponges (regenerated cellulose + fibres + porophores) .
II.3.6- Other methods .
II.3.6.1. Density measurement
II.3.6.2. Water absorption
II.3.6.3. Shrinkage .

REFERENCES

Rsum du chapitre III
Structure des solutions aqueuses de NaOH et de cellulose/NaOH avec et sans additifs. Limite de
dissolution de la cellulose

CHAPTER III
STRUCTURE OF AQUEOUS NAOH AND CELLULOSE-NAOH SOLUTIONS WITH AND WITHOUT ADDITIVES. LIMIT OF
CELLULOSE DISSOLUTION .

III.1- Bibliography: phase diagram of aqueous sodium hydroxide solution ...
III.1.1- Phase diagrams .....
III.1.1.1. Binary phase diagram ...
III.1.1.2. Ternary phase diagram
III.1.1.3. Methods for building phase diagrams. The limits of the DSC analysis .
III.1.2- State of the art on the structure of NaOH aqueous solutions .
III.1.2.1. Pure sodium hydroxide
III.1.2.2. Sodium hydroxide hydrates .

III.2- Structure of cellulose solvents: NaOH/water, NaOH/urea/water and NaOH/ZnO/water ..
III.2.1- NaOH/water binary system ..
III.2.1.1. DSC melting thermograms of NaOH/water solutions ..
III.2.1.2. Peak at lower temperature: melting of the eutectic mixture
III.2.1.3. Peak at higher temperature .
III.2.1.4. Proportions between NaOH and water ..
III.2.1.5. Conclusions on the phase diagram and structure of NaOH/water solutions ...
III.2.2- NaOH/urea/water ternary system
III.2.2.1. Urea/water binary phase diagram ..
III.2.2.2. NaOH/urea/water ternary phase diagram .
III.2.3- NaOH/ZnO/water ternary system
III.2.3.1. Viscosity of dilute NaOH/ZnO/water solutions ..
62
63
63
63

64
64
65
67
67

69
69
69
71
72
74
74
75
76
77
77
78
79
79
80
81
81
83
84
85
85
85
85

86

87

91

93
93
93
94
95
97
97
98

101
101
101
102
105
105
107
108
108
113
118
119
Table of contents
_____________________________________________________________________________________________________
- 3 -
III.2.3.2. NMR study ..
III.2.3.3. DSC results
III.2.3.4. Conclusions on the NaOH/ZnO/water system ..
III.2.4- Conclusions on the structure of cellulose solvents: NaOH/water, NaOH/ZnO/water
and NaOH/urea/water ...

III.3- Structure of cellulose solutions in NaOH/water, NaOH/ZnO/water and
NaOH/urea/water
III.3.1. Cellulose/NaOH/water ternary solutions
III.3.1.1. DSC results on Avicel/NaOH/water
III.3.1.2. Peak at low temperature. Limit of cellulose dissolution ..
III.3.1.3. Peak at high temperature .
III.3.1.4. Influence of cellulose origin on the structure of cellulose/NaOH/water
solutions ..
III.3.1.3. Conclusions .
III.3.2. Cellulose/NaOH/urea/water solutions .
III.3.2.1- DSC results on Avicel/NaOH/6%urea/water .
III.3.2.2. Peak of NaOH eutectic .
III.3.2.3. Peak of urea eutectic
III.3.2.4. Peak of free ice melting
III.3.2.5. Conclusions
III.3.3. Cellulose/NaOH/ZnO/water solution ...
III.3.4. Conclusions on the structure of cellulose solutions in NaOH/water with and without
additives ..

III.4- Influence of freezing on the thermal properties of the cellulose/NaOH mixtures

III.5- Conclusions .

REFERENCES

Rsum du chapitre IV
Ecoulement et glification des solutions aqueuses de cellulose/NaOH : Influence des additifs ..

CHAPTER IV
FLOW AND GELATION OF AQUEOUS CELLULOSE/NAOH SOLUTIONS: INFLUENCE OF ADDITIVES .

IV.1. Influence of ZnO and urea on the flow behaviour of cellulose/7.6NaOH aqueous
solutions
IV.1.1. Influence of the presence of ZnO or urea in the solvent on the flow of
cellulose/7.6NaOH aqueous solutions ...
IV.1.2. Conclusions

IV.2- Gelation and dissolution of cellulose in the presence of ZnO
IV.2.1. Influence of the preparation conditions on gelation of cellulose/7.6NaOH/water
solutions in the presence of ZnO
IV.2.2. Influence of ZnO concentration on gelation and cellulose dissolution ..
IV.2.3. Dissolution of cellulose in a solvent with a constant molar ratio ZnO:NaOH=1:10 .
IV.2.4. Influence of the presence of ZnO on gelation kinetics .
IV.2.5. Influence of ZnO on cellulose intrinsic viscosity ...
IV.2.6. Discussion on the influence of ZnO on gelation of steam exploded cellulose in
7.6NaOH aqueous solutions
IV.2.7. Conclusions

IV.3. Gelation and dissolution of cellulose in the presence of urea; comparison between
urea and ZnO
IV.3.1. Influence of the preparation conditions on cellulose dissolution and gelation in the
presence of urea and of urea concentration .
IV.3.2. Influence of cellulose DP and origin on the dissolution and gelation in the presence
of urea and ZnO .
IV.3.3. Discussion on the influence of urea on gelation and dissolution of steam exploded
cellulose in 7.6NaOH aqueous solutions and comparison with the role of ZnO .
IV.3.4. Conclusions ...............

IV.4. Influence of freezing on rheological properties of cellulose/NaOH aqueous solutions
IV.5- Tests on gelation and dissolution of cellulose in other alkali solvents: KOH and NaOH
120
121
125

126

127
127
127
128
132

132
134
135
135
136
137
137
138
138

139

140

144

147

149

151

153

153
159

160

161
163
165
165
167

169
171

172

172

175

179
181

181

Table of contents
_____________________________________________________________________________________________________
- 4 -
with other additives
IV.5.1. Testing cellulose dissolution in potassium hydroxide ..
IV.5.2. Addition of surfactants ..
IV.5.3. Other additives: salts and oxides
IV.5.3. Conclusions

IV.6. Conclusions ...

REFERENCES

Rsum du chapitre V
Proprits mcaniques dobjets rgnrs prpars partir de solutions aqueuses de cellulose/NaOH

CHAPTER V.
MECHANICAL PROPERTIES OF REGENERATED OBJECTS PREPARED FROM CELLULOSE-NAOH AQUEOUS
SOLUTIONS

V.1. Summary on sponge preparation: viscose and NaOH processes ...
V.1.1. Viscose process ..
V.1.2. NaOH process .

V.2. Properties of sponges prepared with the NaOH process ..
V.2.1. Wet samples made from regenerated 5cellulose/7.6NaOH aqueous solutions, without
reinforcing fibres and porophores .
V.2.1.1. Tensile stress results .
V.2.1.2. Influence of regeneration bath parameters
V.2.1.3. Influence of solution thermal treatment ..
V.2.1.4. Influence of additives .
V.2.1.5. Conclusions
V.2.2. Wet samples made from regenerated 5cellulose/7.6NaOH aqueous solutions with
reinforcing fibres without porophores .
V.2.2.1. Influence of added reinforcing fibres ...
V.2.2.2. Influence of fibres treatment on the rupture stress ...
V.2.2.3. Influence of freezing on rupture stress and adhesion between fibres and
matrix ...
V.2.2.4. Influence of the thermal history of fibre treatment with 8%NaOH aqueous
solution on the mechanical properties of wet regenerated samples .
V.2.2.5. Conclusions .
V.2.3. Wet samples made from regenerated 5cellulose/7.6NaOH aqueous solutions with
reinforcing fibres and porophores ...
V.2.3.1. Rupture stress ..........................................................
V.2.3.2. Absorption ...
V.2.3.3. Shrinkage ....
V.2.4. Conclusion on the role of the freezing step ....

V.3. Conclusions

REFERENCES

CONCLUSIONS

APPENDIXES
183
183
185
187
188

189

190

193

195

197
197
199

201

202
202
203
203
205
205

206
206
207

208

211
213

213
214
215
217
219

221

222

223

227

Rsum du travail de thse
_____________________________________________________________________________________________________
- 5 -

R
R
E
E
S
S
U
U
M
M
E
E
D
D
U
U
T
T
R
R
A
A
V
V
A
A
I
I
L
L
D
D
E
E

T
T
H
H
E
E
S
S
E
E

CONTEXTE DE LETUDE
La cellulose est un polymre naturel de la famille des polysaccharides. Elle est forme lors de la
photosynthse. La cellulose est le principal constituant des parois cellulaires des vgtaux et peut
donc tre extraite de lherbe, des feuilles, des tiges (essentiellement la paille), des arbres. Le coton
est trs pur en cellulose, jusqu 98% de la composition totale. Pour la prparation des ptes de
cellulose, le bois reste la source majoritaire dautant plus que les arbres sont aussi exploits dans
ce but. La cellulose est biodgradable sous laction des bactries.
Mais la cellulose est difficile mettre en oeuvre. Elle ne peut tre fondue car elle se dgrade avant
davoir atteint une fusion. Pour fabriquer des fibres, des films ou des objets tridimensionnels, la
cellulose doit tre soit dissoute dans des solvants directs, soit transforme chimiquement et tre
ensuite mise en forme.

Actuellement, le procd industriel le plus couramment utilis pour mettre en forme des objets
en cellulose pure est le procd dit viscose . La cellulose est dabord active par une solution
aqueuse de 18-20% dhydroxyde de sodium (NaOH) afin de faciliter la pntration du solvant. Le
sulfure de carbone (CS
2
) ragit alors avec les chanes de cellulose et forme le xanthate de
cellulose, soluble dans NaOH dilu. Le mlange ainsi obtenu est appel viscose et peut tre
mis en forme. La viscose est ensuite rgnre afin de retrouver la cellulose. Des objets varis
sont ainsi obtenus : fibres, films minces et objets tridimensionnels. Ce procd est trs efficace et
totalement matris. Cependant, lutilisation du CS
2
et le dgagement des composs soufrs issus
des ractions chimiques le rendent fortement polluant. A lheure actuelle, les rejets en phase
aqueuse (oxydes de soufre) sont totalement recycls tandis que les rejets atmosphriques (H
2
S) ne
reprsentent plus que 10% du CS
2
utilis. Mais les rglementations environnementales deviennent
de plus en plus contraignantes et les techniques de traitement de la pollution sont coteuses.
_____________________________________________________________________________________________________
- 6 -
Pour des raisons la fois conomiques et environnementales, les fabricants de produits finis
cellulosiques cherchent des alternatives pour mettre en forme la cellulose, le but tant de
supprimer ltape de transformation de la cellulose par le CS
2
et de trouver un solvant direct de la
cellulose non polluant et peu coteux.

Depuis le XIX
e
sicle, les industriels utilisent des solutions aqueuses dhydroxyde de sodium,
diffrentes concentrations, afin de modifier laspect et les proprits des fibres de cellulose. Dans
lindustrie textile, par exemple, les traitements la soude permettaient damliorer la brillance des
tissus ou labsorption des colorants.
En 1939, Sobue [SOB1939] montra que 7-10% de NaOH/eau pouvait tre un solvant direct de
la cellulose. De nos jours, plusieurs quipes de recherche dissolvent directement la cellulose dans
des solutions aqueuses dhydroxyde de sodium (7-10%NaOH/eau, en dessous de 0C) mais les
proprits des produits finis sont souvent infrieures celles obtenues par le procd viscose .
Le problme est que ni la dissolution de la cellulose dans 7-10%NaOH/eau, ni les proprits des
solutions cellulose/NaOH/eau ne sont bien comprises.

Deux partenaires industriels ont particip ce projet de recherche. Innovia Films Ltd, en
Grande-Bretagne, qui fabrique des films cellulosiques pour lemballage et Spontex, en France,
produisant des ponges cellulosiques.

Le but de ce travail est dexplorer la possibilit de prparer des films et ponges cellulosiques par
le procd soude tout en approfondissant les aspects plus fondamentaux de la dissolution de la
cellulose dans les solutions aqueuses peu concentres dhydroxyde de sodium. Le premier objectif
tait donc de comprendre la structure des solvants de la cellulose solution dhydroxyde de
sodium avec ou sans oxyde de zinc ou ure et la structure des solutions de cellulose dans ces
mmes solvants. Ensuite, nous avons tudi les proprits rhologiques des solutions de cellulose
en dterminant leur comportement lcoulement et les conditions de glification, en temps et en
temprature. Linfluence des additifs, ZnO et ure, fut galement tudie. Pour finir, nous avons
test mcaniquement des objets de cellulose rgnre et tudi linfluence des conditions de
prparation des chantillons et notamment linfluence de ltape de conglation (solution de
cellulose restant 20 heures -20C) et des bains de rgnration.

Problmatique 1 : Les films obtenus partir du procd soude sont moins rsistants que les
films produits partir du procd viscose . Au cours des essais exprimentaux sur le
procd soude , il apparat que seulement une assez faible quantit de cellulose peut tre
dissoute dans les solutions NaOH/eau. Pour des raisons lies la fois aux proprits
mcaniques et la faisabilit conomique du procd, il faut augmenter la quantit de
cellulose mise en solution. La question est donc de dterminer la concentration maximale de
cellulose qui peut rellement tre dissoute dans une solution donne de NaOH/eau.
Etude exprimentale : Ltude des proprits thermiques des solutions de cellulose permettra de
mieux comprendre la structure des solutions de cellulose et les interactions qui interviennent
_____________________________________________________________________________________________________
- 7 -
au cours de sa dissolution entre les chanes de cellulose, les molcules de NaOH, leau et les
additifs.

Problmatique 2 : Pour la fabrication des films, les solutions de cellulose doivent tre les plus
homognes possible, toutes particules fibres non dissoutes, impurets dun diamtre
suprieur ~8m doivent tre filtres pour viter la dchirure du film au moment de son
tirage. En outre, il se pose un problme de glification lors du procd pour la fabrication
des films, et ce dautant plus que la concentration en cellulose est leve. Les solutions de
cellulose dans la soude sont des systmes instables qui glifient, dautant plus rapidement que
la temprature est leve au dessus de la temprature ambiante. De ce fait, si un dbut de
glification se produit avant la mise en forme du film, le morceau de gel devient alors un
point de rupture du film.
Plan dexprience : Nous tudierons en dtail les proprits rhologiques des solutions de
cellulose afin de (i) connatre leur comportement lcoulement, (ii) dterminer les conditions
de glification et (iii) tudier linfluence dadditifs tels que oxyde de zinc et ure sur la
rhologie des solutions de cellulose/NaOH/eau.

Problmatique 3 : Les ponges prpares partir du procd soude doivent tre congeles
plusieurs heures -20C pour avoir des proprits assez proches de celles des ponges
viscose . Un procd avec un sjour prolong trs basse temprature nest pas viable
conomiquement. Il est donc ncessaire dtudier linfluence de la conglation sur les
proprits finales dobjets de cellulose rgnre.
Plan dexprience : Une ponge cellulosique est en fait un composite poreux ayant une matrice
de cellulose rgnre et des fibres de renfort de cellulose naturelle (coton, lin, bois). Ltude
de la rsistance la rupture et de linfluence de la conglation sur les proprits mcaniques se
fera tape par tape, c'est--dire sur la matrice de cellulose rgnre seule, puis sur le
composite matrice+fibres et pour finir sur le produit fini matrice+fibres+porophores, crant
la macroporosit ncessaire aux proprits dabsorption.

Ces trois problmatiques industrielles structurent notre thse, ainsi que ce rsum, en constituant
les chapitres III, IV et V. Le chapitre I est consacr ltude bibliographique tandis que le
chapitre II prsente les matriaux et mthodes utiliss tout au long de cette tude.

Une brve description des expriences ralises et des rsultats obtenus est prsente ci-dessous.

_____________________________________________________________________________________________________
- 8 -
I- LITTERATURE : LA CELLULOSE ET SA DISSOLUTION DANS LES SYSTEMES
HYDROXYDE ALCALIN/EAU
La cellulose est un polymre linaire constitu dunit anhydroglucose (AGU). Les groupes
hydroxyles prsents sur chaque motif, et donc tout au long de la chane, sont facilement
accessibles pour crer des liaisons hydrogne intra- et intermolculaires. Par ces liaisons
hydrogne, les chanes de cellulose ont une forte tendance sagrger pour former des
arrangements cristallins trs ordonns. Ils sont au nombre de quatre : la cellulose I, ou cellulose
native, la cellulose II, la cellulose III et la cellulose IV. Les structures cristallines II, III et IV
sobtiennent par traitement chimique, seule la cellulose I se trouve dans les membranes des
cellules vgtales ltat natif. Les chanes de cellulose native sorganisent dabord sous forme de
micro fibrilles qui sagglomrent leur tour dans les macro fibrilles pour donner finalement la
fibre de cellulose.
Lorganisation des fibrilles (diamtre, densit, texture) varie suivant lorigine de la cellulose
(coton, bois), ce qui explique des diffrences de comportement au contact des solvants
[CUI2006] par exemple. Lors de lextraction de la cellulose, des traitements chimiques et
mcaniques sont galement appliqus pour sparer la cellulose des autres constituants du bois
(lignine, hmicelluloses). Ces traitements modifient la structure des fibres de cellulose et
influent donc sur les proprits de ces dernires.

Comme nous lavons expos dans le contexte de ltude, la soude, relativement concentre, est
utilise depuis de trs nombreuses annes pour modifier les proprits des textiles (brillance,
absorption des colorants), et ce, grce au gonflement des fibres de cellulose puis leur
transformation de la cellulose I la cellulose II : cest le phnomne appel mercerisation
[MER1903]. Il apparat galement que la mercerisation peut se produire plus faible
concentration en NaOH (~9%) si la temprature est faible. Davidson [DAV1934] et Sobue
[SOB1939] montrent que la cellulose peut tre dissoute dans la soude mais dans un domaine de
temprature et de concentration en NaOH trs troit (7-10%NaOH/eau, en dessous de 0C).
Les solutions de cellulose obtenues ne sont pas stables et glifient dans le temps et en fonction de
la temprature [ROY2003]. Dautres systmes alcalins comme KOH/eau et LiOH/eau sont
galement tudis. Si le premier ne permet quune faible solubilit de la cellulose, le deuxime est
trs efficace mais beaucoup plus cher que la soude. La plupart des tudes se consacreront donc
la dissolution de la cellulose dans la soude, solvant alcalin efficace et peu onreux.

Peu de temps aprs la dcouverte de la mercerisation, lajout dadditifs fut tudi pour amliorer
laction de NaOH sur les fibres de cellulose. Loxyde de zinc est dj mentionn cette poque
[MER1903] et apparat dans la littrature de faon sporadique [DAV1937] [WO 02/22924]. Par
contre, aucune explication sur son rle nest propose. Lajout dure est tudi depuis seulement
quelques annes [LAS1998] [ZHO2000] mais de faon trs dtaille. Non seulement il est montr
que lure amliore nettement la dissolution de la cellulose et la stabilit des solutions [ZHO2000]
mais certains auteurs tentent galement dexpliquer ces phnomnes [CAI2005] [CAI2006].

_____________________________________________________________________________________________________
- 9 -
II- MATERIAUX ET METHODES EXPERIMENTALES
La dissolution de la cellulose dans NaOH/eau, avec ou sans additif, seffectue -6C pendant
deux heures sous agitation. Une fois prpare, les solutions peuvent tre gardes un mois environ
+5C. Nanmoins, avant chaque prlvement, elles devront tre re-homognises car les fibres
non dissoutes sdimentent.

Ltude des proprits thermiques des solutions peut se diviser en deux parties, ltude des
solvants NaOH/eau, ure/eau, NaOH/ure/eau et NaOH/ZnO/eau et celle des solutions
de cellulose. Afin de dterminer les temprature et enthalpie de fusion de chaque composant
prsent dans la solution, nous avons utilis un appareil DSC 7 (PERKIN-ELMER) permettant de
descendre basse temprature (-65C), les vitesses de refroidissement et de chauffe utilises tant
de 1C/min. La corrosivit des solutions (prsence de soude) a ncessit lutilisation de capsule
inox plaque en or. Dans la plupart des cas, les tempratures et enthalpies de fusion ont t
calcules directement par le logiciel de DSC. Dans le cas de pics superposs, une dconvolution
de pics tant ncessaire, nous avons utilis le logiciel PeakFit 4.12.
Pour dterminer la structure des solvants, nous avons galement effectu des analyses de
diffraction de rayons X (XPERT-PRO de PANalytical) et des analyses de RMN. Ces dernires
ont t ralises lInstitut Fraunhofer Glm, en Allemagne.

Les proprits rhologiques des solutions semi dilues de cellulose (de diffrents DP, traitements
et origines) dans 7,6NaOH/eau, avec ou sans additif (oxyde de zinc et lure, diffrentes
concentrations) ont largement t tudies laide dun rhomtre Gemini (BOHLIN) quip
doutils cne plan et dune platine Peltier. Cette dernire permettant des variations rapides de la
temprature, nous avons pu dterminer le point de glification en fonction de la temprature et
du temps. Le comportement lcoulement a galement t tudi par des mesures de viscosit
en fonction du taux de cisaillement.
Quant ltude des solutions dilues de cellulose, elle a t effectue en utilisant un viscosimtre
capillaire Ubbelohde quip dune burette automatique de dilution (Lauda).

La rsistance la rupture de films de cellulose rgnre, de composites cellulose rgnre +
fibres de renfort et de produits finis (cellulose rgnre + fibres de renfort + porophores crant
la porosit) a t mesure laide dune machine ZWICK Z2.5, en mode de traction. Les
chantillons sont dans leur tat humide et la mesure seffectue lair ambiant.

Pour finir, tout au long de cette tude, des observations microscopiques ont t ralises que ce
soit au microscope optique ou au microscope lectronique balayage (MEB, SEM en anglais).

_____________________________________________________________________________________________________
- 10 -
III- STRUCTURE DES SOLUTIONS DE CELLULOSE ET LIMITE DE DISSOLUTION
Nous prsentons tout dabord quelques aspects thoriques sur les diagrammes de phase binaires
et ternaires et lutilisation des thermogrammes de DSC pour les construire. Une construction trs
prcise des diagrammes de phase partir des pics de fusion obtenus par DSC ncessite de trs
nombreux essais variation de la masse m de lchantillon et de la vitesse de chauffe v afin de
pouvoir extrapoler les rsultats m=0 et v=0. Le but de notre tude tant seulement de connatre
les phases prsentes dans le systme, leur composition et leur proportion, nous avons ralis les
mesures pour une seule vitesse de chauffe et une seule masse.

Ltude de structure des solvants et des solutions de cellulose a essentiellement t base sur des
analyses de DSC. Les systmes ont t refroidis trs basse temprature (-65C) afin quils soient
totalement cristalliss, puis remonts lentement (+1C/min) temprature ambiante en
enregistrant les pics de fusion des diffrents composants.

Dans un premier temps, nous avons analys des solutions aqueuses dhydroxyde de sodium
contenant entre 0 et 20% en poids de NaOH. A partir des thermogrammes de fusion, nous avons
pu tracer la partie du diagramme de phase 0-20%NaOH/eau qui correspond parfaitement celui
trouv dans la littrature [PIC1893] [ROL1964]. Dans ce domaine de concentrations en NaOH,
le systme forme un mlange eutectique compos dun hydrate de sodium NaOH,5H
2
O et de
quatre molcules deau. Cette adquation entre lexprience et la littrature nous a permis
dutiliser une mthode de calcul, base sur les enthalpies de fusion, permettant de dterminer la
proportion de chaque phase en solution.
Pour une solution de 7,6%NaOH/eau on aura tout dabord cristallisation de leau libre puis du
mlange eutectique NaOH,5H
2
O+4H
2
O quand la temprature diminue, les tempratures de
fusion tant respectivement -4C/-5C et -33,4C.

Lanalyse du systme 0-100%ure/eau montre un comportement eutectique simple, le mlange
eutectique tant compos de 1ure+8H
2
O.
Laddition de 0-25% dure dans une solution de 7,6%NaOH/eau ne modifie ni la temprature ni
lenthalpie de fusion de leutectique NaOH,5H
2
O+4H
2
O : lure ne ragit pas avec NaOH. Par
ailleurs, quand la concentration en ure augmente, le pic de fusion du compos ure se dcale
vers les plus hautes tempratures, tandis que celui de leau libre se dcale vers les basses
tempratures. Les deux pics se rejoignent pour une concentration totale en ure de 18%. Un tel
systme est en fait compos de leutectique NaOH,5H
2
O+4H
2
O et de leutectique ure+8H
2
O.
Il ny a plus deau libre. Tant quil y a suffisamment deau pour les deux mlanges eutectiques, ils
se forment indpendamment. Si leau vient manquer, chaque NaOH sentoure de ses 9
molcules deau puis leau restante entoure lure, lure en excs restant dans son tat pur.

Lors de sa dissolution en milieu fortement basique, loxyde de zinc ragit avec OH
-
pour former
des hydroxydes de zinc. Ces ractions chimiques provoquent donc une modification des
solutions et par consquent des changements au niveau du diagramme de phase NaOH/eau. Plus
la quantit en ZnO augmente, plus les temprature et enthalpie de fusion de leutectique
_____________________________________________________________________________________________________
- 11 -
NaOH diminuent. On a ainsi dtermin que trois NaOH pour un ZnO ne participent plus
leutectique et que les composs de zinc forms lors de la dissolution sont fortement hydrats
(jusqu 25H
2
O).

La structure des solvants NaOH/eau, NaOH/ure/eau et NaOH/ZnO/eau tant dsormais
connue, nous y avons ajout la cellulose afin de dterminer quelles interactions ont lieu lors de sa
dissolution.

Lajout de 0,5-7,6% de cellulose dans une solution de 7,6-9% de NaOH/eau nentrane pas de
diminution de la temprature de fusion de leutectique NaOH : la prsence de cellulose ne
modifie pas le diagramme de phase du solvant NaOH/eau. Par contre, lenthalpie diminue
proportionnellement la quantit de cellulose ajoute et atteint zro quand la quantit de
cellulose est gale la quantit de NaOH prsent en solution. Nous en avons donc dduit que (i)
la dissolution de cellulose dans NaOH ncessite au moins quatre molcules de NaOH qui se lient
une unit anhydroglucose (AGU) et (ii) la limite de dissolution de la cellulose est atteinte quand
m
cell
=m
NaOH
.
Quant la temprature de fusion de la glace libre, elle augmente lgrement avec la cellulose, ce
qui sexplique par un changement de lenvironnement molculaire de leau d la prsence de
cellulose. Lenthalpie de leau libre reste constante quand la concentration en cellulose augmente
ce qui signifie que la quantit deau libre en solution est constante. Nous en avons donc conclu
que les NaOH lis la cellulose sont entours des 9 molcules deau formant initialement le
mlange eutectique.

Des analyses DSC similaires ont t ralises sur des solutions de cellulose dans NaOH/eau
contenant 6% dure. Les valeurs denthalpie de fusion de leutectique NaOH suivent la mme
courbe que celles obtenues prcdemment pour des solutions de cellulose dans NaOH/eau, sans
ure. La prsence dure ne permet pas daugmenter la limite de dissolution de la cellulose qui
reste telle que m
cell
=m
NaOH
, chaque AGU sentourant au minimum de 4 NaOH et 36 molcules
deau.
Quant au pic de leutectique ure , il ne change ni en intensit, ni en temprature. Nous en
concluons donc que les chanes de cellulose ne ragissent pas avec lure.

Nous avons vu que lajout de ZnO dans le solvant diminue ou supprime la temprature de fusion
de leutectique NaOH . Lors de lajout de cellulose, ce pic napparat pas, il nest donc pas
possible de dterminer ni la concentration limite en cellulose pouvant tre dissoute dans une
solution donne de NaOH/ZnO/eau, ni la quantit de NaOH li chaque AGU.

La dernire section de ce chapitre concerne linfluence de la conglation sur la structure des
solutions, modification structurale pouvant tre dtecte par une analyse thermique.
Aprs un sjour prolong -28C, un pic apparat 0C, temprature de fusion de la glace. Ceci
signifie quune certaine quantit deau a cristallis de manire tout fait indpendante. Dun point
de vue thermodynamique, cette eau pure ne peut provenir ni du mlange eutectique
_____________________________________________________________________________________________________
- 12 -
NaOH,5H
2
O+4H
2
O libre en solution ni de leau initialement libre. La seule possibilit est que
cette eau pure provienne de leau lie la cellulose. A partir des enthalpies de fusion, nous avons
pu calculer quune molcule deau par AGU a t dtache. Mais tant donn que cette eau est
dgage dans un systme en quilibre thermodynamique, elle cristallise immdiatement et
indpendamment des autres constituants du systme, donnant ainsi un pic la temprature de
fusion de leau pure.
Des rsultats tout fait quivalents ont t observs pour des solutions de
cellulose/NaOH/ZnO/eau.

IV- ECOULEMENT ET GELIFICATION DES SOLUTIONS AQUEUSES DE
CELLULOSE/NAOH. INFLUENCE DES ADDITIFS
Ltude bibliographique a montr que dune part la cellulose ne se dissout que dans des
conditions prcises de concentration en NaOH et de temprature [SOB1939] et que dautre part,
une fois dissoute, les solutions ne sont pas stables et glifient en fonction du temps et de la
temprature [ROY2003]. Il apparat galement que lajout dadditifs tels que ure et oxyde de zinc
amliore la dissolution de la cellulose et la stabilit des solutions [ZHO2000]. Le but de ce
chapitre est donc dtudier le comportement des solutions lcoulement ainsi que dterminer les
conditions de glification, et ce en faisant varier de nombreux paramtres tels que DP,
concentration et origine de la cellulose, temprature et additifs.
Les solutions ont t tudies (i) en rgime semi dilu en mode continu afin de mesurer la
viscosit des solutions en fonction du taux de cisaillement, (ii) en mode dynamique pour
dterminer le point de glification et (iii) en rgime dilu pour mesurer la viscosit intrinsque de
la cellulose dans diffrents solvants. Paralllement aux rsultats rhologiques, de nombreuses
observations microscopiques (microscope optique) ont t effectues pour avoir des donnes
qualitatives sur la dissolution des fibres de cellulose.

Par les mesures de viscosit en fonction du taux de cisaillement, nous avons montr que le
comportement lcoulement des solutions de cellulose explose la vapeur est identique dans
les trois systmes de solvant tudis 7,6%NaOH/eau, 7,6%NaOH/6%ure/eau et
7,6%NaOH/0,7%ZnO/eau. Ce ne sont pas des solutions molculairement disperses. Leur
comportement se rapproche plutt de celui dune suspension. Ce rsultat concorde bien avec le
rsultat prcdent savoir quen solution les chanes de cellulose sont entoures dune grande
quantit de molcules dhydroxyde de sodium et deau.

Les observations microscopiques ont montr que la prsence doxyde de zinc ou dure amliore
la dissolution de la cellulose puisque la quantit de fibres non dissoutes diminue. Les mesures de
viscosit intrinsque dans les diffrents solvants ont rvl que cette amlioration de la
dissolution des fibres est due une amlioration de la qualit de solvant de la faon suivante
7,6%NaOH/eau est moins bon que 7,6%NaOH/6%ure/eau qui est moins bon que
7,6%NaOH/0,7%ZnO/eau.
_____________________________________________________________________________________________________
- 13 -
Par ailleurs, quel que soit le solvant utilis, la viscosit intrinsque de la cellulose diminue quand la
temprature augmente. Ceci signifie que les chanes de cellulose se contractent et que la qualit du
solvant diminue. Par consquent les interactions cellulose-cellulose sont favorises ce qui conduit
la glification des solutions en rgime semi dilu et la compactisation des molcules en rgime
dilu.
Pour des solutions de mme concentration en cellulose et en NaOH, la dtermination du point
de glification en fonction de la temprature donne ncessairement lordre suivant
T
gel
(cellulose/NaOH/eau) < T
gel
(cellulose/NaOH/ure/eau) < T
gel
(cellulose/NaOH/ZnO/eau).

Lajout de ZnO dans 7,6%NaOH/eau est dautant plus efficace (augmentation de la temprature
de glification) que sa concentration est leve, jusqu 1,5% qui correspond la limite de
solubilit de ZnO dans 7,6%NaOH/eau. Au-del de cette valeur, les proprits restent
constantes. Nous en concluons donc que ZnO amliore la dissolution de la cellulose et la stabilit
des solutions seulement sil a dabord t dissout dans NaOH. Ceci signifie que ce sont les
complexes forms entre NaOH, ZnO et leau qui permettent damliorer la dissolution des fibres
de cellulose et de retarder la glification. Par ailleurs le mcanisme de glification semble tre le
mme avec ou sans ZnO, qui agit donc uniquement en empchant les interactions cellulose-
cellulose qui apparaissent plus haute temprature.
En ce qui concerne lajout dure dans 7,6%NaOH/eau, nous avons constat que nos rsultats ne
correspondaient pas ceux trouvs dans la littrature. Nous avons donc prpar des solutions de
cellulose de diffrentes origines (diffrence dessence darbre, de traitement des fibres, de DP)
dans 7.6NaOH/eau contenant de 6 20% dure. Nous avons alors mesur les modules G et G
en fonction de la temprature afin de dterminer le point de gel. La temprature de glification et
la concentration optimale en ure varient suivant la cellulose utilise. Une solution de cellulose
explose la vapeur atteint sa temprature maximale de glification 40C pour une
concentration en ure de 6% tandis quune solution de cellulose irradie par un faisceau
dlectron atteint sa temprature maximale presque 60C pour une concentration en ure de 12-
15%. La stabilit des solutions dpend donc fortement de lorigine de la cellulose.
A lheure actuelle, les raisons de telles diffrences restent une question ouverte mais ces rsultats
expliquent pourquoi nous trouvons des rsultats diffrents de ceux rapports dans la littrature.

V- PROPRIETES MECANIQUES DOBJETS REGENERES A PARTIR DE
SOLUTIONS DE CELLULOSE/NAOH/EAU
Ce chapitre concerne la fabrication dobjets cellulosiques par dissolution directe de la cellulose
dans des solutions aqueuses dhydroxyde de sodium. Nous prsentons tout dabord le procd
viscose actuel puis le nouveau procd soude . Ce dernier ncessite une tape de
conglation des blocs dponge afin dobtenir des proprits du produit fini proches de celles
obtenues pour les ponges viscose . Etant donn que ce sjour prolong basse temprature
rend ce procd conomiquement non viable, il est important dtudier linfluence de la
_____________________________________________________________________________________________________
- 14 -
conglation sur les proprits des ponges et de savoir sur quel lment du produit (matrice,
fibres) elle intervient.

Lessentiel de cette tude porte sur des tests de traction permettant de mesurer la rsistance la
rupture des ponges. Une ponge est en fait un matriau composite constitu dune matrice de
cellulose avec des fibres de renfort et de trous. La matrice est compose de cellulose rgnre,
provenant de la solution de cellulose dans NaOH/eau, et de fibres de renfort, ensemble de fibres
de coton, de lin et/ou de bois. Les trous sont forms suite la fusion de cristaux de sel
Na
2
SO
4
,10H
2
O.
Afin de comprendre o et quand agit la conglation, nous avons prpar puis test des
chantillons (i) de matrice de cellulose rgnre seule, (ii) de matrice de cellulose
rgnre+fibres de renfort et (iii) de matrice+fibres+porophores crant la porosit.

Les tests de traction sur la matrice de cellulose rgnre seule montrent que les conditions
utilises initialement (conglation -20C et rgnration dans leau +70C) sont trs
dfavorables la rsistance la rupture. Les meilleures proprits mcaniques sont obtenues
lorsque la solution de cellulose/NaOH/eau est traite temprature ambiante que ce soit au
moment de la glification mais galement lors de la rgnration dans leau. Nous remarquons
toutefois quune rgnration en milieu acide donne de bonnes proprits mcaniques mais ce
mode de rgnration convient essentiellement lors de la fabrication de film mince et non lors de
la fabrication dobjet tridimensionnel telles que les ponges. La conglation nagit donc pas sur la
rsistance mcanique de la matrice de cellulose rgnre. Les chantillons congels de cellulose
rgnre rvlent de nombreuses empreintes en forme daiguille dues la cristallisation puis la
fusion de leau.

Lajout de fibres de renfort augmente la contrainte la rupture mais le rsultat reste mdiocre.
Ladhsion entre la matrice et les fibres est donc amliorer. Les observations microscopiques
ont montr quau cours de la conglation, les fibres de renfort, de DP lev, ont gonfl au
contact de la soude. Dans un premier temps, nous avons pens que ce gonflement pouvait
permettre une certaine pntration de la solution de cellulose et donc amliorer ladhsion
fibres/matrice. Cette hypothse sest rvle fausse, la conglation nintervenant pas au niveau de
ladhsion fibres/matrice.
Lajout de promoteur dadhsion na pas donn de rsultats satisfaisants.

Finalement, nous avons prpar et test les produits finis : matrice de cellulose rgnre + fibres
de renfort + porophores crant la porosit. Cest ce stade seulement que nous avons vu
linfluence de la conglation sur les proprits des ponges. Outre lamlioration de la rsistance
la rupture, la conglation des ponges entrane galement une augmentation de labsorption en
eau et une diminution du rtrcissement des ponges, compar des ponges quivalentes
nayant pas subit dtape de conglation. Ceci est en adquation avec les observations
microscopiques qui rvlent une porosit plus importante des ponges congeles, conduisant
une meilleure absorption.
_____________________________________________________________________________________________________
- 15 -
Nous avons galement observ que, aprs un premier schage lair ambiant, un grand nombre
de pores se sont ferms ce qui explique un retrait non ngligeable des chantillons congels ou
non.

Le manuscrit se termine par des conclusions gnrales et des annexes qui donnent une
description dtaille de quelques techniques et rsultats exprimentaux.

- 16 -

Introduction
_____________________________________________________________________________________________________
- 17 -
I
I
N
N
T
T
R
R
O
O
D
D
U
U
C
C
T
T
I
I
O
O
N
N

Cellulose is a natural polymer which is the structural substance present in all plants, allowing their
stalk and their protection. This polysaccharide is created by the photosynthesis from carbon
dioxide and water, using the sunlight as the energy source. Cellulose can be extracted from grass,
leaves, stems, tree trunks and brunches. Wood is the most often used source since trees are
planted especially for this purpose. Nevertheless, properties of the wood pulp strongly vary as a
function of age and growing conditions of trees and these parameters are difficult to control
when timber comes from the wild. Moreover, it is interesting to note that pulp manufacturers
plant sufficient trees to at least replace what they use.
At the end of the life cycle, cellulose degrades readily under bacterial action forming compost or
peat and ultimately degrades further producing methane, carbon dioxide and water. The loop is
closed.

The major problem of cellulose is that it is degrading before melting. Consequently, to process
cellulose and make fibres, films or 3D objects, cellulose has to be either dissolved in direct
solvents or derivatised and then processed.

The most common industrial process of making shaped pure cellulose objects is called the
viscose process: after cellulose activation with aqueous 18-20%NaOH, the carbon disulfide
(CS
2
) is reacting with cellulose chains to form the cellulose xanthate which is then dissolved in
dilute sodium hydroxide. At this stage, the mixture is named the viscose and can be shaped.
Finally, cellulose is regenerated and various final products can be obtained: fibres, films or tri-
dimensional objects. This process is very efficient and totally mastered, but it is very polluting:
because of the use of CS
2
and the numerous chemical reactions, sulphured products are released
creating pollution in atmosphere and in aqueous phases. Although the major part of the pollution
is treated in the plant some pollutants are still released. Moreover, technologies for pollution
treatments are very expensive. Thus for economical and for environmental reasons,
manufacturers of pure cellulose products have been trying to find alternative routes to shape
cellulose. There is a continuous search on environmentally friendly cellulose processes.
Introduction
_____________________________________________________________________________________________________
- 18 -
NMMO, for example, is a good cellulose solvent, but these solutions are not easy to prepare and
to process in the case of film and tri-dimensional object. This solvent is now used in the fibre
manufacture, called the Lyocell process.

For a long time, sodium hydroxide aqueous solutions at various concentrations were used in
cellulose processing, especially to change aspects and properties of cellulose fibres: in the textile
industries. For example, NaOH was largely used to increase the degree of lustre and to improve
the absorption of dyes. In 1939 Sobue showed [SOB1939] that (7-10)%NaOH aqueous solutions
can be a direct solvent of cellulose. Nowadays, several research teams directly dissolve cellulose
in sodium hydroxide aqueous solutions (7-10%NaOH below 0C) but the properties of final
products are often inferior to the ones of similar products made from viscose process. One of
the problem is that neither cellulose dissolution in 7-10%NaOH, nor the properties of
cellulose/NaOH/water solutions are well understood.

This work has two different goals, the first is the fundamental understanding of the behaviour of
the mixing of cellulose in NaOH-based aqueous solutions, and the second is to apply this
knowledge to the manufacturing of films and sponges.

The first goal of the work was to understand the structure of cellulose solvent aqueous solution
of sodium hydroxide with and without ZnO or urea and of cellulose solution in these solvents,
and then to determine how NaOH reacts with the polymer chains. From this fundamental aspect
we can determine the maximal cellulose concentration that can be dissolved in NaOH/water.
The study of the rheological properties of cellulose solutions by determining the flow behaviour
and the conditions of the occurrence of gelation, in time and in temperature, gave information on
the gelation. The influence of additives on rheological properties was also investigated.

The second goal was to investigate two aspects of the manufacturing of objects from this
process. In the case of the preparation of films, the difficulty lies in the amount of cellulose that
can be dissolved in the cellulose solution. The higher is the better, but this causes extreme
difficulties during processing due to gelation. The results obtained above explained why. The
second process is the manufacturing of sponges where a freezing step was shown to be
important. We thus studied the mechanical properties of regenerated objects and the influence of
preparation conditions and especially the influence of the freezing step (keeping cellulose
solution at -20C for 20 hours) and the influence of regeneration baths.

Two industrial partners supported this thesis. Innovia Films Ltd, in Great Britain, is a
manufacturer of cellulose films for wrapping and packaging and Spontex, in France, is a company
producing cellulose sponges.

Introduction
_____________________________________________________________________________________________________
- 19 -
The thesis is divided into five chapters.

The first one gives generalities on cellulose as well as a description of the different techniques
used for its extraction and its activation. We also present the state of the art concerning the
swelling and the dissolution of cellulose in sodium hydroxide/water and the influence of
additives such as zinc oxide or urea on the cellulose solubility.

In the second chapter, materials (cellulose, solvents and additives) and methods (DSC, rheology,
optical and electron microscopy, rheology, viscometry, x-ray scattering) used in this work are
described.

The next three chapters concern the results and their discussion.
Chapter III investigates, using DSC analysis, the structure of the various solvents used in this
study as well as the structure of cellulose solutions in these solvents. As the NaOH/water binary
system is described in details in literature, it enables to check the validity of the methodology
used to calculate NaOH/water phase diagram below 0C at NaOH concentrations used for
cellulose dissolution. The same methodology was applied on other solvents urea/water,
NaOH/ZnO/water and NaOH/urea/water and on cellulose solutions. We also determine the
amount of cellulose that can be dissolved in sodium hydroxide/water and we study the influence
of additives on this limit of cellulose dissolution.

Chapter IV focuses on the rheological properties of cellulose/NaOH/water solutions with and
without additives. The experiments in steady state mode were used to describe the behaviour of
cellulose/NaOH/water solutions under shear. The oscillatory mode allows examination of the
conditions for the occurrence of gelation.

Chapter V tackles the mechanical properties of regenerated wet cellulose samples. The tensile
tests were performed on regenerated pure cellulose and cellulose with the reinforcing fibres and
porophores. The influence of regenerating bath parameters and of the freezing step on the
mechanical properties was studied. We investigated how to improve the sponge manufacture
process while keeping in mind the necessity to have costs as low as possible.

Finally, a general conclusion summarizes the main results of this work and presents some
prospective aspects.

- 20 -

Rsum du chapitre I Etat de lart : La cellulose et sa dissolution dans des solutions aqueuses dhydroxyde de sodium
_____________________________________________________________________________________________________
- 21 -

R RE ES SU UM ME E D DU U C CH HA AP PI IT TR RE E I I
E ET TA AT T D DE E L L A AR RT T: : L LA A C CE EL LL LU UL LO OS SE E E ET T S SA A D DI IS SS SO OL LU UT TI IO ON N D DA AN NS S
D DE ES S S SO OL LU UT TI IO ON NS S A AQ QU UE EU US SE ES S D D H HY YD DR RO OX XY YD DE E D DE E S SO OD DI IU UM M
Les polysaccharides sont des matriaux naturels, renouvelables et biodgradables. Par consquent
leur utilisation dans lindustrie, en tant que matire premire ou en tant quadditif, se dveloppe.
La cellulose, par exemple, est le constituant principal des vgtaux ; sa production naturelle est
donc considrable. Par contre, elle est difficile mettre en uvre et doit subir de nombreux
traitements avant de se retrouver dans le produit fini.
La cellulose ne peut pas tre fondue comme la plupart des polymres et ne peut tre dissoute ni
dans leau ni dans la plupart des solvants. Sa dissolution ncessite lemploi de solvants forts,
souvent dangereux et polluants. De nombreux travaux de recherche portent donc sur la mise en
solution de la cellulose dans des solvants peu coteux, efficaces et non dangereux pour
lenvironnement.

Le but de ce chapitre est de faire un tat de lart sur la cellulose et sa dissolution en milieu alcalin.

Dans un premier temps, nous prsenterons la cellulose du point de vue de son organisation
molculaire et structurale. Limportant rseau de liaisons hydrognes permet non seulement de
nombreuses combinaisons donnant ainsi quatre structures cristallines diffrentes mais galement
une trs forte cohsion entre les chanes, rendant difficile la pntration des solvants. Nous
verrons que lorganisation supramolculaire des chanes de polymre dpend de lorigine de la
cellulose (coton ou bois par exemple). Il existe galement diffrentes mthodes dextraction, de
traitement et dactivation des fibres pour prparer les pulpes de cellulose. Tous ces aspects
influenant sur les proprits des pulpes de cellulose rendent difficile luniversalit des rsultats et
des interprtations.

La deuxime partie de ce chapitre aborde deux techniques de dissolution de la cellulose.
La premire intervient dans le procd viscose , procd actuel de fabrication de films minces
ou dobjets tridimensionnels base de cellulose. Il sagit dune dissolution indirecte : la cellulose
est dabord transforme chimiquement avant dtre dissoute puis mise en forme et enfin
rgnre (retour la structure de la cellulose). La succession de ractions chimiques et
lutilisation de sulfure de carbone (CS
2
) comme solvant de la cellulose le rendent trs polluant.
Ceci a conduit les industriels trouver une alternative ce procd.

Rsum du chapitre I Etat de lart : La cellulose et sa dissolution dans des solutions aqueuses dhydroxyde de sodium
_____________________________________________________________________________________________________
- 22 -
Une deuxime mthode de dissolution des fibres de cellulose est alors tudie. Il sagit, cette fois,
dune dissolution directe de la cellulose dans une solution peu concentre (7-10%) dhydroxyde
de sodium/eau. La soude fut dabord utilise comme agent de gonflement des fibres de cellulose
avant dtre utilise comme un solvant direct de la cellulose. Nous rapportons dans cette partie de
nombreux travaux effectus sur les solutions de cellulose/NaOH/eau. Si les premiers traitent
essentiellement des conditions de dissolution de la cellulose (concentration en NaOH,
temprature, additifs), les suivants cherchent davantage expliquer les phnomnes observs
(gonflement irrgulier des fibres, mcanisme de transformation de la cellulose I la cellulose II,
glification des solutions de cellulose, rle des additifs sur la dissolution des fibres et leur
influence sur la glification).

Chapter I State of the art: cellulose and its dissolution in alkaline hydroxide aqueous solutions
_____________________________________________________________________________________________________
- 23 -

C
C
H
H
A
A
P
P
T
T
E
E
R
R
I
I

I- I
S ST TA AT TE E O OF F T TH HE E A AR RT T
: : C CE EL LL LU UL LO OS SE E A AN ND D I IT TS S
D DI IS SS SO OL LU UT TI IO ON N I IN N A AL LK KA AL LI IN NE E H HY YD DR RO OX XI ID DE E A AQ QU UE EO OU US S
S SO OL LU UT TI IO ON NS S
The general goal of our work is to determine the structure of cellulose solutions in aqueous
sodium hydroxide and to understand the mechanisms of cellulose dissolution and gelation of
solutions. So flow behaviour and gelation properties were investigated as well as the influence of
additives (urea and zinc oxide) on cellulose dissolution and kinetics of gelation. In order to
understand all this, it is necessary to know and understand the structure and properties of
cellulose itself and to review what is known in literature about cellulose behaviour in aqueous
alkaline solutions. Thus, there will be two main parts in this chapter: Cellulose: structure,
morphology and treatment and Cellulose dissolution in aqueous alkali hydroxide solutions.

In the first part, the structure of cellulose and its polymorphs are briefly presented in order to
describe the crystalline structure of cellulose and to better understand specificities of cellulose
materials. For example, we will see how the network of hydrogen bonding changes depending on
polymorph of cellulose I and II which explains differences in stability and reactivity. The supra-
molecular structure of cellulose of different origins is shown to emphasize the complexity of fibre
arrangement. These two points have a great interest because supra-molecular structure and
hydrogen bonding have to be broken to dissolve cellulose. Finally, different treatments of
cellulose are described because we will use microcrystalline cellulose and steam exploded cellulose
and thus we must know what their specificities are.

In the second part, the first section concerns the current industrial process to dissolve cellulose:
the viscose process. It is still used by our partners but as this process generates some pollution,
_____________________________________________________________________________________________________
- 24 -
an alternative route to dissolve cellulose must be found. The others sections deal with another
cellulose solvent based on NaOH aqueous solution. Varying the nature of alkali, the
concentrations, and the temperature enables to better understand what happens when alkali
hydroxide aqueous is added to cellulose fibres. Some explanations on mercerisation mechanism
are given. Finally, the action of some additives to improve cellulose dissolution will be described.
_____________________________________________________________________________________________________
- 25 -
I.1- CELLULOSE: STRUCTURE, MORPHOLOGY AND TREATMENT
Cellulose is one of the many polysaccharides present in nature (part I.1.1-). Its chemical structure
is relatively simple (part I.1.2-) but the arrangement of chains in crystalline (part I.1.3-) and in
supra-molecular structures is very complex and depends on the origin of cellulose (part I.1.4-).
The synthesis of cellulose is difficult; nowadays only chains with about ten units can be obtained.
So native cellulose is used as a raw material, after treatments to extract it from plants and activate
it (part I.1.5-).

I.1.1- Generalities
Polysaccharides are the most abundant organic polymers available world-wide. Photosynthesis
produces approximately 1.5 10
12
tonnes of cellulose, starch and other natural polymers, whereas
the annual production of petroleum polymers is 7-8 times lower. The economical interest of
cellulose as a raw material is thus evident. Humans use 5.10
8
tonnes per year of cellulose as wood,
cotton, paper, textile, wrapping films

Plants are natural composite materials and chemical complexes of cellulose, lignin,
hemicelluloses, and other chemicals. The main ingredients are as follows:
o Cellulose is the main component of plant cell walls. It is the structural substance, which
allows protecting and stalking. Cellulose content varies from 30% to 50% depending on the
source, for example, in straws, hard- and softwood, early- and latewood, algae. Cotton is the
purest natural form of cellulose that reaches more than 98% of the whole composition.
Cellulose represents around 50% of the whole biomass. It can also be produced by some
bacteria and found in skeleton of sea animals (tunicin).
o Lignin is the encrusting substance solidifying the cell wall since lignin resists strongly to
compression. One of its important chemical properties is hydrophobicity. Consequently, cells
become impermeable and can transport water and mineral salts. Lignin is a group of chemical
compounds formed by carbon rings. The proportion of lignin varies from 15% to 30%
depending on species and represents about 20% of the whole biomass.
o Hemicelluloses are the matrix substances and provide a linkage between cellulose and lignin.
Hemicelluloses are shorter chains of polysaccharides. They can be divided into 3 groups:
xylans, mannans, and galactans. Their contents vary from 20% to 30%.
o Other compounds consist of a wide variety of chemicals as resins, fats, silica crystals,
proteins In spite of their small content, extraneous compounds significantly influence the
final properties and characteristics of cellulosic materials.

People have used cellulose in a natural state for several thousand years. However, its processing
and use as a raw material in the chemical industry started 150 years ago, with the discovery of
cellulose derivatives, and has ever since been studied. Although the chemical structure of
cellulose is well-known, there are still many debates concerning its crystalline structure, i.e. the
_____________________________________________________________________________________________________
- 26 -
way in which the cellulose molecules are packed in crystals and how these crystals are assembled
into microfibrils, fibres, cell walls or other cellulose morphologies.

I.1.2- Chemical structure of cellulose
Cellulose is a linear syndiotactic homopolymer of -(14)-D-glucopyranose units in
4
C
1

conformation (Fig.I.1-1).

Non-reducing end group Anhydroglucose unit Reducing end group
(n = value of DP)
O
OH
OH
HO
OH
HO
OH
OH
OH
H
O
Reducing end group with
a hemiacetal function
Reducing end group with
an aldehyde function
#

Fig.I.1-1: Molecular structure of cellulose (on the top) and
the reaction occurring on the reducing end group (on the bottom)
# indicates the anomeric carbon C(OR)(OR)

Cellulose is based on glucose units having taken the shape of 6-membered rings, called pyranoses
or glucopyranoses. They are connected by a single oxygen atom (acetal linkages) between the C1
of one pyranose ring and the C4 of the next ring. During the chemical reaction forming the acetal
linkage, a molecule of water is given off. Consequently, the glucose units in the cellulose polymer
are referred to as anhydroglucose units. The spatial arrangement, or stereochemistry, of these
acetal linkages is very important: when the hydroxyl group at C1 is on the same face of the ring as
the C6 carbon, it is said to be in the configuration where all functional groups are in equatorial
positions. This causes an extended conformation cellulose molecular chain (a semi-rigid
conformation), making it a good fibre-forming polymer.

Each of the anhydroglucose units possesses hydroxyl groups at C2, C3 and C6 positions, capable
of undergoing the typical reactions known for primary and secondary alcohols. Because of these
reactions, a lot of cellulose derivatives, such as methylcellulose (-CH
3
groups substitute -OH
groups), can be obtained.
The two chain ends are chemically different. The non-reducing end has a D-glucopyranose unit
in which the anomeric carbon atom (indicated by # in the Fig.I.1-1) is involved in a glycosidic
linkage, whereas the reducing end has a D-glucopyranose unit in which the anomeric carbon
atom is free. The cyclic hemiacetal function is in equilibrium with an aldehyde function (as shown
Fig.I.1-1), which gives rise to reducing properties, so that the cellulose chain has a chemical
polarity. Determination of the relative orientation of cellulose chains in the three-dimensional
structure has been and remains one of the major problems in the study of cellulose.
#
#
#
#
1
2
3
4
5
6
_____________________________________________________________________________________________________
- 27 -
Because of the equatorial positions of the hydroxyls on the cellulose chain, they protrude laterally
along the extended molecule. This positioning makes them readily available for hydrogen
bonding with the ring oxygen atoms. This hydrogen bonding has a strong tendency to aggregate
chains and to form highly ordered structural entities. Since the chains are usually longer than the
crystalline regions, they are thought to pass through several different crystalline regions, with
areas of disorder in between. The inter-chain hydrogen bonds in the crystalline regions are
strong, giving the resultant fibres good strength and insolubility in most solvents. They also
prevent cellulose from melting. In the less ordered regions, the chains are further apart and more
available for hydrogen bonding with other molecules, such as water. Most cellulose structures can
absorb large quantities of water.

I.1.3- Crystalline structures of cellulose
The hydroxyl groups present in the cellulose macromolecule are likely to be involved in a number
of intra- and intermolecular hydrogen bonds, which may give rise to various ordered crystalline
arrangements. Four main polymorphs of cellulose have been found and they are named as
cellulose I, II, III and IV. Their brief description is given in the following sections and their
crystalline parameters are shown in Tab.I.1-1.

Unit cell (, )
Type
Chains
number in
a unit cell a b c
I

I

II
III1
IV1
IV2
1
2
2
2
2
2
6,74
7,85
8,10
10,25
8,03
7,99
5,93
8,27
9,05
7,78
8,13
8,10
10,36
10,38
10,31
10,34
10,34
10,34
117
90
90
90
90
90
113
90
90
90
90
90
81
96,3
117,1
122,4
90
90
Tab.I.1-1 [ZUG2001]: The parameters of crystalline unit cell of cellulose polymorphs
a, b and c correspond to the axes of unit cell;
, , are the angles between b and c, a and c, a and b , respectively.

I.1.3.1. Native cellulose or cellulose I
Studies on crystalline structure of cellulose started almost one century ago and have generated
many debates. In his recent paper, Zugenmaier [ZUG2001] reported a historical development of
the structure determination of native cellulose. By various techniques, such as X-ray and electron
diffractions, NMR, molecular modelling studies, the global structure has been defined and only
the fine structure is expected to be improved.

_____________________________________________________________________________________________________
- 28 -

Fig.I.1-2 [ZUG2001]: Representation of the model of cellulose I a) onto the a-b plan,
b) intrasheet (100) of molecules through the centre of the unit cell to demonstrate the hydrogen bonding.
a and b axes represent the unit cell vectors as a<b; c is the third axis of the unit cell and corresponds to the fibre axis;
is the angle between a and b

The natural cellulose crystal is metastable and called cellulose I. Cellulose I has a parallel
crystalline structure [ELL1962][GAR1974][SAR1974][MAC1990] and is composed of two
distinct crystalline forms I
and I
[ATA1984], the latter being thermodynamically more stable.

Cellulose can also be a composite of I
, which has a triclinic unit with one chain, and of I
, which
has a monoclinic unit with two chains. Fig.I.1-2 represents (a) the typical arrangement of parallel
chains of cellulose I composed of the monoclinic unit of cellulose I
, and (b) O3-HO5 and

O2-HO6 intramolecular and O6-HO3 intermolecular hydrogen bonds. Between two
molecular layers, only Van der Walls forces occur, there are no hydrogen bonds. The proportion
between the two crystalline phases I
and I
depends on the cellulose origin [ATA1999].

Cellulose produced by primitive organisms (bacteria, algae etc.) is enriched with the I
phase
whereas cellulose of higher plants (woody tissues, cotton, ramie etc.) consists mainly of the I

phase.

I.1.3.2. Cellulose II
The crystalline structure of cellulose II is obtained when the native cellulose is either mercerised
(alkaline treatment) or regenerated (dissolution and then precipitation). Cellulose II has an
antiparallel crystalline structure organised in a monoclinic unit with two chains (Fig.I.1-3a)
[SAR1974][KOL1976][STI1976]. After many controversies, in 90s, Isogai et al. [ISO1989] and
Raymond et al. [RAY1995] demonstrated that all hydroxyl groups of cellulose II (in C6) are
oriented gauche-trans (gt), e.i. O61-C61 is positioned gauche to O51-C51 and trans to C41-C51
(Fig.I.1-3b). They also revealed that the central chain is shifted as compared with the cellulose
chains at the origin of the unit cell [RAY1995]. Moreover, cellulose II has a complex network of
hydrogen bonds inside and between molecular layers, which was precisely analysed by Langan et
al. [LAN1999]. These numerous hydrogen bonds induce a greater stability of cellulose II as
compared with cellulose I.
a) b)
_____________________________________________________________________________________________________
- 29 -

Fig.I.1-3 [ZUG2001]: Representation of the model of cellulose II a) onto the a-b plan,
b) intrasheet (010) of molecules through the origin of the unit cell.
a and b axes represent the unit cell vectors as a<b; c is the third axis of the unit cell and corresponds to the fibre axis;
is the angle between a and b

I.1.3.3. Cellulose III
Liquid ammonia treatment of cellulose gives another polymorph: cellulose III. It can be divided
into cellulose III
I
and III
II
obtained from cellulose I and cellulose II respectively. The
transformation of cellulose I to III
I
and of cellulose II to III
II
is reversible. The unit cell for both
cellulose III
I
and III
II
structures is the same but the meridional reflections differ. The chains
arrangement corresponds to the one of the initial cellulose, that is to say parallel for cellulose III
I

and antiparallel for cellulose III
II
[SAR1976]. In 2001, Wada et al. [WAD2001] reported that the
unit cell of cellulose III
I
contains one chain and has a volume twice as small as the one of the
model proposed by Sarko et al. [SAR1976].

I.1.3.4. Cellulose IV
Heating of cellulose III produces the crystalline structure of cellulose IV, thermodynamically
stable. As well as for cellulose III, the two allomorphs IV
I
and IV
II
originate from cellulose I and
cellulose II respectively. The unit cell is orthogonal [GAR1985].

I.1.4- Supra-molecular structure of cellulose
I.1.4.1. Micro-fibrils
The chemical structure of cellulose (part I.1.2-), with a significant network of intra- and
intermolecular hydrogen bonds, generates chain aggregations and highly ordered structures.
In the plasma membrane of cells, enzymes called cellulose synthetases synthesise cellulose
molecules by glucose addition, the degree of polymerisation DP (= number of anhydroglucose
units) reaching several thousand units. As several molecules are synthesised simultaneously in the
same area, they grow in a parallel way [Web site 1]. These parallel cellulose molecules create
hydrogen bonds between them to form the basic crystalline arrangement of cellulose into the
micro-fibrils (Fig.I.1-4). Micro-fibrils have a diameter in the range of 2-20 nm depending on
a)
b)
_____________________________________________________________________________________________________
- 30 -
cellulose origin, and are agglomerated into fibrils. The latter have diameters in the range of 60-
360 m and the arrangement of micro-fibrils also depends on cellulose source.

Cellulose fibres
Macro-fibrils
Micro-fibrils
Cellulose molecules:
glucose chains

Fig.I.1-4 [Web site 2]: Supramolecular structure of cellulose

I.1.4.2. Wood and cotton cellulose fibres
Wood can be divided into two main classes: hardwood (or deciduous wood) that is generally
composed of packed cells with thick walls (Fig.I.1-5) and softwood (or coniferous wood) usually
composed of large cells with thin walls. As we saw previously (part I.1.1-), cells contain cellulose
and also high proportions of lignin, hemicelluloses, etc.
Cotton is a seed hair, unicellular and thus very pure in cellulose. The cotton hair is practically
cylindrical and contains a central canal known as lumen (Fig.I.1-6). When the hair has been
removed from the seed, the cell collapses into a flat ribbon which twists into an irregular spiral
band [MAR1941].

Despite the differences in their origin, the morphological architecture of fibres presents some
similarities: the organisation in layers of fibrils (Fig.I.1-7). On the other hand, the fibrillar texture
and density change according to species [KLE1998].

_____________________________________________________________________________________________________
- 31 -

Fig.I.1-5: Wood is composed of dead cells.
Cellular walls are a composite material of cellulose fibres
and lignin.

Fig.I.1-6: Cotton flower and cotton fibre cross-section.
The blue circle shows a mature cotton fibre, with a thick
and well-developed wall. The red circle shows an
immature cotton fibre with a thin and poorly developed
wall.

Fig.I.1-7 [KLE1998]: Scheme of the morphological architecture of a cotton and delignified spruce fibres:
(A) cotton fibre (B) delignified spruce wood fibre
C = cuticle (rich in pectins and waxes) ML = middle lamella (mainly lignin)
P = primary wall
S1 = secondary wall (winding layer)
S2 = secondary wall (main body)
R = reversal of the fibril spiral T = tertiary wall
L = lumen W = wart layer

The so-called primary wall (P) forms when cells are individualised and it follows their growth
[AIT1988]. In this primary wall, fibrils of about 10nm in diameter are positioned crosswise to a
layer of about 50nm thickness. In his review, Klemm et al. [KLE1998] reported that this
crosswise positioning possibly impedes a swelling of the inner secondary wall (S).
_____________________________________________________________________________________________________
- 32 -
When the cell has reached its definitive size, the secondary wall grows with fine lamellae inside
the cell. The secondary wall consists of many lamellae of cellulose, which are not separated from
one another by non-cellulosic substances but represent dense and less dense areas of a cellulose
fibre. This wall can be divided into two layers S
1
and S
2
.
The thickness of the S
1
layer is about 100nm and 300nm in the case of cotton and spruce
pulp fibre respectively. The fibrils are aligned parallel and densely packed. The angle between
the microfibril and the cell axis is large (50-70), and the direction of helix may be opposite in
subsequent S
1
layers. This structure in S
1
is called crossed fibrillar texture. Moreover, the S
1

layer can strongly impede swelling of the S
2
layer.
The S
2
layer is the thickest within the layers of secondary wall and contains most of the
cellulose mass. It is a compact region in which a high degree of parallelism of microfibrils
exists. The S
2
has a small microfibrils angle to the cell axis (10-30). In wood, despite this
microfibrillar orientation in S
2
, there are lamellae transitions on its inner and outer surfaces.
Several lamellae show a gradual shift of microfibril angles between S
1
and S
2
and between S
2

and S
3
[HON2001]. In cotton, there are some areas in the secondary wall where cellulose
fibrils change from right-hand to left-hand, and vice-versa, at frequent intervals. One type of
reversal (R) is that in which one set of spiral strands ends and a second system running in the
opposite direction begins (Fig.I.1-7A); the ends of the strands overlap at the reversal
[MAR1941].
The inner layer, closest to the fibre lumen, named S
3
for cotton or T for wood, is comparably
thin and has the fibrils aligned in a flat helix as in the layer S
1
.

I.1.5- Extraction of cellulose and activation
I.1.5.1. Manufacture of cellulosic pulp
The manufacture of cellulosic pulp consists of extracting cellulose fibres from the biomass. All
methods induce a progressive impoverishment of the cell wall, i.e. a decrease of the amount of
lignin, hemicelluloses and other compounds, in order to obtain a high proportion of cellulose. As
a result, individualised cellulose fibres are obtained with given morphological, optical and
mechanical characteristics, depending on the source. Pulps can be classified according to the
extraction technique.

o Mechanical pulps
Wood is crushed in order to tear out fibres. With this technique, the yield is significant since pulp
keeps all components of wood lignin, cellulose and hemicelluloses. The use of steam, high
pressure or chemical agents allows obtaining pulps with different properties. As the resistance of
these mechanical pulps is relatively low, they are especially used for newspapers and magazines.

_____________________________________________________________________________________________________
- 33 -
o Chemical pulps
In this technique, only chemical products are added to separate fibres by dissolving lignin.
According to the compound used, different types of pulp are obtained:
The two alkaline processes: (i) the so-called sulphate or Kraft process and (ii) soda
process. These two processes are relatively similar but the former is cheaper and thus
more often used.
The acid process: so-called sulphite process.
These processes are suitable for soft- and hardwood. The main principle consists in cooking
chips in a chemical reactor with white liquor, a liquid composed of chemicals. Cooking is
occurring from 130C to 180C, under pressure, during 2-5 hours depending on wood species.
After this treatment, chips are in suspension in black liquor, a liquid containing chemicals and
dissolved lignin. After defibration, fibres become individualised and flexible. Then, washing and
bleaching steps are following.
An important feature of the alkaline processes is the recovery and the re-utilisation of the
chemicals in the digester liquor and washings after digestion. This serves the double purpose of
economy and elimination of the trouble of disposing of a polluting effluent [GRA1958]. Pulps
obtained from alkaline process have a good mechanical resistance and thus can be used as they
are in wrapping. Despite of a difficult bleaching, these alkaline pulps are also used for printing
paper.
From sulphite process, pulps have lower mechanical properties, especially in tear strength, but
they are light and easy to bleach. However, for environmental reasons, they are rarely used,
except for very refined papers.

o Mi-chemical pulps
Fibres are obtained from moderate chemical and mechanical treatments. They are especially used
in flute papers since fibres are individualised but some lignin is still present. The presence of
lignin rigidifies fibres and induces a high quality of compression resistance.

Other than from wood, pulps can be made of:
Straw: in specific papers with high transparency and rigidity
Cotton and flax: in very resistant papers as bank notes and cigarette papers
Recycled cellulosic fibres: in newspapers and cardboards.

I.1.5.2. Activation of cellulose
As we saw previously, the density of hydroxyl groups in cellulose molecules is quite high. In this
sense, it should not be surprising if cellulose were completely water-soluble. However, this has
never been realised owing to a very high intra- and intermolecular hydrogen bonds in natural
cellulose. These bonds prevent the dissolution of cellulose fibres in aqueous solvents. A so-called
activation of cellulose improves its dissolution by breaking down intramolecular hydrogen bonds,
by decreasing the DP (degree of polymerisation) of cellulose molecules, by removing some layers
_____________________________________________________________________________________________________
- 34 -
of cellulose fibres and by purifying cellulose. Two examples of treatments to activate cellulose are
given below.

o Microcrystalline cellulose
Microcrystalline cellulose (MCC) is a purified and partially depolymerised cellulose. This cellulose
is obtained by a controlled hydrolysis with a diluted mineral acid solution, at ~105C. The fibrous
structure is thus broken and gives a particular structure. In addition, the DP is lowered down to
200. This change of morphology is due to the preferential hydrolysis of cellulose chains in
amorphous region. So MCC cellulose has a high degree of crystallinity. Further filtration of
cellulose suspension enables to purify the sample. The process is schematically shown in Fig.I.1-8

Fig.I.1-8: Process to obtain microcrystalline cellulose
1- Roll of -cellulose; 2- Chipping; 3- Acid hydrolysis;
4- Filtration and washing to remove impurities and chemicals;
5- Drying on a fluidised bed; 6- packaging
Conditions to produce microcrystalline cellulose, which stay trade secrets, considerably influence
the quality of the final cellulose pulp.

o Steam exploded pulp
The steam explosion method has the potential ability to destroy the inner and outer structure of
any material. The destruction of the inner structure of cellulose corresponds to the breakdown of
O3-HO5 intramolecular hydrogen bonds, which seems to be the key for producing alkali-
soluble cellulose, as demonstrated by Kamide et al. [KAM1984] with solid-state cross-
polarisation/magic angle sample spinning
13
C NMR (CP/MAS/
13
C NMR). The destruction of
the outer structure corresponds to the removing of primary wall of cellulose fibres, which
prevents swelling of secondary wall and thus fibre dissolution.

The steam explosion apparatus was originally designed by Sawada [SAW1985] and the explosion
process is schematised Fig.I.1-9 [YAM1990].
Stage 1: Cellulose is stored at room temperature.
Stage 2: In order to obtain a weight ratio water/cellulose of 1/1, one weight part of cellulose
is immersed in 20 weight parts of distilled water at 25C for 15h; then the excess water is
removed by centrifuging (2000rpm, 5min)
Steam explosion: 200g (dry base) of the so-prepared cellulose is put into the vessel heated at a
desired temperature. The vessel is closed. Then a saturated water vapour controlled to a given
steam pressure P is introduced into the vessel for a given treatment time t (15-300s).
Controlling the time enables to control the final DP.
Stage 3: The vapour input valve is closed and the valve connected to the discharging nozzle is
abruptly opened. Thus the treated cellulose is discharged.
Stage 4: Treated sample is thoroughly washed with water.
_____________________________________________________________________________________________________
- 35 -

Fig.I.1-9 [YAM1990]: Schematic representation of the steam explosion process.

Yamashiki et al. [YAM1990] reported that soft wood pulp is very sensitive to steam explosion
treatment because of the high porosity of fibres. On the other hand, hard wood pulp necessitates
higher pressure to be completely soluble in 9%NaOH/water. Finally, the morphology and
solubility of cotton linter do practically not change after steam explosion treatment.

I.2- CELLULOSE DISSOLUTION IN ALKALINE HYDROXIDE AQUEOUS SOLUTIONS
Cellulose is not a thermoplastic polymer. Thus, cellulose melting is not possible and dissolution is
a necessary step to process cellulose. Because of intra- and intermolecular hydrogen bonds, the
penetration of solvent is difficult and it has to be strong enough to break them. Different types
of solvent can be used; they are largely described and classified by Klemm et al. [KLE1998]. The
oldest process to dissolve cellulose is named viscose process: cellulose is chemically
transformed before being dissolved, this is the xanthation step. But as this process is polluting, an
alternative route is to dissolve cellulose in alkali hydroxide aqueous solutions. After describing the
current industrial process (part I.2.1-), we will present the first studies on this new solvent of
cellulose (part I.2.2.1). Our work was based on these data concerning experimental conditions:
concentrations, temperature, additives. Several authors investigated cellulose/alkali hydroxide
aqueous solutions (parts I.2.2.2 and I.2.2.3) in order to understand the dissolution mechanism.
Some hypotheses are given concerning the mechanism at the level of fibres and microfibrils (part
I.2.2.4), but there are practically no literature dealing with the mechanism at the molecular level.
Concerning additives, it will be shown that the influence of the proportion between NaOH and
urea or thiourea or zinc oxide was studied in details (part I.2.3-). However, the mechanism of the
action of these additives still remains unclear, despites a rather large amount of publications.

Pure cellulose
(Pulp, cotton)
Wetted cellulose
(Pulp, cotton)
Water
Stocked at Troom
Water content = 10%
Water content = 100%
Steam Explosion
As-steam-treated
cellulose
(Pulp, cotton)
Water content = 200%
Water content = 10%
Washed steam-treated
cellulose
(Pulp, cotton)
P = 15 bars
t = X sec.
Stage 1
Stage 2
Stage 3
Stage 4
Washing
_____________________________________________________________________________________________________
- 36 -
I.2.1- Xanthation in viscose process
The viscose process was discovered in 1892 by Cross, Bevan and Beadley and allowed spinning
of viscose fibres. Still nowadays, the step of xanthation of this process is a method widely used in
industry to dissolve cellulose. This process can be divided into three main chemical steps, which
are identical in most of the industrial viscose processes (Fig.I.2-1).

Fig.I.2-1: Scheme of viscose process
(1) formation of alkali-cellulose, (2) xanthation (3) dissolution

(1) Cellulose, prepared from either wood pulp or, less commonly, cotton linters, is treated with
17-20% sodium hydroxide (NaOH), at a temperature in the range of 18-25C, to convert
cellulose to alkali-cellulose.
Cell-OH + NaOH Cell-O
-
,Na
+
+ H
2
O
Cellulose fibres are swelling and become more accessible to other reactants. Then the alkali
cellulose is aged under controlled conditions of time and temperature (between 18 and 30C)
in order to depolymerise the cellulose to the desired degree of polymerisation (DP). The
molecular weight obtained determines the viscosity of viscose solution and cellulose
concentration.

(2) After this maturing step, carbon disulfide (CS
2
), in gas or liquid state, is added to the solution
of alkali-cellulose to react with hydroxyl groups and generate the xanthation reaction:
S
Cell-O
-
,Na
+

+ CS
2
Cell-O-C
S
-
,Na
+
The resulting product is a substance called cellulose xanthate.

(3) Then this alkaline cellulose xanthate is dissolved in dilute sodium hydroxide (~2.7 wt%) and
forms a viscous solution called viscose.
S S
-
,Na
+

Cell-O-C + NaOH + H
2
O Cell-O-C OH + H
2
O
S
-
,Na
+
S
-
,Na
+
(1)
(2)
(3)
_____________________________________________________________________________________________________
- 37 -
The large xanthate substituents on the cellulose force the chains apart, reducing the
interchain hydrogen bonds and allowing water molecules to solvate and separate the chains.

After obtaining viscose, industrial processes vary. The next steps can be: filtration and then
spinning to obtain fibres or thin films, addition of reinforcing fibres and porophores then
moulding to obtain sponges (see Chapter V), etc.
The last step of processes is called coagulation: cellulose is precipitating in regeneration baths
(water, acid or base). The final regenerated cellulose is obtained and CS
2
and sulphured
compounds formed during chemical reactions are going out.

However, viscose process involves some pollution: in the atmosphere, because of CS
2
and H
2
S,
and in water, because of the reducing agents (containing sulphur compounds) which decrease the
oxygen amount in water. Optimisation and improvement in the viscose process result in a total
recycling of the pollutant compounds in the aqueous phase and in a decrease of pollutant
compounds in the atmosphere. Nowadays the quantity of the latter is only 10% of CS
2
used
initially. Unfortunately, technologies for pollution treatment are very expensive. Thus for
economical and for environmental reasons, it is very important to find new cellulose solvents.
The purpose of research laboratories is to step away from the chemical modifications of cellulose
by CS
2
. NMMO, for example, is a good cellulose solvent, but these solutions are not very easy to
prepare and to process. That is why it is interesting to look for alternative routes.

I.2.2- Action of alkaline hydroxide solutions on cellulose fibres
For a long time, it was known that strong alkaline aqueous solutions act on cellulose. Indeed, at
the end of 19
th
century, Mercer [MER1903] discovered that cellulose fibres are swelling in
aqueous solutions of sodium hydroxide and that was drastically changing the properties of the
fibres. Several years later, in 1939, Sobue showed that cellulose can be dissolved in a narrow
range of temperature and concentration of NaOH [SOB1939].

I.2.2.1. Introduction on mercerisation process and cellulose dissolution
Mercerisation is a process, which John Mercer developed between 1844 and 1850. Cotton cloth is
immersed in a strong caustic alkaline solution, and then washed, in order to improve the lustre
and smoothness, for example. Mercerisation also increases the ability to absorb dye, improves the
reactions with a variety of chemicals, the strength and elongation of fibres and the stability of
form. This type of cotton is often compared to silk and is especially used to produce high-quality
fabrics. But this process tended to shrink the cotton cloth. In 1889, Horace Lowe improved this
technique by keeping the material under tension whilst being mercerised and he applied a more
thorough washing process to remove the caustic soda. Thus the mercerisation became a viable
textile process.

During mercerisation, alkaline solution acts on cellulose chains by changing the fine structure,
morphology and conformation of cotton. The native cellulose I crystalline form is transformed
_____________________________________________________________________________________________________
- 38 -
into cellulose II, resulting in a variation in fibre strength and lustre as well as adsorption
properties. In addition, soda fills up almost entirely the central cavity (the lumen) of cotton fibres.
So, the ribbon-like cotton fibres become perfectly cylindrical and loose their convolutions
inducing a smoother and shinier texture [MAR1941]. In mercerising, a decrease in length of
cotton hair is accompanied by an increase in diameter: cellulose fibres are swelling.

According to Mercer [MER1903], mercerisation also occurs at low NaOH concentration
(~9%NaOH) and the properties of cotton are improved. The lower the temperature, the more
effectively the soda acts on the fibrous material. This reveals that the term of mercerisation was
very wide and corresponded to all changes in cellulose fibres due to the action of alkaline
hydroxide aqueous solutions. Several conditions of temperatures and NaOH concentrations are
reported depending on material (cotton, ramie, flax, yarn or cloth) and on process. The most
often used conditions are 18-32% NaOH concentrations at 25-40C. It is important to note that
hardly few minutes (from 20s to 3min) are enough to obtain mercerisation of cellulose.
Nowadays, the term of mercerisation essentially represents the action of concentrated alkali
hydroxides on cellulose.

The NaOH concentration and temperature conditions to swell cellulose fibres are relatively well
determined and slightly change depending on cellulose origin. In 30s, Davidson defined optimal
conditions to dissolve modified cotton. As illustrated Fig.I.2-2, a decrease of temperature
improves hydrocellulose dissolution in sodium hydroxide solutions [DAV1934]
(hydrocellulose was obtained by cellulose hydrolysis with strong acids).

Fig.I.2-2 [DAV1934]: Hydrocellulose solubility versus
NaOH concentration and solution temperature

The maximum solubility is 80% and it occurs at NaOH concentration of 10%, at 5C.
Moreover, the solubility increases when the chain length decreases. Consequently, authors
concluded that unmodified cellulose cannot be dissolved because of the large length of the
macromolecules. But, in 1939, Sobue et al [SOB1939] showed that it is possible to dissolve
_____________________________________________________________________________________________________
- 39 -
cellulose, from natural ramie fibres, in a narrow range of phase diagram: between NaOH
concentrations of 7% and 10% at low temperature (-5C/+1C).

In the 80s-90s, cellulose dissolution in 7-10%NaOH/water was largely studied. Kamide et al.
[KAM1984] revealed that regenerated cellulose can be dissolved in NaOH/water and that the
dissolution depends on the degree of breakdown in O3-HO5 intramolecular hydrogen bonds.
They also showed that the cellulose solubility depends on the DP, the concentration and the
crystallinity of cellulose samples. Some years later [KAM1990], the authors also compared the X-
ray diffraction peak intensities of cellulose I (I
1
) and Na-Cell I (I
2
) as a function of cellulose
concentration in 9%NaOH/water. Fig.I.2-3 represents the results obtained. It appears that up to
9% of steam exploded cellulose (DP being ~340), there is no peak of cellulose I and no peak of
Na-cell I. This means that 9%NaOH/water totally dissolves cellulose when its concentration is
inferior to 9% and that the total dissolution of cellulose I (C
cell
< 9%) in 9%NaOH aqueous
solution gives a solution without conversion of cellulose I to Na-Cell I.

Fig.I.2-3 [KAM1990]: The cellulose concentration dependence of I1 and I2
for cellulose/9%NaOH aqueous system at 4C.

More recently, it was shown that a ground Borregaard cellulose with a DP of 550 is swelling in a
heterogeneous way in contact with a 9%NaOH/water solution [ROY_PhD]: some cellulose
fibres remain non-swollen and non-dissolved, some others are partly dissolved and some others
are swollen with balloons. The formation and the reason of the appearance of these balloons
were largely studied in our laboratory these three past years, on different native cellulose fibres
(cotton, wood) in mixtures of NMMO/water [CUI2006a] and in NaOH/water with or
without additives (ZnO and urea) [CUI2006b]. In good solvent of cellulose, the dissolution is
occurring in four phases. First, the cellulose fibre swells by ballooning, with cellulose being totally
dissolved inside the balloons. This phase gives the beaded structure often observed in relatively
bad cellulose solvent. Secondly, the balloons continue to swell until the burst of the membrane.
_____________________________________________________________________________________________________
- 40 -
Then the unswollen sections (between two balloons) are dissolved from the surface to the centre
of the fibre. Finally, the scraps of the membrane are then dissolved. At this stage, all cellulose is
dissolved [CUI2006a]. In a worse cellulose solvent, the dissolution of cellulose fibres is stopping
at one of these phases. This is the case when 7.6%NaOH/water is used at -5C to dissolve native
cellulose fibres; the balloons are formed but do not burst [CUI2006b]. This paper also revealed
that the presence of urea increases the quality of the alkaline solvent by a faster balloon
formation, with a larger expansion. On the other hand, the presence of ZnO increases the
swelling kinetics but decreases the size of balloons.
When cellulose is dissolved, even partly, in 9%NaOH/water, the solutions obtained are not
stable. Gelation is occurring and the gel time decreases when the temperature increases
[ROY2003].

I.2.2.2. Formation of alkali-celluloses
Mercer [MER1903] listed the influence of NaOH concentrations on cellulose fibre swelling:
3.2% NaOH is without action
6.5% NaOH acts only very slightly and at very low temperature
10% NaOH increases the effect which remains most marked at the lowest temperature
18.8% NaOH greatly increases the effect on cellulose.
This seems to be the outline of the cellulose/NaOH/H
2
O phase diagram (Erreur ! Source du
renvoi introuvable.) plotted by Sobue et al [SOB1939]. Fig.I.2-4 shows the presence of five
different forms of alkali celluloses, or Na-Cell (I, II, III, IV or Q and V).

Fig.I.2-4 [SOB1939]: Phase diagram of the system cellulose/NaOH/water, cellulose being natural ramie fibres.
This graph reveals zones of different sodium celluloses as a function of NaOH concentration and temperature.
The unbroken hatching indicates the regions in which the sodium cellulose concerned is optimally formed.

The formation of these Na-Cell compounds depends on the temperature (from -20C to 100C)
and NaOH concentration (from 0% to 45% by weight) but the boundaries of the various regions
_____________________________________________________________________________________________________
- 41 -
are defined more or less distinctly. For example, below 0C and at about 8 wt% NaOH
concentration, is the location of what the authors called the alkali cellulose Na-Cell Q, Q as
Quellung (swelling in German). Its X-Ray diffraction pattern is very characteristic and reveals a
highly disordered state and a swelling in layers [SOB1939].

Sobue et al. [SOB1939] also investigated the composition of alkali celluloses from the correlation
between two methods: the solid-phase analysis and the examination of volumes. The latter is
based on the difference between the volumes of Na-Cellulose unit cell, determined from X-ray
diffraction patterns, and of cellulose unit cell. This difference in volume can be used as a
measurement for the quantity of the substance incorporated into the lattice by substituting the
required volume for NaOH (32.6
3
) and for H
2
O (29.7
3
). Their conclusions are reported
Tab.I.2-1.

Sodium celluloses Most probable composition Molecular weight Density
Na-Cell I C6H10O5 1NaOH 3H2O 256 1.51
Na-Cell III C6H10O5 1NaOH 2H2O 238 1.51
Na-Cell II C6H10O5 1NaOH 1H2O 220 1.6
Na-Cell V C6H10O5 1NaOH 4.5-5H2O 283-292 1.41-1.46
Na-Cell IV C6H10O5 1H2O 180 1.48
Tab.I.2-1 [SOB1939] : Most probable compositions of sodium celluloses
and the molecular weights and densities calculated therefrom.

I.2.2.3. Hydration of alkali ions
In 1941, Marsh [MAR1941] published a review of different works concerning cellulose swelling
in aqueous solutions of alkaline hydroxides as lithium, sodium, and potassium but also rubidium
and caesium. It appears that all alkaline hydroxides can act on cotton hairs and that the maximum
of swelling is reached at a certain hydroxide concentration depending on the nature of metallic
ion (Fig.I.2-5).

Fig.I.2-5 [HEU1925]: Change in volume of cotton hairs in five alkalis
Swelling is measured in terms of volume percentage between swollen and dried fibres.
_____________________________________________________________________________________________________
- 42 -
An increase of the atomic force of the metallic ion Li < Na < K < Rb < Cs leads to a
decrease of swelling. Bigger is the alkali ion, more difficult is its penetration into cellulose fibres
[PET1948] but more uniform is the swelling [SRE1993]

The shape of the swelling versus alkali concentration curves can be explained by the size of
hydrates [FRE1983]. For low alkaline hydroxide concentrations, a lot of water molecules form
hydrated ion pairs (Fig.I.2-6). The hydrodynamic diameter of this system is too large to easily
penetrate into the macromolecular structure of cellulose. Thus the swelling is relatively low.
When ion concentration increases, the number of water molecules decreases to form solvated
dipole hydrates. The hydrodynamic diameter decreases and ions can penetrate into amorphous
and then well-organised domains. The maximum of swelling is reached.

Low-concentrated
solutions
Concentrated
solutions
Pair of separated ions
Pair of hydrated ions
Solvated hydrated dipole
Hydrated dipole

Fig.I.2-6: Various forms and the corresponding hydrodynamic diameter
of hydrates as a function of ion concentration

Another explanation is based on the necessity of the presence of water to swell cellulose
[LEG1952]. If NaOH concentration is too low, the number of Na
+
is insufficient to bring water
molecules into cellulose chain in order to cleave hydrogen bonding. If the NaOH concentration
is too high, the hydration of alkali ions is insufficient to break hydrogen bonding [KUO2005].

Investigating the solubility of oxycellulose (Oxycellulose being obtained after the action of
an oxidant agent on cellulose) in these alkaline hydroxides, Davidson [DAV1936] showed that
they follow the same order as previously for swelling: KOH < NaOH and LiOH, solubility being
less than 5% and 85-90% respectively (Fig.I.2-7).

Due to considerations of cost and efficacy, only caustic soda is used as alkali in mercerisation, to
change the aspect of cellulose fibres, and is considered for a potential use to dissolve cellulose.

_____________________________________________________________________________________________________
- 43 -

Fig.I.2-7 [DAV1936]: Oxycellulose solubility versus
alkali concentration and type of ion, at 5C.

I.2.2.4. Mechanism of mercerisation
It is known that the sodium hydrate size depends on the NaOH concentration [PIC1893] and
that the cellulose dissolution region is very narrow (Fig.I.2-4). In 1990, from an X-Ray diffraction
analysis, Kamide et al. [KAM1990] proved that NaOH penetrates into the cellulose chains
inducing swelling of fibres. Thus we can think that there is an optimal size of sodium hydrate at
which it can penetrate into cellulose fibres. This hypothesis has been mentioned by Yamashiki et
al. [YAM1988], but the problem with this paper is that the molar ratios do not seem correct.
Nevertheless, according to them, Na
+
does not play an important role in the dissolution of
cellulose. The main parameter is to have a specific structure of alkali hydroxide, which is formed
at 9%NaOH. Na
+
and OH
-
ions, surrounded by their solvation cages, break the intra- and
intermolecular hydrogen bonds in the solid structure of the cellulose. Sodium hydrates interact
with the cellulose chains and give various Na-Cell compounds depending on the experimental
conditions as it appears on the phase diagram (Fig.I.2-4) [SOB1939]. If sodium hydroxide is then
removed from the mixture, the alkali cellulose transforms into cellulose II.

As we saw previously (partI.2.2.1), the irreversible transformation from cellulose I, with a parallel
crystalline structure, to cellulose II, with an antiparallel crystalline structure, at room temperature,
is named mercerisation. Okano, Sarko and Nishimura [OKA1984] [OKA1985] [NIS1987a]
[NIS1987b] studied in details the mechanism and the molecular changes of this transformation
and mentioned the complex Na-Cell I as the first intermediate component. This complex has an
antiparallel structure [OKA1985], very similar, in the chain conformation, to the structure of
cellulose II [NIS1991a]. The structure of Na-Cell Q, mentioned by Sobue [SOB1939], was not
analysed extensively because of temperature conditions (below 0C).

A proposed mechanism of mercerisation is schematised on the Fig.I.2-8.
_____________________________________________________________________________________________________
- 44 -

Fig.I.2-8 [OKA1985]: Possible mechanism of mercerisation

The raw material is native cellulose, or cellulose I, which has a parallel crystalline structure. In
a fibre of cellulose, there are a large number of crystallites in which the chain directions are
parallel but the crystallites are distributed along the fibre axis with equal probability for up
and down directions, as represented in the Fig.I.2-8 1
st
pattern. French et al. [FRE1987]
reported that the convention used by the most authors is that the cellulose chain direction is
defined by the vector from C4 to C1, i.e. from the non-reducing to the reducing ends. If the
angle between the vector C4-C1 and the c-direction (fibre axis) of the unit cell is acute, the
chain is called up, if not, it is called down. Let us note that Gardner and Blackwell
[GAR1974], for example, interchanged this notation.

It was demonstrated that the accessibility of OH groups depends on the crystallinity of
cellulose [TAS1994]. Consequently, highly crystalline cellulose is difficult to mercerise
[CHA1976]. Thus, to begin the mercerisation it is necessary to have an amorphous phase in
the cellulose or some imperfections in the crystalline morphology. During the first step of the
mercerisation, the alkali solution enters the amorphous regions that exist as the interface
regions between the crystallites, where the polymer chains are arranged along both directions
up and down. The penetration of NaOH/H
2
O causes swelling of cellulose fibres
[OKA1984] and in these swollen regions, the polymer chains are expected to be relatively
mobile. Thus very little lateral movements of chain segments enable the formation of an
antiparallel soda-cellulose, the Na-Cell I [OKA1985]. In terms of energy and entropy, the
formation of Na-Cell I is favoured and thus the crystalline regions of cellulose I gradually
diminish in size while the crystallites of Na-Cell I increase simultaneously [PET1948]
[NIS1987a]. Furthermore, as the amount of amorphous regions does not increase during the
conversion steps, the authors concluded on a crystal-to-crystal phase transition occurring
without an intervening amorphous state [OKA1984]. This shows that the conversion from
the parallel cellulose I to antiparallel Na-Cell I proceeds through a side slip of polymer chains
and not a reversal of those.

Na-Cell I contains one NaOH molecule per anhydroglucose unit (AGU) and has a twofold
helical chain conformation with a repeat unit of 10.1. Fig.I.2-9 schematises the various
possibilities of interactions between the cellulose chains and the Na
+
ions. Practically all
authors [KAM1985] [NIS1991a] [TAK1991] [FIN1995] [ISO1997] agree that Na
+
breaks the
intermolecular hydrogen bonds O2_HO6 and that the most probable location of the Na
+

ion is on the OH at C2. The carbon C3 is more resistant to the complexation with the
sodium hydroxide.
_____________________________________________________________________________________________________
- 45 -

Fig.I.2-9 [FIN1995]: Interaction between Na
+
and the cellulose chain.

When the initial conversion step is completed, the Na-Cell I is free to absorb more NaOH
and convert to Na-Cell II, with a threefold helical chain conformation with a repeat unit of
15.4.

Finally, after NaOH has been washed out of the structure, the cellulose chains again revert to
a twofold helical conformation (lower energy structure without NaOH), cellulose II
polymorphic form. Later, Nishimura and Sarko [NIS1991b] added the Na-Cell IV as an
intermediate compound between the Na-Cell II and the cellulose II.

I.2.3- Influence of additives on cellulose dissolution
I.2.3.1. Influence of zinc oxide (ZnO)
Mercer [MER1903] reported that the mercerisation can begin at low NaOH concentrations
(8.5%-9%NaOH) when the temperature is decreased below 0C. He added: It is a remarkable fact,
however, that the addition of zinc oxide to a lye containing 9%NaOH will enable it to mercerise at ordinary
temperature. () it is ineffective if the solution is heated.

In the last part of his article, Davidson [DAV1937] showed that ZnO improves oxycellulose
dissolution and that unmodified cotton can be dissolved at 5C, at a relatively high
ZnO/NaOH ratio (Fig.I.2-10 and Tab.I.2-2). He added that precipitation of zinc oxide takes
place if the ZnO/NaOH ratio is too high and especially for concentrated-cellulose solutions.
The last paper talking about the addition of ZnO into cellulose/alkali hydroxide aqueous
solutions is a Polish patent [WO02/22924]. It revealed that in order to improve the stability of
cellulose solutions, zinc oxide content has to be at least 0.1wt% (0.01N), NaOH concentration
being inferior to 10wt%.

Davidson [DAV1937] also investigated the influence of ZnO on cellulose dissolution in
potassium hydroxide (KOH). As ZnO is difficult to dissolve in KOH, only one ZnO/KOH ratio
was studied and was equal 0.14. Author showed that the addition of ZnO increases the
percentage of oxycellulose dissolved but dissolution remains very low.

_____________________________________________________________________________________________________
- 46 -
Maximum dissolution
ZnO/NaOH
(moles)
NaOH
(N)
NaOH
(%)
ZnO
(%)
%
dissolved
0.0 2.5 10 0 5.2
0.05 2.5 10 1.0 14.5
0.10 2.5 10 2.0 39.2
0.178 2.75 11 4.0 70.1
0.229 2.75 11 5.1 82.6

Tab.I.2-2: Values of NaOH and ZnO concentrations obtained
at the maximum dissolution of cotton linter for a given
ZnO/NaOH ratio, from Erreur ! Source du renvoi introuvable..
Fig.I.2-10 [DAV1937]: Cotton linter solubility, at -5C, as a
function of NaOH concentration and the ZnO/NaOH ratio

Unfortunately, no article gives any explanation concerning the action of ZnO on cellulose
dissolution. We only know that when ZnO is dissolved in alkaline solution, new compounds are
formed. In the literature, it is stipulated that, depending on NaOH concentration, the dissolution
of ZnO in NaOH gives several zinc hydroxides. The determination of the compounds formed
between NaOH and ZnO is not easy and the conclusions differ from one paper to the other. But
most of authors agree that at high pH (~12), the main species are Zn(OH)
3
-
, Zn(OH)
4
2-
and
Zn
2
(OH)
6
2-
[DIR1954] [REI1975].

I.2.3.2. Influence of urea and thiourea
In 1998, Laszkiewicz [LAS1998] reported that the addition of 1% urea in 8.5% NaOH at 5C
improves the solubility of bacterial cellulose having an initial DP of 680. The fraction of
dissolved cellulose is 48.6% with 1% urea and 17.8% without urea. Some years later, a research
group in Wuhan University in China reported that cellulose is totally dissolved in NaOH with
urea [ZHO2000] [ZHO2002a] [ZHO2004] or thiourea [ZHA2002] [WEN2004] and both
additives are removed out completely from the cellulose during the coagulation and washing.

Zhou et al. [ZHO2000] investigated the influence of NaOH/urea solvent composition on
cellulose dissolution. They concluded that the optimal NaOH/urea/water ratios to dissolve
cellulose at 0C are 6/4/90 and 8/2/90. According to them, cellulose solutions in these solvents
are stable and did not form a gel with temperature increase. They also studied the influence of
cellulose molecular weight on the dissolution for different NaOH/urea solvent compositions and
for different cellulose samples (Fig.I.2-11). From their papers, we calculated the DP of each
sample with M
/162, where M
is the cellulose molecular weight given in the article and 162 is

_____________________________________________________________________________________________________
- 47 -
the molar mass of the anhydroglucose unit (AGU); the DP are also shown in Fig.I.2-11.
Solubility S
a
(in percentage) of cellulose was calculated by:

[ ] 100 ) /(
2 1 1
+ = w w w S
a

We plotted the best solubility results NaOH/urea = 6/4, 8/2, 6/0 and 8/0 (see selected
columns in Fig.I.2-11) versus DP (Fig.I.2-12). It is clear that the solubility decreases when the
DP of cellulose samples increases: a result that should be expected.

Linter-1 DP 617
Linter-2 DP 981
Bagasse-1 DP 514
Bagasse-2 DP 448
Alkali-soluble DP 293
cellulose

Fig.I.2-11 [ZHO2000]: Data on 5% cellulose solubility in different NaOH+urea mixtures

0
20
40
60
80
100
0 200 400 600 800 1000 1200
DP
S
a

(
%
)
'6/0' '6/4' '8/0' '8/2' Ratio NaOH/Urea =

Fig.I.2-12: Solubility of cellulose in pure 6%NaOH, in 6%NaOH+4%urea, in pure 8%NaOH and in 8%NaOH+2%urea
as a function of DP, based on literature data from [ZHO2000].
_____________________________________________________________________________________________________
- 48 -
Some years later, the same team studied the influence (i) of the nature of alkali hydroxide and (ii)
of the process to dissolve cellulose [CAI2005]. They classified the solvents according to their
dissolution power on cotton linter: LiOH/urea/water > NaOH/urea/water >>
KOH/urea/water. Moreover, the best NaOH/urea/water solvent composition was found to be
7/12/81.
On one hand, they showed [ZHO2004] [CAI2005] with
13
C NMR studies that (i) NaOH/urea is
a direct solvent of cellulose and (ii) urea and NaOH interact, this interaction probably playing an
important role in the solvation of cellulose. On the other hand, DSC experiments described in
[CAI2005] mentioned urea hydrates and NaOH hydrates, no new NaOH/urea compound
was found. They added that only NaOH hydrates are linked to cellulose and that the interactions
between urea and cellulose decrease the self-association of cellulose chains and thus enhance the
stability of solution. But the nature and mechanism of cellulose/urea interactions was not
explained and is not revealed by NMR. Finally, from viscometry in dilute state, they concluded
that in such aqueous solutions, cellulose exists as semi-flexible chains [ZHO2004].

Zhang et al. [ZHA2002] studied the influence of thiourea on cellulose dissolution. They
determined that the best dissolution occurs when the NaOH/thiourea/water composition is
6/5/89. Cellulose solutions obtained are transparent and colourless. This new direct solvent of
cellulose is better than 6%NaOH/4%urea/water. Indeed, for cotton linter DP620, the solubility
in the thiourea solvent is 2 or 3 times superior to the solubility in the urea solvent and the
intrinsic viscosity is also higher in 6%NaOH/5%thiourea (245g.cm
-3
) than in 6%NaOH/4%urea
(150g.cm
-3
). They also concluded that thiourea prevents the self-association of cellulose
molecules in NaOH aqueous solutions leading to improve cellulose solubility.

This group of researchers essentially investigated regenerated cellulose membranes prepared from
NaOH/urea or NaOH/thiourea aqueous solutions. Consequently, the influence of coagulation
conditions on pore size, water permeability and mechanical properties was largely studied.
Whatever the nature of regeneration baths is, adjusting the acid concentration and the
coagulation time enable to control properties of the membranes [ZHO2002a][ZHO2002b].
However, only few papers concern mechanism of cellulose dissolution or solutions gelation.
Weng et al. [WEN2004] studied gelation of 4-6%cellulose/6%NaOH/5%thiourea aqueous
solutions, as a function of temperature. They noted that after gelation, G always increases and
never reaches a plateau. Consequently, the gelation process does not involve chemical cross-links
but probably results from the formation of a condensed network, accompanied by a micro-phase
separation. Moreover, increasing cellulose concentration shifts gelation point towards lower
temperatures by increasing the chance of collision of the cellulose chains. Another evidence of
physical interactions is that
13
C NMR spectra of solution and gel states are similar indicating that
there is no change in macromolecular structure during the sol-gel transition.
They proposed the following dissolution and gelation mechanisms of cellulose/NaOH/thiourea
(or urea) aqueous solutions (Fig.I.2-13). Because of strong intra- and intermolecular hydrogen
bonds, cellulose has a high tendency to aggregate and self-associate (Fig.I.2-13a). The addition of
NaOH/thiourea (or urea)/water solution allows breaking these intermolecular hydrogen bonds
to create a significant ion-pair interaction. Then water or thiourea (or urea) molecules can form
_____________________________________________________________________________________________________
- 49 -
hydrogen bonds with the released OH groups on the cellulose chains leading to a high swelling
of cellulose fibres in alkaline aqueous solution (Fig.I.2-13b) and then dissolution. All solvent
molecules surround the cellulose chains to prevent its self-association (Fig.I.2-13c). At a lower
temperature, this system is relatively stable whereas increasing temperature breaks the overcoat
layer. So the hydroxyl groups of cellulose are exposed to each other inducing the gel formation
caused by the self-association force of cellulose macromolecules.

Fig.I.2-13 |CAI2005]: Schematic dissolution process of the cellulose in LiOH/urea and NaOH/urea aqueous solutions
pre-cooled to -10C : (a) cellulose bundle in the solvent, (b) swollen cellulose in the solution,
(c) transparent cellulose solution

I.2.3.3. Other additives
In his works, Mercer reported the use of some additives to prevent shrinkage of cellulose fibres
as silicate or aluminate of soda, ethylic or methylic alcohol, oxalic acid [MER1903]. But at that
time, only findings were published and no explanation of the action of additives on cellulose
solutions was given.

At room temperature and at low NaOH concentrations (<10%NaOH), Legrand [LEG1952]
investigated the influence of alcohol addition on Na-Cellulose formation. It appears that when
the alcohol concentration increases, and thus the water amount decreases, the Na-Cell formation
is favoured. But on the other hand, if no water is present in the solution, native cellulose is not
transformed: this seems to confirm the significant role of water in cellulose swelling,
mercerisation and dissolution.
Moreover, gelatine, glycol, glycerol decrease the swelling of cotton [MAR1941] and penalize the
Na-Cell formation [LEG1952], whereas acetone and ethylenediamine favour it.

_____________________________________________________________________________________________________
- 50 -
I.3- CONCLUSION
In the part I.1-, the structure of cellulose was described with its dense network of intra- and
intermolecular hydrogen bonding which explains the difficulties to dissolve cellulose in most of
solvents. There is a large diversity of cellulose samples:
the crystals with four possible structures,
the supra-molecular arrangements depending on origin of cellulose (various species of
wood, cotton),
the treatment to extract pure cellulose (alkaline or acid processes) that is changing the
physical and chemical nature of the fibres,
the treatment to activate cellulose fibres.

In the part I.2-, we reviewed literature concerning cellulose dissolution in alkali hydroxide
aqueous solutions. Sodium hydroxide is the cheapest and the most effective alkali hydroxides and
thus has been largely studied for many years. It appears that NaOH can act as a swelling agent of
cellulose in a large domain of NaOH concentrations and temperatures. But, in a narrow range of
NaOH concentrations (7-10wt%) and temperatures (-6C-+4C), NaOH is a direct solvent of
cellulose. The mechanism of cellulose mercerisation was described in terms of chain arrangement
and formation of Na-Cell intermediates, the latter containing some NaOH and water molecules.

A large amount of publications reveal that cellulose/NaOH/water solutions are not stable in
time, even after the addition of zinc oxide, urea or thiourea. But in these papers, the best
conditions to improve cellulose dissolution and increase the stability of solutions are investigated
but no clear explanation on the reasons of gelation and action of additives are given.

_____________________________________________________________________________________________________
- 51 -
R RE EF FE ER RE EN NC CE ES S
[AIT1988] AITKEN Y., CADEL F., VOILLOT C. Constituants fibreux des ptes papier et carton.
Pratique de l'analyse. Grenoble : Centre Technique de l'Industrie du Papier Cartons et Celluloses /
Ecole Franaise de Papeterie et des Industries Graphiques (1988), Xp.

[ATA1984] ATALLA R.H., VANDERHART D.L. Native cellulose: a composite of two distinct
crystalline forms. Science (1984), 223, p.283-284

[ATA1999] ATALLA R.H., VANDERHART D.L. The role of solid state 13C NMR
spectroscopy in studies of the nature of native celluloses. Solid State NMR (1999), 15 , p.1-19

[CAI2005] CAI J., ZHANG L. Rapid dissolution of cellulose in LiOH/urea and NaOH/urea
aqueous solutions. Macromol. Biosci. (2005), 5, p.539-548

[CHA1976] CHANZY H., ROCHE E. Fibrous transformation of Valonia cellulose I into
cellulose II Appl. Polym. Symp. (1976), 28, 701

[CUI2006a] CUISSINAT C., NAVARD P. Swelling and dissolution of cellulose Part I: free
floating cotton and wood fibres in N-methylmorpholine-N-oxide water mixtures. Macrom. Symp.
To appear

[CUI2006b] CUISSINAT C., NAVARD P. Swelling and dissolution of cellulose Part II: free
floating cotton and wood fibres in NaOH-water-additives systems. Macrom. Symp. To appear

[DAV1934] DAVIDSON G.F. The dissolution of chemically modified cotton cellulose in
alkaline solutions. Part I: In solutions of NaOH, particularly at TC below the normal. J. Text.
Ind. (1934), 25, p.T174-196

alkaline solutions. Part II: A comparison of the solvent action of solutions of Lithium, Sodium,
Potassium and tetramethylammonium hydroxides. J. Text. Ind. (1936), 27, p.T112-130

alkaline solutions. Part III: In solutions of Sodium and Potassium hydroxyde containing dissolved
Zinc, Bryllium and Aluminium oxides. J. Text. Ind. (1937), 28, p.T27-44

[DIR1954] DIRKSE T.P., POSTMUS C., VANDENBOSCH J.R., VANDENBOSCH R. A
study of alkaline solutions of zinc oxide. J. Am. Chem. Soc. (1954), 76, p.6022-6024

[ELL1962] ELLIS K.C., WARWICKER J.O. A study of the crystal structure of cellulose I. J.
Polym. Sci. (1962), 56, p.339-357

[FIN1995] FINK H.P., WALENTA E., KUNZE J., MANN G. Wide angle X-Ray and solid
state C-NMR studies of cellulose alkalisation. In: KENNEDY J.F. et al. eds. Cellulose and cellulose
derivatives: Physic-chemical aspects and industrial applications. Woodhead publishing Ltd (1995)

_____________________________________________________________________________________________________
- 52 -
[FRE1983] FREYTAG, DONZE; Wood and Cellulose: Alkali treatment of cellulose fibres
Handbook of fibres Sciences and Technology Vol.1 (1983)

[FRE1987] FRENCH A.D., ROUGHEAD W.A., MILLER D.P. X-Ray diffraction studies of
ramie cellulose I. In: ATTALA R.H. The structures of cellulose characterisation of the solid states. ACS
Symposium Series 340. Washington, DC: American Chemical Society (1987)

[GAR1974] GARDNER K.H., BLACKWELL J. The structure of native cellulose. Biopolymers
(1974), 13, p.1975-2001

[GAR1985] GARDINER E.S., SARKO A. Packing analysis of carbohydrates and
polysaccharides. 16. The crystal structures of cellulose IV
I
and IV
II
. Can. J. Chemistry (1985), 63,
p.173-180

[GRA1958] GRANT J. Chapter 7. The acid sulphite process. Description and chemistry. The
nature and chemistry of the process / Chapter 9. The alkaline processes. Theory, general
description and chemistry. In: GRANT J. Cellulose pulp and allied products. (Incorporated the third
edition of Wood pulp). London: Leonard Hill [Books] Ltd (1958)

[HEU1925] HEUSER E., BARTUNEK R.; Cellulosechem. (1925), 6, 19

[HON2001] HON D.N.S., SHIRAISHI N. Wood and cellulosic chemistry. New York . Basel: Marcel
Dekker, Inc. (2001), 512p

[ISO1989] ISOGAI A., USUDA M., KATO T., URYU T., ATALLA R.H. Solid-state CP/MAS
13C NMR study of cellulose polymorphs. Macromolecules (1989), 22, p.3168-3172

[ISO1997] ISOGAI A. NMR Analysis of cellulose dissolved in aqueous NaOH solutions. Cellulose
(1997), 4, p.99-107

[KAM1984] KAMIDE K., OKAJIMA K., MATSUI T., KOWSAKA K. Study on the solubility
of cellulose in aqueous alkali solution by deuteration IR and
13
C NMR. Polym. J. (1984), 16 (12),
p.857-866

[KAM1985] KAMIDE K., OKAJIMA K., KOWSAKA K. Determination of intramolecular
hydrogen bonds and selective coordination of sodium cation in alkalicellulose by CP/MASS C13
NMR. Polym. J. (1985), 17 (5), p.707-711

[KAM1990] KAMIDE K., YASUDA K., MATSUI T., OKAJIMA K., YAMASHIKI T.
Structural change in alkali-soluble cellulose solid during its dissolution into aqueous alkaline
solution. Cellulose Chem. Technol (1990), 24, p.24-31

[KLE1998] KLEMM D., PHILIPP B., HEINZE T., HEINZE U., WAGENKNECHT W.
Comprehensive cellulose chemistry: Volume 1, Fundamentals and analytical Methods. Weinheim: WILEY-
VCH (1998)

[KOL1976] KOLPAK F. J., BLACKWELL J. Determination of the structure of cellulose II.
Macromolecules (1976), 9, p.273-278

_____________________________________________________________________________________________________
- 53 -
[KUO2005] KUO Y.N., HONG J. Investigation of solubility of microcrystalline cellulose in
aqueous NaOH. Polym. Adv. Technol. (2005), 16, p.425-428

[LAN1999] LANGAN P., NISHIYAMA Y., CHANZY H. A revised structure and hydrogen
bonding system in Cellulose II from a Neutron fiber diffraction analysis. J. Am. Chem. Soc. (1999),
121, p.9940-9946

[LAS1998] LASZKIEWICZ B. Solubility of bacterial cellulose and its structural properties. J.
Appl. Polym. Sci (1998), 67, p.1871-1876

[LEG1952] LEGRAND C., GRUND A. Formation des alcali-celluloses en milieu
hydroalcoolique. J. Polymer Sci. (1952), 9 (6), p.527-530

[MAC1990] MACCHI E.M. The polarity of chain packing in native cellulose. A meridional
electron diffraction analysis on Valonia fibres. Macromol. Chem. (1990), 191, p.2217-2226

[MAR1941] MARSH J.T. Mercerising. London: Chapman & Hall Ltd (1941), 458p

[MER1903] The Editors of "Dyer and Calico Printer". Mercerisation: a practical and historical manual.
Vol.I. London: Heywood and company LTD (1903), 240p

[NIS1987a] NISHIMURA H., SARKO A. Mercerization of cellulose. III. Changes in crystallites
sizes. J. Appl. Polym. Sci (1987), 33, p.855-866

[NIS1987b] NISHIMURA H., SARKO A. Mercerization of cellulose. IV. Mechanism of
mercerization and crystallites sizes. J. Appl. Polym. Sci (1987), 33, p.867-874

[NIS1991a] NISHIMURA H., OKANO T., SARKO A. Mercerization of cellulose. V. Crystal
and molecular structure of Na-cellulose I. Macromolecules (1991), 24, p.759-770

[NIS1991b] NISHIMURA H., SARKO A. Mercerization of cellulose. VI. Crystal and molecular
structure of Na-cellulose IV. Macromolecules (1991), 24, p.771-778

[OKA1984] OKANO T., SARKO A. Mercerization of cellulose. I. X-Ray diffraction evidence
for intermediate structures. J. Appl. Polym. Sci (1984), 29, p.4175-4182

[OKA1985] OKANO T., SARKO A. Mercerization of cellulose. II. Alkali-cellulose intermediates
and a possible mercerisation mechanism. J. Appl. Polym. Sci (1985), 30, p.325-332

[PET1948] PETITPAS T. Etude de l'alcali-cellulose: Variations de structure de la cellulose dans
les lessives alcalines. Compte-rendu Laboratoire central des Services chimiques de l'Etat (Paris) (1948),
226, p.139-147

[PIC1893] PICKERING S.U. The hydrates of sodium, potassium and lithium hydroxides. J.
Chem. Soc. (1893), 63, p.890-909

[RAY1995] RAYMOND S., HEYRAUD A., TRAN QUI D., KVICK ., CHANZY H. Crystal
and molecular structure of b-D-celloteraose hemihydrate as a model of cellulose II. Macromolecules
(1995), 28, p.2096-2100
_____________________________________________________________________________________________________
- 54 -
[REI1975] REICHLE R.A., McCURDY K.G., HEPLER L.G. Zinc hydroxide: Solubility product
and hydroxy-complex stability constants from 12.5-75C. Can. J. Chem (1975), 53, p.3841-3845

[ROY_PhD] ROY C. Etude de mlanges de cellulose dans des solutions aqueuses de soude. Thse Doct.
Sciences et gnie des matriaux, Sophia-antipolis : Ecole des Mines de Paris (2002), 213p.

[ROY2003] ROY C., BUDTOVA T., NAVARD P. Rheological properties and gelation of
aqueous cellulose-NaOH solutions. Biomacromolecules (2003), 4, p.259-264

[SAR1974] SARKO A., MUGGLI R. Packing analysis of carbohydrates and polysaccharides. III
Valonia cellulose and cellulose II. Macromolecules (1974), 7, p.486-494

[SAR1976] SARKO A., SOUTHWICK J., HAYASHI J. Packing analysis of carbohydrates and
polysaccharides. 7. Crystal structure of cellulose III
I
and its relationship to other cellulose
polymorphs. Macromolecules (1976), 9, p.857-863

[SAW1985] SAWADA T.; Proc. Int. Symp. Wood and pulping chemistry (1985) Vancouver, Canada.

[SOB1939] SOBUE H., KIESSIG H., HESS K. The cellulose-sodium hydroxide-water system as
a function of the temperature. Z. Physik. Chem. B (1939), 43, p.309-328

[SRE1993] SREENIVASAN S., BHAMA IYER P., PATEL G.S. Studies on the swelling of
cotton fibers in alkali metal hydroxides. III. Structure-Property relations in fibers swollen at 0C.
J. Appl. Polym. Sci. (1993), 48, p.393-404

[STI1976] STIPANOVICH A. J., SARKO A. Packing analysis of carbohydrates and
polysaccharides. 6. Molecular and crystal structure of regenerated cellulose II. Macromolecules
(1976), 9, p.851-857

[TAK1991] TAKAHASHI M., OOKUBO M., TAKENAKA H. Solid state 13C NMR spectra
analysis of alkalicellulose. Polymer J. (1991), 23 (8), p.1009-1014

[TAS1994] TASKER S., BADYAL J.P.S., BACKSON S.C.E., RICHARDS R.W. Hydroxyl
accessibility in celluloses. Polymer (1994), 35 (22), p.4717-4721

[WAD2001] WADA M., HEUX L., ISOGAI A., NISHIYAMA Y., CHANZY H., SUGIYAMA
J. Improved structural data of cellulose III
I
prepared in supercritical ammonia; Macromolecules
(2001), 34, 1237-1243

[WEN2004] WENG L., ZHANG L., RUAN D., SHI L., XU J. Thermal gelation of cellulose in a
NaOH/Thiourea aqueous solution. Langmuir (2004), 20 (6), p.2086-2093

[WO 02/22924] MIKOLAJCZYK W., STRUSZCZYK H., URBANOWSKI A., WAWRO D.,
STAROSTKA P. Process for producing fibres, film, casings and other products from modified soluble cellulose.
Poland, Patent n WO 02/22924 (21 mars 2002)

[YAM1988] YAMASHIKI T., KAMIDE K., OKAJIMA K., KOWSAKA K., MATSUI T.,
FUKASE H. Some characteristic features of dilute aqueous alkali solutions of specific alkali
_____________________________________________________________________________________________________
- 55 -
concentration (2,5mol/l) which possess maximum solubility power against cellulose. Polym. J.
(1988), 20 (6), p.447-457

[YAM1990] YAMASHIKI T., MATSUI T., SAITOH M., OKAJIMA K., KAMIDE K.,
SAWADA T. Characterisation of cellulose treated by the steam explosion method. Part 1:
Influence of cellulose resources on changes in morphology, degree of polymerisation, solubility
and solid structure. British Polymer Journal (1990), 22, p.73-83

[ZHA2002] ZHANG L., RUAN D., GAO S. Dissolution and regeneration of cellulose in
NaOH/Thiourea aqueous solution. J. Polym. Sci.: Part B (2002), 40, p.1521-1529

[ZHO2000] ZHOU J., ZHANG L. Solubility of cellulose in NaOH/Urea aqueous solution.
Polym. J (2000), 32 (10), p.866-870

[ZHO2002a] ZHOU J., ZHANG L., CAI J., SHU H. Cellulose microporous membranes
prepared from NaOH/urea aqueous solution. J. Membrane Sci (2002), 210, p.77-90

[ZHO2002b] ZHOU J., ZHANG L., SHU H., CHEN F. Regenerated cellulose films from
NaOH/urea aqueous solution by coagulating with sulphuric acid. J. Macromol. Sci. - Physics (2002),
B41(1), p.1-15

[ZHO2004] ZHOU J., ZHANG L., CAI J. Behavior of cellulose in NaOH/Urea aqueous
solution characterized by light scattering and viscometry. J. Polym. Sci.: Part B (2004), 42, p.347-
353

[ZUG2001] ZUGENMAIER P. Conformation and packing of various crystalline cellulose fibers.
Prog. Polym. Sci (2001), 26, p.1341-1417

[Web site 1]
http://www.snv.jussieu.fr/bmedia/paroi/synth-cell.htm

[Web site 2]
http://ici.cegep-ste-foy.qc.ca/profs/gbourbonnais/pascal/fya/chimcell/notesmolecules/glucides_3.htm

Rsum du chapitre II Matriaux et mthodes exprimentales
_____________________________________________________________________________________________________
-57-

R RE ES SU UM ME E D DU U C CH HA AP PI IT TR RE E I II I
M MA AT TE ER RI IA AU UX X E ET T M ME ET TH HO OD DE ES S E EX XP PE ER RI IM ME EN NT TA AL LE ES S
Ce chapitre dcrit les produits, les dispositifs et les mthodes utiliss dans ce travail.

Diffrentes celluloses ont t utilises, les deux principales tant une cellulose microcristalline et
une cellulose de bois explose la vapeur afin de diminuer le DP et de faciliter sa dissolution
dans la soude. Quel que soit le type de cellulose utilis, la dissolution seffectue de la mme
manire : deux heures de mlangeage -5C dans une solution aqueuse de 7,6%-9% dhydroxyde
de sodium pralablement refroidie -5C. Au cours de ltude bibliographique, nous avons vu
que lajout dadditif peut amliorer la dissolution des fibres de cellulose. Nous avons donc
travaill avec de loxyde de zinc et de lure mais galement avec quelques surfactants.
Une ponge cellulosique est un composite poreux constitu dune matrice de cellulose rgnre,
provenant de la solution cellulose/NaOH/eau dcrite prcdemment, et de fibres de renfort
naturelles (gnralement du coton, du lin ou des fibres de bois). Des cristaux de sulfate de
sodium (Na
2
SO
4
,10H
2
O) sont incorpors en grande quantit dans le mlange, leur fusion crant
la macro porosit des ponges. Nous donnons une description rapide des fibres de renfort
utilises au laboratoire et des cristaux de sel. La mthode de prparation des ponges est
galement prsente et reprend les donnes fournies et utilises au laboratoire pilote Spontex.

Nous verrons dans les chapitres suivant que de nombreuses techniques danalyse physico-
chimique ont t utilises que ce soit pour dterminer les proprits thermique, rhologique,
mcanique et autres. Dans ce chapitre, nous prsentons la fois leur principe de base et les
conditions exprimentales appliques nos chantillons.

Les proprits thermiques des solvants utiliss dans cette tude ainsi que celles des solutions de
cellulose ont t analyses par un appareil de DSC (Differential Scanning Calorimetry). Les
thermogrammes nous ont permis de mesurer les temprature et enthalpie de fusion de chaque
phase prsente dans le systme et ainsi dterminer la structure des solutions, en dessous de 0C.

La rhologie des solutions a t tudie sur de nombreuses solutions de cellulose. Nous avons fait
varier le DP, la concentration et lorigine de la cellulose, mais galement les systmes de solvant
en modifiant la nature et la concentration des additifs. Le but de ltude rhologique est de mieux
comprendre le comportement des solutions de cellulose lcoulement (mesure de viscosit en
fonction du taux de cisaillement), et ce pour diffrentes tempratures. Les solutions ntant pas
Rsum du chapitre II Matriaux et mthodes exprimentales
_____________________________________________________________________________________________________
-58-
stables thermodynamiquement, nous avons dtermin les conditions de glification des solutions
en temps, en temprature et en fonction des additifs prsent dans le systme de solvant. Le point
de gel est dfini comme tant le point dintersection des deux modules G et G. La rhologie en
mode dilu a t ralise plusieurs tempratures et nous a permis de classer les diffrents
solvants en fonction de leur pouvoir de solvant sur la cellulose dune part et dtudier linfluence
de la temprature sur cette qualit de solvant dautre part.

Les produits de cellulose rgnre ont t analyss mcaniquement par des tests de traction
allant jusqu rupture de lprouvette. Le but de ces essais est de comparer les diffrentes
conditions de glification et de rgnration de la cellulose sur la rsistance mcanique afin
damliorer les proprits finales des ponges.

Des observations microscopiques, que ce soit laide dun microscope optique ou dun
microscope lectronique balayage (MEB SEM en anglais), ont t ralises tout au long du
projet pour tudier qualitativement la solubilit de la cellulose en fonction du solvant, linfluence
de la conglation sur laspect des fibres de renfort et de la porosit des ponges finales. Le
microscope optique reli une camera nous a permis de visualiser laction de la soude sur les
fibres de cellulose.

Chapter II Materials and methods
_____________________________________________________________________________________________________
-59-

C
C
H
H
A
A
P
P
T
T
E
E
R
R
I
I
I
I

II- II
M MA AT TE ER RI IA AL LS S A AN ND D M ME ET TH HO OD DS S

In the first chapter, the diversity of cellulose pulps was shown as well as the numerous studies on
the optimal concentrations of the different components (cellulose, sodium hydroxide, additives).
Consequently, in order to minimize misinterpretations in the discussion of results, a well-detailed
description of all experimental conditions is given. This second chapter is divided in three main
parts Materials, Solution preparation and Methods.

The first part deals with the materials used: cellulose pulps, solvent and additives. We will give
some details about the origin of cellulose and its DP, the suppliers of each components, and their
purity.

The second one presents the procedure to dissolve cellulose in NaOH/water,
NaOH/ZnO/water and NaOH/urea/water solutions. Some results are given to explain our
choice in sample preparation. This concerns the drying of initial cellulose pulp to avoid the
problem of humidity, the storage conditions and the necessity, or not, to filter solutions before
rheological measurements.

The last one describes the various methods used in this study. As DSC (Differential Scanning
Calorimetry) analyses were largely performed, not only the conditions of sample preparation are
described but also the principle of DSC apparatus and the method used to determine melting
temperature and enthalpy. For the other techniques (X-ray diffraction analyses, rheology,
microscopic observations, mechanical tests), essentially details on sample preparation are
presented.

-60-

_____________________________________________________________________________________________________
-61-
II.1- MATERIALS
II.1.1- Cellulose samples
In this study, several types of cellulose were used (Tab.II.1-1). The main ones were
microcrystalline cellulose Avicel from FMC Corporation and steam exploded cellulose
Borregaard prepared by Innovia Films Ltd (B342, B389, B407 and B500). Some work (Chapter
IV) was also performed on non-steam exploded cellulose from Borregaard and cellulose samples
provided by Lenzing: L307 sample obtained with electron beam degradation and L198 sample
obtained from cellulose regeneration from alkali solutions. A detailed description of each
cellulose type is given in the following sections.
Name Type of cellulose
(Appendix I)
Origin Treatment DP
Avicel Cellulose I Wood but origin
is unknown
MCC 170
B342 Cellulose I Spruce and pine Steam explosion 342
B530 Cellulose I Spruce and pine Elemental Chlorine
Free bleaching (ECF)
530
L307 Cellulose I Eucalyptus Degradation by
electron beam
307
L198II Cellulose II Eucalyptus Alkali cellulose 198
Tab.II.1-1: Main data concerning the pulps of cellulose used in this study.
Names consist of the supplier initial and the DP value.

II.1.1.1. Avicel cellulose
Avicel cellulose is a microcrystalline cellulose (MCC). This cellulose is very pure; there is no lignin
and no hemicelluloses and it consists of pure cellulose micro-crystals (see more details in Chapter
I, part I.1.5.2). Such a cellulose sample has a relatively low degree of polymerisation (DP).
Here the grad used was Avicel PH101 with a DP of 170.

II.1.1.2. Borregaard cellulose
Cellulose extraction from spruce and pine softwood was performed in Borregaard ChemCell:
Steam-exploded samples: The steam explosion treatment had been made by Innovia Films
Ltd. As we saw previously (Chapter I, part I.1.5.2), this treatment enables to activate cellulose
fibres by decreasing the DP of cellulose. Consequently, cellulose dissolution is improved as
compared with untreated samples. In this study samples of different DPs were used: 342,
389, 407 and 500. They will be named B342, B389, B407 and B500 respectively.
Non-steam exploded samples: The B530 cellulose sample was treated using the sulphite
pulping process and ECF bleaching (Elemental Chlorine Free bleaching using chlorine
dioxide as chemical). It is a high quality dissolving pulp containing around 92% of cellulose,
the rest being moisture and very low level impurities (hemicelluloses). Its DP is 530.

_____________________________________________________________________________________________________
-62-
II.1.1.3. Other cellulose samples
Two other cellulose pulps used were kindly provided by Lenzing. The initial pulp was Solucell
1175 with DPv=412. It was treated in two ways:
o By electron beam irradiation (60kGy), which gave cellulose I with DP307 (sample L307).
o Regenerated from alkali solution, which gave cellulose II with DP198 (sample L198
II
).

II.1.2- Solvents and additives
Cellulose was dissolved in alkali hydroxide aqueous solutions, with or without additives.

II.1.2.1. Alkali hydroxides
The main alkali used was sodium hydroxide NaOH; potassium hydroxide KOH was also tested
for some experiments. NaOH and KOH were pellets from the supplier VWR; purities were 98%
and 85% respectively. Water was distilled in the lab.

II.1.2.2. Additives
Two main additives were added to solutions in order to improve cellulose dissolution. They were
zinc oxide ZnO and urea CO(NH
2
)
2
. ZnO was powder from the supplier VWR; its purity was
98%. Urea was pellets from Fisher; its purity was 99%.

II.1.2.3. Surfactants
In the process of sponge fabrication, surfactants are added to improve properties of final
product. The surfactants used are non-ionic and can be dissolved in water. Two of them are
fluoropolymers Zonyl FS and Zonyl FSN and the third one is an alkylpolyol, named AP in
the following.
The exact structure of these surfactants is not known.
In cellulose solutions, the surfactant concentration is 1% of cellulose content, which corresponds
to 0.05% in a 5% cellulose solution.

II.1.3- Other components used in sponge manufacture
II.1.3.1. Reinforcing fibres
The fibres were either cotton fibres, with a length of 20mm, or fluff. The latter is a short wood
fibre which is cheaper than cotton. Fluff is a wood pulp as the one for paper manufacturer, with
some differences in order to make easier the defibration.
The fluff fibres provided by Weherhauser under the reference CF416, with a DP of 1100.
These two fibres should not be dissolved in 7.6NaOH/water, even at -6C. According to
Yamashiki [YAM1990], cellulose can be dissolved in 8-9%NaOH if its DP is inferior to 400. In
_____________________________________________________________________________________________________
-63-
our case, we partly dissolved cellulose with DP of 500, but cellulose with higher DP, as the
Raywhite HJ cellulose with a DP of 1044 (provided by the supplier Rayonier), are effectively not
dissolved in 7.6NaOH/water (Fig.II.1-1).

Fig.II.1-1: non-dissolved fibres of Rayonier DP1044 cellulose in a 5cellulose/7.6NaOH/water solution.
Scale bar represents 100m

II.1.3.2. Adhesion promoters
Three DYNASYLAN
adhesion promoters, from Hls France, were used:

0.5% and 1% of GLYMO (Glycidyloxypropyltrimethoxysilane)
1% AMEO (Aminopropyltriethoxysilane)
1% CPTMO (Chloropropyltrimethoxysilane)

II.1.3.3. Na
2
SO
4
,10H
2
O crystals
The Na
2
SO
4
,10H
2
O crystals are used as porophores in the sponges. The size of these crystals is
between less than 1mm and 4mm. The distribution of diameters generates different porosities
which are adjusted as a function of properties of the final product.
The Na
2
SO
4
,10H
2
O crystals added at the beginning of the industrial process are recycled at the
end and re-introduced in the process.

_____________________________________________________________________________________________________
-64-
II.2- SOLUTIONS PREPARATION
The description of the proportions between the components in the solutions that will be
studied and the corresponding notations used in the following chapters will be presented.
The reason of such a notice is that systems studied are of a complex composition, like a
mixture of cellulose, NaOH, urea, zinc oxide and water, and confusions in concentration
determination should be avoided.

The notion of concentrations will be used only for binary solvents, like NaOH/water or
urea/water. In all other cases we shall use and show in the notations the amount of
grams of the substance in 100 g of solution. In other words, Xcellulose/YNaOH/water
means that in 100 g of solution there are X grams of cellulose, Y grams of NaOH and
(100XY) of water. Another example is as follows: Xcellulose/YNaOH/Zurea/water
means that in 100 grams of solution there are X grams of cellulose, Y grams of NaOH, Z
grams of urea and (100-X-Y-Z) grams of water. For the binary systems concentration in
weight per cent, calculated as % 100
water i
i
M M
M
Y
+
= (M
i
and M
water
being component i
and water weights, respectively), coincides with the amount of grams Y of this
component in 100 g of solution: Ycomponent/water.

II.2.1- Drying of steam exploded cellulose
Steam exploded cellulose was provided in a wet state. The water content was from 10% for B500
to 47% for B342. Moreover, because of changes in the humidity in the lab, water content in given
wet cellulose pulp could vary. To avoid any mistake in the cellulose concentration when
preparing solutions, cellulose was dried in an oven at +60C during 10 hours in order to obtain a
dry pulp.

Two solutions were prepared. The first one consists in 5 dry B342 cellulose/7.6NaOH/water,
the second one in 5 wet B342 cellulose/7.6NaOH/water. It is clear that in the wet cellulose
solution, water content is taken into account when weighting cellulose.

Quick tests were performed to study the influence of cellulose drying on rheological properties of
5 B342 cellulose/7.6NaOH/water solutions. On one hand, viscosity measurements (Fig.II.2-1)
reveal that drying slightly increases viscosity of solutions, for the three temperatures studied. On
the other hand, moduli versus temperature measurements (Fig.II.2-2) show that gelation point is
reached later with wet cellulose (T
gel
=31C) than with dry cellulose (T
gel
=28C). It seems that
water content in cellulose pulp influences results of rheological tests, even slightly. Thus, in order
to remove this parameter, cellulose was dried in all our experiments.

_____________________________________________________________________________________________________
-65-
0,1
1
10
100
0,01 0,1 1 10 100
shear rate (s-1)

(
P
a
.
s
)
40C
20C
10C
dry
dry
wet
wet

Fig.II.2-1: Viscosity versus shear rate, at 10C, 20C and 40C,
for 5 B342 cellulose/7.6NaOH/water, with dry (full symbols) and wet cellulose (open symbols)

0,01
0,1
1
10
100
1000
0 10 20 30 40 50
TC
G
'

a
n
d

G
'
'

(
P
a
)
Tgel ~ 28C
G' = G'' = 0,4 Pa
Tgel ~ 31C
G' = G'' = 0,3 Pa
G'
G''

Fig.II.2-2: G and G versus temperature measurements and determination of gelation point when G=G
for 5 B342 cellulose/7.6NaOH/water, with dry (full line) and wet (dashed line) cellulose

II.2.2- Dissolution procedure
Cellulose content was varied from 0.5g to 7.6g in 100g of solution depending on experiments.
The content of alkali will be in the middle of the region of the phase diagram where there is
dissolution, i.e. 7.6NaOH/water. 7.6KOH/water was also tested as cellulose solvent using the
same procedure as for cellulose/NaOH dissolution, described below. Zinc oxide, urea or
surfactants were added to 7.6NaOH/water to improve cellulose dissolution.
_____________________________________________________________________________________________________
-66-
The dissolution process was as follows:

1. Preparation of solvent: Three concentrated solvents were prepared to be
further mixed with swollen-in-water cellulose. The content of NaOH was 7.6g
in most of cases; details will be given in each part.

NaOH/water: 12%NaOH/water and 20%NaOH/water solutions were
prepared. The first solution was used to dissolve cellulose in
NaOH/water and in NaOH/ZnO/water. The second was used to dissolve
cellulose in the presence of urea.
NaOH/ZnO/water: Zinc oxide was added to the alkali hydroxide (12%)
and it has to be fully dissolved. ZnO content in 100g of the finale
cellulose solution was varied from 0.4g to 3.9g.
Urea/water: a 50%urea/water solution was prepared. Urea content in
100g of the finale cellulose solution was varied from 2g to 20g.

2. Water was added to dry cellulose to make fibres slightly swollen. Water
content was adjusted to obtain 100g of solution.

3. Solvents and swollen-in-water cellulose were cooled down:
Cellulose/water to +5C
NaOH/water and NaOH/ZnO/water to -6C
Urea/water to +5C

4. When the components are cooled, NaOH/water (or NaOH/ZnO/water) was
added to the swollen-in-water cellulose and mixed thoroughly with a spatula.
In the case where urea was used, 50%urea/water solution was added to
cellulose/NaOH/water mixture.

5. Mixtures were placed into the thermobath at -6C and mixed at ~1000 RPM
using a rotary overhead mixer, for 2 hours.

6. Ready solutions were stored at +5C.

As we will see in the next chapter, cellulose/NaOH/water solutions are complex: some cellulose
fibres are not dissolved and gelation occurs. Thus a procedure of storage and treatment of
solutions before measurements has to be clearly determined in order to have reproducible results.
Because of gelation of solutions, the storage conditions must be studied. And because of the
presence of non-dissolved fibres, the influence of filtration must be investigated. Both aspects are
described in sections II.2.3 and II.2.4; the most suitable conditions for having reproducible
results will be selected.

_____________________________________________________________________________________________________
-67-
II.2.3- Storage
It is known [YAM1996] [ROY2003] that cellulose/NaOH solutions are gelling in time and with
temperature increase. Consequently, it was important to check if there is any influence of the
storage time on the viscosity of solutions. The aim of this section was to determine the time for
which cellulose solutions remain stable at +5C.

To do this, viscosity versus shear rate measurements were performed on a StressTech rheometer
(see part II.3.3.1) on 5 B500 cellulose/7.6NaOH/water solutions kept at +5C during different
storage time: t=0 (fresh solution), t=1, 2, 4 and 28 days. Temperatures of viscosity measurements
were 0C and 20C.
An example of this test at 0C is shown Fig.II.2-3. The results show that if solutions are stored at
+5C, there is no storage time influence on their viscosity (within the experimental errors) during
at least one month. Nevertheless, the solutions need to be homogenised because of
sedimentation of non-dissolved fibres. Otherwise results become not reproducible.

0,1
1
10
0,01 0,1 1 10 100 1000
shear rate(s-1)

(
P
a
.
s
)
t=0 (fresh solution)
t=1 day
t=4 days
t=28 days

Fig.II.2-3: Influence of storage time at +5C on the viscosity of 5 B500 cellulose/7.6NaOH/water solution.
Measurements have been performed at 0C
(at 20C the viscosity values are very close and do not appear for a better visibility)

Conclusion: Cellulose/NaOH/water solutions were stored at +5C and used within one month.

II.2.4- Filtration
We will see in the following that cellulose is rarely totally dissolved in NaOH/water: a small
amount of non-dissolved fibres often remains after dissolution. The question is if the presence of
non-dissolved fibres influences properties of solutions. Cellulose solutions were filtered and the
rheological properties of filtered or not solutions were compared. The objective was to decide if
filtration is needed before detailed rheological studies.
_____________________________________________________________________________________________________
-68-
As the Borregaard cellulose solutions are the ones containing the larger amount of non-dissolved
fibres, filtration tests were performed on these Borregaard cellulose solutions.
The first question was if the presence of non-dissolved fibres changes the results of intrinsic
viscosity. Does the presence of non-dissolved fibres influence the intrinsic quality of solvent of
cellulose?

Firstly, a 0.5 B342 cellulose/7.6NaOH/water solution was prepared by diluting an initial 5 B342
cellulose/7.6NaOH/water solution with 7.6NaOH/water. Secondly, the dilute cellulose solution
was filtered using a Schott filter n3 with pore size = 17-40m. A water pump helped the
solution to go through the filter.
Dilute filtered or not cellulose solutions were introduced in a capillary viscometer to determine
intrinsic viscosity of solutions. Results are shown on Fig.II.2-4. This fine filtration induces a
slight decrease in reduced viscosity but the intrinsic viscosity is approximately the same for
filtered or not cellulose solutions.

y = 10024x + 103,25
y = 13423x + 101,99
100
120
140
160
180
0 0,002 0,004 0,006
cellulose concentration (g/cm3)

r
e
d

(
c
m
3
/
g
)
without filtration
with filtration (Schott nb3)

Fig.II.2-4: Reduced viscosity versus cellulose concentration to determine the intrinsic viscosity, at +25C,
of a B342 cellulose/7.6NaOH/water solution, without filtration (in blue) and with a fine filtration (in yellow)
Intrinsic viscosities are indicated on the graph (circled values)
Error bars correspond to 5%.

The second question was if the presence of non-dissolved fibres influences gelation: faster
kinetics or increase of the gel modulus. Consequently, gel points were determined for filtered or
not solutions and results were compared.

5 B342 cellulose/7.6NaOH/water and 5 B342 cellulose/7.6NaOH/0.7ZnO/water solutions are
rather viscous and cannot be filtered with the Schott filters because of the jamming of pores.
Indeed, it seems that the low pressure in the flask induces gelation and jams the filter. Thus a
special set-up for filtering concentrated solutions was provided by Innovia Films Ltd.
G and G moduli versus temperature were measured for filtered or not 5 B342
cellulose/7.6NaOH/water and 5 B342 cellulose/7.6NaOH/0.7ZnO/water solutions. Results of
_____________________________________________________________________________________________________
-69-
gel points were compared and are summarised in the Tab.II.2-1. The gel point is defined as being
the intersection point of both moduli, when G=G. Thus at the gel point corresponds two
coordinates (x,y) where x corresponds to the gelation temperature T
gel
and y to G value at gel
point (more details on determination of T
gel
and G
gel
will be precise in section II.3.3.1).
Whatever the system is, the gel point is the same with or without filtration of the non-dissolved
fibres.
in NaOH in NaOH/0.7%ZnO
without filtration 30C ; 2.8 Pa 46C ; 1.9 Pa
with filtration 31.5C ; 3.4 Pa 45C ; 1.9 Pa
Tab.II.2-1: Summarised results of the rheological measurements versus temperature
on 5 B342 cellulose/7.6NaOH/water solutions with or without 0.7ZnO,
with or without filtration of non-dissolved fibres.

From these results, the presence of non-dissolved fibres does not influence either the intrinsic
viscosity (filter pores varied from 17 to 160 m) or the gelation and its kinetics.

Thus, in the forthcoming experiments, the solutions will not be filtered.

II.3- METHODS
II.3.1- Differential Scanning Calorimetry (DSC)
The instrument used was a PERKIN-ELMER DSC-7 composed of two thermally insulated
ovens, in an inert atmosphere (N
2
). The sample was placed in a stainless steel screwed capsule,
gold-plated or not (the nature of capsule will be given depending on cases but the procedure is
the same). The reference consists of an identical empty capsule.
The main principles of DSC measurements and determination of important characteristics
(melting temperature, melting enthalpy) are given below.

II.3.1.1. Main principles
In a material, first order chemical or physical transformations, such as changes of phase or of
crystalline structure, induce temperature variations: either heat absorption (endothermic
phenomenon) or heat emission (exothermic phenomenon). Calorimetric techniques can measure
these heat exchanges.

Differential Scanning Calorimetry (DSC) is a technique for measuring the necessary energy to
establish a nearly zero temperature difference between a substance and an inert reference
material, which follow the same temperature control program. In this study, we used a power-
compensation DSC (Fig.II.3-1): the temperatures of the sample and reference are controlled
independently using separate, identical furnaces. The temperatures of the sample and reference
are made identical by varying the power input to the two furnaces; the energy required to do this
is a measure of the enthalpy or heat capacity changes in the sample relative to the reference.
_____________________________________________________________________________________________________
-70-

Fig.II.3-1: Diagram of a power-compensated DSC

The basic principle of this technique is that when the sample undergoes a physical
transformation, the sample needs more (or less) heat than the reference to maintain both at the
same temperature. For example, during an endothermic process, such as melting, the sample
absorbs a certain amount of heat. Consequently, the sample requires a more significant heat flow
than the reference to increase its temperature at the same rate as the one of reference. By
observing the difference in heat flow between the sample and reference, Differential Scanning
Calorimeters are able to measure the amount of energy absorbed or released during such
transitions. DSC can thus be used to determine the temperature range and/or measure the
enthalpies of glass transitions, first order transitions, calorific capacity. DSC gives information to
build phase diagrams, to investigate kinetic of crystallisation, and in some cases to measure the
purity of samples.

The heating elements are kept very small (weighing less than 1 gram) in order to ensure that
heating, cooling, and thermal equilibration can occur as quickly as possible. The sample and
reference are located above their respective heaters. The temperatures of sample Ts and the one
of reference Tr are monitored using electronic temperature sensors located just beneath them.
Generally platinum resistance thermometers are used due to its high melting point.

Electronically, the instrument consists of two temperature control circuits [DEA1995] (Fig.II.3-2)

1) The average temperature control circuit (in blue) is used to monitor the progress of the
temperature control program set by the operator.
o The average temperature (Ts+Tr)/2 is compared with the programmed temperature
Tp.
o A power generator sends an electrical power P to the sample and reference heaters so
that (Ts+Tr)/2=Tp. This power P maintains the two heaters in the given temperature
program.

2) The differential temperature control circuit (in red) is used to determine the relative
temperatures of the sample and reference, and to adjust the power going to the respective
heaters so as to maintain both at the same temperature, when a disequilibria appears
(exothermic or endothermic phenomenon) in the sample. This adjustment is continuous
and automatic.
o The difference between temperatures Ts and Tr is controlled.
_____________________________________________________________________________________________________
-71-
o The second power generator sends an auxiliary power P to the heater with the lower
temperature so that Ts=Tr (in reality, the generator sends +P/2 in one hand and -
P/2 in other hand to decrease the response time of the apparatus).

3) The output of the differential temperature control circuit (in green) is used to generate the
DSC curve:
o X-axis represents the average temperature Tp=(Ts+Tr)/2
o Y-axis represents the power P

SAMPLE

Ts
REFERENCE

Tr
T = Ts-Tr
Power
generator
(Ts+Tr)/2
Power
generator
Temperature
control
Ps Pr
P/2 P/2
P P
P
T

Fig.II.3-2 [NAV1980]: Electrical diagram of DSC instruments

II.3.1.2. Temperature measurement
Measuring the temperature difference between the sample and the platinum resistance
thermometer located in the heater allows determining the sample temperature. This difference is
due to various thermal resistances present in the heater-sample-outer system. These thermal
resistances induce heat leakage (shown on Fig.II.3-3) and thus temperature variations.

_____________________________________________________________________________________________________
-72-

Fig.II.3-3: Various thermal resistances present in DSC

The various thermal resistances, which influence the system, are:
o Re: the sample thermal resistance
o Rf: the leakage thermal resistances due to thermometer wires
o R: the heater/outer thermal resistance
o Ro: the sample/heater thermal resistance

R and Rf are identical for the two ovens and, if the sample weight is low, Re << Ro.
Consequently, Ro is the only one thermal resistance to consider. Ro is the sum of two thermal
resistances, which are the sample/caps and caps/heater thermal couplings. The P power, sent
through the Ro thermal resistance, generates a temperature difference T between terminals of
Ro. This can be written as follows:
) / ).( / .( ) / .( . dt dT dT dW Ro dt dW Ro P Ro T = = =

= Tp Cp Ro T
E
. .
CpE represents the sample calorific capacity. E = S or L if
the sample is in solid or liquid state respectively.

II.3.1.3. Melting temperature measurement
A melting thermogram reveals the melting temperature Tf and gives the melting enthalpy H.

In the case of a pure substance, melting is isothermal. Tf is determined at the onset of the peak
and corresponds to the intersection between the straight line
= Tp Cp P
S
. (CpS is the calorific
capacity of the sample in solid state) and the straight line of peak increase. The melting thermogram of
distilled water is an example (Fig.II.3-4).

_____________________________________________________________________________________________________
-73-
-20 -10 0 10 20
TC
P
Endothermic
reaction
Exothermic
reaction

Fig.II.3-4: Melting thermogram of distilled water
Heating rate = +1C/min

In the case of a binary system where the two components are not miscible, the DSC
thermograms reveal two peaks. The first one allows defining the solidus, it is sharp and the
melting temperature is determined as in the case of a pure substance: at the onset of the peak.
The second peak allows defining the liquidus, the melting occurs in a range of several degrees and
Tf is determined at the end of the peak and corresponds to the intersection between the straight
line
= Tp Cp P
L
. (CpL is the calorific capacity of sample in liquid state) and the straight line of peak
decrease. The melting thermogram of 7.6% NaOH aqueous solution is an example (Fig.II.3-5).

-50 -40 -30 -20 -10 0 10
TC
P

Fig.II.3-5: Melting thermogram of 7.6%NaOH aqueous solution
Heating rate = +1C/min

= Tp Cp P
S
.
T
f

= Tp Cp P
S
.
T
f(end)

= Tp Cp P
L
.
_____________________________________________________________________________________________________
-74-
II.3.1.4. Melting enthalpy measurement
Integrating the melting peak gives the melting enthalpy. It can be shown that the enthalpy of
transition can be expressed as follows:
KA H = H is the enthalpy of transition
A is the area under the curve
K is the calorimetric constant
The calorimetric constant K varies from instrument to instrument and depends on experimental
conditions. K can be determined by analyzing a well-characterized sample with known enthalpies
of transition [PUN1994].

II.3.1.5. Calibration and experimental conditions
Before quantitative measurement, the calorimeter has to be calibrated to obtain the calorimetric
constant K and to determine an exact temperature scale. This calibration is performed by using
very pure substances with well-known melting enthalpies and with melting temperatures near the
experimental conditions.
s
s
A
H
K

=
HS is the melting enthalpy of the well-characterized pure substance
AS is the area of DSC melting peak of this well-characterized pure substance

In studies of an ordinary polymer or metal, indium is the most often used as a reference; its
melting temperature is 156.6C. As our experiments are performed between -60C and +20C, it
is better to use a reference substance with a melting temperature below 0C. Moreover,
calibration has to be performed in more or less the same conditions as the experiments, that is to
say in the same capsules and with the same heating rate.

Stainless steel gold-plated screwed capsules have been used because of the corrosive action of
high concentrated NaOH-water solutions on aluminium capsules. These stainless steel capsules
(Perkin-Elmer B018-2901 and B018-2902) withstand an internal pressure of 150 atmospheres
and temperature up to +400C. To carry out a perfect capsule locking, a gold-plated copper
seal is added and capsule is locked with wrenches.
The stainless steel capsules have a high thermal resistivity: heat flows are slow. Thus, in order to
be near the thermodynamical equilibrium, it is necessary to apply low heating and cooling rates.
So heating and cooling rates were equal to 1C/min for all experiments of this work.

DSC apparatus was calibrated with n-decane (to determine K) and distilled water (to fix the
temperature scale). The melting temperature of n-decane is -29.66C; its melting enthalpy is
202.090J/g. The melting temperature of distilled water is naturally considered as being 0C.
Moreover, in order to check the temperature scale, n-dodecane with a melting
temperature at -9.65C was analysed.

_____________________________________________________________________________________________________
-75-
II.3.1.6. Software for peaks deconvolution: PeakFit 4.12
Each solid phase present in a system gives a melting peak when its melting temperature is
reached. As enthalpy is a function of weight (enthalpy unit = J/g), each melting peak depends on
the amount of the corresponding solid phase. Thus, we can deduce that measurement of melting
enthalpies gives information on the contribution of each phase in the system.

In the case of a mixture, several peaks may appear and each of it has to be characterised. Fig.II.3-
5 shows the melting peaks of the NaOH/water eutectic mixture and of water in
7.6%NaOH/water solution (such thermograms will be explained in chapter III); here both peaks
are well separated from each other and can be easily characterised. For the case of urea/water
solution, the situation becomes more complicated (Fig.II.3-6a) and with the increase of urea
concentration, water peak is practically hidden inside melting peak of urea/water eutectic
mixture (Fig.II.3-6b). In order to obtain melting enthalpies of each peak, a peak deconvolution
program was used: PeakFit 4.12 from Systat. An example of similar thermograms treatment with
this program can be found in literature [PRA1995].

DSC thermograms and peaks deconvolution
6% urea aqueous solution
0
0,2
0,4
0,6
0,8
1
1,2
1,4
-30 -25 -20 -15 -10 -5 0 5 10
TC
H
e
a
t

f
l
o
w

(
W
/
g
)
experimental points
eutectic peak
ice peak
sum
0
0,2
0,4
0,6
0,8
1
1,2
1,4
-30 -25 -20 -15 -10 -5 0 5 10
TC
H
e
a
t

f
l
o
w

(
W
/
g
)
experimental points
eutectic peak
ice peak
sum
a)
b)

Fig.II.3-6: DSC melting thermograms and peaks deconvolution for a) 6% urea + H2O and b) 20%urea + H2O
_____________________________________________________________________________________________________
-76-
First of all, the equations describing the peak shape must be chosen. On DSC thermograms
(Fig.II.3-6a), the melting peak of water (at higher temperature) is very asymmetric and can not be
defined by simple Gaussian or Lorentzian functions. Several functions were tested and an
Exponentially Modified Gaussian function (EMG) was chosen because of the best fit and
possibility to vary the asymmetric coefficient (Eq.1).

(
(
+
|
|
\
|
|
|
\
|

+ =
3
3
3
2
2
1
3
1
2
3
2
2
3
0
. 2 . 2 2
exp
2 a
a
a
a
a
a x
erf
a
x a
a
a
a
a
y Eq. 1

Where x corresponds to the temperature and where the coefficients a
0
, a
1
, a
2
and a
3
are adjustable
parameters:
a0 is the peak area
a1 is the peak centre
a2 is the peak width (>0)
a3 is the distortion (0)

Fig.II.3-6 reveals the comparison between experimental points and calculated curves. The fit is
good, the same result was obtained for other thermograms. Thus, the choice of the EMG
function is correct and will be used in the following for peaks deconvolution.
Such deconvolutions were performed for all experiments. After having fitted the four variables,
the software calculates the percentage of each peak in comparison with the total area. As DSC
analysis gives the total enthalpy, we can determine the contribution of each phase and calculate
the melting enthalpies of the different compounds.

II.3.2- XR diffraction
In our study, X-rays diffraction was only used to check:
If the solid phase, which crystallises in binary system containing more than 30% of urea +
water (see Chapter III), is pure urea.
The nature of cellulose samples tested in this work. Are they cellulose I or cellulose II?

The diffractometer system was a XPERT-PRO (from PANalytical). The electron voltage was
40kV, the current was 30mA and the anticathode was made of Cu, so the wavelength K
1
was
1.5406 . The diffraction patterns were obtained by a continuous single scan from 5 to 60 with
0.04 steps.

The preparation of samples consisted in deposing the material (in liquid or solid state) on the
holder. Pure urea and cellulose samples were in solid state pellets and fibres respectively.
Moreover, as experiments were carried out at room temperature, the 70%urea+water sample was
a mixture of crystals and solution. It was the only sample for which the surface was difficult to
prepare. The surface has to be very smooth and in the continuity of the borders of the support.
_____________________________________________________________________________________________________
-77-
When the surface is irregular, the diffraction peaks can be slightly shifted in comparison with the
theoretical ones. If the studied sample is unknown, translation of some peaks can induce serious
mistakes in the determination of nature, crystallographic structure, and elementary unit of crystal.
In our case, the components present in the samples were known, so the irregularity on surface of
the 70%urea+water sample was not a problem.

The diffraction spectrum consists in a network of peaks or spots (depending on the detector
nature). Each trace corresponds to a series of lattice planes (hkl), which is characteristic of the
crystallographic structure. So, the crystalline parameters can be determined from the Braggs law:
n d
hkl
= sin 2
dhkl is the length between two consecutive lattice plane and
1
2
c
l
b
k
a
h
d
hkl
+ + =
h, k and l represent the Miller indexes of the considered lattice plane and a, b and c
represent the parameters of crystalline unit.
is the angle between the sample surface and the incident ray.
is the wavelength of X-rays used.
n diffraction order

i incident radiation (hkl) lattice plane
d scattered radiation Braggs angle

For native cellulose and cellulose II, the lattice parameters are given in Appendix I.

II.3.3- Rheology
The flow and gelation of semi-dilute and dilute cellulose solutions were studied using different
rheological methods. They are briefly described in this section.

II.3.3.1. Semi-dilute cellulose solution
Rheological measurements of semi-dilute cellulose solutions were performed on a:
StressTech rheometer (Reologica Instruments) equipped with a thermobath to control the
temperature. The tested solutions were equilibrated at a given temperature for 30 minutes
before starting the viscosity versus shear rate measurements.
The geometry used was a Couette double gap, it is made of two concentric cylinders (both
gaps = 2mm). The volume of cellulose solution was about 24ml.
_____________________________________________________________________________________________________
-78-
Gemini 150 rheometer (Bohlin Instruments) equipped with a Peltier temperature control.
With such a temperature control, temperatures can vary in a quick and accurate way, up to 60
degrees per minute in the interval from -30C to 180C. The sample remained 60s at
temperature before measurement.
The geometry used was a cone and plane (60/2), made of stainless steel. About 3ml of
solution was placed on the plate in order to fill the gap perfectly (Fig.II.3-7). A solvent trap
was used to avoid evaporation.

Fig.II.3-7: Scheme to fill the gap between cone and plate correctly

Steady state mode
The measurements of viscosity (Pa.s) versus shear rate (s
-1
) were performed in isothermal mode
at different temperatures from 0C to +40C. The shear rate varied from 0.05s
-1
to 100s
-1
.

Oscillatory mode
The storage modulus G and the loss modulus G were measured as a function of time or of
temperature.
The kinetics of gelation was studied at various constant temperatures from +10C to +40C.
To be in linear regime (Appendix 2), the frequency was fixed at 0.1Hz and the stress at
0.01Pa.
The gelation temperature was determined from a sort of an express method where G and
G were measured as a function of temperature. Keeping the experimental conditions always
the same, this method allows a quick comparison of the gelation process of different cellulose
solutions.
Temperature varied from +10C to +80C with a heating rate of +5C per minute. We chose
a rapid heating rate in order to neglect the influence of time.
The rheometer was in the auto-stress mode: it fits stress during the measurement to maintain
a constant strain of 1%.
The frequency was fixed at 1Hz.

II.3.3.2. Dilute cellulose solution
The intrinsic viscosities of cellulose solutions were studied using an Ubbelohde capillary
viscometer equipped with an automatic dilution burette (from Lauda). The viscometer was
immersed in a water bath at different constant temperatures from +12C to +40C.
The initial concentration of a cellulose solution was 0.5% obtained with 10 times dilution of a 5%
cellulose solution. In order to obtain the viscosity versus cellulose concentration dependence, 7
or 8 dilutions were made. The relative viscosity
rel
is calculated as a ratio between the kinematic
viscosities of the solution
Solution
and the solvent
Solvent
(Eq. 3):
_____________________________________________________________________________________________________
-79-
Solvent
Solution rel
= Eq.2
The intrinsic viscosity [] is determined as follows:
|
|
\
|
=
c
rel
c
1
lim ] [
0
Eq.3
where c represents the polymer concentration.

The intrinsic viscosity is the viscosity of solution at infinite dilution and is usually obtained when
plotting
|
|
\
|
=
c
rel
red
1
dependence as a function of concentration. Graphically, the intrinsic
viscosity corresponds to the point of intersection between the straight line passing through the
reduced viscosity points
red
and the y-axis

For a given polymer in a given solvent, a decrease of the intrinsic viscosity means that the
intramolecular links are favoured and the solvent quality decreases.

II.3.4- Microscopy
II.3.4.1. Optical Microscopy (OM)
The observations of cellulose solutions were performed with a transmission optical microscope
Leitz Metallux 3 with unpolarised light in order to observe the swelling of cellulose fibres and the
state of cellulose dissolution.
Qualitative observations of cellulose solutions were carried out at room temperature.
Cellulose + solvent (solvent being 7.6NaOH/water or 7.6NaOH/0.7ZnO/water or
7.6NaOH/6urea/water) fresh solution was placed between two glass plates. In this case,
solution had been prepared before the observation and the results were time-independent.
The study of cellulose dissolution mechanism was performed at -6C: the optical microscope
is equipped with a LINKAM TMS 91 hot stage. A few dry cellulose fibres were placed
between two glass plates slightly shifted (Fig.II.3-8). The solvent contained in a pipette was
introduced by capillary forces between the two plates. The whole process of swelling and
dissolution was recorded with CCD camera on DVD recorder and then analysed.

Solvent
Cellulose
fibres

Fig.II.3-8: Samples preparation for optical microscopy at -6C.

_____________________________________________________________________________________________________
-80-
II.3.4.2. Scanning Electronic Microscopy (SEM)
Scanning Electronic Microscopy (SEM) was used to observe the porosity of regenerated cellulose
samples and the reinforcing fibres (cotton and fluff) in their initial state and after treatment in
NaOH solutions. SEM has two main advantages in comparison with optical microscopy: to
obtain greater magnification (up to 150000 times instead of ~1000 with OM) and to observe
irregular samples because of an important depth of field.

Electron gun sends an electron beam to the sample surface. When the electron beam strikes the
target, it is reflected and the resulted incident radiation is picked up by the SEM sensors. This is
repeated very rapidly on several points until covering the whole sample surface. In other words,
the electron beam scans the whole sample and data are analysed to build the picture.

A non-conducting material reflects very badly the electron beam and it is very difficult to obtain a
picture. So, either the sample is covered by very fine gold particles or the microscope is used with
the environmental mode. Indeed, the latter allows observing non-conducting samples due to the
presence of a gaseous atmosphere in the microscope chamber. A Peltier temperature control is
used for monitoring the relative humidity in the microscope chamber.

The instrument used was a Philips XL-3000 Scanning Electron Microscope.

Environmental mode
The temperature was fixed at +8C and the chamber pressure at 2.5 mbar, giving a relative
humidity around the sample of approximately 30%. The voltage was fixed at 15kV and a GSE
detector was used.
The samples were placed in the vacuum chamber, without other treatment.
The problem of environmental mode is a relatively low resolution.

High vacuum mode
This mode allows obtaining greater magnification with a good resolution because of high vacuum
in the microscope chamber. Samples have to be metallised with gold particles.
SE detector was used and voltage was fixed at 12kV or 15kV depending on experiments.

_____________________________________________________________________________________________________
-81-
II.3.5- Mechanical property: tensile test
The mechanical properties of regenerated cellulose films, of regenerated cellulose + reinforcing
fibres and of final product (regenerated cellulose + re-enforcing fibres + porophores) were
studied using the tensile mode on a ZWICK Z2.5 machine (Fig.II.3-9).

The experiments were performed at room temperature. The sample was fixed between two
holders or grips (see Fig.II.3-9b). The lower grip stayed immobile and the higher grip moved
until rupture. The tensile force was measured. The rupture force F was considered as being 80%
of maximal tensile force. On the forthcoming graphs, one point corresponds to 3-4
measurements at least. Error bars show the standard deviations.

G
G
S

(a) (b)

Fig.II.3-9: Picture of the Zwick Z2.5 machine and scheme of the tensile test.
G: grips; S: specimen of regenerated cellulose; F: the tensile force.

The rupture stress F
R
is determined as follows:
S
F
F
R
=
FR is the rupture stress (daN/cm)
F is the tensile force when rupture appears (daN)
S is the initial specimen section (cm)
Units were chosen to easily compare with the Spontex values.

II.3.5.1. Tensile test on films of regenerated cellulose
Tensile tests were performed on films of regenerated cellulose. This means that cellulose
solutions were prepared from the previous procedures (part II.2-), then underwent various
thermal treatments and finally they were placed in a regeneration bath to obtain the final
products. After regeneration, cellulose has the crystalline structure of cellulose II, as expected.

_____________________________________________________________________________________________________
-82-
The procedure was as follow:

1. Cellulose solution was prepared from the procedure described above (part
II.2-). Cellulose was the B389 or B407 pulp and the solutions consisted in
5cellulose/7.6NaOH/water, 5cellulose/7.6NaOH/0.7ZnO/water or
5cellulose/7.6NaOH/6urea/water.

2. Cellulose solution was stored at +5C up to the disappearance of air
bubbles.

3. About 10g of the solution was poured in a plastic Petri box (=10cm).

4. Petri box was kept at various temperatures:
20 hours at +25C
1.5 hours at +70C
1.5 hours at +70C, then 20 hours at +25C
20 hours at -20C
20 hours at -20C, then 20 hours at +25C.

5. Immediately after the thermal treatment, a non-solvent was added (water or
1N acetic acid) into the Petri box at the temperature of regeneration (+5C,
+25C or +70C for water and +25C for acetic acid).
NB: In order to make the regenerated sample as homogeneous as possible
(in terms of thickness, surface, without holes, etc), the non-solvent was
carefully added ON the sample. Putting the sample in the regeneration bath
induced cracks and holes.

6. Regeneration bath was changed several times until it becomes pH neutral
(checked with a pH-paper).

After regeneration, tensile dumbbell-like specimens were cut. The dimensions were:
Effective length = 17mm
Width = 5mm
Thickness = 1-1.5mm
After the preparation of the specimen, it was immediately put back
into regeneration bath to avoid drying in the air. The specimens were
removed from the bath just before performing the tensile test. As the
experiment lasted for about one minute, it was not necessary to take
into account the humidity of ambient air.
The speed of the top grip is fixed at 5mm/min.

17mm
5mm
_____________________________________________________________________________________________________
-83-
II.3.5.2. Tensile test on regenerated cellulose + reinforcing fibres
In order to approach industrial conditions, but varying different parameters, reinforcing fibres
were added to cellulose solution. The fibres were either cotton with a length of 20mm or fluff.
Reinforcing fibres, cotton and fluff, have high DP (>1100). The dissolved cellulose + reinforcing
fibres system was regenerated in water before mechanical tests.

The sample preparation was as follows:

1. Cellulose solution was prepared from the procedure described above (part
II.2-). Cellulose was the B389 or B407 pulp and the solutions consisted in
5cellulose/7.6NaOH/water.
NB: Stirring lasts 1.5 hour instead of 2 hours, according to the Spontex
procedure.

2. 0.2g of AP surfactant was added to the 100g cellulose solution and mixed
10 minutes.

3. 2g of reinforcing fibres (cotton or fluff) were impregnated with
8%NaOH/water at T
room
or at -6C (20 hours). These fibres were added to
the 100g solution (still at -6C) and mixed for 1.5 hour, at -6C.

4. The system of dissolved cellulose + fibres was poured in a mould and
different thermal treatments were applied:
a. 2 hours ageing at +5C and then 20 hours freezing at -20C
b. 20 hours freezing at -20C (without ageing)
c. 20 hours gelation at +25C
d. 2 hours gelation at +70C

5. Immediately after the thermal treatment, water at 70C or 25C was added
on the cellulose system.

6. Regeneration bath was changed several times until it becomes pH neutral.

After regeneration, two types of specimens were cut:
o tensile dumbbell-like specimens, as previously
o tensile parallelepiped specimens (for practical purposes). The dimensions were:
Width = 10mm
Thickness = 2-2.5mm
The speed of testing is also fixed at 5mm/min

_____________________________________________________________________________________________________
-84-

II.3.5.3. Tensile test on sponges (regenerated cellulose + fibres + porophores)
To reproduce the Spontex conditions, porophores were added to the previous system. The
porophores are crystals of Na
2
SO
4
10H
2
O, which melt above +40C.
A large quantity of crystals was included to the cellulose system in the amount of 3 to 4 times of
the weight of cellulose/fibres/NaOH/water system. In order to avoid an important temperature
increase of cellulose during the addition of porophores, Na
2
SO
4
was cooled down at +5C,
before being added to the system.

In this study, only fluff was used because of a better homogenisation with solution. Cotton fibres
are enrolling and sticking around the mixer.

The sample preparation was as follows:

1. Solution of cellulose + fluff fibres was prepared as previously: steps from 1
to 3.

2. Na
2
SO
4
crystals were pre-cooled at +5C, added to cellulose solution +
reinforcing fibres and mixed about 15 minutes.

3. The system of cellulose solution + fibres + porophores was poured in a
mould and different thermal treatments were applied:
a. 20 hours freezing at -20C (without ageing)
b. 20 hours gelation at +25C

4. Immediately after the thermal treatment, water at +70C was added on the
cellulose system.

5. Regeneration bath was changed several times until it becomes pH neutral.

After regeneration, tensile parallelepiped specimens were cut. The dimensions were:
Width = 25mm
Thickness = 10mm
The speed of testing is fixed at 300mm/min, to be near the Spontex procedure.

_____________________________________________________________________________________________________
-85-
II.3.6- Other methods
Several methods were used to analyse the properties of sponges. Thus regenerated samples
consisted of cellulose + fluff fibres + Na
2
SO
4
crystals.

II.3.6.1. Density measurement
The density is one of the main properties of sponges since it influences all their properties
(rupture stress, water absorption). The density is in g/dm
3
.
w
d
V
m
d =
d represents the density of the sponge (g/dm
3
)
md is the weight of the dry sample (g)
Vw is the volume of the wet sample (dm
3
)
To determine the density, a piece of wet sponge was cut and the volume was measured. The
sample was placed in an oven at 105C during 2 hours until a constant weight and then in a
dessicator during 15 minutes to cool down without trapping the ambient humidity. Finally, the
sample weight was measured.

II.3.6.2. Water absorption
The absorption capacity represents the quantity of water that a sponge can absorb.
d
d w
m
m m
Abs
) (
.

=
Abs. is the absorption capacity
mw represents the weight of wet sample
md is the weight of dry sample
The sample weight in dry state was measured and then it was placed in distilled water at room
temperature during a few minutes. The sample was taken at an extremity and drained during 1
minute. Finally the sample was weighted to obtain the wet weight.

II.3.6.3. Shrinkage
Shrinkage represents the percentage of surface loss between a wet sample, which has never been
dried, and a sample dried for 2 hours in the oven and then swollen in water again.
w
dw w
A
A A
Shr

= .
Shr. is the shrinkage of the sponge sample
Aw is the surface of a never-dried-wet sample
Adw is the surface of a dried-and-swollen again
sample.
To determine the shrinkage, the surface areas of never-dried-wet and of dried-and-swollen
samples were measured.

_____________________________________________________________________________________________________
-86-
[DEA1995] DEAN J.A. The Analytical Chemistry Handbook. New York: McGraw Hill, Inc. (1995),
p15.1-15.5

[NAV1980] NAVARD P., MONASSE B. Analyse calorimtrique diffrentielle compensation de puissance
(DSC). Ecole des Mines de Paris, Sophia-Antipolis, France (1980), 39 p.

[PRA1995] PRANKERD R.J., ELSABEE M.Z. Thermal analysis of chiral drug mixtures: the
DSC behavior of mixtures of ephedrine HCl and pseudoephedrine HCl enantiomers.
Thermochimica Acta (1995), 248, p.147-160

[PUN1994] PUNGOR E. Chapter 15: differential scanning calorimetry (DSC) In: A Practical
Guide to Instrumental Analysis. Boca Raton, Florida: CRC Press (1994), 400p.


[WIN1986] WINTER H.H., CHAMBON F. Analysis of linear viscoelasticity of a crosslinking
polymer at the gel point. J. Rheol. (1986), 30 (2), p.367-382

[YAM1990] YAMASHIKI T, MATSUI T, SAITOH M., OKAJIMA K, KAMIDE K, SAWADA
T. Characterisation of cellulose treated by the steam explosion method. Part 1: Influence of
cellulose resources on changes in morphology, degree of polymerisation, solubility and solid
structure. British Polymer Journal (1990), 22, p.73-83

[YAM1996] YAMANE C, SAITO M, OKAJIMA K Seni gakkaishi (1996), 52 (6), 310-317
Traduit du japonais - Mthode de prparation industrielle de cellulose-solution aqueuse de soude caustique
haute solubilit

Rsum du chapitre III Structure des solutions aqueuses de NaOH et de cellulose/NaOH avec et sans additifs.
Limite de dissolution de la cellulose
_____________________________________________________________________________________________________
-87-

R RE ES SU UM ME E D DU U C CH HA AP PI IT TR RE E I II II I
S ST TR RU UC CT TU UR RE E D DE ES S S SO OL LU UT TI IO ON NS S A AQ QU UE EU US SE ES S D DE E N NA AO OH H E ET T D DE E
C CE EL LL LU UL LO OS SE E/ /N NA AO OH H A AV VE EC C E ET T S SA AN NS S A AD DD DI IT TI IF FS S. .
L LI IM MI IT TE E D DE E D DI IS SS SO OL LU UT TI IO ON N D DE E L LA A C CE EL LL LU UL LO OS SE E
Ce chapitre prsente une tude dtaille des proprits thermiques des solvants dune part et des
solutions de cellulose dautre part, tude base essentiellement sur des analyses de DSC.
Lobtention des thermogrammes de fusion allant de -65C +10C permet de dterminer les
tempratures et enthalpies de fusion de chaque phase prsente dans le systme, de tracer les
diagrammes de phase et de dterminer les interactions ayant lieu dans les solutions.

La premire partie de cette tude a port sur lanalyse de solutions NaOH/eau ayant des
concentrations en NaOH comprise entre 0 et 20%. Le trac du diagramme de phase dans ce
domaine de concentrations en NaOH a t effectu partir des tempratures de fusion obtenues
par DSC et correspond parfaitement celui rapport dans la littrature [PIC1893] [ROL1964].
Entre 0 et 20%NaOH, quand tout le systme est solide (T<-33,4C), nous sommes en prsence
(i) dun mlange eutectique comprenant un hydrate de sodium NaOH,5H
2
O et 4 molcules deau
et (ii) de glace libre. La temprature de fusion de leutectique est constante (-33,4C) pour toute
concentration en NaOH tandis que celle de leau diminue quand la concentration en NaOH
augmente. Quant lenthalpie dun composant X, sa valeur est proportionnelle la quantit de
composant X cristallis dans le systme. En effet, les fractions deutectique et de glace
dtermines partir du diagramme de phase par la rgle des segments inverses correspondent aux
fractions calcules partir des enthalpies de fusion (f
eut
=H
eut, X%NaOH
/H
eut, pure
et f
ice
=H
ice,
X%NaOH
/H
ice, pure
). Cette mthode de calcul partir des enthalpies de fusion tant valide par
ltude des systmes binaires NaOH/eau, elle sera utilise non seulement pour les autres systmes
de solvant mais galement pour les solutions de cellulose.

Avant dtudier le systme ternaire NaOH/ure/eau, il nous a fallu dterminer le comportement
du systme binaire ure/eau et tracer son diagramme de phase. Il savre que ce dernier a un
comportement de mlange eutectique simple. Leutectique est constitu de 1 ure et 8 molcules
deau, ce qui correspond une concentration en ure de 29,4%, il fond -11,4C.
Dans le systme ternaire, la concentration en NaOH est fixe 7,6% et celle de lure varie entre
0 et 25%. Il apparat que la temprature de fusion de leutectique NaOH ne diminue que trs
lgrement et que son enthalpie de fusion est constante quelle que soit la concentration en ure.
Ceci signifie que lure ne ragit pas avec les molcules de NaOH. Une augmentation de la
_____________________________________________________________________________________________________
-88-
concentration en ure entrane un dcalage du pic de fusion de lure vers les plus hautes
tempratures et celui de la glace libre vers les plus basses tempratures. Ils se rejoignent pour un
systme constitu de 7,6%NaOH et 18%ure, ce qui signifie quil ny a plus deau libre en
solution. En effet, cette composition, il y a tout juste assez deau pour former simultanment et
de faon indpendante les deux mlanges eutectiques NaOH+9H
2
O et ure+8H
2
O. Au-del de
cette concentration, il ny a plus suffisamment deau pour les deux eutectiques qui rentrent alors
en comptition. Le pic de leutectique NaOH tant toujours constant pour des concentrations
en ure suprieures 18%, ceci signifie quil est prioritaire sur lure, probablement du fait de sa
forte polarit. Lure en excs cristallise donc dans sous forme pure.

Nous avons trouv dans la littrature [DIR1954][REI1975][BEN1999] que la dissolution de
loxyde de zinc a lieu en milieu basique par raction des ions OH
-
avec ZnO pour former des
hydroxydes de zinc. Ces ractions chimiques entranent une modification de la solution et il nest
donc pas tonnant de constater que la prsence doxyde de zinc dans le systme NaOH/eau
modifie le diagramme de phase : les temprature et enthalpie de fusion de leutectique NaOH
diminuent proportionnellement la quantit de ZnO prsente en solution. Nous en avons conclu
que chaque ZnO ragit avec 3 NaOH qui ne participent plus leutectique. Les complexes ainsi
forms sont fortement hydrats puisquils peuvent tre entours jusqu 25 molcules deau,
valeur dtermine partir du pic de glace libre.

Lajout de cellulose dans NaOH/eau ne modifie pas les tempratures de fusion du mlange
eutectique NaOH+eau et de la glace libre : la prsence de cellulose ne modifie pas le diagramme
de phase du solvant NaOH/eau.
Par contre, lenthalpie de fusion du mlange eutectique diminue proportionnellement la
concentration en cellulose et atteint zro quand la masse de cellulose est gale la masse
dhydroxyde de sodium. Nous en avons donc conclu que la limite de dissolution de la cellulose
dans NaOH/eau est atteinte quand m
cell
=m
NaOH
et que la dissolution de la cellulose ncessite au
moins quatre molcules de NaOH qui se lient chaque motif de polymre (AGU). Le rapport
entre lenthalpie de fusion de leutectique dtermine pour une concentration en cellulose donne
et lenthalpie de fusion de leutectique pur (H
eut, Y%cell
/H
eut, pur
) permet de calculer la fraction
deutectique qui a cristallis dans le systme et den dduire la quantit de NaOH li la cellulose,
et ce, pour chaque concentration de cellulose. Les rsultats montrent que pour des concentrations
en cellulose suprieures 1,5-2%, le rapport NaOH/AGU est gal 4. Par contre, pour de
faibles concentrations en cellulose, le rapport NaOH/AGU peut atteindre 20, ou encore 50
partir de mesures de turbidit [KUO2005]. Cette transition autour de 2% de cellulose correspond
la transition rgime dilu / rgime semi-dilu rapporte dans la ref. [ROY2003]. Il nest certes
pas tonnant de trouver des valeurs un peu plus importantes en rgime dilu d la diminution
de lencombrement strique mais un tel cart entre les deux types de solutions (4/20 ou
encore 4/50) reste, lheure actuelle, une question ouverte.
En ce qui concerne le pic de fusion de la glace libre, son enthalpie reste constante quelle que soit
la teneur en cellulose. La quantit deau libre en solution nest pas modifie par la prsence de
_____________________________________________________________________________________________________
-89-
cellulose, ce qui signifie que les 9 molcules deau cristallisant dans le mlange eutectique
NaOH+eau sont galement lies la cellulose.
Chaque AGU est donc entoure dun minimum de 4NaOH et 36H
2
O.

Une analyse similaire a t effectue sur des solutions de cellulose dans 7,6NaOH/6ure/eau. Les
thermogrammes de DSC ainsi obtenus donnent des rsultats tout fait semblables aux
prcdents, savoir (i) pas de diminution des tempratures de fusion (eutectique NaOH,
eutectique ure et glace libre), la prsence dure ne modifie le diagramme de phase du solvant,
dans les domaines de concentrations tudis, (ii) pas de variation des enthalpies de fusion de la
glace libre et de leutectique de lure, la quantit deau libre en solution reste constante et la
cellulose ne se lie pas avec lure mais (iii) diminution de lenthalpie de fusion du mlange
eutectique NaOH+eau proportionnelle la teneur en cellulose, la cellulose pige une certaine
quantit de NaOH. De plus, la courbe enthalpie de leutectique en fonction de la concentration
en cellulose pour des solutions de cellulose dans 7,6NaOH/6ure/eau se superpose celle
obtenue pour des solutions de cellulose dans 7,6NaOH/eau. Ceci signifie que la limite de
dissolution de cellulose est atteinte pour m
cell
=m
NaOH
, que la dissolution de la cellulose ncessite
au moins quatre NaOH entour de ses 9 molcules deau. Lure ne permet pas daugmenter la
quantit de cellulose pouvant tre dissoute dans un systme NaOH/eau donn.

Nous avons vu que la prsence doxyde de zinc diminue la temprature de fusion du mlange
eutectique NaOH+eau, il y a donc modification du systme de solvant. Les thermogrammes de
fusion des solutions de cellulose/NaOH/ZnO/eau ne rvlent pas le pic de fusion de
leutectique, il nest donc pas possible de dterminer la quantit de NaOH li la cellulose. Par
ailleurs, lenthalpie du pic de fusion de la glace libre reste constant mme aprs ajout de cellulose :
comme prcdemment pour les solutions de cellulose/NaOH/eau et cellulose/NaOH/ure/eau,
les solutions de cellulose/NaOH/ZnO/eau ont la mme quantit deau libre quelle que soit la
concentration en cellulose.

Pour finir, nous avons tudi linfluence dune tape de conglation sur les proprits thermiques
des solutions de cellulose/NaOH/eau, avec ou sans ZnO. Dans les deux cas, un sjour prolong
-28C entrane la cristallisation deau pure, dont la temprature de fusion est de 0C. Dun point
de vue thermodynamique, ces molcules deau ne peuvent pas se dtacher du mlange eutectique
NaOH/eau. De plus, tant donn que leau libre a dj cristallis dans le systme, la seule
possibilit est que cette eau pure soit dtache de la cellulose. A partir des rapports
denthalpies de fusion, nous avons pu dterminer que dans les deux cas, cellulose/NaOH/eau et
cellulose/NaOH/ZnO/eau, une molcule deau par AGU est dtache lors dun sjour prolong
-28C.

Cette analyse approfondie nous a permis de dterminer la structure des solvants NaOH/eau,
ure/eau, NaOH/ure/eau et NaOH/ZnO/eau et celle des solutions de cellulose dans ces
mmes solvants.

-90-

Chapter III Structure of aqueous NaOH and cellulose/NaOH solutions with and without additives. Limit of cellulose dissolution
_____________________________________________________________________________________________________
-91-
C
C
H
H
A
A
P
P
T
T
E
E
R
R
I
I
I
I
I
I

III- III
S ST TR RU UC CT TU UR RE E O OF F A AQ QU UE EO OU US S N NA A
O OH H A AN ND D
C CE EL LL LU UL LO OS SE E/ /N NA AO OH H S SO OL LU UT TI IO ON NS S W WI IT TH H A AN ND D
W WI IT TH HO OU UT T A AD DD DI IT TI IV VE ES S. .
L LI IM MI IT T O OF F C CE EL LL LU UL LO OS SE E D DI IS SS SO OL LU UT TI IO ON N
. .
Chapter III is devoted to a detailed study of the structural organisation of cellulose in the
aqueous sodium hydroxide solutions with and without additives. The main goal is to understand
what are the proportions between cellulose and solvent molecules linked to it and, as a
consequence, what is the limit of cellulose dissolution in aqueous sodium hydroxide solutions.
The role of additives like ZnO and urea that are known to help cellulose dissolution is
investigated. The changes in cellulose/NaOH solutions induced by freezing are also considered.
Most of the study is performed using DSC technique accompanied by viscometric measurements.
This chapter is divided into four main parts.

In the first part, entitled Bibliography: phase diagram of aqueous sodium hydroxide solution,
we will first describe some theoretical aspects concerning binary and ternary phase diagrams, and
especially how to build them from DSC analyses. Then, we will review the literature on the
structure of NaOH/water: it is known that several sodium hydrates can be formed depending on
NaOH concentration. The number of water molecules surrounding sodium hydroxide ions
increases when the NaOH concentration decreases. The authors of two papers [PIC1893]
[COH1960] gave the composition of sodium hydrates and the melting temperatures of different
compounds formed.

The second part concerns the structure of cellulose solvents: NaOH/water, NaOH/urea/water
and NaOH/ZnO/water. As the NaOH/water system was known, we shall compare our DSC
results obtained in the region of 0-30%NaOH with the literature and thus validate the
_____________________________________________________________________________________________________
-92-
methodology used. Then urea/water binary system will be studied and its phase diagram will be
built. This was the necessary step in the understanding of the structure of NaOH/urea/water
ternary system. Finally, the influence of zinc oxide on NaOH/water solutions will be
investigated.
The third part concerns the structure of cellulose solutions in NaOH/water, NaOH/urea/water
and NaOH/ZnO/water. We will study the influence of the addition of cellulose in the solvents
studied in the previous part. This will allow us to understand the structural organisation of
cellulose solutions and to determine the limit of cellulose dissolution in NaOH/water and in
NaOH/urea/water. A large part of this study is performed on Avicel cellulose; other cellulose
samples will be tested to study the influence of cellulose origin on the structure of solutions.

The last part deals with the influence of freezing on the thermal properties of cellulose/NaOH
mixtures.

_____________________________________________________________________________________________________
-93-
III.1- BIBLIOGRAPHY: PHASE DIAGRAM OF AQUEOUS SODIUM HYDROXIDE
SOLUTION
III.1.1- Phase diagrams
III.1.1.1. Binary phase diagram
Let us consider two components that are completely miscible in the liquid state but with their
solid phase fully immiscible [KON2001]. The melting point of the first component is always
lowered when adding a second one. This can be shown on a temperature versus mixture
composition diagram (Fig.III.1-1, graph). From such a binary phase diagram, at constant
pressure, the following parameters can be measured:
the compositions of solid and liquid phases from solidus and liquidus curves, respectively
the relative amounts of each phase in the whole system from level rule or level principle.
Three examples are given on Fig.III.1-1 for a, b and c systems.

The eutectic point, named e, is where the two solubility/melting curves of the solid phases meet
(Fig.III.1-1, graph). The composition x
2e
of the eutectic point is fixed at a given pressure. The
eutectic crystal is a mixture of the crystals of both components that were formed simultaneously.
Its morphology is very fine.
At either side of e and below T
e
, there are the crystals of the eutectic mixture plus an excess of
pure 1 or 2 crystals (depending on the composition of the initial mixture see Fig.III.1-1, table)
that have slowly crystallised upon cooling.

Component 1 Component 2
x2c x2a
d
q
r

Fig.III.1-1 [KON2001]: Binary phase diagram under constant pressure (left-hand graph)
Explanation on how to read such a diagram (right-hand table)

X-axis corresponds to component 2 fraction in the mixture.

Tm1 = melting temperature of a pure first component
Tm2 = melting temperature of a pure second component
Te = melting temperature of eutectic component
x2e = eutectic composition

Explanation

System a b c
Composition x2a x2e x2c
Temperature
T1 The whole system is
a liquid (l) with the
composition x2a
The whole system is
composition x2e
The whole system is
composition x2c
T3 There are two
phases:
A liquid phase (l)
composition d,
quantity r/(r+q)
A solid phase (s1)
pure component 1,
quantity q/(r+q)

There are two
phases:
A liquid phase (l)
composition d,
quantity r/(r+q)
A solid phase (s2)
pure component 2,
quantity q/(r+q)
Te The liquid (l)
crystallises to form
the eutectic
compound,
composition x2e, very
fine crystals of pure
components 1 and 2

T4 Two crystals:
Pure solid 1,
eutectic compound
composed of fine
crystals of
components 1 and 2
only eutectic
compound
Two crystals
Pure solid 2, eutectic
compound

_____________________________________________________________________________________________________
-94-
III.1.1.2. Ternary phase diagram
The representation of the equilibrium diagrams of a ternary system needs a three dimensional
model. Two dimensions are required to show the variation in composition, and a third dimension
is needed for the temperature variable. The latter is often drawn on the vertical axis and the
former on the horizontal plane (Fig.III.1-2). The three-dimensional model is a right-angled prism,
the sides representing the three binary systems [HUM1953].
Generally, isothermal cross-sections are useful to study the equilibrium of a ternary system. But
vertical cross-sections are interesting to reveal how a binary system is affected by the addition of
a third component.

Fig.III.1-2 represents the liquidus surface of a ternary phase diagram of three components A, B,
and C. These components are fully miscible in the liquid state and making two by two eutectic
mixtures. We can see that the addition of a component to the binary liquid phase decreases the
melting temperatures of the binary eutectic.
Example:
In the front of this figure, the diagram of the binary system containing the components A and B
is represented; the component C is not present, C
C
= 0%.
When the component C is added at a constant concentration C
C
=X%, the binary phase diagram
of components A and B is determined by the red plane. We can see that all the melting
temperatures decrease, even the eutectic one: T
0
mA
becomes T
mA
, T
0
mB
becomes T
mB
, and Te
AB

becomes Te
AB
(C
C
=X).

T
mA

(C
C
=X)

T
mB

(C
C
=X)

Te
AB

(C
C
=X)

X

X

A
B
C
AB
AB
AB
A

B

C

Fig.III.1-2 [KON2001]: Ternary isobaric phase diagram
showing the melting-point depressions of each of the
components caused by the addition of the other two to
the liquid phase.

T
0
mA, T
0
mB, and T
0
mC: melting temperatures of pure
components A, B, and C respectively;
eAB: the eutectic mixture of components A and B; TeAB its
melting temperature; eAB its projection on the base triangle;
TmA and TmB: the melting temperatures of components A and
B, when X% of the component C is added;
TeAB(CC=X): the melting temperature of eutectic eAB when
X% of the component C is added to A and B components;
Q: the quadruple point for the considered pressure; Q its
projection on the base triangle.

_____________________________________________________________________________________________________
-95-

Fig.III.1-3 [HUM1953]: Vertical cross-section of a ternary phase diagram
containing the components A, B and C.
The concentration of C is constant and equals 5%,
A and B concentrations vary from 0 to 95%.

Fig.III.1-3 is a vertical cross-section of a ternary phase diagram, where the concentration of C
remains constant (here C
C
=5%) and where A and B concentrations change. Consequently, the
points x and y represent two systems containing 5%C+95%B and 5%C+95%A respectively. The
point l on the liquidus curve shows that a system containing 5%C+50%A is exactly on the
liquidus surface at the temperature concerned. However, the point s is not the solidus point
corresponding to the liquidus point l. The reason for this is that all points on this figure represent
systems containing 5%C, but there is no reason why a liquid containing 5%C should be in
equilibrium with a solid containing the same percentage of C.
Thus, such a diagram only shows the phases present in the system at different temperatures. This
is the aspect which is interesting in our study of solvent systems. But such a diagram cannot be
used to deduce the compositions of mixtures in equilibrium with one another. Vertical sections
are thus easily misleading, because after working with binary phase diagram, there is an instinctive
tendency to imagine that horizontal lines may be drawn to give the compositions of various
phases.
In ternary systems, compositions of different phases can be determined by studying the
isothermal sections. This will not be described because a complete investigation of ternary
solvent system is not our topic.

III.1.1.3. Methods for building phase diagrams. The limits of the DSC analysis
The building of phase diagrams can be divided into two parts:

A) The first one consists in plotting the liquidus and solidus curves.
For example, the liquidus can be easily determined by solubility measurements.
Another classical method is the use of calorimetry, where a large amount of material is slowly
heated or cooled in order to be always in equilibrium conditions. Temperatures and heat
_____________________________________________________________________________________________________
-96-
exchanges are then used to build the phase diagram. It is a very slow method requiring a lot of
sample, but it is the only exact manner to plot the phase diagrams.

Differential Scanning Calorimetry (DSC) is often used to measure enthalpies of phase transitions
and their corresponding temperatures (part II.3.1) and thus to build phase diagram (Fig.III.1-4).
It is a very fast method, using only a very small amount of material. The drawback of it is that
several corrections must be applied if one wishes to have correct temperature values. Outside the
classical calibration corrections, one major difficulty is to obtain a correct temperature for the
melting of material phases like crystalline ones. To be correct, all measurements should be done
by extrapolating the data to zero heating or cooling rate and to zero mass. This is very time
consuming and usually it is not done. However, without such measurements there are some
difficulties in exploiting the DSC data. We give below (Fig.III.1-5) one example where we show
that we are making an error when using DSC to plot the liquidus if we are not using the
extrapolation at zero mass and zero heating rate.

Te

Fig.III.1-4: Solid/liquid equilibrium in a binary eutectic-forming system as observed by DSC.
Curve 1: Melting of pure components 1 and 2; Curve 2: Eutectic melting;
Curve 3: Melting of a mixture with an excess of solid 2 (in comparison with the eutectic composition)

T
T
3
T
4
T
1
T
2
Sample 1, with a mass m1
Sample 2, with a mass m2

Fig.III.1-5: Example of DSC traces showing the error when the extrapolation at zero mass
and heating rate is not used

For the samples of two different masses (m
2
> m
1
), the end of the melting peaks is shifted
towards higher temperatures for a higher-mass sample, both for a pure substance melting at a
single temperature or for a substance having a large melting peak. This is due to the time needed
for heat transfer inside the sample, considering that on a DSC trace such as the one shown here
(Fig.III.1-5), the horizontal axis that is usually labelled as the temperature axis is in reality a time
_____________________________________________________________________________________________________
-97-
axis. Each time that a substance is melting at a single temperature, the temperature is constant,
while time is flying. In the Fig.III.1-5, while points T
3
and T
4
are given in the DSC trace as two
different temperatures, the sample is in fact staying for a certain time between these two points at
the temperature T
3
, which is the melting temperature.

This will have a small impact on our data since it will slightly shift all temperatures of the liquidus
at high water content of the mixtures by less than 1C. The lower is the water content, the
smaller is the peak of the melting of ice, and the smaller is the error. The real and DSC-
determined liquidus shapes are as shown below in Fig.III.1-6.
It must be noted that the values of the enthalpy of the peaks measured by DSC do not depend
on mass or heating rates.

T
A= 100%
B= 0%
%A
%B
real liquidus
liquidus measured by DSC

Fig.III.1-6: Real and DSC-determined liquidus,
when the extrapolations at zero mass and heating rate are not used

In our study, the goal was not to plot very exact phase diagrams that would have taken a lot of
time, but to understand the structure of cellulose solutions and proportions between the
components. For this reason, we did not use the extrapolation technique to measure precisely
temperature, but we always used the same range of heating rates and masses. As we will see later,
this had no influence on our work.

B) The second one consists in determining the nature of the crystalline solid phases; X-Ray
diffraction analysis is the most often used method.

III.1.2- State of the art on the structure of NaOH aqueous solutions
III.1.2.1. Pure sodium hydroxide
Anhydrous sodium hydroxide is a white and translucent solid, strongly hygroscopic. Thus it easily
catches the humidity of atmosphere and, in that case, also absorbs carbon dioxide. Also called lye
or caustic soda, sodium hydroxide is a strong chemical base and, thus, its aqueous solutions
violently react with acid.
We saw previously (part I.2.2) that cellulose can be dissolved in relatively low NaOH
concentrated solutions, at low temperature.
_____________________________________________________________________________________________________
-98-
The anhydrous sodium hydroxide has two crystalline structures and , at low and high
temperature respectively. However, as far as NaOH is strongly hygroscopic, NaOH is essentially
found as hydrated forms. In the forthcoming, crystalline structures of anhydrous sodium
hydroxide will not be described, only the sodium hydroxide hydrates will be detailed.

III.1.2.2. Sodium hydroxide hydrates
Several papers [PIC1893][COH1960][ROL1964] report the formation of stable sodium hydrates
in which the number of water molecules depends on the NaOH concentration and on the
solution temperature.
In 1893, Pickering [PIC1893] identified 8 hydrates of sodium hydroxide:
The monohydrate (NaOH,H
2
O)
The dihydrate (NaOH,2H
2
O)
The trihydrate (9NaOH,28H
2
O which can also be written NaOH,3.11H
2
O). Pickering
[PIC1893] as well as Rollet and Cohen [ROL1964] are doubtful concerning this formula
but experiment of solubility for the former and thermal analyses for the latter do not
allow to make a choice between the NaOH,3H
2
O and NaOH,3.11H
2
O formulae.
The 3.5hydrate (NaOH,3.5H
2
O)
Two tetrahydrates (NaOH,4H
2
O and NaOH,4H
2
O )
The pentahydrate (NaOH,5H
2
O)
The heptahydrate (NaOH,7H
2
O)

Pickering [PIC1893] also investigated other alkaline hydroxides as KOH and LiOH and plotted
the liquidus curves for these three binary systems.
Some years later, Antropoff and Sommer [ANT1926] added the part of liquidus concerning the
anhydrous sodium hydroxide.
From thermal analyses and solubility studies, Cohen-Adad [COH1960], Rollet et al. [ROL1964]
checked the Pickering results, which turned out to be very precise [PIC1893]. They added the
solidus curves and thus completed the binary phase diagrams for NaOH/water, KOH/water,
LiOH/water, RbOH/water, and CsOH/water.

Fig.III.1-7 represents the NaOH/water binary phase diagram, plotted by Rollet and Cohen-Adad,
with the well-determined NaOH hydrates in crystallised states.

_____________________________________________________________________________________________________
-99-
Te
E
E
Te

Fig.III.1-7 [COH1960]: Binary phase diagram of NaOH/water system
with stable and metastable NaOH hydrates in crystallised state.

Y-axis corresponds to the temperature and X-axis to the mixture composition given in NaOH
weight percentage (x=0% corresponds to pure water on the left of the graph)

As we can see, this phase diagram is complex: it contains stable and metastable hydrates and
several eutectic mixtures. We can note that:
o x=24% corresponds to the stable crystalline sodium hydrate NaOH,7H
2
O
o x=31% corresponds to the metastable crystalline sodium hydrate NaOH,5H
2
O
_____________________________________________________________________________________________________
-100-
The points E and E (in the detailed part of the Fig.III.1-7) correspond to eutectic mixtures,
formed by ice and sodium hydrates NaOH,7H
2
O and NaOH,5H
2
O, respectively. According to
Rollet and Cohen-Adad [ROL1964], NaOH,7H
2
O is a stable hydrate obtained after a rapid drop
in temperature, in liquid air or dry ice, followed by an isothermal step at a higher temperature,
and NaOH,5H
2
O is a metastable hydrate, obtained when the temperature decreases relatively
slowly and that the thermal analysis is performed just after the crystals formation. They also
reported that the NaOH/water binary system forms the metastable equilibria very easily.

When ions are in solution, they have different solvated forms depending on their concentration:
in a concentrated solution, the two ions form a hydrated dipole, whereas in a dilute solution, they
are separated (Fig. I.2-6).
In the case of sodium hydroxide in water, Bartunek [BAR1954] determined the existence of
separated ions for NaOH concentrations inferior to 6% and Yamashiki [YAM1988] observed
separated ions for NaOH concentration around 9%. But whatever the concentration limits are,
the ions Na
+
and OH
-
are surrounded by a cage of water molecules. In 1949, Bockris
[BOC1949] defined a primary and a secondary cage of solvation. In the primary cage of
solvation, water molecules are strongly linked to the ions Na
+
and OH
-
(but it is not a covalent
link). In the secondary cage of solvation, water molecules are less linked to the ions and thus can
be easily exchanged with the free solvent molecules.

Different techniques (NMR, ultrasonic velocity, ionic mobility) were used to measure the
number of water molecules in different cages but the results are not in agreement. According to
Yamashiki [YAM1988], from
23
Na and
1
H NMR and ultrasonic velocity measurements at +4C,
for a 9% NaOH solution, the primary and secondary cages contain 8 and 23 water molecules
respectively (Fig.III.1-8), that is to say 31 water molecules on the whole. From
23
Na NMR, Kunze
[KUN1985] determined a total of 22 water molecules for a 9% NaOH concentration at 30C.
The comparison between these two figures is not totally possible because the temperature is very
different and we know that the structure of hydrates is strongly dependent of temperature. But
these results give an idea on the significant hydration state of Na
+
and OH
-
ions.

Fig.III.1-8 [YAM1988]: Schematic representation of a 9% (w/w) NaOH aqueous solution at +4C
Na
+
and OH
-
ions are in the centre of the cage,
the grey part represents the primary solvation cage
and the white part is the secondary cage.

_____________________________________________________________________________________________________
-101-
The cage structure and the number of water molecules linked to the ions determine the diameter
of the sodium hydrate. This parameter has been thought to be important for the cellulose
dissolution since the sodium hydrates must penetrate into the cellulose fibres.

III.2- STRUCTURE OF CELLULOSE SOLVENTS: NAOH/WATER,
NAOH/UREA/WATER AND NAOH/ZNO/WATER
The goal of this part is to precisely investigate the structure of solvent systems and to plot their
phase diagrams. Binary systems NaOH/water and urea/water will be first studied and then
ternary systems NaOH/urea/water and NaOH/ZnO/water will be investigated in order to
observe the influence of the addition of a third component. The knowledge of phases present in
solution and of the proportions between the species in this or that phase allows a better
understanding of the interactions between urea and NaOH and/or water, and between ZnO and
NaOH and/or water.

We should like to remind here (see also part II.2) how the proportions between the
components will be given in the following.
The notion of concentrations will be used only for binary solvents, like NaOH/water or
urea/water. In all other cases we shall use and show in the notations the amount of
grams of the substance in 100 g of solution. In other words, Xcellulose/YNaOH/water
means that in 100 g of solution there are X grams of cellulose, Y grams of NaOH and
(100XY) of water. Another example is as follows: Xcellulose/YNaOH/Zurea/water
means that in 100 grams of solution there are X grams of cellulose, Y grams of NaOH, Z
grams of urea and (100-X-Y-Z) grams of water. For the binary systems concentration in
weight per cent, calculated as % 100
water i
i
M M
M
Y
+
= (M
i
and M
water
being component i
and water weights, respectively), coincides with the amount of grams Y of this
component in 100g of solution: Ycomponent/water.

III.2.1- NaOH/water binary system
The binary phase diagram of NaOH/water system has been largely studied in literature (part
III.1.2.2). This part revisits it in order to validate our experimental results as well as the
methodology used to calculate the composition of each phase.

III.2.1.1. DSC melting thermograms of NaOH/water solutions
Thermal analyses with DSC consist in determining melting peaks for pure NaOH/water
solutions (pure means without urea, without ZnO and without cellulose). We used a
concentration of NaOH varying from 7.6wt% to 20wt%. The DSC thermograms are given on
_____________________________________________________________________________________________________
-102-
7.6 -33.2 -5.4
8 -33.2 -5.9
9 -33.6 -7.7
12 -33.5 -11.7
20 -33.3 no peak
NaOH, %
Tm (onset)
eutectic peak, C
Tm (end)
ice peak, C
Fig.III.2-1; the corresponding melting temperatures are reported in Tab.III.2-1 and the melting
enthalpies in Tab.III.2-2 and Tab.III.2-3.

The phase diagram of NaOH/water solutions (Fig.III.1-7 and Fig.III.2-2) shows that at low
sodium hydroxide concentration (<30%NaOH in water) a eutectic mixture and ice should melt at
lower and higher temperatures, respectively [PIC1893][COH1960]. In the following sections,
each peak is analysed in order to find the type of the melting compound and its composition.

0
0,4
0,8
1,2
1,6
-50 -40 -30 -20 -10 0
TC
Heat Flow (W/g)
20%
12%
9% 7.6%
8%

Fig.III.2-1: DSC melting thermograms of NaOH/water solutions.
CNaOH = 7.6%, 8%, 9%, 12% and 20%
Curves are shifted vertically for a better visibility.

Tab.III.2-1: Melting temperatures determined from DSC thermograms (Fig.III.2-1)

III.2.1.2. Peak at lower temperature: melting of the eutectic mixture
The melting temperature of the peak at low temperature is around -33.4C. The temperature
variation with the NaOH concentration is not negligible, but it is due to the inherent difficulty of
finding the proper beginning of melting with the DSC technique (part II.3.1). In addition, the
melting temperatures have not been measured at heating or cooling rates and mass of sample
extrapolated to zero, which induces a small temperature error (see part III.1.1.3). The peak
corresponds to the melting of the eutectic mixture. It is in the range of temperatures given in the
literature (-34C [COH1960]).

_____________________________________________________________________________________________________
-103-
A) What is the molecular composition of the eutectic mixture which crystallises during the
DSC experiments?
The peak at higher temperature disappears when the NaOH concentration is 20%. This means
that only eutectic is present in the solution, in agreement with the NaOH/water phase diagram
described in ref. [PIC1893][COH1960]. Fig.III.2-2 also represents NaOH/water phase diagram
and indicates the terms used in this report. Thus 20%NaOH/80%H
2
O in weight per cent is
the composition of the pure eutectic. Taking into account the molar weight of water and sodium
hydroxide, 18 and 40, respectively, we can also write:
m
NaOH
= 20% (m
NaOH
+ m
H2O
)
40 = 0.2(40 + n
H2O
* 18)
n
H20
~ 9
Thus, the eutectic mixture, melting at -33.4C, has a total composition of 1NaOH and 9H
2
O in
moles.

B) What is the structure of this eutectic mixture?
Let us consider the detailed structure of the eutectic, which is a mixture of two crystals.
In a recent paper [ROY2001], it is reported that the eutectic mixture is composed of the stable
crystalline sodium hydrate NaOH,7H
2
O and water. But, in fact, the melting temperature of the
eutectic containing the stable hydrate + water is -29C whereas the melting temperature of the
eutectic containing the metastable crystalline hydrate NaOH,5H
2
O + water is -34C [COH1960],
see Fig.III.1-7. In addition, Rollet and Cohen-Adad [ROL1964] reported that a low decrease of
temperature leads to the formation of metastable sodium hydrates. Because we used stainless
steel capsule, the temperature decrease rate is low (part II.3.1.5) which confirms the formation of
the eutectic metastable crystalline hydrate NaOH,5H
2
O + water.
From X-Ray diffraction pattern obtained by Roy [ROY_PhD], it is not possible to discriminate
both hydrate crystals since the experimental diffracting angle was 16.4, value very close to
literature data: NaOH,7H
2
O at 16.22 and NaOH,5H
2
O at 16.26 and 16.53 [MOO1994].

We conclude that the peak at low temperature corresponds to the melting of the eutectic mixture
composed of two crystals: the crystalline hydrate NaOH,5H
2
O and ice, the composition being
NaOH,5H
2
O + 4H
2
O, in order to obtain the ratio 1NaOH/9H
2
O.
_____________________________________________________________________________________________________
-104-

Liquid
E
Eutectic compound
NaOH,5H
2
O + 4H
2
O
f
hyd.
f
ice

hydrate+ice
L+ice
Tx
~-34
Solidus
f
liq.

f
eut.

f
ice

X%
Liquidus
H
2
O

Fig.III.2-2: Detailed part of NaOH/water phase diagram (Fig.III.1-7 [COH1960]) for 0-31% NaOH concentrations
fice : fraction of free ice, fliq: fraction of liquid
at T~-34C, fliq = feut with feut: fraction of the eutectic compound, function of the NaOH concentration
fhyd and fice fractions of sodium hydrate and ice, respectively, in the eutectic compound

On the Fig.III.2-2, the left-hand axis of the phase diagram corresponds to water. At the right-
hand side, the vertical line corresponds to the crystalline pentahydrate NaOH,5H
2
O having a
molar weight of 130g/mol. The composition of this crystalline hydrate is 30.8%NaOH and
69.2% water, in weight per cent.
For information, the level rule applied on this part of the binary NaOH/water phase diagram
enables to determine the proportion of each phase present in the eutectic mixture. Knowing that
the NaOH concentration at the eutectic point is 20%, we obtain:
o Hydrate fraction = f
hyd
= 20/30.8 = 65%
o Ice fraction = f
ice
= (30.8-20)/30.8 = 35%
The eutectic compound is a mixture of 35% ice and 65% crystalline NaOH,5H
2
O hydrate, in
weight per cent.

To conclude, the eutectic mixture can be defined:
o By the nature of crystals: NaOH,5H
2
O and ice, with one mole of NaOH,5H
2
O and
four moles of water
o By the proportion of crystals: 65% sodium hydrate + 35% H
2
O
o By the proportion of components: 20% NaOH + 80% H
2
O
Its melting temperature is ~33.4C.

_____________________________________________________________________________________________________
-105-
III.2.1.3. Peak at higher temperature
The melting peak at higher temperature changes with NaOH concentration (Fig.III.2-1). Higher
is the NaOH concentration; lower are the melting temperature and the intensity of the peak. This
peak disappears at 20% NaOH. It corresponds to the melting of free ice. In agreement with the
phase diagram, this peak is very large because of the gradual melting of crystallised water.

III.2.1.4. Proportions between NaOH and water
The peak at low temperature (-33.4C) corresponds to the melting of the eutectic mixture
containing 20% NaOH. Thus the DSC thermogram of the 20%NaOH solution gives the melting
enthalpy of pure eutectic: H
eut, pure
= 187 J/g (Tab.III.2-2).
The peak at higher temperature corresponds to the melting of free ice. The melting enthalpy of
pure crystallised water: H
ice,pure
= 358 J/g (Tab.III.2-3) was measured with distilled water. This
value is in the acceptable agreement with the data of literature (333.5 J/g [WEA1983]).
Applying the level rule on the NaOH/water phase diagram, it is possible to calculate the fraction
of eutectic mixture f
eut, calc.
and the fraction of free ice f
ice, calc.
at any NaOH concentration (X% on
the graph Fig.III.2-2). When the NaOH concentration C
NaOH
is lower than 20%, we obtain:

20
. ,
NaOH
calc eut
C
f = Eq. 1

20
20
,
NaOH
calc ice
C
f

= Eq. 2
Since we know the melting enthalpies of the pure eutectic H
eut,pure
and pure ice H
ice,pure
, it
comes that the experimental fractions of eutectic f
eut,exp.
and ice f
ice,pure
, at any NaOH
concentration, are:

pure eut
NaOH eut
eut
H
C H
f
,
. exp ,
. exp ,
) (
= Eq. 3

pure ice
NaOH ice
ice
H
C H
f
,
. exp ,
. exp ,
) (
= Eq. 4

The fractions, determined from DSC thermograms and calculated from the phase diagram, are
compared Tab.III.2-2 and Tab.III.2-3 for the eutectic and ice peaks, respectively.

CNaOH %
Heut,exp.(CNaOH), J/g
DSC
feut,exp.
Eq.3
feut,calc.
Eq.1
7.6 72 0.39 0.38
8 73 0.39 0.40
9 83 0.44 0.45
12 111 0.59 0.6
20 187 = Heut,pure 1 1
Tab.III.2-2: Melting enthalpy of eutectic mixture Heut,exp.(CNaOH) determined from DSC thermogram (Fig.III.2-1;
fractions of eutectic in NaOH+water solution at a given NaOH concentration CNaOH from 7.6% to 20%,
determined from DSC feut,exp. and calculated from phase diagram level rule feut,calc.
_____________________________________________________________________________________________________
-106-
CNaOH %
Hice,exp.(CNaOH), J/g
DSC
fice,exp.
Eq.4
fice,calc.
Eq.2
%
0 358 = Hice,pure 1 1
7.6 197 0.55 0.62 11
8 188 0.53 0.60 12
9 173 0.48 0.55 13
12 133 0.37 0.4 7
Tab.III.2-3: Melting enthalpy of free ice Hice,exp.(CNaOH) determined from DSC thermogram (Fig.III.2-1);
fractions of free ice in NaOH+water solution at a given NaOH concentration C
NaOH
from 0% to 13%,
determined from DSC fice,exp. and calculated from phase diagram level rule fice,calc..
Column 5 is the relative difference between the two experimental and calculated values.

Tab.III.2-2 and Tab.III.2-3 show that fractions calculated from phase diagram level rule and
determined from DSC thermograms coincide within the experimental errors, in agreement of the
phase diagram plotted in Fig.III.2.2. The NaOH and water mixtures behave as normal eutectic-
type mixtures. All the components are fully identified, at all concentrations under consideration;
their state (crystalline or liquid) and fractions are known from the phase diagram.
Thus the experimental data obtained and methodology used can be applied for the analysis of
urea/water binary system as well as for the analysis of ternary phase diagrams
NaOH/urea/water, NaOH/ZnO/water, cellulose/NaOH/water solutions and we will apply the
same consideration to four-body mixtures (cellulose/NaOH/water/urea).

Revision of the previous work
In the previous work on NaOH/water solutions [ROY_PhD][ROY2001], the comparison of
experimental and calculated ice fractions gave differences superior to 20%, even reaching 64%. In
order to explain this discrepancy, the presence of a third phase amorphous water was
suspected.

Fig.III.2-3: Scheme of two different calculations of melting enthalpy of ice.
The red area is the correct one.
_____________________________________________________________________________________________________
-107-
However, from the phase diagram (Fig.III.2-2), it is not possible to have three phases and the
thermograms used by Roy [ROY_PhD][ROY2001] were re-examined. It turned out that the peak
area of the ice was strongly underestimated due to error on the base line. The enthalpy was
calculated as a blue area shown on Fig.III.2-3 while the peak is larger: it starts at low temperature
(shown by on Fig.III.2-3), just at the beginning of the peak of the eutectic mixture, in
accordance with the phase diagram. The real enthalpy values are greater than the ones measured
by Roy (area in red is larger than in blue).
We can note that in our case (Tab.III.2-3), the relative difference between experimental and
measured values is about 10%. The fraction of free ice calculated from phase diagram and
determined from enthalpy measurements are in good agreement, within the experimental errors.
This means that there is no third component like amorphous water.

III.2.1.5. Conclusion on the phase diagram and structure of NaOH/water solutions
The mixture of sodium hydroxide and water in the 0-31%NaOH range is a simple solution of
two components: NaOH and water. They are miscible in liquid state and immiscible in solid
state. In our experimental conditions, sodium hydroxide and water form a eutectic mixture
composed of the sodium pentahydrate and water molecules in the following proportions:
NaOH,5H
2
O+4H
2
O. This eutectic mixture crystallises around 33.4C.
Consequently, at very low temperature (T<-33.4C), the NaOH/H
2
O mixture is composed of
two crystalline phases (Fig.III.2-4):
o The crystalline hydrate: NaOH,5H
2
O
o Ice in two locations:
free ice forming rather large crystals (if C
NaOH
< 20%)
ice with the crystalline hydrate in the form of tiny crystals to form the eutectic
mixture.

From the phase diagram, the calculations of the fractions of the various components presented at
a given composition and temperature show a good agreement with the experimental data. The
system is at equilibrium at all moment in the experimental conditions that have been used.
It is thus possible to schematically represent the structure at temperatures below -33.4C as
below:

Eutectic
mixture

Fig.III.2-4: Scheme of the structure of crystallised NaOH/water system at CNaOH<20% and temperature below -33.4C.
Note that the shape of the eutectic crystals is not known. It is presented here in the form of spheres for the sake of drawing.

_____________________________________________________________________________________________________
-108-
III.2.2- NaOH/urea/water ternary system
The aim of this section is to understand how NaOH and urea are reacting when they are mixed
in aqueous solutions. Firstly, we will determine the 0-100%urea/water binary phase diagram.
Secondly, NaOH/urea/water ternary system will be investigated: NaOH concentration being
fixed at 7.6wt% and urea concentrations varying from 0% to 25%. A section of the phase
diagram will be drawn.

III.2.2.1. Urea/water binary phase diagram
DSC experiments consist in determining melting peaks for urea/water solutions, urea
concentrations varying from 0% to 100%. This study is divided into two parts: from 0% to 30%
of urea and from 30% to 100%.

From 0% to 30% of urea
DSC melting thermograms of 0-30%urea aqueous solutions are represented on the Fig.III.2-5.
The melting temperature and enthalpy of each peak are reported in the Tab.III.2-4 and Tab.III.2-
5, respectively. In the Tab.III.2-4, the term solidus (DSC) corresponds to the onset of the peak
at low temperature and the term liquidus (DSC) to the end of the peak at higher temperature,
both determined with our DSC analyses. The measurements of the liquidus are reported in
literature from the data on urea solubility in water [RAB1978][WEA1983]. We can note that
DSC values and Handbook values are relatively close and differ because of the difference in
the method of determination.

0
0,4
0,8
1,2
1,6
2
-16 -12 -8 -4 0 4
TC
H
e
a
t

F
l
o
w

(
W
/
g
)
6%
30%
20%
16%
12%
8%

Fig.III.2-5: DSC melting thermograms for urea/water solutions
Curea = 6%, 8%, 12%, 16%, 20%, and 30%

_____________________________________________________________________________________________________
-109-
%urea solidus
(DSC) (DSC) (Handbook)
6 -11,3 0,8 -1,9
8 -11,3 -0,4 -2,6
12 -11,4 -1,3 -4,0
16 -11,2 -2,9 -5,4
20 -11,6 -4,5 -7,0
30 -11,2 -11,4
40 -11,4 0
50 -11,4
60 -11,4 39 35
70 -11,5 64 56
80 85 80
100 137 135
liquidus

Tab.III.2-4: Temperatures of solidus and liquidus curves.
DSC values correspond to the melting temperatures obtained by DSC experiments.
Handbook values correspond to solubility temperatures reported in handbooks [RAB1978][WEA1983]

The urea/water DSC thermograms (Fig.III.2-5) are similar to the NaOH/water DSC
thermograms (Fig.III.2-1). Apart for 30% urea, all DSC thermograms reveal two melting peaks
more or less well-separated:
o The first one, at low temperature, has the same position whatever is the urea
concentration from 6% to 30%. At the onset of the melting peak, the temperature is
about -11.3C (Tab.III.2-4, column 2). This peak should correspond to the melting of a
eutectic mixture.
o The second one, at higher temperatures, is shifted towards low temperatures when urea
content increases (Tab.III.2-4, column 3 and 4). We deduced that this peak probably
corresponds to the gradual melting of free ice.
The peak of free ice disappears at about 30% of urea. This means that the eutectic mixture
should contain 30% of urea in the whole.
From these data, the 0-30%urea part of the phase diagram can be drawn.

We can see here the effect of the phenomena described in part III.1.1.3 concerning the influence
of using non-zero heating rate and mass. The peak at high water content is finishing above 0C.
The error here is about 1C.

30% of urea
As the molar masses of urea CO(NH
2
)
2
and water are 60g/mol and 18g/mol respectively,
30% of urea in weight corresponds to a molar ratio n
H2O
/n
urea
= 7.8. Thus, the total eutectic
composition is probably 1 urea and 8 H
2
O. This can be checked using the above-described
methodology in the case of NaOH/water binary system.
Melting enthalpies of urea eutectic mixture and ice are reported in Tab.III.2-5. Since the two
peaks overlap, the enthalpy of each peak was deconvoluted with PeakFit software, described in
the part II.3.1.6. All the thermograms with their deconvolution treatment are shown in the
Appendix 3.

_____________________________________________________________________________________________________
-110-
DSC
% urea Total eutectic ice
0 358 = Hice,pure
6 325 51 275
8 323 67 256
12 319 111 208
16 311 155 156
20 308 212 96
29,4 290 = Heut,pure
30 291
Enthalpies
PeakFit

Tab.III.2-5: Melting enthalpies of eutectic mixture and free ice.
The total enthalpy was determined with DSC software from the thermogram.
Eutectic and ice enthalpies were calculated with PeakFit software as described in part II.3.1.6

The calculated and measured fractions of eutectic mixture and free ice are compared in order to
check the existence and the composition of the urea/water eutectic mixture. Calculations were
performed with the hypothesis that eutectic mixture is reached at 29.4% urea, which corresponds
to the composition 1 urea + 8 water molecules.
The equations are described below and values are reported in the Tab.III.2-6.

4 . 29
. ,
urea
calc eut
C
f = Eq. 5

pure eut
urea eut
eut
H
C H
f
,
. exp ,
. exp ,
) (
= Eq. 6

4 . 29
4 . 29
. ,
urea
calc ice
C
f

= Eq. 7

pure ice
urea ice
ice
H
C H
f
,
. exp ,
. exp ,
) (
= Eq. 8

% urea
feut,exp.
Eq.6
feut,calc
Eq.5
%
fice,exp.
Eq.8
fice,calc.
Eq.7
%
6 0,17 0,20 15 0,77 0,80 4
8 0,23 0,27 16 0,72 0,73 2
12 0,38 0,41 6 0,58 0,59 2
16 0,53 0,54 2 0,44 0,46 4
20 0,73 0,68 8 0,27 0,32 16
Tab.III.2-6: Comparison of fractions of eutectic mixture and free ice,
determined from DSC thermograms feut,exp. and fice,exp. and calculated from
phase diagram level rule (eutectic having 29.4%urea) feut,calc. and fice,calc.

Tab.III.2-6 shows that fraction of eutectic mixture and free ice calculated from phase diagram
and measured from DSC thermograms are similar within the experimental errors. This is
validating the hypothesis of a eutectic mixture containing 29.4wt% of urea, which corresponds to
a molar urea/water ratio of 1/8. This eutectic melts at -11.4C. But knowing the exact structure
of eutectic mixture needs an X-Ray diffraction analysis, as we will see below.

_____________________________________________________________________________________________________
-111-
From 30% to 100% of urea
DSC analyses were also performed on 30-100%urea/water solutions. Typical thermogram is
represented on the Fig.III.2-6 and the melting temperatures are reported in the Tab.III.2-4.
For urea aqueous solutions, two peaks are seen. The first one is still at -11.4C, as in the case of
0-30%urea/water solutions; thus corresponding to the melting of the same eutectic mixture. The
second one is shifted towards higher temperature. Pure urea gives a single peak at 137C.

TC
~ -11

Fig.III.2-6: Scheme of DSC melting thermograms for urea aqueous solutions with 30% < Curea < 100%.

In order to check the crystalline structure of the component which melts at high temperature for
an aqueous solution containing more than 30% of urea, and thus the composition of the eutectic
mixture, we carried out X-Ray diffraction analyses.
We tested pure urea and the 70%urea/water system. X-ray diffraction patterns were compared
and are represented on the Fig.III.2-7.

0
20000
40000
60000
80000
100000
120000
10 20 30 40 50 60
2Theta ()
I (cts)
100% urea
70% urea

Fig.III.2-7: X-Ray diffraction patterns for pure urea and for a 70%urea aqueous solution.

_____________________________________________________________________________________________________
-112-
The two patterns of 100%urea and 70%urea show the same diffraction peaks. Nevertheless, we
can note a slight shift of some peaks, especially at high value of 2 Theta. This could be explained
by an irregular surface of the 70%urea/water sample. We concluded that the peak at higher
temperature corresponds to the melting of pure urea.
Consequently, the eutectic mixture contains two crystals: pure urea and ice, without crystalline
hydrate. The urea/water binary system has a simple eutectic behaviour.

Conclusion on the urea/water binary phase diagram
DSC and X-Ray scattering enable us to draw the whole urea/water binary phase diagram
(Fig.III.2-8).
-20
0
20
40
60
80
100
120
140
0 10 20 30 40 50 60 70 80 90 100
% urea
T
C
Eutectic compound
extrapolation
pure ice
pure urea
Literature data

Fig.III.2-8: Binary phase diagram of urea/water system

The urea/water system has a typical eutectic behaviour:
o The eutectic compound is a mixture of two crystals: pure urea + ice. Its melting
temperature is -11.4C and its composition is 1 urea + 8H
2
O.
o From 0% to 29.4% of urea, the liquidus corresponds to the melting of pure ice.
o From 29.4% to 100% of urea, the liquidus corresponds to the melting of pure urea.

At low temperature (<-11.4C), the system has two crystalline phases but the structure and the
size of urea and ice crystals depends on the urea concentrations.
o Below 29.4%urea, the system is composed of pure urea and ice in two locations: free ice
and ice with pure urea to form the eutectic mixture (Fig.III.2-9a).
_____________________________________________________________________________________________________
-113-
o 29.4% is the eutectic composition; the system only contains the eutectic mixture, i.e.
crystals of pure urea and crystals of ice (Fig.III.2-9b).
o Above 29.4%urea, the system is composed of ice and urea in two locations: free urea and
urea with ice to form the eutectic mixture (Fig.III.2-9c).

a)

b)

* eutectic mixture

c)
Fig.III.2-9: Schematic structure of crystallised urea/water system below -11.4C,
a) Curea<29.4% ; b) Curea=29.4% ; c) Curea>29.4%

III.2.2.2. NaOH/urea/water ternary phase diagram
Now, as NaOH/water and urea/water binary phase diagrams are known, the NaOH/urea/water
ternary system can be investigated. The compositions of mixtures studied were in the range from
7.6NaOH/6urea/water to 7.6NaOH/25urea/water. In other words, 100g of solution
7.6NaOH/Zurea/water contained 7.6g of NaOH, from Z = 6 to 25 g of urea and from 86.4 to
69.4 g of water, respectively. DSC melting thermograms were plotted and are represented on the
Fig.III.2-10. In this range of urea concentrations, five melting peaks are seen. A given
NaOH/urea/water system reveals some of them depending on the urea concentration.

0
0,5
1
1,5
2
2,5
-50 -40 -30 -20 -10 0
TC
H
e
a
t

F
l
o
w

(
W
/
g
)
0 urea
12 urea
6 urea
18 urea
20 urea
14 urea
25 urea

Fig.III.2-10: DSC melting thermograms for 7.6NaOH/Z urea/water solutions (Z= 0, 6, 12, 14, 18, 20 and 25).
DSC traces were shifted for a better visibility
*
*
*
1
2
3
3
3
4
4
1
5
5
_____________________________________________________________________________________________________
-114-
Appendix 4 shows the DSC thermograms for 7.6NaOH/Z urea/water solutions (Z= 6, 12, 14,
and 18) with deconvolution treatment.

Peak n1
The peak n1 corresponds to the melting of the metastable eutectic mixture NaOH,5H
2
O+4H
2
O
(named NaOH eutectic mixture in the following). Tab.III.2-7 column 3 shows that as soon as
urea is added, the melting temperature is depressed from about -33.4C in binary NaOH/water
solution, without urea to about -35.6C with urea. This slight decrease of eutectic melting
temperature in ternary system compared with a binary system is due to the presence of the third
component (part III.1.1.2). On the other hand, we can note that this melting peak remains
constant for all urea concentrations studied because the NaOH/water eutectic mixture is never
crystallizing in presence of a liquid.

Peak n2
This peak appears at about -30C but it is not always present. This one corresponds to the
melting of the stable eutectic mixture NaOH,7H
2
O+2H
2
O studied previously (part III.1.2.2). It
seems, but we do not know why, that the presence of urea in a 7.6NaOH/water makes easier the
crystallisation of this stable eutectic.

With the DSC7, it is not possible to obtain only a stable NaOH eutectic mixture. Consequently,
its pure melting enthalpy is unknown. In the next parts, we will use the same enthalpy for both
eutectic melting, i.e. 187J/g. The low intensity of the peak of the stable eutectic in comparison
with the one of metastable eutectic mixture (Fig.III.2-10 for 7.6NaOH/water solutions
containing 14, 18 and 25 grams of urea) will not bring a lot of error.

7.6NaOH/
Z urea/water
Heut NaOH,
J/g
Tm NaOHeut
(onset)
Tm ureaeut
(peak)
Tm free comp
(end)
0 72 -34.2 - -6.9
6 72 -35.9 -25.7 -8.6 free
12 71 -36.0 -20.6 -12.5 ice
14 70 -35.5 -19.2 -12.9
18 71 -35.3 -17.3 -16.3
20 74 -35.5 -17.8 -14.1 free
25 69 -35.6 -18.5 -6.2 urea

Tab.III.2-7: Melting enthalpy of NaOH eutectic and melting temperatures of NaOH eutectic, urea eutectic, free
compounds (ice and urea) for 7.6NaOH / Z urea / water solutions (Z= 0, 6, 12, 14, 18, 20, and 25)

The melting temperature of the NaOH eutectic is taken at the onset of the peak, the one of the
urea eutectic at the maximum of the peak (the onset is impossible to be measured) and the one of
ice at the end of the peak to determine the liquidus (with the same comments concerning the
drawback of using a finite mass of sample and a finite heating rate).

_____________________________________________________________________________________________________
-115-
As the melting enthalpy and the melting temperature of NaOH eutectic mixtures (Tab.III.2-7
column 2 and 3) remain constant with an increase of urea concentration, we can conclude that
urea neither disrupt the formation of NaOH eutectic nor modify its composition, which
is still 1NaOH+9H
2
O. During the time of experiment, urea does not react with NaOH.
Consequently, the species in solution are:
NaOH eutectic mixtures, in metastable and/or stable state.
Urea compounds which remain to be determined
and water

Peaks n3 and n4
The peak n3 appears as soon as urea is added but its location changes depending on urea
concentration. This peak should correspond to the melting of a urea compound whose the nature
has to be determined.
The temperature of the melting peak n4 decreases when urea concentration increases from 0 to
14 g of urea in 100 g of solution. This peak could be attributed to the gradual melting of free ice.

From DSC data obtained we may assume that urea does not react with NaOH molecules and
does not modify the NaOH eutectic. Thus we can expect that in solution in the presence of
NaOH urea and water behave as in the binary system: they form a eutectic mixture containing 1
urea and 8 water molecules. In that case, the peaks n3 and n4 should represent the melting of
urea eutectic mixture and the melting of free ice, respectively.

Plotting the melting temperature of the peaks as a function of urea concentration (Fig.III.2-11)
shows that these two peaks n3 and n4 are joining at 7.6NaOH/18urea/water. If our hypothesis
is correct, this means that at this composition there are the two NaOH and urea eutectic mixtures
coexisting but there is no free water in solution.

Calculation to determine the nature of the urea compound
Let us consider a solution containing 7.6g NaOH + 18g urea + 74.4g water. For 100g of mixture,
we have:
M
H2O
=18g/mol; 74.4 g H
2
O 4.13moles of water in the whole solution
M
NaOH
=40g/mol; 7.6 g NaOH 0.19 moles of NaOH which corresponds to 0.19 moles of
NaOH eutectic mixture.
M
urea
=60g/mol; 18 g urea 0.3 moles of urea in the solution

The NaOH eutectic mixture is not modified by the addition of urea and it contains 1NaOH and
9H
2
O molecules
0.19*9 = 1.71 moles of water molecules are trapped in the NaOH eutectic mixture.
It remains 4.13-1.71 = 2.42 moles of free water.

_____________________________________________________________________________________________________
-116-
(i) The urea eutectic mixture is composed of 1urea and 8H
2
O molecules
0.3*8 = 2.4 moles of water are trapped in the urea eutectic
It only remains 0.02 moles of free water in the whole solution.
(ii) After the formation of NaOH eutectic mixture, 100g of solution contains:
18 g of urea
2.42 mol*18 = 43.56g of free water
Consequently, the local concentration of urea is [18/ (18+43.56])*100 = 29.2g which corresponds
to the composition of the urea eutectic mixture.

These calculations allow concluding that:
Both NaOH and urea eutectic mixtures are formed independently and thus are
competing for water
At 7.6NaOH/18urea/water there is no free water, only NaOH and urea eutectic
mixtures.

Peak n5
For a 7.6NaOH/20urea/water solution, we can observe a small peak at the end of the eutectic
peak. For a 7.6NaOH/25urea/water solution, we can guess that there is a peak at about -6C.
These two phenomena are reproducible. Since this peak cannot be due to NaOH, it is due to the
melting of strongly depressed urea. This is the clear sign that upon competing with water, NaOH
is winning over urea.

Conclusion on the 7.6NaOH/urea/water ternary phase diagram
From the melting temperatures reported in the Tab.III.2-7, we can draw a part of the ternary
phase diagram of the 7.6NaOH/urea/water system (Fig.III.2-11).

The figure shows the crystalline phases present in the system at a given temperature and a given
urea concentration, with 7.6g of NaOH in 100g of solution. It is not possible to calculate the
proportion of each phase from such a phase diagram, as we saw in the theoretical part of this
chapter (part III.1.1.2).

_____________________________________________________________________________________________________
-117-
-60
-50
-40
-30
-20
-10
0
0 5 10 15 20 25
mass of urea in 100g of solution
TC

Fig.III.2-11: Ternary phase diagram for 7.6NaOH / (0-25)urea / water solution
full triangle: melting of NaOH eutectic mixture; open triangle: melting of the urea eutectic mixture;
full square: melting of free ice; open square: melting of free urea

We saw that NaOH and urea are forming both their normal eutectic mixture with water
NaOH,5H
2
O+4H
2
O and urea+8H
2
O respectively. This means that they are not interacting but
they are competing for water. In 100 g of solution:
at m
urea
< 18 g, there is enough water for both compounds to feel independent.
at m
urea
= 18 g, there is no free water in solution, all water molecules are linked either to
the NaOH eutectic mixture or to the urea eutectic mixture.
at m
urea
> 18 g, there is not enough water for both eutectics. Because of a strong polarity
of NaOH (pairs of ions), the water molecules surrounding the Na
+
and OH
-
ions in
solution are more strongly linked to NaOH molecules than water molecules surrounding
urea molecules. Consequently, the NaOH eutectic mixture is thermodynamically easier to
form than the urea eutectic mixture. This explains that the melting peak of NaOH
eutectic mixture appears for all the urea concentrations studied and that, at urea
concentrations higher than 18 g in 100 g of solution, pure urea is present in the system.
NaOH, urea and water behave as in their binary situations.

All is liquid
NaOH + urea + water
Ice + liquid
Crystalline urea + liquid
Everything crystallised
Liquid +
+ Crystalline urea: in urea eutectic
+ Ice: in urea eutectic and free ice
+ Crystalline urea: in urea eutectic and free crystals
+ Ice: in urea eutectic
+ Crystalline urea: in urea eutectic
+ Ice: in NaOH and urea eutectics and free ice
+ Crystalline urea: in urea eutectic and free crystals
+ Ice: in NaOH and urea eutectics
_____________________________________________________________________________________________________
-118-
III.2.3- NaOH/ZnO/water ternary system
The goal of this section is to understand how sodium hydroxide and zinc oxide are reacting when
they are put together in water.

From literature [LIU1998], we know that ZnO can be dissolved in acid or strongly basic aqueous
solutions (Fig.III.2-12). Moreover, dissolving ZnO, in solid state (s), induces some reactions with
OH
-
to form zinc hydroxyl ions; the major species at high alkalinity are Zn(OH)
3
-
and Zn(OH)
4
2-

(Fig.III.2-13) [DIR1954] [REI1975] [BEN1999], according to the following equilibrium
equations:
Zn(OH)
3
-
ZnO
(s)
+ H
2
O + OH
-

Zn(OH)
4
2-
ZnO
(s)
+ H
2
O + 2OH
-

Fig.III.2-12 [LIU1998]: Influence of pH on solubility of tin and zinc

Fig.III.2-13 [REI1975]: Graph of fraction f of Zn(II) existing as
Zn
2+
, Zn(OH)
-
, Zn(OH)2, Zn(OH)3
-
, and Zn(OH)4
2-

over a range of pH at 25C.

As soon as ZnO is added to a NaOH/water solution, the above reactions occur to totally
dissolve ZnO crystals. Consequently, the structure of the alkaline solutions should be modified
_____________________________________________________________________________________________________
-119-
by the presence of ZnO. In the forthcoming sections, we will compare the two aqueous solvents
NaOH/water and NaOH/ZnO/water by viscosimetry, NMR, and DSC.
The mixtures that we used contained up to 20 g of NaOH (without ZnO, this corresponds to the
eutectic mixture) and up to 1.84 g of ZnO in 100 g of solution.

III.2.3.1. Viscosity of dilute NaOH/ZnO/water solutions
We measured the viscosity of the NaOH/water and NaOH/ZnO/water solutions during their
dilution with water with an Ubbelohde capillary viscometer (see description in II.3.3.2). The
initial compositions were 12 g of NaOH and 1.1 g of ZnO in 100 g of solution. As water is the
solvent, ZnO concentration decreases upon dilution in the same way as NaOH concentration
and the ratio ZnO/NaOH is constant. That is why, on Fig.III.2-14, the relative viscosity is
plotted versus NaOH concentration.
1
1,2
1,4
1,6
1,8
2
0 5 10 15
NaOH concentration (g/cm3)
r
e
l
a
t
i
v
e

v
i
s
c
o
s
i
t
y
(a)
1
1,2
1,4
1,6
1,8
2
0 5 10 15
r
e
l
a
t
i
v
e

v
i
s
c
o
s
i
t
y
(b)
1
1,2
1,4
1,6
1,8
2
0 5 10 15
r
e
l
a
t
i
v
e

v
i
s
c
o
s
i
t
y
(c)
1
1,2
1,4
1,6
1,8
2
0 5 10 15
r
e
l
a
t
i
v
e

v
i
s
c
o
s
i
t
y
(d)
1
1,2
1,4
1,6
1,8
2
0 5 10 15
r
e
l
a
t
i
v
e

v
i
s
c
o
s
i
t
y
(e)
1
1,2
1,4
1,6
1,8
2
0 5 10 15
r
e
l
a
t
i
v
e

v
i
s
c
o
s
i
t
y
(f)

Fig.III.2-14: Relative viscosity versus NaOH concentration, comparison of
NaOH/water solution ( ) and NaOH/ZnO/water solution (----)
(a) 12C; (b) 15C; (c) 20C; (d) 25C; (e) 30C and (f) 40C.
Initial compositions: 12NaOH/water and 12NaOH/1.1ZnO/water where NaOH/ZnO=cst for all measurements.

Whatever the temperature is, from +12C to +40C, the relative viscosities of the mixtures
versus NaOH concentration curves are not linear. This means that the solvents are not simple
solutions but that some intermolecular associations and interactions take place, in agreement with
the fact that there is a complex solvation of the NaOH ions [BAR1954] [KUN1985] [YAM1988].
We can note that the relative viscosity of NaOH/water is the same as NaOH/ZnO/water,
within the experimental errors. We can conclude that the formation of zinc hydroxyls, during
_____________________________________________________________________________________________________
-120-
ZnO dissolution in NaOH/water, does not significantly influence the main non linear effect of
viscosity due to the interactions between NaOH and water.

III.2.3.2. NMR study
This work was performed in the group of Dr Hans-Peter Fink in the Fraunhofer Institute in
Glm, Germany. The full report is in Appendix 5.
1
H and
23
Na nuclei were analysed by NMR in NaOH/water with compositions of (6-
10)NaOH/water, with or without the addition of (0.2-1.5)g of ZnO. Temperature was varied
between -20C and 20C. Two parameters were studied, the chemical shift and the spin lattice
time T1 which expresses the mobility of the nuclei, influenced by the environment.
The purpose of the work was to investigate the hydration behaviour of the Na
+
ions influenced
or not by temperature variation and the presence of ZnO.

23
Na
The relaxation time T1 of
23
Na decreases with temperature due to the variation of viscosity and
with NaOH concentration (also due to the viscosity increase). Addition of ZnO is slightly
decreasing the relaxation time, this effect being more pronounced with increasing ZnO
concentration and solution temperature (Fig.III.2-15 and Appendix 5).
The chemical shift for
23
Na is increasing with decreasing temperature or increasing NaOH, also
due to viscosity changes. The chemical shift is not affected by the presence of ZnO.

6 7 8 9 10
5
10
15
20
25
30
35
40
45
20 C
10 C
0 C
-10 C
-20 C
NaOH concentration [%]
2
3
N
a

r
e
l
a
x
a
t
i
o
n

t
i
m
e

T
1

[
m
s
]

0,0 0,4 0,8 1,2 1,6
5
10
15
20
25
30
35
40
45
2
3
N
a

r
e
l
a
x
a
t
i
o
n

t
i
m
e

T
1

[
m
s
]
ZnO concentration [%]
20 C
10 C
0 C
-10 C
-20 C

for the lyes for 8% lye with ZnO
Fig.III.2-15: Results of
23
Na relaxation times

1
H
The experiments with
1
H show that the relaxation time T1 decreases with increasing NaOH
concentration, which may reflect either a change of viscosity/mobility or a difference in the
free/bound water mole fraction. There is no influence of the addition of ZnO.

_____________________________________________________________________________________________________
-121-
Conclusion on the NMR analysis
The conclusion of the authors is that there is no change in the hydration of Na
+
ions in the
temperature/concentration range explored since any change should give non linear characteristic
of the relaxation times or the chemical shifts as a function of NaOH concentration. Only linear
regimes were found.
As far as ZnO is concerned, the experiments do not show any influence of the addition of this
compound. There is no major change like the formation of a complex between ZnO and NaOH.
This is in agreement with the viscosity measurements reported above.

III.2.3.3. DSC results
As ZnO reacts with OH
-
to form

zinc hydroxyl ions and as ZnO is not soluble in water, the
number of Na
+
/OH
-
ion pairs should decrease. Consequently, the formation of the NaOH
eutectic mixture should be disrupted by the addition of ZnO and DSC thermograms should
reveal a smaller melting peak of NaOH eutectic than the one of NaOH solutions without ZnO.
Fig.III.2-16 shows DSC thermograms for 20NaOH/(0-1.84)ZnO/water solutions. Without
ZnO, this NaOH concentration corresponds to the eutectic mixture and there is no free water
crystallising. This figure shows a constant decrease of melting temperature (measured at the onset
of the peak) and of melting enthalpy of the NaOH eutectic peak when ZnO concentration
increases (Tab.III.2-8), and no melting of ice.

DSC melting thermogram for
20% NaOH + ZnO aqueous solutions
0
0,4
0,8
1,2
1,6
-50 -40 -30 -20 -10 0
TC
H
e
a
t

f
l
o
w

(
W
/
g
)
+1.84 ZnO
+1.1 ZnO
+0.7 ZnO
without ZnO
Melting temperature
-36
-35
-34
-33
0 0,5 1 1,5 2
ZnO
T
C
Melting enthalpy
150
160
170
180
190
0 0,5 1 1,5 2
ZnO

H
,

J
/
g

Fig.III.2-16: DSC melting thermograms for 20NaOH /x ZnO/water solutions (x= 0, 0.7, 1.1, and 1.84).
This melting peak corresponds to the melting of the metastable NaOH eutectic mixture.
The changes of this NaOH eutectic peak (melting temperature and melting enthalpy) versus ZnO concentration are
represented on inset graphs.

_____________________________________________________________________________________________________
-122-
Peak of the NaOH eutectic mixture
Melting temperature
If we look at the melting temperature of the eutectic peak (Fig.III.2-16 and Fig.III.2-8), we can
note that the decrease is constant when ZnO concentration increases. This means that the
NaOH eutectic mixture is crystallising in the presence of liquid which contains at least ions
involving Zn. According to the equilibrium reactions of ZnO dissolution in alkali solutions and
Fig.III.2-13, these ions could be Na
+
/Zn(OH)
3
-
and 2Na
+
/Zn(OH)
4
2-
.

Melting enthalpy
The constant decrease of melting enthalpy (Fig.III.2-16 and Tab.III.2-8) confirms that some
NaOH molecules react with ZnO. According to the literature [DIR1954] [REI1975] [BEN1999],
at pH>12, the major species in solution are Zn(OH)
3
-
,Na
+
and Zn(OH)
4
2-
,2Na
+
(Fig.III.2-13),
thus one or two NaOH should be linked to ZnO.

The number of NaOH having participated to the ZnO dissolution can be determined from the
decrease of the peak since these NaOH molecules cannot participate to the crystallisation of the
eutectic mixture NaOH,5H
2
O+4H
2
O:

pure eut
ZnO eut
eut eut in not NaOH
H
m H
f f
,
. exp ,
.
) (
1 1
= =

ZnO NaOH
ZnO NaOH
eut in not NaOH
m M
M m
f ZnO NaOH
=

/ Eq. 9
where m
NaOH
=20g, m
ZnO
is the mass of ZnO in solution, M
NaOH
=40g/mol and M
ZnO
=81.4g/mol.

The results are reported in the Tab.III.2-8.

xZnO in
20NaOH/water
Tmeut
(C)
Heut (J/g)
NaOH/ZnO
Eq.9
H2O/ZnO
Eq.11
0 -33.3 187=Heut pure
0.7 -34.3 178 2.8 14
1.1 -34.5 171 3.2 20
1.84 -35.2 157 3.5 25
Tab.III.2-8: Melting temperature and enthalpy of the NaOH eutectic peak (columns 2 & 3)
as a function of ZnO concentration in 20NaOH/xZnO/H2O solutions (x= 0, 0.7, 1.1, and 1.84),
columns 4 & 5 give the calculated ratios NaOH/ZnO and H2O/ZnO necessary to dissolve ZnO.

We see that about 3NaOH are linked to each ZnO whereas the literature is suggesting a value of
one and two.
From these values, an extrapolation of the eutectic melting enthalpy to zero gives a limit of ZnO
dissolution in 20NaOH/water at about 12g ZnO in 100g of solution. This concentration
corresponds to a calculated molar ratio (eq.9) NaOH/ZnO=3.4. This result does not correspond
to our solubility experiments since we measured that ZnO can be dissolved in NaOH up to a
_____________________________________________________________________________________________________
-123-
maximal molar ratio NaOH/ZnO=10 (i.e. about 4g of ZnO in 100g of solution containing 20g
of NaOH).

Peak of free ice
We know that the 20NaOH/water solution without ZnO is only composed of the NaOH
eutectic and that there is no free water. In that case, the DSC melting thermogram has just one
melting peak at -33.4C.
Dissolving ZnO in highly basic solution induces reaction with NaOH. So water molecules
initially surrounding NaOH should be released and thus should crystallise as free water. This is
not the case since no melting peak of free ice appears on the DSC thermograms of solutions
containing ZnO (Fig.III.2-16). All the water molecules present in solution are linked to other
species that prevented them to freeze as ice.

We also studied mixtures with a lower NaOH content (12NaOH/water), where the melting peak
of free ice was observed when ZnO is added. For the 12NaOH/ZnO/water solutions, the
NaOH eutectic could not crystallise most probably because the lowest temperature of the DSC
oven was not low enough to induce crystallisation. Consequently, only the melting peak of free
ice appears (Fig.III.2-17) and we can note that its melting temperature and enthalpy remain
constant at all ZnO concentrations. This means that the quantity of free water is always the same,
even after ZnO addition.

0
0,2
0,4
0,6
0,8
1
1,2
-50 -40 -30 -20 -10 0
TC
H
e
a
t

f
l
o
w

(
W
/
g
)
without ZnO
+ 0.22 ZnO
+ 0.64 ZnO
+ 1.1 ZnO
H (J/g)
124
120
121
121
Tm = -33.5C
H = 108 J/g
Tm (C)
-11.3
-11
-11
-10.8

Fig.III.2-17: DSC melting thermograms of 12NaOH/xZnO/water solutions (x=0, 0.22, 0.64, 1.1)
with melting enthalpy and temperature of free ice.
The melting peak of NaOH eutectic does not appear because the temperature of the cooler was not low enough.

We will calculate in the following how many water moles are surrounding the ions involving
ZnO. This calculation will be performed for the 20NaOH/ZnO/water solution since it was the
only one showing a decrease of the NaOH eutectic peak, thus allowing calculating the number of
NaOH moles involved in a reaction with ZnO.
_____________________________________________________________________________________________________
-124-
The total number of moles of water n
tot.w
in solution is:

18
80 100
2
.
ZnO
O H
ZnO NaOH
w tot
m
M
m m
n

=
|
|
\
|
=
The number of moles of water linked to NaOH to form the eutectic mixture n
eut.w
is:
5 . 4
) (
9
,
. exp ,
.

=
|
|
\
|
=
pure eut
ZnO eut
NaOH
NaOH
eut w eut
H
m H
M
m
f n
where m
NaOH
=20g, m
ZnO
is the mass of ZnO in solution, M
H2O
=18g/mol, M
NaOH
=40g/mol and
M
ZnO
=81.4g/mol.

According to the dissolution reactions, whatever the zinc hydroxyl ion formed is, one water
molecule reacts with one ZnO and thus cannot participate to the eutectic crystallisation:

ZnO
ZnO
ZnO w dissol
M
m
n n = =
.

Thus we can determine how many water molecules n
surr.w
are surrounding the ion pairs. These
water molecules are not free and can not participate to the melting peak of free ice.

w eut w tot w dissol w surr
n n n n
. . . .
= +
5 . 4
) (
18
80
,
. exp ,
.

= +
pure eut
ZnO eut
ZnO
ZnO w surr
H
m H
m
n n Eq. 10
Thus, if Eq.10 is divided by n
ZnO
, the ratio H
2
O/ZnO is obtained and corresponds to the quantity
of water necessary to dissolve ZnO in alkali solutions.

ZnO
ZnO w surr
n
n n
ZnO O H
+
=
.
2
/ Eq. 11
The values are reported in the Tab.III.2-8. We can note that the number of moles of water linked
to ZnO increases with ZnO content.
If we assume that 25 moles of water per ZnO is necessary for the ZnO dissolution in
NaOH/water, it is possible to calculate the dissolution limit since the number of water molecules
surrounding ZnO has to be equalled to the number of water molecules not crystallised in NaOH
eutectic mixture. We obtain the following equation:

w ZnO w eut w tot
n n n
. . .
= Eq. 12
where n
ZnO.w
is the number of water molecules involved in ZnO dissolution
ZnO
ZnO
M
m lim
25 = and
n
eut.w
the number of water molecules crystallised in NaOH eutectic mixture
eut NaOH
n
.
9 =
( )
|
|
\
|
= =
ZnO
ZnO
NaOH
NaOH
ZnO with NaOH tot NaOH
M
m
M
m
n n
lim
4 . 3 9 9
.. . .
, where n
NaOH.with.ZnO
represents the
number of NaOH molecules involved in ZnO dissolution, 3.4 being the ratio NaOH/ZnO
determined from the extrapolation of melting enthalpy.
_____________________________________________________________________________________________________
-125-
Substituting these expressions in the Eq.12 gives:

ZnO
ZnO
ZnO
ZnO
NaOH
NaOH
O H
ZnO NaOH
M
m
M
m
M
m
M
m m lim
25
lim
4 . 3 9
lim 100
2
=
|
|
\
|

O H
NaOH
NaOH
NaOH
O H ZnO
ZnO
M
m
M
m
M M
m
2 2
100
9
1 6 . 5
lim

=
|
|
\
|

Thus, g m
ZnO
2 . 4 lim = in 100g of 20NaOH aqueous solution.

This calculation suggests that ZnO can be dissolved in 20NaOH/water solution up to about
4.2g. This concentration corresponds to a 0.052mol of ZnO in 0.5mol NaOH. The calculated
limit of ZnO dissolution is reached when NaOH/ZnO=10 which corresponds to our
experimental data, with the assumption that 25 moles of water are necessary for dissolving ZnO.

III.2.3.4. Conclusion on the NaOH/ZnO/water system
Intermolecular interactions take place in NaOH/water and NaOH/ZnO/water systems which
are reflected by the non-linear dependence of the relative viscosity upon dilution. It seems that
ZnO does not influence the structure of the NaOH/water complexes, at least seen from
viscometric measurements and NMR analyses.

On the other hand, DSC analyses reveal significant changes in NaOH/water system when ZnO
is added:
During dissolution, ZnO reacts with NaOH to form zinc hydroxides, NaZn(OH)
3
and
Na
2
Zn(OH)
4
according to literature. As a consequence, some NaOH molecules are not
anymore able to form the NaOH eutectic mixture and its melting peak decreases. About
3NaOH/ZnO is involved in the dissolution.
The quantity of water crystallizing as free ice does not change with the addition of ZnO:
the melting enthalpy of free ice remains constant. Some water molecules are involved in
the dissolution reactions 1H
2
O/ZnO and some other are surrounding the ions in
solution. The total ratio H
2
O/ZnO increases up to 25, when ZnO concentration
increases from 0 to 1.8g in 100g of solution.

From these results we also determined that ZnO needs not only Na
+
/OH
-
ions but also a large
quantity of water to be dissolved. With the assumption that 25 water molecules are needed for
dissolving one molecule of ZnO, the limit of ZnO dissolution is reached for a NaOH/ZnO
molar ratio equals to 10, value in agreement with our solubility experiment.

Finally, the NaOH eutectic mixture is crystallising in presence of a liquid containing zinc
hydroxides, some sodium hydroxide and water molecules. That is why the NaOH eutectic
melting temperature decreases when ZnO content increases. This may also explain why, in some
cases, the NaOH eutectic peak does not appear when we analyse solutions containing ZnO.
_____________________________________________________________________________________________________
-126-
III.2.4- Conclusions on the structure of cellulose solvents: NaOH/water,
NaOH/urea/water and NaOH/ZnO/water
The NaOH/water binary phase diagram obtained from DSC experiments is in agreement with
the ones reported in literature [PIC1893] [COH1960] [ROL1964].

The NaOH/water and urea/water binary systems have both typical eutectic behaviours:
o Below the eutectic composition (NaOH content lower than 20g and urea content lower
than 29.4g, in 100g of solution), free ice is crystallising and its melting temperature
decreases with the increase of NaOH or urea concentrations.
o The eutectic mixture is melting at a constant temperature for all concentrations. For
sodium hydroxide aqueous solutions, the eutectic mixture is composed of NaOH,5H
2
O
and 4H
2
O and melts at -33.4C. For urea aqueous solutions, the eutectic compound is a
mixture of 1urea and 8H
2
O, it melts at -11.4C.
o At concentrations higher than the eutectic composition, the components that are
crystallising are NaOH,5H
2
O and urea for sodium hydroxide aqueous solutions (from
20g NaOH to 31g NaOH in 100g of solution) and for urea aqueous solution (higher than
29.4g of urea in 100g of solution) respectively. Their melting temperature increases with
the increase of NaOH and urea concentrations.

The 7.6NaOH/urea/water ternary system has a particular behaviour. Actually, the two NaOH
and urea eutectic mixtures are forming independently and are competing for water molecules.
Because of the strong polarity of sodium hydroxide, NaOH is attracting water more than urea.

In the NaOH/ZnO/water ternary system, some NaOH and water molecules react with ZnO to
dissolve it and to form zinc hydroxides. The NaOH molecules not involved in ZnO dissolution
are crystallising in eutectic mixture. Its melting temperature decreases with the increase of ZnO
content because the NaOH eutectic is crystallising in presence of liquid containing zinc
hydroxides and its surrounding molecules (water and probably NaOH).
We determined that during dissolution ZnO is trapping about 3NaOH and 25 water molecules.

_____________________________________________________________________________________________________
-127-
III.3- STRUCTURE OF CELLULOSE SOLUTIONS IN NAOH/WATER,
NAOH/UREA/WATER AND NAOH/ZNO/WATER
III.3.1- Cellulose/NaOH/water ternary solutions
III.3.1.1. DSC results on Avicel/NaOH/water
The goal of this part is twofold: (i) to determine the concentration limit of cellulose that can be
dissolved in sodium hydroxide and (ii) to propose a mechanism of cellulose dissolution by
determining the amount of NaOH and H
2
O linked to cellulose. To do this, we will use
microcrystalline cellulose, pure and of low DP.

In part III.2.1-, we studied the NaOH/water solvent from DSC analyses. We described the
methodology used to determine the fraction of each compound present in the system from its
melting enthalpy and we plotted the phase diagram. We will base our research on the same
technique and approach.

Most of the studies were performed on cellulose/7.6NaOH/water systems. Cellulose content
was varied from 0 to 7.6g in 100g of solution. In order to check the influence of NaOH
concentration, cellulose/8NaOH/water solutions were also tested. In all cases studied the
condition of being inside the domain of the phase diagram [SOB1939] where cellulose can be
dissolved in NaOH/water was fulfilled.

An example of DSC melting thermograms for solutions of different cellulose concentrations is
given in Fig.III.3-1. Melting temperature and enthalpy of each peak for the curves shown and for
some other cellulose concentrations are presented in Tab.III.3-1. The analysis of the peak at
lower and higher temperatures is discussed below.

0
0,4
0,8
1,2
1,6
-40 -30 -20 -10 0
TC
Heat Flow (W/g)
7.6
0
0.5
4
5
2

Fig.III.3-1: DSC melting thermograms of Avicel cellulose/7.6NaOH/water solutions.
Cellulose content is indicated on the graph. 0 corresponds to a 7.6NaOH/water solution without cellulose.
_____________________________________________________________________________________________________
-128-
eutectic peak ice peak Avicel
X g in a 100g solution Tm (onset), C Heut, J/g Tm (end), C Hice, J/g
0 -32.9 72 -5.1 195
0.5 -34.9 50 -4.8 205
1 -33.7 43 -4.1 200
2 -33.5 41 -4.3 205
3 -33.2 30 -3.9 210
4 -33.4 29 -3.7 209
5 -33.5 22 -3.9 217
7.6 No peak - -3.8 178
Tab.III.3-1: Melting temperatures and enthalpies of the peaks at lower and higher temperatures
for Avicel/7.6NaOH/water solutions.

eutectic peak ice peak Avicel
X g in a 100g solution Tm (onset), C Heut, J/g Tm (end), C Hice, J/g
0 -33.2 74 -5.9 187
1 -34.6 52 -5.9 180
3 -33.9 34 -5.2 175
5 -33.8 25 -4.7 178
for Avicel/8NaOH/water solutions.

III.3.1.2. Peak at low temperature. Limit of cellulose dissolution
Whatever are the concentrations of Avicel cellulose and of the solvent, the melting temperature
of the peak at lower temperature is constant (Tab.III.3-1 and Tab.III.3-2), within the
experimental errors. It practically coincides with the melting temperature of the eutectic mixture
in pure aqueous sodium hydroxide solution of 7.6NaOH/water and 8NaOH/water (without
cellulose) but with a systematic shift of about 1C towards lower temperatures. This means two
things. First, the same eutectic mixture NaOH,5H
2
O + 4H
2
O is present in cellulose/sodium
hydroxide aqueous solutions in this region of temperatures and cellulose concentrations, in (7.6-
8)NaOH/water, owing to the fact that the temperature shift is very small. The presence of
cellulose is not perturbing the composition of solvent system. Second, because of the addition of
cellulose, the molecular environment of the NaOH eutectic mixture has changed, even slightly.
Consequently, crystallization and melting of the eutectic mixture are disrupted which induces a
slight shift in the melting temperature. A similar behaviour was observed after the addition of
urea in NaOH/water solution (see part III.2.2.2). Moreover, we will see in the next paragraph
that this change due to the presence of cellulose is also visible in the melting peak of pure water.

The results obtained mean that we are not here in the presence of a ternary phase diagram where
each of the components makes eutectic mixtures two by two. Cellulose does not form either
covalent bond or eutectic mixture with water and NaOH. Nevertheless, the presence of cellulose
_____________________________________________________________________________________________________
-129-
is drastically changing the amount of the NaOH/water eutectic mixture. The influence of
cellulose concentration on the eutectic peak can be seen from the variation of enthalpy values:
the enthalpy decreases down to zero with the increase of cellulose concentration (see Fig.III.3-1
curve in bold corresponding to m
cell
=7.6g does not show any eutectic peak at -33.4C). This
means that cellulose interacts with sodium hydroxide. The higher is the cellulose concentration,
the smaller is the amount of eutectic mixture that can crystallise and then melt at -33.4C. Since
NaOH is present only in the eutectic mixture, the decrease of its melting enthalpy allows the
calculation of the number of NaOH molecules linked to cellulose and thus not able to participate
to NaOH,5H
2
O crystal fraction of the eutectic mixture.

The eutectic melting enthalpies (Tab.III.3-1 and Tab.III.3-2) enable (i) to determine the limit of
cellulose dissolution and (ii) to calculate how many NaOH are linked to the cellulose, for all the
cellulose concentrations.

Limit of cellulose dissolution
Fig.III.3-2 represents the dependence of eutectic melting enthalpy on cellulose concentration, for
cellulose/7.6NaOH/water solutions.
The fact that H
eut
is reaching zero at a certain Avicel cellulose concentration means that all
sodium hydroxide molecules have been trapped by the cellulose chains. This corresponds to the
dissolution limit since there is no more NaOH molecules able to solvate any additional cellulose
chain that could be brought in the mixture. At this dissolution limit, we can calculate the
proportion between cellulose anhydroglucose unit (AGU) and NaOH molecules. In this case, in
100g of cellulose/NaOH/water solution, there are 7.6g of cellulose and 7.6g of NaOH when the
eutectic peak disappears. As the molar mass of NaOH is 40 while the molar mass of an
anhydroglucose unit is 162, it gives that four NaOH molecules are associated to each AGU.

0
10
20
30
40
50
60
70
80
0 2 4 6 8
m(cell), g
Heut, J/g

Fig.III.3-2: Eutectic melting enthalpy (J/g) versus cellulose concentration
for the Avicel cellulose/7.6NaOH/water system.
The dashed line is given to guide the eyes.
_____________________________________________________________________________________________________
-130-
The other data obtained for Avicel/8NaOH/water solutions (Tab.III.3-2) as well as the ones
reported in the ref. [ROY2001], where Avicel cellulose was dissolved in 9%NaOH/water, give
very similar results. In order to compare the results for all these systems and to avoid the
influence of NaOH concentration, we plotted reduced melting enthalpy versus reduced cellulose
concentration (Fig.III.3-3). The reduced eutectic enthalpy corresponds to the enthalpy of the
eutectic peak at X% of cellulose, H
eut,X%cell
, divided by the one at m
cell
=0, H
eut,0cell
. The reduced
cellulose concentration corresponds to the cellulose weight divided by the one of NaOH.

0,0
0,2
0,4
0,6
0,8
1,0
1,2
0 0,5 1
m(cell)/m(NaOH)
reduced enthalpy
0,0
0,2
0,4
0,6
0,8
1,0
1,2
0 0,5 1
m(cell)/m(NaOH)
reduced enthalpy

Fig.III.3-3: Reduced melting enthalpy versus reduced cellulose concentration mcell/mNaOH
for Avicel/7.6NaOH/water ( ) and for Avicel/8NaOH/water ( ).
Inset: the results for Avicel in 9%NaOH/water from [ROY2001] were added ( )
The lines are given to guide the eyes

For the three systems studied, the decrease of the reduced melting enthalpy is reaching zero at
m
cell
/m
NaOH
=1.
This means that the maximum Avicel cellulose concentration for having a proper dissolution
is equal to the sodium hydroxide concentration in weight per cent. At this concentration, all
the NaOH molecules are associated with the cellulose. Consequently, the dissolution of
further added cellulose should not be possible. Above this concentration, cellulose can not
fully dissolve and could form cellulose aggregates.

Number of NaOH and H
2
O molecules linked to cellulose

Tab.III.3-1 and Tab.III.3-2 allow the calculation of the proportion between AGU and NaOH for
the other cellulose concentrations. The fraction of eutectic linked f
eut.linked
to cellulose is as follows:
_____________________________________________________________________________________________________
-131-

cell eut
cell X eut cell eut
linked eut
H
H H
f
% 0 ,
% , % 0 ,
Eq. 13

Thus the amount of eutectic (or NaOH) linked to cellulose, in moles, is given by

cell
AGU
NaOH
NaOH
linked eut NaOH
m
M
M
m
f AGU n /
,
= Eq. 14

The amount of NaOH molecules linked to one anhydroglucose unit as a function of reduced
cellulose concentration m
cell
/m
NaOH
for the system Avicel/7.6NaOH/water and for
Avicel/8NaOH/water is shown in Fig.III.3-4. At m
cell
/m
NaOH
> 0.4 the proportion between
AGU and NaOH molecules is constant and equal to the limit of cellulose dissolution which is
four NaOH per AGU. At low cellulose concentrations, m
cell
/m
NaOH
< 0.25, up to 20 NaOH
molecules seem to be linked to one anhydroglucose unit. The change of the slope in Fig.III.3-4 as
well as in Fig.III.3-3 may be caused by the transition of cellulose solution from the dilute to semi-
dilute state. In ref. [ROY2003], the overlap concentration for Avicel in 9%NaOH/water was
reported to be C
cell
1%. Fig.III.3-4 shows that the slope changes at C
cell
1.5-2% (here
C
cell
=m
cell
). It may be possible that due to the sterical hindrance caused by the transition from
dilute to the semi-dilute state, the number of NaOH molecules linked to cellulose strongly
decreases.

0
5
10
15
20
25
0 0,5 1
m(cell)/m(NaOH)
nNaOH / AGU

0
20
40
60
0 0.5 1
m
cell
/m
NaOH
n
NaOH
/ AGU

Fig.III.3-4: Number of NaOH molecules linked to one anhydroglucose unit (AGU), calculated from melting enthalpy data,
versus cellulose/NaOH weight ratio for cellulose/7.6NaOH/water ( ) and cellulose/8NaOH/water ( )
Inset: the same data (squares) and the ones from ref. [KUO2005] for NaOH concentrations of 4.5, 6.2 and 7.9%.
The dashed line is given to guide the eyes.

The same proportion between NaOH and anhydroglucose unit corresponding to the limit of
cellulose dissolution, four NaOH per AGU, can be obtained from the study of Avicel solubility
in NaOH aqueous solutions of different concentrations [KUO2005]. Using turbidimetry, authors
_____________________________________________________________________________________________________
-132-
measure the amount of undissolved cellulose as a function of cellulose and NaOH concentration.
We recalculated their data as the weight of NaOH per one gram of dissolved cellulose, converted
the obtained figures into mole ratios and plotted together with our data as a function of
m
cell
/m
NaOH
. The result is presented in the inset of Fig.III.3-4 : at higher cellulose concentrations,
the results of ref. [KUO2005] coincide with the ours and show that at least 4 NaOH molecules
are needed to dissolve one AGU. At low cellulose concentrations, m
cell
/m
NaOH
< 0.25, the
amount of NaOH linked to one anhydroglucose unit calculated from ref. [KUO2005] is twice
higher than obtained by us using DCS technique. In general, it is not clear how and why so many
NaOH molecules can be bound to one anhydroglucose unit. However, this is not an artefact
since the proportion (20-50)NaOH per one AGU at cellulose concentrations around 1-2% was
obtained independently by different methods.

III.3.1.3. Peak at high temperature
Fig.III.3-1, Tab.III.3-1 and Tab.III.3-2 show that the temperature of the melting peak at higher
temperatures is slightly increasing when increasing the cellulose concentration. Again, this change
is small, as in the case of the eutectic peak. This increase of the melting temperature of ice with
increasing cellulose concentration reflects the fact that the molecular environment of the
crystallizing water is changed when cellulose is trapping NaOH.

The proportion between water, NaOH and cellulose molecules can be calculated as follows. In
cellulose/NaOH/water solutions, H
2
O is present in three forms or compounds:
as a free water frozen to ice below the liquidus curve,
in the eutectic mixture that is not linked to cellulose and
in the mixture linked to cellulose through NaOH.
As far as the amount of free water is not changing with adding cellulose (see H
ice
in Tab.III.3-1
and Tab.III3.-2), it means that the number of water molecules linked to NaOH that are bound to
cellulose is the same as in the eutectic mixture. The number of water molecules linked to one
anhydroglucose unit is then straightforward to calculate: it is a product of n
NaOH
x 9; n
NaOH

calculated according to Eq.14 and nine being the amount of water molecules in the eutectic
mixture.

III.3.1.4. Influence of cellulose origin on the structure of cellulose/NaOH/water solutions
This part is a preliminary quantitative study on the influence of cellulose origin on the structure
of cellulose/NaOH/water solutions. The nature of cellulose pulp highly influences the cellulose
dissolution and the stability of solutions. Using DSC analysis with the methodology described
previously, it should be possible to determine the limit of cellulose dissolution and thus how
many NaOH and water molecules are linked to the cellulose chain for other cellulose than Avicel.

The cellulose samples used in this part are:
steam-exploded Borregaard cellulose with a DP of 407 (B407)
_____________________________________________________________________________________________________
-133-
irradiated Lenzing cellulose with a DP of 307 (L307), before the irradiation treatment it is
called Solucell 1175 (part II.1.1)
regenerated Lenzing cellulose II with a DP of 198 (L198
II
), this cellulose also comes from
the Solucell 1175.

Fig.III.3-5 represents the DSC melting thermograms of 5cellulose/7.6NaOH/water solutions,
cellulose being L198
II
, L307 and B407. The results were compared with the one of Avicel
cellulose (DP170), described above. The values of melting temperatures and enthalpies are listed
in Tab.III.3-3.

0
0,2
0,4
0,6
0,8
1
-40 -30 -20 -10 0
TC
Heat flow (W/g)
Avicel
B407
L307
L198II

Fig.III.3-5: DSC melting thermograms of 5cellulose/7.6NaOH/water solutions.
Cellulose origins are indicated on the graph.

eutectic peak ice peak
5cellulose/7.6NaOH/water
Tm (onset), C Heut, J/g Tm (end), C Hice, J/g
Avicel -33.5 22 -3.9 217
L198II -33.7 / -29.2 33 -4.7 187
L307 -34.1 24 -5.3 226
B407 -33.8 30 -5.6 189
for cellulose of various origins in 5cellulose/7.6NaOH/water system.

First of all, we can note that a peak appears at -29.2C, especially for the L198
II
cellulose solution.
This peak corresponds to the melting of the stable NaOH,7H
2
O+2H
2
O eutectic mixture (see
part III.1.2.2 above). We assume that the melting enthalpies of the stable and the metastable
eutectic mixtures are relatively close which enables to determine the total melting enthalpy, H
eut
,
of eutectic mixtures, stable and/or metastable.
Each DSC melting thermogram reveals a peak at low temperature, -33C/-34C, corresponding
to the melting of the metastable NaOH,5H
2
O+4H
2
O eutectic mixture. The temperature is
approximately the same for each cellulose pulp. On the other hand, its enthalpy is changing from
one cellulose origin to another. According to the enthalpy values, it seems that cellulose pulps can
_____________________________________________________________________________________________________
-134-
be divided into two groups: Avicel/L307 and L198
II
/B407. At the same cellulose concentration,
the former is trapping more sodium hydroxide (lower enthalpy) than the latter. It clearly means
that L198
II
and B407 are not dissolved well.

Concerning the peak at higher temperature, we can see that the melting temperature of ice
decreases when the cellulose DP increases. Let us note that for pure 7.6NaOH/water solution,
without cellulose, the ice melting temperature is -5.1C. This decrease is again the sign that
cellulose of higher DP are not fully dissolving (for Avicel, the higher the cellulose concentration,
the lower the melting temperature of the ice).
Moreover, regarding the values of enthalpies, the two previous cellulose groups are appearing
again.

It would have been very interesting to perform the same analyses as previously for
Avicel/NaOH/water solutions on cellulose of different origins, treatments and DP. We could
determine:
If some cellulose is more or less easily accessible to fix NaOH and water.
If the amount of linked and free NaOH and water molecules is a significant parameter
influencing the dissolution, solution properties and gelation.

DSC experiments and careful analyses of experimental data allowed understanding the
thermodynamic behaviour and structure of cellulose/sodium hydroxide aqueous solutions at
temperatures below 0C, in the region of cellulose dissolution.

It was possible to determine the limit of cellulose dissolution in NaOH/water as being at least
four NaOH molecules per one anhydroglucose unit (corresponding to equal cellulose and NaOH
concentrations in weight). Water molecules forming the eutectic mixture in solution are still
around the NaOH linked to the cellulose. If the concentration of cellulose is higher than this
limit, it will not be dissolved. Because cellulose can be dissolved only in a narrow range of sodium
hydroxide concentrations, from 7 to 10%, this means that maximal amount of cellulose that can
be dissolved in NaOH/water solutions is 10%.

This result is compatible with what is found experimentally when trying to prepare
cellulose/NaOH/water solutions for processing. It is a strong limitation if one wishes to use
such a solvent in commercial applications since the amount of polymer is rather low. It has two
bad side effects. One is the cost of processing. The second is the fact that the properties of the
final products are strongly correlated with the amount of polymer material in the solution. In
addition, this dissolution limit in NaOH/water was determined for a low DP cellulose sample.
Increasing the molecular weight is decreasing the ability to dissolve cellulose. As shown in the
Chapter IV, a lot of insoluble parts are present in the solutions with higher DP pulps.

_____________________________________________________________________________________________________
-135-
III.3.2- Cellulose/NaOH/urea/water solutions
III.3.2.1. DSC results on Avicel/NaOH/6urea/water
Let us first recall the main conclusions of the ternary NaOH/urea/water mixtures:
When adding urea to a NaOH/water mixture, both the melting of water and the melting
of the urea/water eutectic that is crystallizing are changed because of the presence of the
NaOH/water liquid mixture that will form the NaOH eutectic.
At a certain urea concentration (18g in 100g of solution), the melting peaks of urea
eutectic and ice are joining. There is no more free ice; there are only NaOH and urea
eutectic mixtures.
NaOH and urea are both forming their normal eutectic mixture meaning that the
components are not interacting.

In this part, we will investigate a set of Avicel cellulose solutions in 7.6NaOH/6urea/water,
cellulose concentrations being from 0 to 5g in 100g of solution. In other words, a 100g solution
contains X g of cellulose (X varying from 0 to 5), 7.6g of NaOH, 6g of urea and the rest being
water. An example of DSC melting thermograms is represented on Fig.III.3-6 and melting
temperatures and enthalpies are reported on Tab.III.3-4. The main features of these DSC melting
curves are similar to the ones reported in literature [CAI2005]

0
0,2
0,4
0,6
-40 -30 -20 -10 0
TC
Heat Flow (W/g)
5
3

Fig.III.3-6: DSC melting thermograms of Avicel cellulose/7.6NaOH/6urea/water solutions.
Cellulose content is indicated on the graph.

The addition of cellulose does not change the shape of the thermograms as compared to the case
without cellulose. The DSC melting curves present three peaks. They were described in part
III.2.2.2:
The peak at the lowest temperature, at about -35C, corresponds to the melting of the
NaOH/water eutectic mixture,
_____________________________________________________________________________________________________
-136-
The next one, at about -25C, corresponds to the melting of the urea/water eutectic mixture,
The peak at the highest temperature, at about -7C, corresponds to the melting of free ice.

As far as the melting peaks are not well separated, the PeakFit software (see details part II.3.1.6)
was used to determine the melting enthalpy of each compound crystallised and melted in the
system.

NaOH eutectic peak urea eutectic peak ice peak Avicel, X g in
100g of solution Tm (onset), C Heut, J/g Tm (peak), C Hurea, J/g Tm (end), C Hice, J/g
0 -35.9 72 -25.7 30 -8.6 160
1 -35.8 41 -24.8 34 -7.2 176
3 -35.7 35 -24.0 29 -6.5 175
5 -35.6 22 -23.7 26 -6.4 172
Tab.III.3-4: Melting temperatures and enthalpies of the peaks for different contents of cellulose in 7.6NaOH/6urea/water

In order to calculate the proportions between the interacting compounds, each peak will be
considered separately.

III.3.2.2. Peak of NaOH eutectic
Taking into account the experimental errors, we can say that whatever the concentration of
cellulose is, the melting temperature of the NaOH/water eutectic mixture is similar to the one
obtained for a pure 7.6NaOH/6urea/water (without cellulose). This means that the addition of
cellulose into an aqueous solution of sodium hydroxide and urea does not disrupt the formation
of the NaOH/water eutectic mixture, as it was in the case of cellulose/NaOH/water solutions.
As previously seen for cellulose/NaOH/water solutions without additives, there is a decrease of
the eutectic melting enthalpy with cellulose concentration: during the dissolution process,
cellulose is trapping some NaOH molecules that can not participate anymore to the
crystallisation of the eutectic mixture. Fig.III.3-7 represents reduced melting enthalpy versus
reduced cellulose concentration for cellulose/7.6NaOH/water (taken from Fig. III.3.2) and for
cellulose/7.6NaOH/6urea/water solutions.
_____________________________________________________________________________________________________
-137-
0,0
0,2
0,4
0,6
0,8
1,0
1,2
0 0,5 1
m(cell)/m(NaOH)
reduced enthalpy

Fig.III.3-7: Reduced melting enthalpy of the NaOH eutectic versus reduced cellulose concentration
for Avicel / 7.6NaOH / water ( , Fig.III.3-2) and for Avicel / 7.6NaOH / 6urea / water ( )
The full line is given to guide the eyes.

Whatever the solvent system is, the decrease of the eutectic melting enthalpy is the same.
The addition of urea does not change the interactions between cellulose and NaOH. This
is a very important result.

III.3.2.3. Peak of urea eutectic
Whatever the concentration of cellulose is, the peak of the urea/water eutectic mixture is seen
(Fig.III.3-6). We can note that the melting temperature of urea eutectic is increasing with
cellulose content (Tab.III.3-4). An explanation of this increase of melting temperature is a change
in the molecular environment of urea and water due to the addition of cellulose.

The values of the melting enthalpy of the urea eutectic are scattered, but we can assume that
there is no influence of the cellulose content on the melting enthalpy of the urea eutectic peak.
The enthalpy of the urea eutectic does not depend on cellulose concentration. The
addition of cellulose does not change the urea-water interactions and cellulose does not
interact with urea.

III.3.2.4. Peak of free ice
We can note that the melting temperature of ice is increasing with cellulose content (Tab.III.3-4).
An explanation of this increase of melting temperature is a change in the molecular environment
of water due to the addition of cellulose.
Since the melting enthalpy of free ice is approximately constant for all cellulose concentrations,
we conclude that the amount of free ice remains the same with and without cellulose.

_____________________________________________________________________________________________________
-138-
Three main facts are obtained from these series of experiments and their interpretations:
1. NaOH and urea do not interact when mixed together with water; their eutectic mixtures
are independent (part III.2.2.2).
2. Urea and cellulose do not interact when mixed with NaOH and water (urea-water
eutectic is not changed).
3. The interactions between NaOH hydrates and cellulose are not changed by the addition
of urea (the decrease of the eutectic peak of NaOH with cellulose content is the same
with or without urea)

This clearly challenges the use of urea as a promoter for cellulose dissolution and, as will be seen
in the next chapter, as a retardation agent for solution gelation. We will elaborate hypothesis on
the influence of urea in the Conclusion part of this chapter.

III.3.3- Cellulose/NaOH/ZnO/water solutions
As far as we did not have enough time to test all cellulose concentrations in 7.6NaOH/0.7ZnO
aqueous solutions, we only performed experiment on solution containing 5g of cellulose Avicel.
The DSC thermograms are represented on Fig.III.3-8 and melting temperatures and enthalpies
are reported in Tab.III.3-5.

0
0,2
0,4
0,6
-40 -30 -20 -10 0
TC
Heat Flow (W/g)
without ZnO
with 0.7 ZnO

Fig.III.3-8: DSC melting thermograms of 5Avicel /7.6NaOH/water and
of 5Avicel /7.6NaOH/0.7ZnO/water solutions.
Curves are shifted vertically for a better visibility

_____________________________________________________________________________________________________
-139-
NaOH eutectic peak ice peak

Tm (onset), C Heut, J/g Tm (end), C Hice, J/g
7.6NaOH/water -32.9 72 -5.1 195
5Avicel/7.6NaOH/water -33.5 22 -3.9 217
7.6NaOH/0.7ZnO/water - - -5.5 186
5Avicel/7.6NaOH/0.7ZnO/water - - -4.9 181
Tab.III.3-5: Melting temperatures and enthalpies of the melting peaks for 5Avicel/7.6NaOH/water
and for 5Avicel /7.6NaOH/0.7ZnO/water.

As far as the temperature of DSC apparatus does not decrease enough to obtain the melting peak
of the NaOH eutectic in the presence of ZnO (Fig.III.3-8), we cannot conclude either on the
amount of NaOH linked to cellulose or on the limit of cellulose dissolution when ZnO is present
in the solvent system.

Nevertheless, the DSC thermogram shows the melting peak of free ice.
As in the case of cellulose/NaOH/water solution, the melting temperature of free ice slightly
increases when cellulose is added. This was previously explained by some changes in the
molecular environment of the crystallizing water when cellulose is trapping NaOH.
Regarding the melting enthalpy of free ice, we can see that it does not change within the
experimental error. This means that the amount of free water remains the same with and without
cellulose and, thus, that cellulose traps NaOH surrounded by the water molecules forming the
eutectic mixture.

III.3.4- Conclusions on the structure of cellulose solutions in NaOH/water
with and without additives
In NaOH/water and in NaOH/urea/water, the addition of cellulose does not change the melting
temperatures of the eutectic mixtures and of free ice, or very slightly only showing some changes
in the molecular environment due to the presence of cellulose. Cellulose does not modify the
phase diagrams of solvent systems.
On the other hand, the enthalpy of the NaOH eutectic is decreasing proportionally to the
cellulose concentration and is reaching zero when the ratio m
cell
= m
NaOH
. This means that
cellulose can be dissolved in 7-9NaOH/water as well as in 7.6NaOH/6urea/water if at least four
NaOH are linked to each anhydroglucose unit. Consequently, in these two solvents the limit of
cellulose dissolution is reaching when m
cell
= m
NaOH
.
For all the cellulose contents from 0 to 7.6, the melting enthalpy of free ice remains constant.
The amount of free water does not change with the addition of cellulose and thus each
anhydroglucose is surrounded by at least 36 water molecules initially forming the NaOH eutectic.

We saw in the study of solvents that the addition of ZnO in NaOH/water decreases the melting
temperature of the NaOH eutectic. As we work in the limit of the DSC apparatus, the addition
of cellulose leads to the disappearance of this peak. Thus it is possible neither to determine the
limit of cellulose dissolution nor the amount of NaOH linked to each AGU.
_____________________________________________________________________________________________________
-140-
For cellulose/NaOH/ZnO/water, only the melting peak of free ice appears and its enthalpy
remains constant, within the experimental errors, when cellulose is added. The amount of free
water stays the same whatever the cellulose content is.

III.4- INFLUENCE OF FREEZING ON THE THERMAL PROPERTIES OF THE
CELLULOSE/NAOH MIXTURES
In his PhD thesis, Roy [ROY_PhD] showed that after freezing of cellulose/NaOH/water
solution at -20C for 8 hours, a pure substance melting around 0C appeared after thawing.
Freezing seems to be an important step in the preparation of sponges from cellulose/NaOH
aqueous solutions, as it will be shown in details in Chapter V. That is why the transformations
occurring in the solution during freezing should be studied and understood.

The temperature of the new melting peak at 0C suggests that pure water appears in the system
during freezing. The enthalpy value of this peak varied from experiment to experiment and was
not reproducible. We decided to examine again this phenomenon, trying to find a way to obtain
reproducible results. We know that non-dissolved fibres or at least non dissolved parts of fibres
are present in cellulose/NaOH/water solutions and that these fibres precipitate relatively rapidly
after the solution preparation. This is varying cellulose concentration in the supernatant. In order
to always have the same cellulose solutions in the measuring cell, we homogenised the mixture
right before all the experiments and solution was taken and immediately put into the DSC cell.

5g B342 cellulose was dissolved in 7.6NaOH/water and in 7.6NaOH/0.7ZnO/water as
described in Chapter II, total solution weight being 100g. After homogenisation, about 20mg of
the solution were put into DSC caps. Three thermal treatments were performed:
The temperature decreased down to -28C, there was no isothermal step and the
temperature immediately increased up to 10C.
The temperature decreased down to -28C, a 2 hours isothermal step was performed
at -28C, and then the temperature increased up to 10C.
The temperature decreased down to -28C, a 20 hours isothermal step was performed
at -28C, and then the temperature increased up to 10C.

The results obtained after freezing and thawing were reproducible (Fig.III.4-1 and Fig.III.4-2).

_____________________________________________________________________________________________________
-141-
0
0,2
0,4
0,6
-30 -20 -10 0
TC
Heat Flow (W/g)
0H at -28C
20H at -28C
2H at -28C

Fig.III.4-1: Melting DSC thermograms for 5B342 cellulose/7.6NaOH/water solutions after freezing at -28C and
immediate thawing (0H) and after keeping solution at -28C for 2 hours (2H) and 20 hours (20H)
(each experiment was performed twice on different solutions)

0
0,2
0,4
0,6
-30 -20 -10 0
TC
Heat Flow (W/g)
0H at -28C
20H at -28C
2H at -28C

Fig.III.4-2: Melting DSC thermograms for 5B342 cellulose/7.6NaOH/0.7ZnO/water solutions after freezing at -28C and
immediate thawing (0H) and after keeping solution at -28C for 2 hours (2H) and 20 hours (20H)
(each experiment was performed twice on different solutions)

Time at -28C,
hours
Tm1, C
of free ice
H1, J/g
of free ice
Tm0,C
of water released
H0, J/g
of water released
0 -5.5 156 - -
2 -5.0 160 - -
2 -5.2 162 - -
20 -5.5 144 0.7 1.8
20 -5.4 147 0.2 2.3
Tab.III.4-1: Temperatures and enthalpies of melting peaks of free ice and of water released
for a 5B342 cellulose /7.6NaOH/water solution, after freezing and keeping for 0, 2 and 20 hours at -28C.
_____________________________________________________________________________________________________
-142-
Time at -28C,
hours
Tm1, C
of free ice
H1, J/g
of free ice
Tm0,C
of water released
H0, J/g
of water released
0 -5.5 155 - -
2 -5.5 156 0.6 Very low
2 -5.2 153 0.8 Very low
20 -5.6 147 -1.3 / 0.7 1.8
20 -5.4 145 0.4 / 1.0 1.9
Tab.III.4-2: Temperatures and enthalpies of melting peaks of free ice and of water released
for a 5B342 cellulose/7.6NaOH/0.7ZnO/water solution, after freezing and keeping 0, 2 and 20 hours at -28C.

During freezing and keeping the solution at -28C for several hours, a phase separation is
occurring.
A 5B342 cellulose/7.6NaOH aqueous solution left at -28C for 20 hours reveals a single and
sharp peak at 0C (Fig.III.4-1). We can say that pure water appeared. These results are in
agreement with the ones obtained by Roy [ROY_PhD].
We can note that this peak is not present after 2 hours at -28C.
A 5B342 cellulose/7.6NaOH/0.7%ZnO aqueous solution left at -28C for 20 hours reveals
two melting peaks, at -1.3C and 0.7C in the first experiment and at 0.4C and 1C in the
second experiment (Fig.III.4-2). The presence of ZnO does not prevent the phase separation
but it seems that two liquids are released separately: probably pure water, as in the case of
cellulose in pure NaOH/water, and a very dilute liquid containing NaOH and/or ZnO in
water.
Moreover, after only 2 hours at -28C, we can see that the phase separation is also occurring
since the melting peaks are present but the enthalpies are too low to be quantified. It seems
that the presence of ZnO in the solvent system makes easier the occurrence of phase
separation.

In the case of cellulose/7.6NaOH/water, we determined that a 20hours isothermal step at -28C
induces a release of pure water. Consequently, using the melting enthalpies of the released water
(Tab.III.4-1) and of pure crystallised water (H
ice pure
= 358J/g), the amount of released water per
one AGU can be calculated.
100
18
162
/
.
.
cell
released water
AGU released water
m
f
n n
=
where
pure ice
released water
released water
H
H
f
.
.
.
= is the fraction of water released and m

cell
is cellulose weight in
100g of solution.

The result is that one water molecule per AGU is released. The reason that there is water
molecule that is detached from water-linked-to-AGU is as follows. The water crystallizing in
time under cooling at -28C could come, in general, from three places in solution where water
_____________________________________________________________________________________________________
-143-
molecules are present: free eutectic mixture, eutectic mixture linked to cellulose and non-frozen
water (according to the phase diagram, at this temperature there is a small amount of a non-
frozen water). On one hand, according to thermodynamic reasoning, water belonging to free
eutectic cannot be detached, so the released water cannot come from eutectic mixture. On
the other hand, as far as the addition of cellulose does not perturb NaOH/water phase diagram,
water in non-frozen water is not modified. The only possibility is thus water-bound-to-cellulose
together with the NaOH ions. We have shown above that there are at least 36 such water
molecules surrounding each AGU. About one is leaving to form the ice melting at 0C. There is
no depression of the melting temperature probably because this water is diffusing in a system in
thermodynamic equilibrium. We may suppose that because cellulose/NaOH/water is not a very
stable organisation, water molecules are detaching from cellulose and moving towards ice that
serves as nuclei for new crystals. Because of low temperatures this diffusion is slow, that is why
several hours are needed for free water to form crystals.
During the freezing step, 1 H
2
O/AGU is released, which is not much in comparison with
the 36 H
2
O molecules surrounding AGU initially. The addition of ZnO does not change
the amount of water released

_____________________________________________________________________________________________________
-144-
III.5- CONCLUSIONS
We performed an extended investigation, mainly based on DSC technique, of the thermodynamic
behaviour of cellulose in NaOH/water solutions with and without additives. This study allowed
understanding the structure and proportions between the components interacting in the system
and determination and explanation of the limit of cellulose dissolution in NaOH/water and in
NaOH/urea/water solutions.

Cellulose in NaOH/water

The phase diagram of NaOH/water in the 0-30% region is of the eutectic type. Our DSC
experiments are in agreement with the phase diagram reported in the literature.

According to the known phase diagram of cellulose/NaOH/water system [SOB1939], cellulose
can be dissolved when mixed with NaOH/water in the 6-10% of NaOH/water below 0C. DSC
experiments reveal that the addition of Avicel cellulose does not change either the temperature of
the NaOH eutectic mixture or the temperature of melting ice. This means that cellulose does not
disrupt the formation of the NaOH eutectic and does not directly interact with water.
Nevertheless, the intensity of the NaOH eutectic peak decreases when cellulose concentration
increases. Thus, it is possible to calculate the amount of NaOH molecules linked to each AGU
and to deduce the maximal amount of cellulose that can be dissolved in a given NaOH/water
solution: at least four NaOH should be bound to one AGU to provide dissolution. This
proportion expressed in weight ratio gives m
cell
/m
NaOH
1 as a limit of cellulose dissolution. In
these limiting conditions, about 36 water molecules are bound to one AGU.
An interesting feature is that the amount of NaOH that is not able to participate to the eutectic
due to the presence of cellulose is increasing when decreasing the cellulose concentration, or in
other words, the amount of NaOH linked to AGU increases with the decrease of cellulose
concentration. We do not have a simple explanation of this phenomenon, and it is not an artefact
since the same can be deduced from turbidity experiments on cellulose dissolution [KUO2005].

As far as cellulose can be dissolved in 6% - 10% aqueous NaOH, the maximum of Avicel
cellulose that can be dissolved in sodium hydroxide is 10%, which can be a blocking point for
solutions processing.

For 5cellulose/7.6NaOH/water solutions, the melting enthalpy of the NaOH eutectic depends
on the cellulose origin. The most difficult to dissolve was cellulose B407 and L198
II
.

_____________________________________________________________________________________________________
-145-
Cellulose in NaOH/urea/water

Urea/water forms a simple eutectic phase diagram.
When mixed with NaOH/water, urea and NaOH do not interact. They each behave as if they
were alone, when enough water is present in the system: both NaOH and urea form their eutectic
mixture with water, NaOH,5H
2
O+4H
2
O and urea+8H
2
O. As soon as the concentrations are
such that water is missing, NaOH is attracting water at the expense of urea.

When cellulose is added to NaOH/urea/water, the interaction between NaOH and cellulose is
the same as if urea were not present. This means that urea is not linked to cellulose. The question
is why urea is efficient for helping the dissolution and delaying gelation (see Chapter IV). A
possible hypothesis is that cellulose dissolution is possible when the NaOH hydrates have a given
size and structure that is found in the region of 6-10%NaOH below 0C. In these conditions, a
lot of liquid water is present in NaOH/water solution. The role of urea is to bind a part of this
water which is not favourable for the cellulose dissolution.

Cellulose in NaOH/ZnO/water

ZnO reacts with NaOH to form zinc hydroxides, NaZn(OH)
3
and Na
2
Zn(OH)
4
. As a
consequence, some NaOH molecules are linked to ZnO and the melting temperature of the peak
NaOH/water eutectic mixture decreases and becomes not measurable when ZnO is added. The
limit of ZnO dissolution is reached when NaOH/ZnO molar ratio becomes equal to 10.

As far as there is no NaOH/water eutectic peak in cellulose/7.6NaOH/0.7ZnO/water and free
ice peak is not influence by the presence of either cellulose or ZnO, the analysis of the
proportions between the components was not possible. The role of ZnO in cellulose dissolution
is supposed to be the same as the one of urea.

Influence of freezing on the thermal properties of the cellulose/NaOH mixtures

The addition of cellulose in NaOH/water does not change the NaOH/water binary phase
diagram, plotted by Pickering [PIC1893] and Rollet et al. [ROL1964]. Consequently, at -20C,
with or without cellulose, a large amount of water (f
ice
~ 0.49) is crystallised decreasing the
amount of liquid water in the system and thus increasing the NaOH concentration of the liquid
up to 15%NaOH. The fact to be outside the domain of cellulose dissolution [SOB1939] may
explain why after freezing the cellulose solution is a weak gel (see Chapter IV).

Secondly, when keeping cellulose/NaOH/water solution below -20C for several hours, free
water appears that crystallises and then melts at 0C. If freezing occurs for short time, no peak
appears at 0C. Approximately one water molecule per one AGU is detached from cellulose
and crystallises.
_____________________________________________________________________________________________________
-146-
The influence of the freezing step on the final properties of regenerated cellulose is more
probably due to the large amount of crystallised water (crystallisation following the NaOH/water
binary phase diagram), which changes the environment of cellulose (decreasing of liquid water
and increasing of NaOH concentration), than the very low amount of released water.

_____________________________________________________________________________________________________
-147-
[ANT1926] ANTROPOFF A., SOMMER R., Z. Physik Chem. (1926), 123, 164

[BAR1954] BARTUNEK R.; Papier (1954), 7, 254

[BEN1999] BENEZETH P, PALMER D.A, WESOLOWSKI D.J. The solubility of zinc oxide
in 0,03m NaTr as a function of TC, with in situ pH measurement. Geochim. Cosmochim. Acta
(1999), 63 (10), p.1571-1586

[BOC1949] BOCKRIS J. Ionic solvation. Quaterly reviews, Chemical Society (1949), 3, p.173-180

aqueous solutions. Macromolecular Bioscience (2005), 5, p.539-548

[COH1960] COHEN-ADAD R., TRANQUARD A., PERONNE R., NEGRI P., ROLLET A-
P. Le systme eau-hydroxyde de sodium. Compte rendus de l'acadmie des sciences (1960), 251, p.2035-
2037

[DIR1954] DIRKSE T.P., POSTMUS C., VANDENBOSCH J.R., VANDENBOSCH R. A
study of alkaline solutions of zinc oxide. J. Am. Chem. Soc. (1954), 76, p.6022-6024

[HUM1953] HUME-ROTHERY W., CHRISTIAN J.W., PEARSON W.B. Chapter 29: Ternary
systems: general principle. In: Metallurgical equilibrium diagrams and experimental methods for their
determination. reprinted 1953. London: Unwin Brothers Limited (1952) 306p., The institute of
physics

[KON2001] KONINGSVELD R., STOCKMAYER W.H., NIES E. Chapter 3: Binary systems,
Bakhuis Roozeboom diagram, Chapter 6: Ternary and multicomponent systems. In: Polymer phase
diagrams. A textbook. New York: Oxford University Press, Inc. (2001) 341p.

[KUN1985] KUNZE J., EBERT A., LANG H., PHILIPP B. Na-NMR spektroskopische
Untersuchungen zur Hydratation von Natriumhydroxid in wBriger Lsung. Z. phys. Chemie
(1985), 266, p.49-58


[LIU1998] LIU Y., PIRON D.L. Study of tin cementation in alkaline solution. J. Electrochem. Soc.
(1998), 145 (1), p.186-190

[MOO1994] MOOTZ D., RTTER H., WISKEMANN R. Die Kristallstrukturen von
NaOH.3,5H2O und NaOH.7H2O. Eine Przisierung. Z. anorg. allg. Chem. (1994), 620, p.1509-
1513

[PIC1893] PICKERING S.U. The hydrates of sodium, potassium and lithium hydroxides. J.
Chem. Soc. (1893), 63, p.890-909

_____________________________________________________________________________________________________
-148-
[RAB1978] RABINOVICH V.A., KHAVIN Z.Ya. Handbook on Chemistry (in Russian), 2nd
edition, "Khimiya", Leningrad division (1978) 392 p.

[REI1975] REICHLE R.A., McCURDY K.G., HEPLER L.G. Zinc hydroxide: Solubility product
and hydroxy-complex stability constants from 12.5-75C. Can. J. Chem (1975), 53, p.3841-3845

[ROL1964] ROLLET A-P., COHEN-ADAD R. Les systmes "eau-hydroxyde alcalin". Revue de
chimie minrale (1964), 1, p.451-478

[ROY2001] ROY C., BUDTOVA T., NAVARD P, BEDUE O. Structure of cellulose - soda
solutions at low temperatures. Biomacromolecules (2001), 2, p.687-693


[ROY_PhD] ROY C. Etude de mlanges de cellulose dans des solutions aqueuses de soude. Thse Doct.
Sciences et gnie des matriaux, Sophia-antipolis : Ecole des Mines de Paris (2002), 213p.

[SOB1939] SOBUE H., KIESSIG H., HESS K. The cellulose-sodium hydroxide-water system as
a function of the temperature. Z. Physik. Chem. B (1939), 43, p.309-328

[WEA1983] WEAST R.C., ASTLE M.J., BEYER W.H. Handbook of chemistry and physics. 64
th

edition. Boca Raton, Florida: CRC Press Inc. (1983-1984)

[YAM1988] YAMASHIKI T, KAMIDE K, OKAJIMA K, KOWSAKA K, MATSUI T,
FUKASE H. Some characteristic features of dilute aqueous alkali solutions of specific alkali
concentration (2,5mol/l) which possess maximum solubility power against cellulose. Polym. J.
(1988), 20 (6), p.447-457

Rsum du chapitre IV Ecoulement et glification des solutions aqueuses de cellulose/NaOH : Influence des additifs.
_____________________________________________________________________________________________________
- 149 -

R RE ES SU UM ME E D DU U C CH HA AP PI IT TR RE E I IV V
E EC CO OU UL LE EM ME EN NT T E ET T G GE EL LI IF FI IC CA AT TI IO ON N D DE ES S S SO OL LU UT TI IO ON NS S A AQ QU UE EU US SE ES S
D DE E C CE EL LL LU UL LO OS SE E/ /N NA AO OH H : : I IN NF FL LU UE EN NC CE E D DE ES S A AD DD DI IT TI IF FS S. .
Le but de ce chapitre est dtudier le comportement lcoulement des solutions de cellulose, de
dterminer les conditions de glification. Nous tudierons galement linfluence des additifs tels
que oxyde de zinc et ure sur les proprits rhologiques des solutions de cellulose. Pour finir,
nous verrons brivement linfluence dune tape de conglation -20C sur ltat du systme.

Dans la premire partie, nous avons ralis des mesures de viscosit en fonction du taux de
cisaillement sur des solutions de cellulose/7,6NaOH/eau et cellulose/7,6NaOH/0,7ZnO/eau
essentiellement, et ce, pour diffrents DP et concentrations en cellulose et pour des tempratures
comprises entre 0C et 40C. Quelques mesures ont galement t effectues sur des solutions
aqueuses de cellulose/7,6NaOH/6ure.
Nous avons ainsi dtermin que, pour les trois systmes alcalins tudis, les solutions de cellulose
nont pas un comportement de solutions molculairement disperses mais que leur
comportement se rapproche de celui dune dispersion. Ce rsultat nest pas tonnant puisque
nous avons vu dans le chapitre III que chaque unit anhydroglucose de la chane de polymre est
entoure, au minimum, de 4 NaOH et 36 molcules deau.

Par des mesures de viscosit intrinsque, nous avons pu classer les solvants en fonction de leur
qualit de solvant de la cellulose. Il savre que 7,6NaOH/eau est moins bon que
7,6NaOH/6ure/eau qui est son tour moins bon solvant que 7,6NaOH/0,7ZnO/eau. Ce
rsultat est en adquation avec les observations microscopiques que nous avons ralises.
Une augmentation de la temprature dessai conduit une diminution de la viscosit intrinsque
et donc une diminution de la qualit du solvant. Quand la temprature augmente, les interactions
cellulose-cellulose sont donc favorises au dtriment des interactions cellulose-solvant, ce qui
entrane la glification des solutions, phnomne dj rapport dans la littrature [ZHO2000]
[ROY2003].

Les solutions de cellulose ne sont pas stables et glifient avec le temps et la temprature. Les
conditions de glification tant mal connues, nous avons dtermin le point de gel des solutions
de cellulose en variant de nombreux paramtres (temprature, DP et concentration en cellulose)
et notamment la prsence ou non dadditifs.
Une augmentation rapide de la temprature donne un point de gel des solutions de cellulose aux
alentours de 30C. La prsence doxyde de zinc ou dure augmente cette temprature de
Rsum du chapitre IV Ecoulement et glification des solutions aqueuses de cellulose/NaOH : Influence des additifs.
_____________________________________________________________________________________________________
- 150 -
glification de plusieurs degrs suivant la nature et la concentration de ladditif et la nature de la
cellulose.
En ce qui concerne la prsence doxyde de zinc, la temprature de glification augmente quand la
teneur en ZnO augmente jusqu une concentration limite de 1,5ZnO dans 7,6NaOH/eau. Le
rapport NaOH : ZnO est alors de 10 et correspond la limite de solubilit de ZnO dans NaOH.
De plus, lajout de ZnO solide aprs mlangeage de la cellulose avec la soude donne la mme
temprature de glification que des solutions de cellulose ne contenant pas du tout de ZnO. Ceci
signifie que ZnO nest efficace que sil est pralablement dissout dans NaOH/eau avant dtre
ajout la cellulose. Etant donn que ZnO et NaOH ragissent pour former des hydroxydes de
zinc, nous pensons que la formation de ces complexes NaOH-ZnO-eau est favorable la
dissolution des fibres de cellulose et la stabilit des solutions.
Quant lure, nous avons vu que sa prsence augmente nettement la temprature de glification
des solutions. Par contre, sa teneur optimale dans 7,6NaOH/eau dpend de la cellulose (origine,
traitement). A lheure actuelle, les raisons de telles diffrences restent une question ouverte.
Dans tous les cas, quels que soient les mcanismes qui interviennent, ure et ZnO retardent la
glification des solutions.

Pour finir, nous avons ralis quelques mesures rhologiques sur des solutions congeles de
cellulose/NaOH/eau. Si on compare une solution ayant subit un sjour prolong -20C et une
solution frachement prpare, on constate que la conglation diminue la qualit de solvant de
NaOH/eau (diminution de la viscosit intrinsque) en favorisant les interactions cellulose-
cellulose allant jusqu la glification des solutions (augmentation de la viscosit et du module G).
Par contre, la prsence doxyde de zinc minimise considrablement ces effets. Aprs conglation,
G a certes augment mais trs faiblement tandis que la viscosit reste pratiquement inchange
sur tout le domaine de taux de cisaillement tudi.
Comme pour la glification par hausse de la temprature, ZnO retarde galement la glification
basse temprature.

Chapter IV Flow and gelation of aqueous cellulose/NaOH solutions: Influence of additives
_____________________________________________________________________________________________________
- 151 -
C
C
H
H
A
A
P
P
T
T
E
E
R
R
I
I
V
V

IV- iv
F FL LO OW W A AN ND D G GE EL LA AT TI IO ON N O OF F A AQ QU UE EO OU US S
C CE EL LL LU UL LO OS SE E/ /N NA AO OH H S SO OL LU UT TI IO ON NS S: : I IN NF FL LU UE EN NC CE E O OF F
A AD DD DI IT TI IV VE ES S
As shown in Chapter I [SOB1939], sodium hydroxide can be a solvent of cellulose in a narrow
range of temperature (-6C-0C) and NaOH concentration (6%-10%). Moreover, according to
the literature [YAM1990], the DP of cellulose samples that can be dissolved in NaOH has to be
inferior to about 400 for steam exploded wood pulp but this value also depends on the cellulose
origin. These very restrictive conditions show the difficulties of cellulose processing. In addition,
even after dissolution, cellulose solutions are not stable and gelation occurs as a function of time
and temperature [ROY2003] [ROY_PhD]. Gelation kinetics and influence of temperature were
studied for microcrystalline Avicel cellulose dissolved in 9%NaOH/water [ROY2003]. Gelation
can be either an advantage or a drawback depending on the industrial process.

Literature reveals that the addition of some compounds such as urea and ZnO improves cellulose
dissolution and stability of solutions [WO 02/22924] [ZHO2000]. The goal of this chapter is to
study the influence of ZnO and urea on the flow and gelation of cellulose/NaOH/water
solutions and to understand the mechanisms of solution gelation. This chapter is divided into
four parts Influence of ZnO and urea on the flow behaviour of cellulose/NaOH aqueous
solutions, Gelation and dissolution of cellulose in the presence of ZnO, Gelation and
dissolution of cellulose in the presence of urea. Comparison between urea and ZnO, and Tests
on gelation and dissolution of cellulose in other alkali solvents: KOH and NaOH with other
additives

The first part corresponds to the study of the flow behaviour of semi-dilute cellulose solutions.
Viscosity versus shear rate measurements were performed on cellulose/NaOH aqueous
_____________________________________________________________________________________________________
- 152 -
solutions, with or without ZnO or urea, by changing parameters as DP and concentration of
cellulose, and solution temperature.

The second and third parts investigate the influence of additives ZnO and urea as well as the
influence of cellulose origin on (i) cellulose dissolution by microscopic observations and (ii)
gelation of solutions by rheological measurements in dynamic mode. The optimal concentration
of both additives for cellulose dissolution was determined. Finally, the role of these two additives
on gelation and dissolution of cellulose is discussed.

The fourth part is devoted to the analysis of other additives, like ZnSO
4
, MgCl
2
, and CaSO
4
on
the gelation of a 5cellulose/7.6NaOH/water solution and on cellulose dissolution.

_____________________________________________________________________________________________________
- 153 -
IV.1- INFLUENCE OF ZNO AND UREA ON THE FLOW BEHAVIOUR OF
CELLULOSE/7.6NAOH/WATER SOLUTIONS
The goal of this section is to investigate the flow behaviour of cellulose/7.6NaOH/water
solutions with and without additives (a detailed study on ZnO presence in the solvent and a few
tests on the presence of urea). Viscosity-shear rate dependence was measured on cellulose
solutions in which many parameters were varied:
The temperature: from 0C to 40C.
The DP of cellulose: DP 342 and DP 500 (Borregaard steam exploded pulp, B342 and
B500, respectively)
The cellulose content: from 0.8g to 5g in 100 g of solution.
The solution composition: cellulose/7.6NaOH/water, cellulose/7.6NaOH/0.7ZnO/
water.

Two other cellulose samples were tested B407 and Avicel but only at 10C and with a
content of 5g. The solutions were Avicel/7.6NaOH/water, B407/7.6NaOH/water, and
B407/7.6NaOH/4urea/water.
4g of urea in a 100g solution containing 7.6g of NaOH was the optimal urea content to
dissolve cellulose in 6-8NaOH/urea/water reported in the ref. [ZHO2000].

IV.1.1- Influence of the presence of ZnO or urea in the solvent on the flow of
cellulose/7.6NaOH/water solutions
This section is devoted to the influence of the presence of additives in the solvent on the flow
behaviour of cellulose/7.6NaOH/water solutions. The main results compare
cellulose/7.6NaOH/water and cellulose/7.6NaOH/0.7ZnO/water solutions, only few results
concern the presence of urea in the system. The solutions were sheared in steady-state mode in a
Couette double gap rheometer.
0,001
0,01
0,1
1 10 100 1000
shear rate (s-1)

(
P
a
.
s
)
0C
30C
20C
10C

Fig.IV.1-1: Viscosity-shear rate dependence of 1.7 B342 cellulose/7.6NaOH/water solutions (full symbols)
and 1.7 B342 cellulose/7.6NaOH/0.7ZnO/water solutions (open symbols) at 0C, 10C, 20C, and 30C
_____________________________________________________________________________________________________
- 154 -
At low cellulose concentrations (Fig.IV.1-1), the viscosity-shear rate curves of B342
cellulose/7.6NaOH/water and B342 cellulose/7.6NaOH/0.7ZnO/water coincide.
The viscosity is constant when the shear rate increases: at low cellulose concentration the flow is
Newtonian. The small increase of viscosity at low shear rate is due to the effect of the rheometer
which reaches its maximum sensitivity. The temperature increase leads to a viscosity decrease (at
least from 0C to 30C). This phenomenon can be explained by an increase of chain mobility and
is classically observed for most other systems. It is quantified through the activation energy,
which is determined from the Arrhenius law in the following way:

) / (
exp .
RT E
a
A =

RT
E
A
a
+ = ln ln
Where : the viscosity (Pa.s); it was arbitrarily measured at
1
10

= s
A: a constant
E
a
: the activation energy (J/mol)
R: the constant of perfect gas = 8.32 J/mol/K
T: the absolute temperature (K)
The activation energy can be calculated from the slope E
a
/R of the lines ln versus 1/T.
Fig.IV.1-2 shows the results for pure water, for aqueous sodium hydroxide solutions, with or
without 0.7ZnO, and for cellulose solutions 1.7 B342 cellulose/7.6NaOH/water, 1.7 B342
cellulose/7.6NaOH/0.7ZnO/water and 2.8 B342 cellulose/7.6NaOH/0.7ZnO/water. Let us
note that it is not possible to determine E
a
for 2.8 B342 cellulose/7.6NaOH/water solution since
the flow is not Newtonian anymore. This second behaviour will be discussed in the forthcoming.
Nevertheless, if we compare 1.7 B342 cellulose/7.6NaOH/water (Fig.IV.1-2, full symbols) with
1.7B342 cellulose/7.6NaOH/0.7ZnO/water (Fig.IV.1-2, open symbols), the presence of 0.7ZnO
in the solvent does not change the viscosity-temperature dependence, within the experimental
error (Fig.IV.1-2, table).
-8
-7
-6
-5
-4
-3
-2
0,0031 0,0032 0,0033 0,0034 0,0035 0,0036 0,0037
1/T
l
n

water
NaOH
NaOH+ZnO
1.7 B342
2.8 B342
0.8 B500

Fig.IV.1-2: Viscosity versus inverse temperature for water, pure solvents (7.6NaOH/water and 7.6NaOH/0.7ZnO/water)
and for cellulose/7.6NaOH/water, cellulose/7.6NaOH/0.7ZnO/water.
Solid lines correspond to solutions in 7.6NaOH/water, dashed lines to solutions in 7.6NaOH/0.7ZnO/water.
Content and DP of cellulose are indicated on the graph
The values of the corresponding activation energy E
a
(kJ/mol) are given in the Tab.IV.1-1.

_____________________________________________________________________________________________________
- 155 -

An increase of cellulose concentration as well as of DP of cellulose leads to an increase of the
activation energy E
a
. As the activation energy is expressing the difficulty of motions of polymer
chains and so is connected with the density of entanglements, the increase of the activation
energy with concentration and DP of cellulose was an expected result. This phenomenon
observed for steam exploded cellulose solutions is in contradiction with the results obtained for
Avicel cellulose solutions [ROY2003]. There, the authors showed that the activation energy was
independent on the cellulose concentration and concluded that cellulose/NaOH aqueous
solutions are not real solutions.

Comparing 1% DP950 cellulose in NMMO/water [GAV2005] with 0.8 B500 (DP500) cellulose
in 7.6NaOH/water shows a large difference with respectively 48kJ/mol and 22.6kJ/mol. This
latter value is close to the one of the solvent (19kJ/mol). It suggests either that NMMO/water is
a complexing solvent that is hampering strongly the motion of the cellulose chains or that the
interactions between the solvent and cellulose in the case of NaOH solutions are weak, leading to
a collapsed state, in a sense similar to what [ROY2003] found.

With the increase of cellulose concentration and, as we will see, temperature, the cellulose
solutions behave in a different way. The flow is non-Newtonian; the shear rate increase leads to a
more or less pronounced decrease of viscosity (Fig.IV.1-4) which is characteristic of a
pseudoplastic behaviour. The flow index n was calculated in the shear-thinning region as
follows:

n
A

= .
Where
: the shear rate (s

-1
)
: the viscosity (Pa.s)
n: the flow index
A: a constant

Cellulose content

0 cellulose
pure solvent
0.8
B500
1.7
B342
2.8
B342
water 18.0
7.6NaOH/water 19.0 22.6 23.1
7.6NaOH/
0.7ZnO/water
18.4 23.0 24.2 26.2
Tab.IV.1-1: Values of activation energy Ea (kJ/mol)
determined from Fig.IV.1-2

10
20
30
0 1 2 3
m(cell), g
Ea (kJ/mol)
B500
B342

Fig.IV.1-3: Activation energy Ea versus cellulose content for
B342 ( ) and B500 ( ) cellulose in 7.6NaOH/water (full
symbols) and in 7.6NaOH/0.7ZnO/water (open symbols)

_____________________________________________________________________________________________________
- 156 -
For each cellulose solutions, the flow index n was calculated and the results are reported in
Tab.IV.1-2 and in Tab.IV.1-3.

Viscosity-shear rate curves of cellulose/7.6NaOH/water, cellulose/7.6NaOH/0.7ZnO/water
solutions are represented on Fig.IV.1-4. Viscosity measurements were performed as a function of
temperature and content and DP of cellulose; the trend is similar in all cases.
At low temperatures (0-20C), the viscosity is decreasing with shear rate. The non-linearity is
more pronounced with increasing either content or molar mass of cellulose, as should be
expected.
Then, at a certain temperature (~30C) and at low shear rate, there is a sudden increase of the
viscosity. As far as we know that cellulose solutions are gelling with temperature increase
[ROY2003], we can say that this sudden increase of viscosity is a direct indication of solution
gelation. Simultaneously, an indirect indication of the gelation is a strong shear thinning of
cellulose solutions, quantified by the flow index n (Tab.IV.1-2). At a low shear rate, the
characteristic time of the flow is large and the gel is reforming easily leading to a thick suspension
of very large elastic gel particles. This induces a high viscosity. At high shear rates, the
characteristic time of the flow is small and the gel is not reforming easily, i.e. is broken in small
pieces. Consequently, the system behaves as a suspension and the viscosity is low.

Moreover, we can note that the direct and indirect indicators of gelation the sudden increases
of viscosity (Fig.IV.1-4) and flow index n (Tab.IV.1-2) respectively appear at higher cellulose
concentration and temperature for a solution containing ZnO compared with the same cellulose
solution without ZnO. For example, a 4 B342 cellulose/7.6NaOH/0.7ZnO/water solution is not
forming a gel, even at 40C, whereas the same solution in 7.6NaOH/water is gelling at 30C. We
can thus conclude that the presence of ZnO in the NaOH/water solvent delays the occurrence
of gelation.
The cellulose/7.6NaOH/0.7ZnO/water systems have Newtonian flow for higher cellulose
concentrations than cellulose/7.6NaOH/water systems (Tab.IV.1-2), we can thus conclude that
the presence of ZnO in sodium hydroxide improves cellulose dissolution.

a) 2.8 B342 b) 4 B342
0,01
0,1
1
1 10 100 1000
shear rate (s-1)

(
P
a
.
s
)

0,01
0,1
1
1 10 100 1000
shear rate (s-1)

(
P
a
.
s
)

_____________________________________________________________________________________________________
- 157 -
c) 2.3 B500 d) 3.8 B500
0,01
0,1
1
1 10 100 1000
shear rate (s-1)

(
P
a
.
s
)

0,1
1
10
1 10 100 1000
shear rate (s-1)

(
P
a
.
s
)

Fig.IV.1-4: Viscosity-shear rate dependence of cellulose/7.6NaOH/water (solid lines)
and cellulose/7.6NaOH/0.7ZnO/water (dashed lines), at 0C ( ), 10C ( ), 20C ( ), 30C ( ) and 40C ( ).
Cellulose concentrations and DP are a) 2.8 B342, b) 4 B342, c) 2.3 B500, and d) 3.8 B500.

x B342 /7.6NaOH/water x B342 /7.6NaOH/0.7ZnO/water
TC X=1.7 x=2.8 x=4 x=1.7 x=2.8 x=4
0 0.04 (N) 0.07 (PP) 0.08 (PP) 0.03 (N) 0.02 (N) 0.04 (N)
10 0.03 (N) 0.07 (PP) 0.07 (PP) 0.02 (N) 0.01 (N) 0.03 (N)
20 0.03 (N) 0.09 (PP) 0.12 (PP) 0.02 (N) 0.01 (N) 0.03 (N)
30 0.03 (N) 0.33 (G) 0.22 (G) 0.01 (N) 0.006 (N) 0.05 (PP)
40 0.52 (G) 0.43 (G) 0.003 (N) 0.008 (N) 0.07 (PP)

x B500 /7.6NaOH/water x B500/7.6NaOH/0.7ZnO/water
TC X=0.8 x=2.3 x=3.8 x=0.8 x=2.3 x=3.8
0 0.05 (N) 0.10 (PP) 0.18 (PP) 0.03 (N) 0.05 (PP) 0.11 (PP)
10 0.03 (N) 0.11 (PP) 0.19 (PP) 0.03 (N) 0.05 (PP) 0.12 (PP)
20 0.04 (N) 0.15 (PP) 0.23 (G) 0.03 (N) 0.06 (PP) 0.11 (PP)
30 0.008 (N) 0.27 (G) 0.53 (G) 0.005 (N) 0.07 (PP) 0.36 (G)
40 0.01 (N) 0.31 (G) 0.54 (G) 0.02 (N) - 0.48 (G)
Tab.IV.1-2: Values of the flow index n, which quantifies the shear thinning, for B342-B500 cellulose /7.6NaOH/water
and B342-B500 cellulose/7.6NaOH/0.7ZnO/water solutions, at different cellulose contents and temperatures.
The values followed by (N) correspond to a Newtonian flow, (PP) to a pseudoplastic behaviour and (G) to solution in gel state

_____________________________________________________________________________________________________
- 158 -
Viscosity as a function of the shear rate at 10C
for 5% cellulose / 7,6% NaOH aqueous solution
0,1
1
10
0,1 1 10 100
shear rate (s-1)

(
P
a
.
s
)
AVICEL
B342
B407
B500

0,1
1
10
150 250 350 450
cellulose DP

(
P
a
.
s
)

Fig.IV.1-5: Viscosity-shear rate dependence, at 10C, of 5cellulose/7.6NaOH/water (solid lines),
5cellulose/7.6NaOH/0.7ZnO/water (dashed lines) and 5cellulose/7.6NaOH/4urea/water (full line with crosses)
Inset: Viscosity, arbitrarily measured at 1s
-1
, versus DP of cellulose for 5cellulose/7.6NaOH/water ( ),
5cellulose/7.6NaOH/0.7ZnO/water ( ) and for cellulose/7.6NaOH/4urea/water ( )
The names (including the DP) of cellulose samples are directly indicated on the graph

Cellulose In NaOH In NaOH +ZnO In NaOH +urea
Avicel (DP170) 0.08 0.03 -
B342 0.15 0.07 -
B407 0.22 0.10 0.17
B500 0.25 0.16 -
Tab.IV.1-3: Values of the flow index n , at 10C
for cellulose/7.6NaOH/water, cellulose/7.6NaOH/0.7ZnO/water and
cellulose/7.6NaOH/4urea/water solutions with cellulose having different DP
Cellulose content is 5g in 100g of solution.

Fig.IV.1-5 represents the viscosity-shear rate dependence, at 10C, for
5cellulose/7.6NaOH/water, 5cellulose/7.6NaOH/0.7ZnO/water and 5cellulose/7.6NaOH/
4urea/water solutions with cellulose having different DP. 10C was chosen because at this
temperature gelation is very slow (more than 300 hours, [ROY2003]) and thus can be neglected
as compared with the duration of experiment. As expected, a DP increase induces a viscosity
increase. Indeed, a DP increase involves an increase in the size of macromolecules and of
functional groups and thus an increase of molecular interactions in the solution and thus a
decrease of the motion of polymer chain.
The shear thinning was quantified and the values of flow index n are reported on Tab.IV.1-3.
_____________________________________________________________________________________________________
- 159 -
Tab.IV.1-2 and Tab.IV.1-3 show that the flow index n depends on the DP and concentration of
cellulose, the temperature and the solvent. When the cellulose solutions are neither Newtonian
nor in gel state, the flow index n remains between 0.05 and 0.25, depending on cellulose
concentration and DP, and on solvent and temperature. These values correspond to the ones
found in literature for Avicel cellulose/9%NaOH/water solutions [ROY2003].

IV.1.2- Conclusions
The flow of two cellulose samples B342 and B500 dissolved in 7.6NaOH/water and in
7.6NaOH/0.7ZnO/water was studied for cellulose contents from 0.8g to 5g in 100g of solution,
at temperatures from 10C to 40C. The flow of two other cellulose samples Avicel and B407
was studied but only for 5g of cellulose at 10C. B407 cellulose was also dissolved in
7.6NaOH/4urea/water.

At low cellulose contents, m
cell
< 2g in 100g of solution, the flow is Newtonian whatever the
solvent is. The increase of solution temperature leads to the decrease of viscosity reflecting a
classical solution behaviour.
The increase of cellulose content leads to a weak shear thinning that is less pronounced when
cellulose is dissolved in 7.6NaOH/0.7ZnO/water than in 7.6NaOH/water. In other words, the
presence of ZnO shifts the Newtonian behaviour towards higher cellulose concentration. We
relate this with the improvement of cellulose dissolution in the presence of ZnO that will be
discussed together with the influence of additives on cellulose gelation and will be confirmed in
the next section by microscopic observations.

The increase of solution temperature at cellulose contents higher than 2g leads to a sharp
increase in solution viscosity and strong shear thinning, when cellulose is dissolved in
7.6NaOH/water. As it will be shown in the following, it is related to solution gelation and gel
breaking under flow, respectively. This phenomenon is much less pronounced for cellulose
dissolved in 7.6NaOH/0.7ZnO/water: the increase of viscosity followed by the shear thinning
occurs at higher cellulose concentrations and higher temperatures. This will be studied in details
with oscillatory measurements. Higher the cellulose concentration, lower the temperature of
gelation.

The activation energy values reported above for Borregaard steam-exploded cellulose correspond
to the ones in ref [ROY2003], where a weak influence of Avicel cellulose concentration was
noticed. On the contrary, with Borregaard steam-exploded cellulose, it appears that the activation
energy increases with Borregaard cellulose concentration as well as with the presence of ZnO in
NaOH/water.
The fact that at low concentration the flow is Newtonian, reaching a non-Newtonian regime at
high concentration, is typical of polymer solution, despite that the flow index is very small.
Nevertheless, the fact that the activation energy of rather concentrated cellulose solutions is very
close to the ones of the solvent is suggesting that the cellulose chain is not participating a lot to
the flow as a single chain undergoing internal conformational changes. This suggests a
_____________________________________________________________________________________________________
- 160 -
suspension-like behaviour, as was also noted by [ROY2003]. We can make the assumption that
the polymer chains are aggregating as soft nano gel-like objects, giving rise to a weak non-
Newtonian behaviour (no strong conformational changes during flow) and a small contribution
to the activation energy of the mixture.

IV.2- GELATION AND DISSOLUTION OF CELLULOSE IN THE PRESENCE OF ZNO
Gelation is due to the formation of more or less strong links between polymer chains in order to
form an infinite structure. Two types of gel can be defined, chemical gel (formation of covalent
connections) and physical gel (formation of a network using secondary forces like hydrogen,
ionic or hydrophobic bonds, helical formation). With the former, gelation is irreversible. The
latter are easy to form but they are thermodynamically stable only in a given range of conditions.
A gel is a solid state body. During gelation, the material is going from a liquid state to a solid
state. Gelation is a continuous process which can be studied with rheological measurements. The
steady-state mode [WIN1986] enables to determine the gel point but not the properties of the
final gel since it is breaking under shear. The dynamic mode [TUN1982] [WIN1986] [FRE1996],
which will be used in this work, allows not only to determine the gel point but also to study in
details the gelation kinetics, the mechanical properties of the gel, and the eventual changes
occurring in time.

The time-moduli dependence as well as the temperature-moduli dependence will be the most
often used experiments to investigate the gel formation of cellulose solutions. The gel point will
be considered to correspond to the intercept point between both moduli, in time or in
temperature.

In this section the influence of ZnO on gelation will be studied in details using dynamic rheology.
Gelation kinetics, temperature of gel formation and gel strength at the gel point are investigated
as a function of the dissolution procedure and ZnO concentration. The optimal concentration of
ZnO in cellulose alkali solutions for delaying gelation in time and temperature is determined. The
rheological results are accompanied by visual observations of the dissolution of cellulose fibres,
performed by optical microscopy.

IV.2.1- Influence of the preparation conditions on gelation of cellulose/
7.6NaOH/water solutions in the presence of ZnO
In order to quantify the influence of additives on gelation, the gelation temperature T
gel
was
determined from G and G measurements as a function of temperature (see details in Chapter
II), and is defined as being the intercept point of both moduli.
First of all, different methods of solution preparation were tested to dissolve B342 cellulose. 100g
of solution contains 5g of cellulose, 7.6g of NaOH, 0.7g of ZnO and water; the preparation
methods were as follows:
_____________________________________________________________________________________________________
- 161 -
Method a Innovia Films Procedure the main method used in this work (see details in Chapter II):
1.1%ZnO is dissolved in 12%NaOH/water; 63.33g of this NaOH/ZnO/water system is
cooled down at -6C.
31.67g of water is added to 5g of cellulose pulp; the cellulose/water system is cooled down at
5C.
The solvent system is added to the cellulose pulp, mixing thoroughly with a spatula
The system is placed at -6C and mixed with a rotary overhead mixer, for 2 hours
Method b:
63.33g of 12%NaOH/water are cooled down at -6C and then added to the cellulose pulp
(5g cellulose + 31.67g water), mixing thoroughly with a spatula
0.7g of ZnO is put in 5cellulose/7.6NaOH/water, at 20C
The system is placed at -6C and mixed with a rotary overhead mixer, for 2 hours
Method c:
63.33g of 12%NaOH/water are cooled down at -6C and then added to the cellulose pulp
(5g cellulose + 31.67g water), mixing thoroughly with a spatula
0.7g of ZnO is put in 5cellulose/7.6NaOH/water, at 6C
The system is mixed with a rotary overhead mixer, for 2 hours, at -6C.

The gelation temperature for these different preparation methods are shown in Fig.IV.2-1 and
compared with 5cellulose/7.6NaOH/water solution prepared from the Innovia Films procedure
(noted NaOH on the graph).

0
20
40
60
80
NaOH (a) (b) (c)
TC gel
0,1
1
10
G'=G'' (Pa)

Fig.IV.2-1: Gelation temperature Tgel ( ) and strength of the gel when G=G ( ) of
5 B342 cellulose/7.6NaOH/0.7ZnO/water mixture for three methods to add ZnO,
compared to the case without ZnO noted NaOH on the graph

Gelation temperature and strength of gel obtained in methods b and c are approximately the
same as cellulose/7.6NaOH/water without ZnO (NaOH on the graph). This means that the
addition of ZnO in cellulose/NaOH/water solutions does not improve their stability with these
b and c methods. On the other hand, gelation is delayed if using method a. ZnO must be
dissolved in 12%NaOH solution and then this solvent added to cellulose to influence the
properties of cellulose solutions. The microscopic observations (Fig.IV.2-2) showed that in b
_____________________________________________________________________________________________________
- 162 -
procedures ZnO is not well dissolved and the ZnO particles stay in suspension, while in method
a all ZnO was dissolved.

Let us try to explain the influence of the preparation method on gelation temperature. It must be
linked to ZnO degree of dissolution. It should be noted that the dissolution of ZnO in aqueous
solutions is rather difficult and strongly depends on pH values [LIU1998]. It is not soluble at all
in water and the dissolution increases with the medium becoming either strongly basic or acid.
Since all the four mixtures of Fig.IV.2-1 have the same pH (about 12), pH is not the reason
explaining the reported differences. Most probably these differences can be explained by a
competition of different components during dissolution, for example, the competition between
ZnO and cellulose for water or for NaOH. The competition for NaOH seems to be the most
probable since we have seen in the previous part (Chapter III) that the dissolution of ZnO in
alkali solution forms zinc hydroxides.

a) 5 B342 cellulose / 7.6NaOH / water solution

b) 5 B342 cellulose / 7.6NaOH / 0.7 ZnO / water solution
Method a: ZnO dissolved in NaOH/water
c) 5 B342 cellulose / 7.6NaOH / 0.7ZnO / water solution
Method b: ZnO added to cellulose/NaOH/water

Fig.IV.2-2: Observations with optical microscope of 5 B342 cellulose / 7.6NaOH / water (a),
5 B342 cellulose/7.6NaOH/0.7ZnO/water, prepared from method a (b), and from method b (c)

When cellulose is added first to NaOH/water, it dissolves with a coating of hydrated ions around
the cellulose chain. Adding ZnO in this solution is not changing the organisation of Na ions
around the chain, and thus not changing the dissolution state. On the contrary, the addition of
ZnO to NaOH/water before the introduction of cellulose is creating new chemical species, as
seen in Chapter III with the temperature decrease or the disappearance of the NaOH/water
eutectic peak. This new solvent is better for dissolving cellulose than pure NaOH/water.
In the following, the Innovia Films dissolution procedure (method a) will be used.
_____________________________________________________________________________________________________
- 163 -
IV.2.2- Influence of ZnO concentration on gelation and cellulose dissolution
In this section, the influence of ZnO content on the stability of cellulose solution (or shift of
gelation temperature) and cellulose dissolution is studied. In 100g of solution, NaOH and B342
cellulose contents were fixed at 7.6g and 5g. ZnO amount varied from 0g to 3.9g.

As in the previous section, G and G were measured as a function of temperature and gelation
temperature and G at gel point were plotted versus ZnO concentration (Fig.IV.2-3).

0
20
40
60
80
0 1 2 3 4
m(ZnO), g
TC gel
0,1
1
10
G'=G'' (Pa)

Fig.IV.2-3: Influence of ZnO content on gelation temperature Tgel ( ) and strength of the gel when G=G ( )
of 5 B342 cellulose/7.6NaOH/X ZnO/water solutions (X = 0, 0.4, 0.7, 1, 1.5, 2.2 and 3.9)

This graph shows that there are two regions of ZnO content:
The first one, from 0 to around 1.5g ZnO, reveals an increase of the gelation temperature
with the increase of ZnO content. For example, the gelation temperature of 5 B342
cellulose/7.6NaOH/water without ZnO is about 30C whereas with 1.5g ZnO it is 65C.
Simultaneously, the gel strength decreases by a factor 5.
Above around 1.5g ZnO, the gelation temperature is constant (65-70C) whatever the ZnO
content is.
Let us compare the results obtained with cellulose dissolution in the presence and absence of
ZnO and also with ZnO solubility.

The solutions were observed with an optical microscope and the following pictures were
obtained (Fig.IV.2-4).

The first picture (Fig.IV.2-4a) corresponds to the solution of 5 B342 cellulose/7.6NaOH/water
without ZnO and on the others pictures the quantity of ZnO is increased. If we look at the
pictures from a) to c), it is clear that the number of non-dissolved fibres or parts of fibres
decreases with the increase of ZnO concentration. Pictures c and d are very similar, showing a
good dissolution of cellulose. Picture d has more dark points that are non-dissolved ZnO
_____________________________________________________________________________________________________
- 164 -
aggregates. Whatever the ZnO concentration is, the cellulose dissolution is not complete.
Fig.IV.2-4 shows that small cylinder-like parts of the fibres remain non-dissolved and un-swollen.
These probably correspond to the un-swollen sections of the fibres as described in [CUI2006a]
and [CUI2006b] (see Chapter I).

a)

b)
c)

d)

Fig.IV.2-4: Observation with optical microscope of 5 B342 cellulose/7.6NaOH/water solutions containing
a) 0g ZnO, b) 0.7g ZnO, c) 1.5g ZnO, and d) 2.2g ZnO

From these microscopic observations, we can say that the cellulose dissolution is better when
adding ZnO. But it seems that there is a threshold ZnO content above which the cellulose
dissolution is not improved anymore (Fig.IV.2-4 c and d). This limit is around 1.5g of ZnO. In
addition, we can note that in 7.6NaOH/1.5ZnO/water, the ratio NaOH/ZnO is 10, which
corresponds to the limit of ZnO solubility determined in Chapter III. Experimentally, we
observed that above 1.5g of ZnO in 7.6NaOH/water, some ZnO particles are not dissolved and
stay in suspension.

If correlating visual observations with the results obtained on gelation temperature, one can
conclude that with the increase of ZnO content up to 1.5, zinc oxide is completely dissolved in
7.6NaOH/water and higher is the ZnO content, higher the gelation temperature and lower the
gel strength. Above 1.5ZnO, zinc oxide is not fully dissolved and thus the gelation temperature
remains constant. The non-dissolved ZnO is not playing any role.

To conclude, we can say that the optimum ZnO content giving the maximum of cellulose
dissolution and the best solution stability (highest gelation temperature) is linked to the maximum
of ZnO solubility in the NaOH/water mixture.
_____________________________________________________________________________________________________
- 165 -
IV.2.3- Dissolution of cellulose in a solvent with a constant molar ratio
ZnO:NaOH = 1:10
As demonstrated in the previous section, the best dissolution of Borregaard cellulose is reached
with the 7.6NaOH/1.5ZnO/water solvent, which corresponds to a ZnO:NaOH molar ratio of
1:10. The goal of this section is to increase ZnO concentration keeping the optimal ratio
ZnO:NaOH = 1:10, hoping that this could improve cellulose dissolution because of keeping
ZnO dissolved. To do this, we fixed the ratio ZnO:NaOH = 1:10 and increased NaOH and ZnO
concentrations simultaneously.

The results were not satisfactory since 15NaOH/3ZnO/water did not dissolve cellulose at all
even at 6C. The reason is that as soon as NaOH concentration is higher than 10% it reaches
the upper limit of the cellulose dissolution domain [SOB1939]. The presence of ZnO does not
increase the range of NaOH concentration able to dissolve cellulose.

IV.2.4- Influence of the presence of ZnO on gelation kinetics
In the previous sections, the evolution of G and G as a function of temperature was studied in
order to rapidly determine gelation temperature and thus to compare the solvents of cellulose in
terms of solutions stability (gelation). In this section, the gelation kinetics of
5cellulose/7.6NaOH/water and of 5cellulose/7.6NaOH/0.7ZnO/water solutions was
investigated at different temperatures (Fig.IV.2-5). Frequency and stress were fixed at 0.1Hz and
0.01Pa, respectively, in order to be in the linear visco-elastic regime. The gel point is defined as
being the intercept between the G and G curves [WIN1986].

Fig.IV.2-5 is an example of curves obtained. The gel point is directly indicated on the graphs. The
results on gelation time as a function of temperature and solvent used are summarised in
Fig.IV.2-6.

The most common trends of moduli evolutions are as follows.
At the beginning of the experiment, at +10C/+20C, the viscous component, G, is higher than
the elastic one, G. This reflects a classical liquid-like behaviour of a polymer solution. At a
certain time, depending on temperature and presence of ZnO, both moduli start to increase. G is
increasing faster than G, the two parameters G and G crossing at a certain time. This time
defines the gelation time t
gel
. At this time, the solution is a mixture of a gel phase and a sol phase.
Above this gelation time, the solution is more elastic than viscous.

With gelation proceeding further, a strong data scattering of the elastic and viscous components
can be seen. The scattering of the viscous component G is more pronounced and appears at a
shorter time than the scattering of G. This scattering is probably due to a phase separation
during gelation.
_____________________________________________________________________________________________________
- 166 -
d)
0,01
0,1
1
10
100
0 500 1000 1500 2000 2500
Time (min)
G
'

a
n
d

G
'
'

(
P
a
)
0,01
0,1
1
10
t
a
n

tgel ~ 1h (~66 min)
G'=G" ~ 0.55 Pa
c)
0,001
0,01
0,1
1
10
100
1000
0 1000 2000 3000 4000
Time (min)
G
'

a
n
d

G
'
'

(
P
a
)
0,01
0,1
1
10
t
a
n

tgel ~ 42h (~1.8d)
G'=G" ~ 0.19 Pa
a)
0,01
0,1
1
10
0 1000 2000 3000 4000
Time (min)
G
'

a
n
d

G
'
'

(
P
a
)
a
n
d

t
a
n

no gelation
at the end of 3 days
G'
G''
tan
b)
0,01
0,1
1
10
0 1000 2000 3000 4000
Time (min)
G
'

a
n
d

G
'
'

(
P
a
)
a
n
d

t
a
n

tgel ~ 58h (~2.5d)
G'=G" ~ 0.6 Pa

Fig.IV.2-5: Rheological measurement G ( ), G ( ) and tan ( ) as a function of time for
a) 5 B342 cellulose/7.6NaOH/0.7ZnO/water, at 10C; b) 5 B342 cellulose/7.6NaOH/water, at 10C;
c) 5 B342 cellulose/7.6NaOH/0.7ZnO/water, at 20C; d) 5 B342 cellulose/7.6NaOH/water, at 20C
f = 0.1 Hz and = 0.01 Pa

Gelation time versus temperature
5% cellulose DP342 / NaOH with or without ZnO
aqueous solution
0
1000
2000
3000
4000
5000
6000
5 10 15 20 25 30 35 40 45
TC
t(gel), min
no gelation at the
end of 3 days
tgel < 2min tgel < 2min

Fig.IV.2-6: Gelation time versus temperature and values of the gel point
for 5 B342 cellulose/7.6NaOH/water ( ) and 5 B342 cellulose/7.6NaOH/0.7ZnO/water ( ) solutions

_____________________________________________________________________________________________________
- 167 -
Without ZnO, the increase of G starts right at the beginning of the experiment, even at 10C
(Fig.IV.2-5, b). The solution is gelling very quickly at 20C and higher. The presence of ZnO
delays the start of gelation. At 10C, 20C and 30C, there is a noticeable period of time from the
beginning of the experiment where the solution is still in its liquid state. With time, even in the
presence of ZnO, gelation starts. Gelation time decreases with the temperature increase for both
solvents (Fig.IV.2-6). It was not possible to measure gelation time of this cellulose solution in the
presence of 0.7ZnO at +10C because of a very slow kinetics. The values of G at the gel point is
much lower when ZnO is present, see table on the Fig.IV.2-6.

The similar shape of the two curves in Fig.IV.2-6 will be discussed together with the other results
in the Discussion section concerning the mechanism of cellulose solution gelation.

IV.2.5- Influence of ZnO on cellulose intrinsic viscosity
Intrinsic viscosity is an important molecular parameter that reflects the size of the macromolecule
in a certain solvent and thus gives the information on solvent thermodynamic quality. The
intrinsic viscosity of Avicel microcrystalline cellulose dissolved in 9%NaOH/water as a function
of temperature was studied in ref [ROY2003]. In this section the intrinsic viscosity of a steam
exploded cellulose (B342) was measured in 7.6NaOH/water with and without 0.7ZnO, also as a
function of temperature.

y = 13423x + 101,99
y = 9845,7x + 131,67
100
120
140
160
180
0,000 0,001 0,002 0,003 0,004 0,005 0,006
Cellulose concentration (g/cm3)
r
e
d
u
c
e
d

v
i
s
c
o
s
i
t
y

(
c
m
3
/
g
)
NaOH without additives
NaOH + ZnO

Fig.IV.2-7: Determination of intrinsic viscosity, at 25C,
for B342 cellulose/7.6NaOH/water ( ) and B342 cellulose/7.6NaOH/0.7ZnO/water ( )
The nature of solvent is indicated close to the curve, intrinsic viscosity is the value in the rectangle.

First, the reduced viscosity (see details in Chapter II) of cellulose solutions was measured as a
function of cellulose concentration during dilution with the corresponding solvent,
7.6NaOH/water or 7.6NaOH/0.7ZnO/water, at different temperatures. An example is shown in
_____________________________________________________________________________________________________
- 168 -
Fig.IV.2-7 for 25C together with the equation corresponding to the least square approximation
for each solution. The intrinsic viscosity is the value of the reduced viscosity at C
cell
0.
Fig.IV.2-8 represents the intrinsic viscosity-temperature dependence for solutions in
NaOH/water, in NaOH/ZnO/water and in NaOH/urea/water.

For a given polymer in a given solvent, a decrease of the intrinsic viscosity means that the
polymer chain is decreasing its hydrodynamic radius. Intramolecular links are favoured and the
solvent quality decreases. On the Fig.IV.2-7, we can see that at 25C the intrinsic viscosity of a
cellulose solution containing ZnO is higher (132cm
3
/g) than the one of cellulose solution in pure
NaOH (102cm
3
/g). This indicates that NaOH/ZnO/water is a better cellulose solvent than
NaOH/water.

The dependence of the intrinsic viscosity of cellulose/7.6NaOH/water in the presence and
absence of 0.7ZnO is plotted as a function of temperature in Fig.IV.2-8. At all temperatures, the
intrinsic viscosity of solutions in the presence of ZnO is higher than the one of solutions without
it. On the same figure, data for cellulose/7.6NaOH/6urea/water solutions are also shown. They
will be discussed in the section dealing with the influence of urea on cellulose gelation and
dissolution.

0,7
0,9
1,1
1,3
1,5
10 15 20 25 30 35 40
TC
[

]

(
d
l
/
g
)
B342 in NaOH+Zn0
B342 in NaOH
B407 in NaOH+urea

Fig.IV.2-8: Intrinsic viscosity (dl/g) versus temperature for B342 cellulose/7.6NaOH/water ( ), B342
cellulose/7.6NaOH/0.7ZnO/water ( ) and B407 cellulose/7.6NaOH/6urea/water ( )

With or without additives, the intrinsic viscosity decreases when temperature increases. The same
result has already been obtained for Avicel/9%NaOH/water solutions [ROY2003]. This is
compatible with the fact that upon increasing temperature, gelation occurs (Fig.IV.2-6). The
solvent is thermodynamically worse when temperature increases. We can also see that the trend
of intrinsic viscosity versus temperature curves is the same for the three studied solvents.

_____________________________________________________________________________________________________
- 169 -
IV.2.6- Discussion on the influence of ZnO on gelation of steam exploded
cellulose in 7.6NaOH/water solutions
The purpose of this part is to summarise the results obtained on the role of zinc oxide on
gelation of cellulose/7.6NaOH/0.7ZnO/water solutions and to compare them with the gelation
of cellulose solutions without ZnO.

The intrinsic viscosity (Fig.IV.2-8) and gelation time (Fig.IV.2-6), both as a function of
temperature, show that the curves with and without ZnO have the same shape. It is thus possible
to plot master curves, i.e. to translate by a certain temperature shift the intrinsic viscosity and the
gelation time to have the curves with and without ZnO coinciding. The master plots are shown
on Fig.IV.2-9 and Fig.IV.2-10.

0,5
0,7
0,9
1,1
1,3
1,5
1,7
0 10 20 30 40
TC
[ ], dL/g

0
500
1000
1500
2000
2500
3000
3500
4000
0 10 20 30 40
TC
t(gel), min

Fig.IV.2-9: same as Fig.IV.2-8 but with a -11C shift of the
intrinsic viscosity of cellulose/7.6NaOH/0.7ZnO/water ( ); the
reference being the cellulose/7.6NaOH/water solution ( ).
Fig.IV.2-10: same as Fig .IV.2-6 but with a -11C shift of the
gelation time of cellulose/7.6NaOH/0.7ZnO/water ( ); the
reference being the cellulose/7.6NaOH/water solution ( ).

The remarkable effect is that these two different phenomena need the same shift factor. When
ZnO is present in the solution there is a high energy barrier to overcome for the chains to start
cross-linking. This is due to the ZnO-NaOH-water complexes that are surrounding the chains of
cellulose and that are slowly weakening with increasing temperature. It is this same effect that is
controlling the main conformation of the cellulose chain, expanding more and having a larger
intrinsic viscosity in the presence of ZnO. It is thus understandable that it is the same shift factor
that is found in the variations of both the gel point and the intrinsic viscosity to bring the results
of cellulose/NaOH/ZnO/water coinciding to the ones of cellulose/NaOH/water.
As G and G moduli depend on the frequency of the measurements, Frey et al. [FRE1996]
proposed to plot the tan versus time dependence at various frequencies for a
12%cellulose/ammonia/ammonium thiocyanate solution. Thus they determined the gel point as
the intercept point of all the curves. Consequently, the gel point is independent of the frequency
used in the experiment and only depends on the materials.
We performed similar experiments: the tan versus time dependence at various oscillation
frequencies was measured for 5 B342 cellulose/7.6NaOH/water solution at 20C and for 5 B342
_____________________________________________________________________________________________________
- 170 -
cellulose/7.6NaOH/0.7ZnO/water solution, at 35C, with a stress of 0.01Pa. The temperature
difference between the two experiments is due to the difference of the gelation kinetics for
cellulose solutions with and without ZnO.

Using this method, the gel point is obtained in less than 1 minute which does not correspond to
the reality. Moreover, there is no distinction between the two solvent systems. This test is not
exploitable in this way. On the other hand, if we determine the gel time as the intersection of
both G and G moduli, it is possible to draw the gel time t gel (Fig.IV.2-11a), as well as the gel
strength G when G=G (Fig.IV.2-11b), versus the frequency of the test. Fig.IV.2-11 a and b
represent the experimental results while c and d correspond to the master plot. The gelation
time obtained for cellulose/NaOH/ZnO/water was multiplied by 9 to coincide with the one
obtained for cellulose/NaOH/water. Similarly, the strength of the gel at the gel point (G value
when G=G) for solution containing ZnO was multiplied by 1.5 to join the one of solution
without ZnO. This could also be helpful to understand the gelation mechanisms.

0
20
40
60
80
100
0 0,2 0,4 0,6
frequency, Hz
t(gel), min

0,0
0,4
0,8
1,2
1,6
2,0
0 0,2 0,4 0,6
frequency, Hz
G' when G'=G''
(gel point)

a) initial data of t (gel) versus frequency

b) initial data of G (when G=G) versus frequency
0
20
40
60
80
100
0 0,2 0,4 0,6
frequency, Hz
t(gel), min

0,0
0,4
0,8
1,2
1,6
2,0
0 0,2 0,4 0,6
frequency, Hz
G' when G'=G''
(gel point)

c) vertical shift of gelation time in NaOH+ZnO.
Here gelation time in NaOH+ZnO is multiplied by 9.
d) vertical shift of G (at the gel point) in NaOH+ZnO.
Here G in NaOH+ZnO is multiplied by 1.5.
Fig.IV.2-11: Comparison of gelation time (on the left-hand side) and strength of the gel when G=G (on the right-hand
side) obtained from G and G versus time measurements at different frequencies ( =0.01Pa)
for 5 B342 cellulose/7.6NaOH/water at 20C and for 5 B342 cellulose/7.6NaOH/0.7ZnO/water at 35C

_____________________________________________________________________________________________________
- 171 -
The master plots (Fig.IV.2-11 c and d) show that the evolution of the gel point (gelation time and
G at G=G) versus frequency is similar in the two cases, cellulose/7.6NaOH/0.7ZnO/ water
and cellulose/7.6NaOH/water. This means that ZnO delays gelation but that the gelation
mechanism is the same one with or without ZnO. The way cellulose chains cross-linked by most
probably forming hydrogen bonds does not depend on the presence of ZnO.

IV.2.7- Conclusions
Summarising the results obtained in Sections IV.2.1 IV.2.6 on the influence of the presence of
ZnO on cellulose dissolution and solution stability (gelation), we can conclude that the increase
of ZnO content from 0 to the limit of ZnO solubility (around 1.5 in 7.6NaOH/water which
corresponds to a ZnO:NaOH molar ratio of 1:10), leads to the following phenomena:
the gelation temperature is shifted towards higher temperatures (+ 30-40C),
gelation time occurs several hours later
and cellulose dissolution is improved.
The higher is the ZnO content up to 1.5g in 100g of solution containing 7.6g of NaOH, the
better is its effect to delay cellulose gelation. Above a ZnO content of 1.5, the excess of zinc
oxide particles are not dissolving. They stay in suspension and ZnO does not influence anymore
either cellulose dissolution or gelation. However, even at the optimal ZnO concentration steam-
exploded cellulose is not completely dissolved.
The improvement of cellulose dissolution is thus effective only if ZnO is completely dissolved in
sodium hydroxide/water, having formed ions such as Zn(OH)
3
-
and Zn(OH)
4
2-
. We saw in
Chapter III that the formation of these ions changes the quantity of free water present in
solution: higher ZnO concentration, higher the number of water molecules surrounding zinc
hydroxide ions (up to a limit of 25 water molecules per one ZnO).

At the same time the mechanism of cellulose gelation in the presence of ZnO is the same as
solution without ZnO (7.6NaOH/water). The fact that the shift factor for gelation time and
intrinsic viscosity is the same suggests that ZnO complexes are coating the cellulose chains in the
same way as NaOH alone is doing. It is just more effective.

The following hypothesis can be made:
The amount of free water molecules present in solution is an important factor influencing
cellulose dissolution. Keeping the NaOH concentration in cellulose dissolution range
defined by Sobue in 1939, lower is the quantity of free water, better are the cellulose
solubility and the solution stability.

_____________________________________________________________________________________________________
- 172 -
IV.3- GELATION AND DISSOLUTION OF CELLULOSE IN THE PRESENCE OF UREA.
COMPARISON BETWEEN UREA AND ZNO
In this section the influence of urea on gelation will be studied using dynamic rheology. The
gelation temperature is determined as a function of the dissolution procedure and urea
concentration. As previously, the rheological results are accompanied by visual observations of
cellulose dissolution in the same solvent performed by optical microscopy.
The optimal concentration of urea in cellulose alkali solutions for delaying gelation is determined
as a function of DP and origin of the cellulose pulp. Results are compared with the one obtained
for cellulose/NaOH/ZnO aqueous solution, described above.

In 100g of solution, the NaOH and cellulose contents were fixed at 7.6g and 5g, respectively; the
urea content was varied from 0 to 20g.

IV.3.1- Influence of the preparation conditions on cellulose dissolution and
gelation in the presence of urea, and of urea concentration
Literature described several procedures allowing a total cellulose dissolution in
NaOH/urea/water and stable solutions, not gelling with the temperature increase [ZHO2000]
[CAI2005] (see part I.2.3). The aim of this section is to find the best procedure of solution
preparation in terms of cellulose dissolution and delay of gelation. The results are compared with
cellulose dissolution and gelation in the presence of 0.7ZnO.

Four methods (see description in the first column of Fig.IV.3-1) were tested to prepare 5 B407
cellulose/7.6NaOH/6-12urea/water solutions. Optical microscopy was used to visualise the
amount of non-dissolved fibres. In the first line of this figure, cellulose solutions prepared with
Innovia Films procedure (IF in the following) without urea are shown as a reference:
5cellulose/7.6NaOH/water (photo a), 5cellulose/7.6NaOH/0.7ZnO/water (photo b) and
5cellulose/7.6NaOH/1.5ZnO/water (photo c). Photos from d to i show cellulose solutions in
the presence of 6g or 12g of urea as a function of the procedure used.

The following conclusions can be made:
Qualitatively, there is no significant improvement in cellulose dissolution with the presence of
urea whereas ZnO clearly decreases the amount of non-dissolved cellulose fibres.
Zhang and Ca procedures (photos g-i) seem to be the least effective: more cellulose
fibres remain non-dissolved as compared with IF and Cemef procedures (photos d-f).
Cellulose solutions with the presence of 12g of urea are gelling immediately at -6C whatever
is the dissolution procedure (photos e, h, and i). That is why the background on the photos is
blurred.

According to the results obtained, Ca procedure was immediately abandoned.

_____________________________________________________________________________________________________
- 173 -

Photo a
5 B407 cellulose/7.6NaOH/water
(IF procedure)

Photo b
5 B407 cellulose/7.6NaOH +
0.7ZnO/water (IF)

Photo c
5 B407 cellulose/7.6NaOH +
1.5ZnO/water (IF)

5 B407 cellulose/
7.6NaOH/6urea/water
5 B407 cellulose/
7.6NaOH/12urea/water
Innovia Films (IF) procedure (described
in Chapter II)

2h pre-cooling:
NaOH/urea aqueous solution, -6C
Cellulose/water, +5C
Mixing all the components
2h stirring at -6C, ~1000rpm

Photo d

Photo e
Cemef procedure

2h pre-cooling:
NaOH aqueous solution, -6C
Urea aqueous solution, +5C
Cellulose/water, +5C
Mixing all the components: first,
cellulose/water with NaOH aqueous
solution and then adding urea
aqueous solution
2h stirring at -6C, ~1000rpm

Photo f

Zhang procedure [ZHA2002]

Mixing: Cellulose/NaOH/urea/water
at room temperature for 5min
4h storage, -6C
1h stirring at +4C, ~1000rpm

Photo g

Photo h
Solution gelling at -6C (preparation)
Ca procedure [CAI2005]

2h pre-cooling:
NaOH/urea aqueous solution, -10C
Mixing with cellulose
30min stirring at room temperature,
~1000rpm

Photo i
Solution gelling at -6C (preparation)
Fig.IV.3-1: Optical micrographs of 5 B407 cellulose/7.6NaOH/water solutions,
with ZnO or urea, prepared with various procedures
_____________________________________________________________________________________________________
- 174 -
In order to decide between the three other procedures (IF, Cemef and Zhang procedures), G
and G of cellulose/7.6NaOH/water solutions in the presence of urea of different contents
versus temperature were measured and gelation temperature T
gel
was determined. The goal is to
choose the preparation procedure that leads to the highest gelation temperature to be used in the
following. Fig.IV.3-2 shows the gelation temperature T
gel
versus urea content for the three
procedures mentioned above: IF, Cemef and Zhang

First of all, the measurement of gelation temperature shows that the addition of urea improves
the stability of solution up to 6g of urea. From Cemef and IF procedure, the solution turns
into a gel at 40C with 6g of urea instead of 30C without it. Moreover, taking into account the
experimental errors, the gelation temperatures are identical for cellulose solutions prepared from
Innovia films or from Cemef procedure.
If using Zhang procedure, urea also increases the gelation temperature. However all values of
gelation temperatures are lower than the ones obtained with IF and Cemef methods. Thus
Zhang procedure was also abandoned.

Secondly, if adding 12g of urea, the solution is gelling immediately after the preparation, for the
three dissolution procedures. In this case the gelation temperature is noted as being -6C.

With 8g of urea, some lumps of non-dissolved cellulose remain after the preparation. Filtration of
such a solution should lead to a significant decrease in cellulose concentration and thus the gel
point was not measured.

-10
0
10
20
30
40
0 2 4 6 8 10 12
m(urea), g
TC gel
lumps of cellulose,
m(cell)<5g
'Zhang'
procedure
'Cemef'
procedure
'Innovia Films'
procedure

Fig.IV.3-2: Influence of urea content and dissolution procedure on gelation temperature
for 5 B407 cellulose / 7.6 NaOH / urea aqueous solutions
Error bars correspond to 5%

_____________________________________________________________________________________________________
- 175 -
In the following sections, the solutions containing urea will be prepared using Cemef
procedure, that is to say that NaOH aqueous solution is pre-cooled at 6C, urea aqueous
solution and cellulose/water are pre-cooled at +5C. All the components are mixed
(cellulose/water with NaOH/water and then with urea/water) and then stirred for 2 hours at
6C.

IV.3.2- Influence of cellulose DP and origin on the dissolution and gelation in
the presence of urea and ZnO
Because of a significant difference between literature and our experiments obtained on cellulose
dissolution in the presence of urea (Cai and Zhang procedures not effective), we tested
cellulose from various origins and different DP. In addition, several urea contents from 0 to 20g
were also used, NaOH content being always fixed at 7.6g. To characterise the cellulose solutions,
G and G versus temperature were measured to determine the gelation temperature. The results
obtained were compared with those of 5cellulose/7.6NaOH/water and
5cellulose/7.6NaOH/0.7ZnO/water solutions.

Gelation temperature was plotted versus DP for several solvent systems (Fig.IV.3-3) and various
cellulose origins (Fig.IV.3-5).

-10
10
30
50
70
100 150 200 250 300 350 400 450 500 550
DP
TC gel
5cell/7.6NaOH
5cell/7.6NaOH/6urea
5cell/7.6NaOH/12urea
5cell/7.6NaOH/16urea
5cell/7.6NaOH/0.7ZnO
5cell/7.6NaOH/0.7ZnO/6urea
Avicel Borregaard
(steam exploded)
L307 B530
-10
10
30
50
70
100 150 200 250 300 350
DP
TC gel
L198II Avicel
L198II: no dissolution
in 7.6NaOH/water
L307

Fig.IV.3-3: Gelation temperature versus cellulose DP for various solvent systems.
Details of cellulose are indicated on the graph.

Several conclusions can be made from the analysis of this graph:
_____________________________________________________________________________________________________
- 176 -
In NaOH/water, NaOH/6urea/water, NaOH/0.7ZnO/water, NaOH/0.7ZnO/6urea/
water:
o Solutions of Avicel microcrystalline cellulose and of all the Borregaard steam
exploded cellulose (from B342 to B500) are reaching the same gelation temperature:
the DP does not influence the occurrence of gelation and gelation temperature
increases when additives are added. The best solution stability (highest gelation
temperature) is reached when cellulose is dissolved in 7.6NaOH/0.7ZnO/6urea/
water.
o Solutions with non-steam exploded Borregaard cellulose (B530) turn into a gel at
lower temperatures than those with steam exploded cellulose. The B530 pulp was
treated using a standard pulping process and elemental chlorine free (ECF) bleaching;
this difference in treatment of initial pulp may explain the results. Moreover, the
fibres of this sample are relatively long and thus more difficult to dissolve than small
cellulose fibres.
o Solutions of Lenzing cellulose I (L307) demonstrate higher gelation temperature than
the one obtained with steam exploded cellulose.
Solutions of cellulose with DP higher than 370 and containing 12g or 16g of urea are gelling
immediately at 6C during the preparation step. Moreover, it seems that the threshold of
DP, for which gelation is occurring or not during the preparation, depends on the cellulose
origin.
Lenzing cellulose II (L198
II
) has an opposite behaviour as compared with the other pulps
studied here. On one hand, 7.6NaOH/water does not dissolve this cellulose (see optical
micrograph a in Fig.IV.3-4) and 0.7ZnO improves cellulose dissolution but gelation
temperature is lower as compared with the other solutions in the same solvent. On the other
hand, higher is the urea content (up to 16g), better the cellulose dissolution is (see Fig.IV.3-4
picture b: less non-dissolved fibres in 7.6NaOH/6urea/water, and picture c: solution is
translucent and no fibres remain non-dissolved in 7.6NaOH/12urea/water) and higher the
gelation temperature is. This significantly different behaviour is not explained yet but the
results obtained are close to some found in literature [CAI2005]. It is reported that 4% of a
cotton linter of DP700 (cellulose from Hubei Chemical Fiber Co Ltd Xiangfan, China
but there is no detail concerning the treatment of fibres) is not dissolved at all in
7%NaOH/water and is totally dissolved (the solution is transparent) in 7%NaOH/12%urea/
water after 60s.

a) b) c)
Fig.IV.3-4: Optical micrographs of a) 5 L198II cellulose/7.6NaOH/water,
b) 5 L198II cellulose/7.6NaOH/6urea/water, c) 5 L198II cellulose/7.6NaOH/12urea/water. Scale bar represents 100m
_____________________________________________________________________________________________________
- 177 -
Fig.IV.3-5 represents the same data on gelation temperature as shown in Fig. IV.3-3 but versus
urea concentration, for different cellulose origins. This way to show the results allows focusing
on other details. We can identify three groups of curves: the first one is including all Borregaard
cellulose samples (Steam exploded or not), the second one is the Avicel cellulose and the last one
is Lenzing celluloses. The latter cellulose samples show very high gelation temperatures. At least
two reasons can be given: cellulose origin and/or treatment eucalyptus origin and electron
beam irradiation treatment in order to decrease the DP. In general, steam exploded treatment
may give more acidic medium that would change the total pH of the solvent and thus the action
of NaOH and urea on cellulose. This hypothesis could be checked by preparing steam exploded
samples from eucalyptus cellulose, with various DP.
-10
0
10
20
30
40
50
60
70
0 5 10 15 20
m(urea), g
TC gel
B530
B500 Avicel
L307
L198
B407

Fig.IV.3-5: Influence of urea content and origin of cellulose on gelation temperature
for 5 cellulose/7.6NaOH/Y urea/water solutions (Y = 0, 6, 12, 16 and 20)

The results presented in Fig.IV.3-3 and Fig.IV.3-5 showed that gelation noticeably depends on
cellulose origin and pulp treatment. Unfortunately, for the time being, we are not able to explain
these differences. As it will be shown in the following paragraphs, dissolution of different
celluloses also occurs in different ways. Nevertheless, this rheological study demonstrated why
our results on the stability/gelation of steam exploded cellulose solutions are different from
literature data: the results strongly depend on cellulose origin and treatment.

The difference in gelation of celluloses was then correlated with the swelling and dissolution of
the same cellulose fibres in the presence of urea and ZnO with the exception of Avicel that is not
composed of fibres.
Dissolution of cellulose was directly observed in time with an optical microscope equipped with a
CCD camera and a DVD recorder. The duration of each experiment was about 5 minutes; the
details are given in Chapter II. This investigation is qualitative since kinetics depends on the
concentration of cellulose fibres that was not controlled and the degree of dissolution was not
measured. Results are summarised in the Tab.IV.3-1 and main pictures, picked up from movies,
are shown (Fig.IV.3-6).
_____________________________________________________________________________________________________
- 178 -
Solvent

Cellulose
Experimental
test
7.6NaOH/water
7.6NaOH/0.7ZnO
/ water
7.6NaOH/6urea/
water
7.6NaOH/12urea
/ water

Visual
observation
Partial dissolution Partial dissolution
Rheology Tgel~21C Tgel~41C
B530
Cellulose I
Microscopy Swelling
Twisting
Quick swelling

Visual
observation
Partial dissolution Partial dissolution Partial dissolution Gelation
Rheology Tgel~30C Tgel~48C Tgel=38C Tgel~-6C
B500
Cellulose I
Steam exploded
Microscopy
Twisting
Swelling
Twisting
Quick swelling
Dissolution
Twisting
Quick swelling
Dissolution
Weak swelling

Visual
observation
Partial dissolution Dissolution Dissolution Dissolution
Rheology Tgel~43C Tgel~61C Tgel=51C Tgel=57C L307
Cellulose I
Microscopy
Swelling with
ballooning
Swelling
Quick
dissolution
Swelling with
ballooning
Slow dissolution

Visual
observation
No dissolution
Complete
dissolution
Rheology - Tgel=68C
L198
Cellulose II
Microscopy
Nothing
observed

Slow dissolution
Tab.IV.3-1: Summary of the observations of cellulose dissolution at +4C for 4 different cellulose pulps and 4 solvents
Movies started to be recorded just before the addition of solvent and were stopped at the end of about 5min.

B500 cellulose/7.6NaOH/water

Twisting Swelling and beginning of dissolution Swollen fibre

L307 cellulose/7.6NaOH/water

Swelling with ballooning
Reference point

Swelling of balloons followed by
dissolution
Fig.IV.3-6: Two examples of cellulose dissolution observed with optical microscope
t = 20s t = 10s t = 65s
t = 27s t = 40s t = 90s
_____________________________________________________________________________________________________
- 179 -
The results presented in Tab.IV.3-1 and Fig.IV.3-6 show that there is a certain difference in the
swelling behaviour between Borregaard and Lenzing celluloses. Steam exploded Borregaard
cellulose fibres twist and swell before dissolution and Lenzing cellulose fibres form balloons that
are swelling and breaking followed by dissolution.
Even qualitatively, it is difficult to compare the kinetics of dissolution of different fibres.

As far as the influence of solvent on swelling and dissolution is concerned, all fibres studied
dissolve better and quicker in the presence of 0.7ZnO. Urea is acting on cellulose in different
ways depending on cellulose origin:
Borregaard fibres are hardly swelling in the presence of 12urea but the presence of 6urea
seems to increase the cellulose dissolution, at least qualitatively.
Lenzing L307 fibres in 7.6NaOH/12urea/water are swelling slowly (we do not know at what
time the swelling begins, but after 5min, duration of observation, nothing is observed) but are
completely dissolving in the same solvent at the end of 2 hours (visual observation).
Cellulose L198
II
is swelling differently from all samples: it is not swelling in 7.6NaOH/water
and not dissolving at all (this was already mentioned in the previous paragraphs concerning
the study on gelation time), and its swelling is very slow (nothing is observed during the 5
first minutes in contact with the solvent) in 7.6NaOH/12urea/water whereas the dissolution
of fibres is complete at the end of 2 hours (Fig.IV.3-4c).

Despite the differences observed in the dissolution of fibres of different origins, the general
mechanism of their swelling is the same: fibres are first swelling with ballooning or not, then
breaking into pieces that are then dissolving. The difference observed between steam exploded
and Lenzing fibres is caused by their treatment: the surface of steam exploded fibres is damaged
and thus balloons that are preserved by the primary wall which plays a role of a membrane
[CUI2006a] [CUI2006b] cannot be formed. The qualitative difference in the swelling kinetics can
also be explained by the presence of a membrane on the fibre surface: if it is intact, swelling
and dissolution is slower as compared with the case of broken membranes in steam exploded
cellulose.

What remains unclear in cellulose dissolution, as well as in cellulose gelation, is why urea is acting
in a different way on fibres of different origins. The results obtained on swelling and dissolution
of sample L198
II
are very different from all the other ones since these fibres were not swelling
and dissolving in 7.6NaOH/water. Gelation temperatures of this cellulose in different solvent
also did not follow the main trend observed for the other samples.

IV.3.3- Discussion on the influence of urea on gelation and dissolution of
steam exploded cellulose in 7.6NaOH/water solutions and comparison with
the role of ZnO
The results obtained on gelation of cellulose/7.6NaOH/water solutions in the presence of ZnO
or urea and on the dissolution of fibres in these solvents show a lot of similarities. In both cases
_____________________________________________________________________________________________________
- 180 -
ZnO and urea of certain content shift the gelation temperature towards higher values which
means that these additives make aqueous cellulose/NaOH solutions more stable. Steam exploded
fibres are dissolving in the presence of 0.7ZnO or 6urea in a similar way: with homogeneous
swelling followed by breaking into pieces and dissolution.

Another proof of the similar action of ZnO or urea on cellulose can be seen in Fig.IV.2-8 where
the intrinsic viscosity of cellulose as a function of temperature in 7.6NaOH/water, in
7.6NaOH/0.7ZnO/water and in 7.6NaOH/6urea/water is presented. The presence of ZnO
improves solvent quality (see Part IV.2.5-) and so does urea. The shape of the curves for all three
solvents is also very similar: the intrinsic viscosity decreases with the increase of solution
temperature. One can thus conclude that the reason of cellulose/7.6NaOH/water solutions
gelation, with or without additives like ZnO or urea, is the decrease of solvent thermodynamic
quality: higher is solution temperature, worse is the solvent quality leading to cellulose-cellulose
and not cellulose-solvent preferential interactions.

What is the mechanism of the action of ZnO or urea on cellulose/NaOH aqueous solutions? In
Chapter III, we studied the structure of NaOH/urea aqueous solutions (without cellulose) and
we demonstrated the formation of a eutectic mixture containing 1urea+8water molecules. This
means that when urea is present in the solvent, some water molecules are trapped by urea
decreasing the amount of free water in the system. A similar effect was observed for ZnO: it can
keep up to 25 water molecules. The hypothesis made at the end of Section IV.2- is compatible
with the results obtained for solutions in the presence of urea: the presence of free water
molecules in cellulose/NaOH aqueous solutions could decrease their stability.

One hypothesis to explain the delay of gelation is as follows. In 7.6NaOH/water, without or with
6urea, the limit of cellulose dissolution is the same and is reached when m
cell
=m
NaOH
. Thus we can
say that urea does not modify the cellulose/NaOH phase diagram, but just take bad water
molecules. The area of cellulose dissolution in NaOH on the phase diagram is very narrow; this
means that solvent has a bad thermodynamic quality and that the solution is not stable. Free
water molecules may act as nuclei to start phase separation. Urea, and maybe ZnO, may improve
solution stability by keeping water molecules. But if these complexes ZnO+25H
2
O and
urea+8H
2
O are not very stable with the increase of temperature, they are broken and free water
appears inducing gelation/phase separation.

Another explanation of the role of urea in 4%cellulose/7%NaOH/12%urea/water is presented
in literature [CAI2005]: urea being surrounded by 8 water molecules locates between cellulose
chains and thus prevents their aggregation. But this hypothesis may not apply to ZnO in sodium
hydroxide aqueous solution. Although the size of zinc compounds surrounded by 25 water
molecules is probably higher than the one of urea+8water, their amount is much too low to
prevent polymer chain aggregation by locating between all chains.

_____________________________________________________________________________________________________
- 181 -
IV.3.4- Conclusions
The addition of urea to NaOH/water improves the quality of cellulose solvent. This explains a
slight increase in cellulose solubility and a shift of gelation temperature towards higher values.
However, the dissolution of steam-exploded cellulose fibres is not complete: parts of fibres
remain non-dissolved.

The optimal urea content to shift the gelation temperature towards the highest values possible
depends on cellulose origin. For Borregaard steam exploded pulp, it is around 6g while for
eucalyptus pulp treated by Lenzing it is around 12-16g. The best solution stability (highest
gelation temperature) was obtained in the presence of both urea and ZnO.

The mechanism of gelation of aqueous cellulose/NaOH solutions in the presence or absence of
urea or ZnO seems to be the same. The additives slightly increase the thermodynamic quality of
the solvent thus enlarging the temperature interval of solution stability by shifting the beginning
of cellulose-cellulose preferential interactions towards higher temperatures. The role of urea or
ZnO could be to bind water molecules, up to a certain temperature, that serve as a driver for the
phase separation.

IV.4- INFLUENCE OF FREEZING ON RHEOLOGICAL PROPERTIES OF
CELLULOSE/NAOH AQUEOUS SOLUTIONS
In ref. [ISO1998][KUO2005], it is reported that freezing is used to dissolve microcrystalline
cellulose in NaOH and thus that freezing improves cellulose dissolution and makes NaOH a
thermodynamically better solvent. In order to check these data, we performed some experiments.

First of all, we prepared steam exploded cellulose solution following the method with freezing
[ISO1998]: One g of microcrystalline cellulose was suspended in 26.9ml of water. 2.5g of NaOH was then
added, and the mixture was shaken to dissolve the NaOH at room temperature, resulting in a suspension of the
cellulose in an 8.5%NaOH solution. The suspension was cooled at -20C and held at temperature until it became
a solid frozen mass. The frozen liquid was then allowed to thaw at room temperature, and was transformed into a
gel-like mass. Water (20.6ml) was added to the gel-like material and, with gentle shaking, a clear cellulose
solution was obtained. The resulting solution contained 2% cellulose in 5% aqueous NaOH.
But proceeding as above-described leads to a system with a precipitate of cellulose. In our case,
we cannot say that the freezing step improves the steam exploded cellulose dissolution.

Then we compared the rheological properties of steam exploded cellulose solutions prepared
with our standard procedure (described in the Chapter II) with and without freezing.
The first experiment is the determination of intrinsic viscosity for frozen and non-frozen
cellulose solutions. The results are shown on the Fig.IV.4-1. We can see that the intrinsic
viscosity of a non-frozen solution is higher than the one of a frozen solution: the freezing step
does not increase the solvent quality of NaOH/water.
_____________________________________________________________________________________________________
- 182 -

y = 12065x + 115,99
R
2
= 0,9998
y = 10164x + 106,24
R
2
= 0,9988
100
120
140
160
180
0 0,001 0,002 0,003 0,004 0,005 0,006
r
e
d
u
c
e
d

v
i
s
c
o
s
i
t
y

(
c
m
3
/
g
)
without freezing
Freezing at -20C
for 20H

Fig.IV.4-1: Determination of intrinsic viscosity (values in rectangles), at 20C,
for B342 cellulose/7.6NaOH/water solutions without freezing ( ) and after freezing at -20C for 20H ( ).

To study the influence of the freezing step on the state of solution, we measured (i) the viscosity
versus shear rate, at 20C, for frozen and non-frozen cellulose/NaOH aqueous solutions,
without or with 0.7ZnO (Fig.IV.4-2) and (ii) the moduli as a function of time, at 15C (Tab.IV.4-
1).
0,1
1
10
100
0,01 0,1 1 10 100
shear rate (s-1)

(
P
a
.
s
)

Fig.IV.4-2: Viscosity versus shear rate at 20C for 5B342 cellulose/7.6NaOH/water ( ) and
for 5B342 cellulose/7.6NaOH/0.7ZnO/water ( )
just after the preparation (full points) and after a freezing step at -20C for 20H (open points)

_____________________________________________________________________________________________________
- 183 -
5B342 cellulose/7.6NaOH/water 5B342 cellulose/7.6NaOH/0.7ZnO/water
Freezing at -20C 0H 20H 0H 20H
G 0.15 70 0.05 10
G 0.40 16 0.2 3
Tab.IV.4-1: Values of G and G moduli for non-frozen (0H) and frozen (20H) 5B342 cellulose/7.6NaOH/water
and 5B342 cellulose/7.6NaOH/0.7ZnO/water

After the freezing step, we can note that cellulose/NaOH aqueous solutions are in the gel state:
G is largely superior to G and the gel is relatively strong (G=70Pa) (see Tab.IV.4-1). The
measurement in steady-state mode reveals an important shear thinning (Fig.IV.4-2). This result is
in agreement with the decrease of the solvent quality after freezing (Fig.IV.4-1).
On the other hand, the presence of ZnO in NaOH/water delays gelation not only after heating
(see part IV.2) but also after freezing: the viscosity versus shear rate curve of frozen solution
almost superimposes the one of non-frozen solution and the strength of gel is relatively low
(G=10Pa with ZnO instead of 70Pa without ZnO)

We can conclude that the freezing step does not improve the cellulose dissolution but induces
the gelation of solutions. As in the case of heating, the presence of 0.7ZnO in 7.6NaOH/water
delays the occurrence of gelation.

IV.5- TESTS ON GELATION AND DISSOLUTION OF CELLULOSE IN OTHER ALKALI
SOLVENTS: KOH AND NAOH WITH OTHER ADDITIVES
This last section describes the results obtained on cellulose dissolution and gelation in other alkali
solvents (KOH) and in 7.6NaOH/water in the presence of other additives (surfactants and salts).
Some of these compounds were cited in literature as possible cellulose solvents. The goal was to
scan different compounds in order to find if indeed they improve the dissolution of steam-
exploded cellulose and shift gelation temperature towards higher values.

IV.5.1- Testing cellulose dissolution in potassium hydroxide
7.6KOH/water was tested as a solvent of 5 B342 cellulose and compared with 7.6NaOH/water.
The procedure to dissolve cellulose in potassium hydroxide is exactly the same as the one used to
dissolve cellulose in sodium hydroxide (see Chapter II), sodium is replaced by potassium.
Microscopic observations were performed to qualitatively investigate dissolution of cellulose
fibres. The pictures obtained are shown on the Fig.IV.5-1.

_____________________________________________________________________________________________________
- 184 -
a)
b)
c)
Fig.IV.5-1: Observation with optical microscope of
a) 5 B342 cellulose in water; b) 5 B342 cellulose/7.6NaOH/water; c) 5 B342 cellulose/7.6KOH/water

Picture a on the Fig.IV.5-1 represents cellulose fibres in pure water. As at room temperature and
standard pressure water does not dissolve cellulose at all, this is the reference to qualitatively
determine the solvent quality.

After the preparation of cellulose in alkali solutions (2 hours mixing at -6C), we can immediately
note that sodium hydroxide is a better cellulose solvent than potassium hydroxide. Indeed, the
aspect and the morphology of solutions are different (Fig.IV.5-1 b and c).
o First of all, 5cellulose/7.6NaOH/water solutions are translucent and the microscopic
observation (Fig.IV.5-1b) reveals that only a few non-dissolved fibres remain (A). Most of
them have undergone swelling (B) or start of dissolution, which corresponds to the more or
less fuzzy zones (C).
o On the contrary, 5cellulose/7.6KOH/water solutions are opaque and after a few minutes a
noticeable sedimentation can be observed. The solution was homogenised and then observed
using optical microscope (Fig.IV.5-1c). The aspect of the fibres is very different from the
initial fibres since they have been broken. No swollen fibre was seen.

We know from literature that a small fraction of cellulose can be dissolved [DAV1936]
[MAR1941] (see part I.2.2): swelling and dissolution of cellulose fibres are lower in KOH than in
NaOH. Even if we only tested one condition of temperature and alkali concentration, literature
stipulates that this is true for all alkali concentrations.
A
B
C
_____________________________________________________________________________________________________
- 185 -

IV.5.2- Addition of surfactants
A surfactant is added during the sponges fabrication to the cellulose solution. Empirically, the
final properties of sponges were shown to be improved in the presence of surfactant. The aim of
this section is to study cellulose dissolution and gelation in 7.6NaOH/water in the presence of
surfactants (AP, Zonyl FS and Zonyl FSN) in order to better understand their role.
The three surfactants used in this study are soluble in water. They were mixed with the
NaOH/water solvent (as in the case of NaOH/ZnO/water) before being added to the cellulose
pulp. The preparation of solutions remains the same as in pure NaOH/water.

a) b)
a) 5 B342 cellulose/7.6NaOH/water; b) 5 B342 cellulose/7.6NaOH/0.05AP/water

Microscopic observations (Fig.IV.5-2) reveal that the addition of 0.05g AP surfactant does not
improve cellulose dissolution. It even seems that the solubility of cellulose slightly decreases.
In order to quantify the influence of surfactants on cellulose dissolution, intrinsic viscosity of
cellulose/7.6NaOH/water in the presence of 0.05ZonylFSN and of 0.05AP was measured at
25C (Fig.IV.5-3). The intrinsic viscosity values were calculated according to a least square
approximation of experimental data by a straight line; they are indicated in the rectangles on
Fig.IV.5-3.

We can see that for cellulose/NaOH/water solutions with or without surfactant, the intrinsic
viscosity is about 100g/cm
3
. This confirms the previous results: the addition of surfactant does
not improve the quality of solvent.

_____________________________________________________________________________________________________
- 186 -
y = 13379x + 96,526
y = 13423x + 101,99
y = 19213x + 96,18
100
120
140
160
180
200
0.000 0.001 0.002 0.003 0.004 0.005 0.006
r
e
d
u
c
e
d

v
i
s
c
o
s
i
t
y

(
c
m
3
/
g
)
+Zonyl FSN
NaOH without additives
+AP

Fig.IV.5-2: Determination of intrinsic viscosity, at 25C, for B342 cellulose/7.6NaOH/water ( ),
B342 cellulose/7.6NaOH/0.05ZonylFSN/water ( ), and B342 cellulose/7.6NaOH/0.05AP/water ( )
The nature of solvent is indicated close to the curve, intrinsic viscosity is the corresponding value in the rectangle.

In addition to the intrinsic viscosity, another important parameter which characterises cellulose
solutions is its gelation temperature. Oscillatory rheological measurements were performed in the
same way as for cellulose solutions described in the previous parts: G and G were recorded as
a function of temperature, gelation point was determined and the results are summarised in the
Tab.IV.5-1.

5 B342 cellulose / Tgel (C) G=G (Pa)
7.6NaOH/water 31.5 2.5
7.6NaOH/0.05AP/water 30 2.4
7.6NaOH/0.05Zonyl FS/water 32 2.7
7.6NaOH/0.05Zonyl FSN/water 28 2.2
Tab.IV.5-1: Gelation temperature and gel strength for
5 B342 cellulose/7.6NaOH/water solutions with and without surfactant.

Whatever the cellulose solution studied here, with and without surfactant, the gelation
temperature and the gel strength are the same, within the experimental error, about 30C and
2.5Pa, respectively. Thus the addition of surfactant does not improve the stability of solutions.

To conclude, we can say that the addition of non-ionic surfactants, at least the ones studied here,
improves neither the cellulose dissolution nor the stability of solutions. These surfactants are thus
acting during another step of the industrial process. For example, during the rapid mixing stage,
the presence of surfactant induces a lot of tiny air bubbles. This increases the total volume of
pores and thus could improve the absorption properties of sponges.

_____________________________________________________________________________________________________
- 187 -
IV.5.3- Other additives: salts and oxides
We saw that the addition of ZnO improves cellulose dissolution and delays gelation of solutions.
However, because ZnO is not soluble in water and thus cannot be removed during regeneration,
its presence may not be acceptable for certain applications. The aim of this part is thus to test if
other additives improve stability of cellulose solutions, in particular other oxides and salts. In
order to compare the results with the ones of cellulose/NaOH/ZnO aqueous solutions, the
preparation of solutions remain the same but ZnO is replaced by ZnSO
4
, MgCl
2
or CaSO
4
.

The optical micrographs of 5 B342 cellulose/7.6NaOH/water solutions in the presence of
ZnSO
4
and MgCl
2
are shown in Fig.IV.5-4 b and c respectively. These compounds do not
improve cellulose dissolution; more fibres than in pure 7.6NaOH/water remain non-dissolved
(Fig.IV.5-4 a). These additives are not efficient to increase the quality of the dissolution.
CaSO
4

seems to improve cellulose dissolution but it is difficult to dissolve in sodium hydroxide
and a lot of crystals stay in suspension (Fig.IV.5-4d). This way would be interesting to investigate
if knowing the limit of CaSO
4
dissolution in a strong basic solution as 7.6NaOH/water and in
pure water also (important for the removal of the chemicals present in the solution during the
regeneration in the form of sponges).

a) b)

c)

d)
a) 5 B342 cellulose/7.6NaOH/water, b) 5 B342 cellulose/7.6NaOH/0.7ZnSO4/water,
c) 5 B342 cellulose/7.6NaOH/0.7MgCl2/water and d) 5 B342 cellulose/7.6NaOH/0.7CaSO4/water
Scale bar represents 100m.

_____________________________________________________________________________________________________
- 188 -

IV.5.4- Conclusions
The goal of this section was to explore other conditions to dissolve cellulose: other alkali media,
other additives.
First of all, we tested potassium hydroxide as solvent of cellulose: using KOH instead of NaOH
is not efficient; this result is in agreement with literature.
We thus explored the way of additives to improve the cellulose solubility in sodium hydroxide.
We studied metallic salts and oxide as well as surfactants but none of them are efficient to
improve cellulose dissolution.

_____________________________________________________________________________________________________
- 189 -
IV.6- CONCLUSIONS
The detailed investigation of rheological properties, in steady-state and dynamic modes as well as
the study of solutions in dilute state, enables us to better understand the behaviour of cellulose in
sodium hydroxide aqueous solutions:

From viscosity versus shear rate measurements, we concluded that the steam exploded
cellulose solutions in NaOH/water are not molecularly dispersed solutions and show a
behaviour close to the one of a suspension. This behaviour was observed for all cellulose
contents and DP studied here as well as with the presence of ZnO and urea.
Although some cellulose fibres still remain non-dissolved, the addition of ZnO and urea
improves cellulose dissolution (microscopic observations). Intrinsic viscosity measurements
revealed that this improvement is due to an increase of the solvent quality in the following
order 7.6NaOH/0.7ZnO/water > 7.6NaOH/6urea/water > 7.6NaOH/water.
The intrinsic viscosity is decreasing when increasing temperature showing that the cellulose
chains are collapsing. The quality of the solvent decreases. This effect favours cellulose-
cellulose interactions and leads to gelation. The fact that the shift factor of intrinsic viscosity
and gelation time is the same is in full agreement with this explanation.
The addition of zinc oxide in 7.6NaOH/water is more and more efficient up to 1.5ZnO,
which corresponds to its limit of solubility. Beyond this concentration, properties remain
constant. ZnO is acting on cellulose dissolution and stability of solution only if it was
dissolved in NaOH before being added to the cellulose pulp. These results show that
complexes between ZnO and NaOH, like the zinc hydroxyl ions Zn(OH)
3
-
and Zn(OH)
4
2-
,
have to be formed to improve cellulose dissolution and to delay gelation.
The gelation mechanisms and physical reasons are the same with and without the presence of
ZnO. The only difference is that the ZnO-NaOH complexes which are formed during
mixing ZnO, NaOH and water are preventing the cellulose-cellulose interactions to start at
temperature higher than without the presence of ZnO.
The addition of urea also delays gelation but the behaviour of solutions is very different
depending on cellulose origin and urea content. If the urea content is too high, the cellulose
solutions are gelling during the preparation at -6C, but this content depends on cellulose
origin: with the Borregaard cellulose pulps, the limit of urea content is 12g whereas with the
Lenzing cellulose pulps it reaches 20g in 100g of solution. The reasons of such differences
stay an open question. These results explain why we found results different from the ones
reported in the literature, since the dissolution and gelation depend on the cellulose
treatment.

We made the hypothesis in Chapter III that the addition of urea improves the quality of the
solvent by trapping the free water. The same was also thought to be true for the addition of
ZnO. The results of this chapter does not allow to confirm or infirm this hypothesis.
_____________________________________________________________________________________________________
- 190 -
aqueous solutions. Macromolecular Bioscience (2005), 5, p.539-548

[CUI2006a] CUISSINAT C., NAVARD P. Swelling and dissolution of cellulose Part I: free
floating cotton and wood fibres in N-methylmorpholine-N-oxide water mixtures. Macrom. Symp.
To appear

[CUI2006b] CUISSINAT C., NAVARD P. Swelling and dissolution of cellulose Part II: free

alkaline solutions. Part II: A comparison of the solvent action of solutions of Lithium, Sodium,
Potassium and tetramethylammonium hydroxides. J. Text. Ind. (1936), 27, p.T112-130

[FRE1996] FREY M.W., CUCULO J.A., KHAN S. A. Rheology and gelation of
cellulose/ammonia /ammonium thiocyanate solutions. J. Polym. Sci. PartB (1996), 34, p.2375-2381

[GAV2005] GAVILLON R., BUDTOVA T. Internal Cemef report (2005)

[ISO1998] ISOGAI A., ATALLA R.H. Dissolution of cellulose in aqueous NaOH solutions.
Cellulose (1998), 5, p.309-319


[LIU1998] LIU Y., PIRON D.L. Study of tin cementation in alkaline solution. J. Electrochem. Soc.
(1998), 145 (1), p.186-190

[MAR1941] MARSH J.T. Mercerising. London: Chapman & Hall Ltd (1941), 458p

[ROY_PhD] ROY C. Etude de mlanges de cellulose dans des solutions aqueuses de soude. Thse de
Doctorat. Sciences et gnie des matriaux, Sophia-antipolis : Ecole des Mines de Paris (2002),
213p.


[SOB1939] SOBUE H., KIESSIG H., HESS K. The cellulose-sodium hydroxyde-water system as
a function of the temperature. Z. Physik. Chem. B (1939), 43 (3), p.309-328

[TUN1982] TUNG C.M., DYNES P.J. Relationship between viscoelastic properties and gelation
in thermosetting systems. J. Appl.Polym. Sci. (1982), 27, p.569-574

[YAM1990] YAMASHIKI T, MATSUI T, SAITOH M., OKAJIMA K, KAMIDE K, SAWADA
T. Characterisation of cellulose treated by the steam explosion method. Part 1: Influence of
_____________________________________________________________________________________________________
- 191 -
cellulose resources on changes in morphology, degree of polymerisation, solubility and solid
structure. British Polymer Journal (1990), 22, p.73-83

[WIN1986] WINTER H.H., CHAMBON F. Analysis of linear viscoelasticity of a crosslinking
polymer at the gel point. J. Rheol. (1986), 30 (2), p.367-382

[WO 02/22924] MIKOLAJCZYK W., STRUSZCZYK H., URBANOWSKI A., WAWRO D.,
STAROSTKA P. Process for producing fibres, film, casings and other products from modified soluble cellulose.
Poland, Patent n WO 02/22924 (21 mars 2002)

[ZHA2002] ZHANG L., RUAN D., GAO S. Dissolution and regeneration of cellulose in
NaOH/thiourea aqueous solution. Journal of Polymer Science: Part B: Polymer Physics (2002), 40,
p.1521-1529

[ZHO2000] ZHOU J., ZHANG L. Solubility of cellulose in NaOH/Urea aqueous solution.
Polymer Journal (2000), 32 (10), p.866-870

Rsum du chapitre V Proprits mcaniques dobjets rgnrs prpars partir de solutions aqueuses de cellulose/NaOH
_____________________________________________________________________________________________________
- 193 -

R RE ES SU UM ME E D DU U C CH HA AP PI IT TR RE E V V
P PR RO OP PR RI IE ET TE ES S M ME EC CA AN NI IQ QU UE ES S D D O OB BJ JE ET TS S R RE EG GE EN NE ER RE ES S P PR RE EP PA AR RE ES S
A A P PA AR RT TI IR R D DE E S SO OL LU UT TI IO ON NS S A AQ QU UE EU US SE ES S D DE E C CE EL LL LU UL LO OS SE E/ /N NA AO OH H
Lobjectif de ce chapitre est dtudier les proprits mcaniques dobjets rgnrs prpars
partir de solutions aqueuses de cellulose/NaOH, avec ou sans additifs, mais galement dtudier
linfluence de la conglation sur les proprits finales des ponges.

Pour ce faire, nous avons procd tape par tape.

Nous avons dabord prpar des solutions de cellulose/NaOH/eau, sans fibres de renfort, ni
porophores, cest la matrice de cellulose rgnre que nous trouvons ensuite dans les ponges.
Aprs glification, nous les avons rgnres. Les chantillons ainsi obtenus ont t tests
mcaniquement dans leur tat humide afin de dterminer leur contrainte la rupture. Nous avons
fait varier de nombreux paramtres, savoir la prsence ou non dadditifs (ure et ZnO), les
conditions de glification (2 heures +70C, 20 heures +25C ou 20 heures -20C), les bains
de rgnration (eau +5C, +25C ou +70C, ou encore acide actique 10%, +25C).
Il apparat que la rgnration haute temprature (+70C) est trs dfavorable la rsistance la
rupture et quun sjour prolong -20C namliore pas les proprits mcaniques de la matrice
de cellulose rgnre. Les chantillons, prpars au Cemef, ayant la plus haute contrainte la
rupture ont t glifis et rgnrs temprature ambiante.
Bien que les deux additifs utiliss dans cette tude amliorent la dissolution de la cellulose et
retardent la glification, leur effet sur les proprits mcaniques est diffrent. La prsence dure
amliore nettement la rsistance la rupture ce qui nest pas le cas pour ZnO. Des observations
microscopiques rvlent la prsence de particule doxyde de zinc dans les chantillons rgnrs
crant probablement des points de rupture.

Lajout de fibres de renfort dans la matrice de cellulose rgnre entrane une augmentation de la
contrainte la rupture. Mais cette amlioration des proprits mcaniques nest pas considrable,
nous avons donc test diffrentes mthodes pour amliorer ladhsion entre les fibres et la
matrice.
La conglation tant une tape ncessaire lobtention de bonnes ponges, nous avons pens
quun sjour basse temprature agissait sur ladhsion fibres/matrice. Avant dintroduire les
fibres de renfort dans la solution de cellulose/NaOH/eau, nous les avons dabord imprgnes de
soude (8%) diffrentes tempratures (+25C, -6C et -20C). Leur observation microscopique
Rsum du chapitre V Proprits mcaniques dobjets rgnrs prpars partir de solutions aqueuses de cellulose/NaOH
_____________________________________________________________________________________________________
- 194 -
montre un gonflement basse temprature uniquement. Mais les tests mcaniques ne rvlent
aucune influence des conditions de pr-imprgnation sur la contrainte rupture. La conglation
nagit pas sur ladhsion fibres/matrice et il nest donc pas ncessaire dimprgner les fibres de
renfort basse temprature.
Lajout de promoteur dadhsion na pas t probant non plus.

Pour finir, nous avons ajout des porophores au systme matrice de cellulose rgnre/fibres de
renfort et effectu des tests de traction sur les chantillons obtenus, toujours dans leur tat
humide. Les conditions de prparation des ponges au Cemef ntant pas optimales, les
proprits mcaniques sont trs faibles. Nous constatons nanmoins que pour des systmes
complets matrice de cellulose rgnre + fibres de renfort + porophores, ltape de conglation
augmente la contrainte la rupture.
De plus, si nous considrons les proprits dabsorption et de retrait, ltape de conglation est
galement primordiale, bien que les proprits restent largement infrieures celles obtenues
pour des ponges viscose. Les observations au MEB des diffrents chantillons montrent une
grande disparit au niveau de leur porosit. En effet, pour des ponges viscose, la microporosit
est trs importante alors quelle est totalement absente dune ponge soude non congele. Quant
lponge soude congele, la microporosit, bien que faible, est prsente. Ces rsultats
concordent tout fait avec les valeurs obtenues pour labsorption. En ce qui concerne le retrait,
on constate, en effet, que la microporosit des ponges soude congeles se ferme aprs schage
lair ambiant.

Ces rsultats ne nous permettent pas encore de comprendre et dexpliquer le rle de la
conglation mais ils nous ont permis damliorer nos connaissances sur linfluence dun sjour
prolong -20C sur les proprits mcaniques et physiques dobjets rgnrs.

Chapter V Mechanical properties of regenerated objects prepared from cellulose/NaOH aqueous solutions
_____________________________________________________________________________________________________
- 195 -

C
C
H
H
A
A
P
P
T
T
E
E
R
R
V
V

V- v
M ME EC CH HA AN NI IC CA AL L P PR RO OP PE ER RT TI IE ES S O OF F R RE EG GE EN NE ER RA AT TE ED D
O OB BJ JE EC CT TS S P PR RE EP PA AR RE ED D F FR RO OM M C CE EL LL LU UL LO OS SE E/ /N NA AO OH H
A AQ QU UE EO OU US S S SO OL LU UT TI IO ON NS S
This chapter concerns the preparation and some mechanical properties of sponges using direct
cellulose dissolution in aqueous NaOH solutions. Being much less polluting than the viscose
process, it is not yet industrialised because of an intermediate energy consuming step: freezing of
cellulose/NaOH/water solution at -20C for several hours. This step is supposed to be necessary
for obtaining mechanical properties of sponges comparable with the ones from viscose process.
Spontex is making sponges from cellulose/NaOH/water solutions (with the freezing step) on
laboratory scale keeping most of the other processing parameters the same as in viscose process.
The goal of this chapter is to determine the mechanical properties of sponges prepared by the
NaOH process in order to study the influence of different processing parameters (treatment of
reinforcing fibres, freezing step, regeneration bath temperature, etc) on the final properties.

The chapter is divided into two parts. The first gives a brief overview of existing sponge
preparation through viscose and NaOH processes. The second part describes the results
obtained on the mechanical properties of sponges made via NaOH process with the accent on
the different processing parameters.

- 196 -
_____________________________________________________________________________________________________
- 197 -
V.1- SUMMARY ON SPONGE PREPARATION: VISCOSE AND NAOH PROCESSES
V.1.1- Viscose process
One of the ways for manufacturing sponges from cellulose is the viscose process. It consists in
the cellulose derivation and then the dissolution. First, cellulose fibres are swollen in a
concentrated aqueous 18-20%NaOH solution in order to improve their accessibility for other
chemicals, as carbon disulfide in the next step of the process: the xanthation. The sodium
xanthate of cellulose can be dissolved in a dilute 2.7%NaOH solution. As a result, a viscous paste
is obtained; it is called viscose. Then, the reinforcing fibres, porophores (salt crystals) and dyers
are added to the viscose in order to obtain the appropriate mechanical and absorption properties
of sponges. The mass is homogenised, aged, regenerated, shaped, and dried.
The following Fig.V.1-1 schematises sponge manufacturing via viscose process.
1
2
3
4
5
6
7
A B

Fig.V.1-1: The different steps to manufacture cellulose sponges from viscose process.

Viscose formation: The step of forming cellulose xanthate and then viscose is shown in
Fig.V.1-2 and described in part I.2.1.

PULP
CARBON
DISULPHIDE
(RECYCLED)
CELLULOSE
XANTHATE
ALKALI
CELLULOSE
CAUSTIC SODA
VISCOSE
DILUTE CAUSTIC SODA

Fig.V.1-2: Scheme of the xanthation step in the viscose process.
The chemical reactions are described in the Chapter I.
1- Cellulose pulp and viscose formation
2- Addition of reinforcing fibres,
porophores and dryers
3- Shaping
4- Coagulation/Regeneration
5- Washing
6- Drying
_____________________________________________________________________________________________________
- 198 -
Addition of reinforcing fibres and porophores (step 2 on the Fig.V.1-1): In order to improve
the final properties of sponges, reinforcing fibres are added to the viscose. They consist of
long fibres such as cotton or flax fibres, or wood pulp with high DP. The reinforcing fibres
are wetted with water or dilute sodium hydroxide solutions; they are then mixed with viscose
(in the wet state). The viscose/reinforcing fibres system is aged to improve adhesion and
homogenisation.
After the ageing step, a large amount of sodium sulphate Na
2
SO
4
, 10H
2
O crystals is added (3-
4 times of the total weight of viscose/reinforcing fibres). The addition of these salt crystals
enables to form the porous structure of sponges after cellulose regeneration and crystals
removal (melting).
Finally, dyers giving the colour of the sponge are added (regenerated cellulose swollen in
water is white).
Shaping (step 3 on the Fig.V.1-1): the viscose/reinforcing fibres/salt crystals system is
shaped, at room temperature.
Coagulation/regeneration (step 4 on the Fig.V.1-1): After the shaping step, the mixture is
coagulated to give the final cellulose product; two different A and B methods are used and
briefly described below. During this step, cellulose is precipitating around the Na
2
SO
4
,10H
2
O
crystals. CS
2
and other sulphured compounds are released. Almost simultaneously, the
Na
2
SO
4
,10H
2
O crystals are melting (at a temperature superior to 40C) forming a porosity
which is directly dependent on the crystals size (between 100m and 3mm)
The method of vapour coagulation (A) consists in immersing the moulds in a sodium sulphate
aqueous solution progressively heated to 100C. By this way, the coagulation/ regeneration
step lasts approximately 8 hours.
For the electrical coagulation method (B), a voltage from 20V to 120V is applied between the
two opposite walls of the mould. The heating of the mixture is due to the joule effects until
boiling. The duration of the electrical coagulation is around 3 hours.
At the end of the coagulation step, the physical and chemical reactions which enable the
manufacture of cellulose sponges are completed.
Washing, bleaching (step 5 on the Fig.V.1-1): The washing and bleaching stages allow
recovering the sodium sulphate present in sponges (due to the melting of Na
2
SO
4
,10H
2
O
crystals) and to destroy the sulphured compounds causing a blackish colouring of the sponge
blocks. Finally, the blocks are de-moulded and spin-dried.
Drying (step 6 on the Fig.V.1-1): The blocks of sponges are dried at 80C, for 24 hours.
Finally, they are cut to obtain the final product and undergo a fungicidal and plastification
treatment with magnesium chloride MgCl
2
. Sponges can then be packaged.

The viscose process of sponge preparation is very efficient but its major problem is the use of
carbon disulfide. This chemical is very polluting as well as the by-products obtained during the
series of chemical reactions of cellulose transformation into viscose.
_____________________________________________________________________________________________________
- 199 -
The viscose process has already been improved to reduce:
The pollution in aqueous phase: The sulphured compounds formed during the process induce a
decrease of the oxygen amount in water. Nowadays, these compounds are completely recycled
and are not discharged in nature.
The atmospheric pollution: It is due to the release of CS
2
and H
2
S. The latter has an extremely bad
smell even at low concentration. In order to decrease pollution, these gases are absorbed on
active coal which enables to recycle 70% of the carbon disulfide used during the whole process.
Nowadays, the gaseous discharge is limited at 10% of the CS
2
used in the process.

However, treatments of pollution are very expensive and thus the viscose process is less and less
economically interesting. That is why less polluting alternative routes are searched. One of the
possible ways is to dissolve cellulose fibres in sodium hydroxide.

V.1.2- NaOH process
This process is only using the sodium hydroxide as chemical. The dissolution, even partial, is
direct and does not generate polluting by-products. The schematic presentation of the steps of
sponge manufacturing via NaOH process is shown on Fig.V.1-3.

Dissolution (step 1 on the Fig.V.1-3): Dissolution is performed according to the procedure
described in part II.2.2: swelling of cellulose in water, mixing with NaOH/water at -5C
reaching resulting concentrations of 5% cellulose in 7-9%NaOH/water. Depending on
cellulose DP, not all fibres may be dissolved.
Addition of reinforcing fibres (step 2 on the Fig.V.1-3): The quantity of reinforcing fibres is
40% of cellulose weight. For example, if 5% cellulose solution was prepared, 2% of fibres
have to be added. Before being added to the cellulose solution, the reinforcing fibres that are
cellulose fibres with high DP superior to 1000 are pre-impregnated for some minutes
with 8% NaOH, at room temperature. Then these fibres are progressively introduced into the
cellulose solution kept at -5C under mixing. The mixing still lasts 1.5 hour. The system
obtained is called the cellulose fibrous solution.
Ageing (step 3 on the Fig.V.1-3): The cellulose mixture undergoes a step of ageing, as in the
viscose process, at 0C, for about 4 hours. During this ageing, gelation probably starts to
occur.
Addition of porophores (Na
2
SO
4
,10H
2
O crystals) (step 4 on the Fig.V.1-3): The cellulose
fibrous solution/crystals weight ratio is 1/3 or 1/4. There are several sizes of crystals and
their proportions define the macro- and micro-porosity of the final product. Because
cellulose/NaOH solution has to be kept at rather low temperatures in order to avoid gelation
and because the amount of crystals is large, they have to be cooled down before being added
to the cellulose system. The whole system is still maintained at -5C and mixed for 30min.
The cellulose mixture is then put into the moulds.
_____________________________________________________________________________________________________
- 200 -

Fig.V.1-3: The different steps to manufacture cellulose sponges from the NaOH process.

Thermal treatments (step 5 on the Fig.V.1-3): The cellulose solution containing reinforcing
fibres and porophores are frozen for 8 hours at -20C. It appears that freezing is an essential
step to obtain sponges with relatively good mechanical properties. After freezing, the system
is thawed at room temperature.
Coagulation/regeneration (step 6 on the Fig.V.1-3): This is the last stage of the process and is
similar to the one described in the viscose process. The moulds are either immersed in a bath
CELLULOSE
PULP
WATER
SODIUM
HYDROXIDE
WATER
at +5C at -5C
5%cellulose in 8%NaOH/water
DISSOLUTION
1.5H at -5C

1- Dissolution
2- Addition of
reinforcing fibres
CELLULOSE REINFORCING FIBRES
(cotton, flax, wood), 40% of initial cellulose
i.e. 2g of reinforcing fibres in a solution with 5g of cellulose
Mixture of cellulose/NaOH solution and fibres,
named cellulose fibrous solution
3- Ageing
POROPHORES: salt crystals (Na2SO4, 10H2O)
3-4 times the weight of mixture
Mixture of cellulose solution,
cellulose fibres and porophores
THERMAL TREATMENTS
Freezing: 20H at -20C
Thawing: 2H at +20C
5- Thermal
treatments
COAGULATION/REGENERATION
in water at +70C
FINAL PRODUCT:
100% CELLULOSE SPONGES
WASHING, then drying and cutting
6- Coagulation/
Regeneration
7- Washing
AGEING
2H at -5C
4- Addition of
porophores
_____________________________________________________________________________________________________
- 201 -
of warm water (70-90C) or connected to a voltage inducing the precipitation of cellulose
around the crystals of Na
2
SO
4
,10H
2
O.
Washing (step 7 on the Fig.V.1-3): The blocks of sponges are washed with cold water (20C)
for about 8 hours. The aim of this step is to remove the soda.

The final product contains only cellulose.

Industrial problem of NaOH process
The sponges obtained with NaOH process have final properties almost similar to the one
obtained with the traditional viscose process. Moreover, because sodium hydroxide can be
recovered, there is practically no pollution. This is the main advantage of this process. However,
the temperatures used during cellulose dissolution and especially during the freezing step at -20C
makes this process industrially not viable. In the next section the properties of sponges prepared
with and without freezing will be described.

V.2- PROPERTIES OF SPONGES PREPARED WITH THE NAOH PROCESS
This section is devoted to the properties of the wet sponges prepared in Cemef following the
NaOH procedure. Wet samples, just after regeneration, were tested supposing that the trends
obtained can be translated to dry sponges. First, the influence of sponge preparation on wet
sponge mechanical properties is studied. Second, the adhesion between reinforcing fibres and
cellulose matrix is investigated for different fibre treatments, keeping in mind that the adhesion is
one of the important factors controlling the mechanical properties of a composite material. A
special attention is paid to the role of the freezing step. In all cases cellulose and NaOH contents
were kept constant: 5g of cellulose and 7.6g of NaOH in 100g of solution.

The measurements of the mechanical properties were performed on sponges using a step by
step approach, i.e. by studying the properties of sponges on different stages of their preparation:
first step: sponges were made only from cellulose/NaOH solution followed by
regeneration;
second step: sponges were made from cellulose/NaOH solution + reinforcing fibres;
followed by regeneration;
third step: sponges were made from cellulose/NaOH solution + reinforcing fibres +
Na
2
SO
4
,10H
2
O crystals (porophores), followed by regeneration.

In each step, different thermal treatments of cellulose/NaOH solutions were performed:
solutions were either gelled or frozen at -20C. When reinforcing fibres were added, their
different treatments were performed in order to vary their adhesion to cellulose matrix.
_____________________________________________________________________________________________________
- 202 -
Regeneration was performed in water baths of different temperatures, an aqueous acid solution
being also tested. Finally, the influence of additives on the mechanical properties of sponges was
investigated.

V.2.1- Wet samples made from regenerated 5cellulose/7.6NaOH/water
solutions, without reinforcing fibres and porophores
V.2.1.1. Tensile stress results
Cellulose was dissolved in NaOH, with or without additives (urea or ZnO), as described in part
II.2.2. The cellulose solution was poured in a Petri box in order to obtain a film with a 1-1.5mm
thickness. Then, after various thermal treatments (gelation at 25C for 20 hours, gelation at 70C
for 2 hours and freezing at -20C for 20 hours), cellulose solutions were regenerated in different
regeneration baths (in water at 5C, 25C and 70C and in 10% acetic acid at 25C). Finally,
tensile tests were performed on wet regenerated cellulose, until the rupture of the sample (see the
description of tensile measurements in part II.3.5). An example of a typical stress-strain curve is
given on Fig.V.2-1. The rupture stress is directly obtained from the graph as the y-axis is in
daN/cm. The values of rupture stress for all samples studied are shown on Fig.V.2-2 and
correspond to the average of three measurements performed on a same solution.

0
0,4
0,8
1,2
1,6
2
0 4 8 12 16 20
Strain (%)
Stress (daN/cm)

Fig.2-1: Example of a tensile test on a regenerated cellulose solution initially containing 5B407
cellulose/7.6NaOH/water; solution was frozen for 20 hours at -20C and regeneration in water at +70C.
The three curves represent three different experiments.
v=5mm/min.

V.2.1.2. Influence of regeneration bath parameters
Whatever is the nature of the solvent NaOH/water, NaOH/urea/water or NaOH/ZnO/water
regeneration in water at 70C considerably decreases the value of the rupture stress (Fig.V.2-2
a, b and c). Regeneration in water at 70C was studied because in the industrial process,
_____________________________________________________________________________________________________
- 203 -
Na
2
SO
4
,10H
2
O crystals that are playing the role of porophores must be melted at this
temperature in order to be removed from regenerated cellulose

Regeneration in 10% acetic acid is relatively efficient to obtain good mechanical properties,
especially for cellulose dissolved in 7.6NaOH/6urea/water (Fig.V.2-2b). But according to our
industrial partners, the kinetics of regeneration in acid baths is very fast and a skin is thus formed
around the blocks of sponges. Consequently, the non-solvent cannot easily penetrate in the block
of sponge and precipitation of cellulose cannot occur anymore. In the manufacturing of cellulose
films, acid non-solvent can be used since the thickness of cellulose is very small.

Regeneration in a water bath at room temperature seems to be the most efficient in terms of the
mechanical properties and energy consumption.

V.2.1.3. Influence of solution thermal treatment
Four different thermal treatments were tested: 2 hours at +70C, 20 hours at +25C, 20 hours at
-20C and 20 hours at -20C followed by 20 hours at +25C. Fig.V.2-2 shows the different values
of rupture stress obtained with these four treatments, with various regeneration bath conditions.
For the solution without additives and whatever is the regeneration method, the best properties
are obtained when the cellulose solution is kept 20 hours at +25C. In this condition, the
cellulose solution is gelling.

The results show that gelation for 20 hours at 25C after a freezing step at -20C does not
increase the rupture stress up to the level of the sample than was gelled in these conditions
without freezing. The state of solution after freezing/thawing is not the same as the one of a
freshly prepared solution (see part IV-4), and thus gelation of frozen sample did not give the
same result as gelation of a non-frozen solution.

Fig.V.2-2a reveals that gelation at room temperature is the most efficient thermal treatment to
resist to the rupture as compared with gelation at +70C and freezing at -20C.

_____________________________________________________________________________________________________
- 204 -
0
1
2
3
4
water at
+5C
water at
+25C
water at
+70C
acetic acid at
+25C
Regeneration bath
Rupture stress
(daN/cm)
20H at -20C
2H at +70C
20H at +25C
20H at -20C then
20H at +25C
a)
Solution thermal treatment
0
1
2
3
4
water at
+5C
water at
+25C
water at
+70C
acetic acid at
+25C
Regeneration bath
Rupture stress
(daN/cm)
20H at -20C
2H at +70C
b)
0
1
2
3
4
water at
+5C
water at
+25C
water at
+70C
acetic acid at
+25C
Regeneration bath
Rupture stress
(daN/cm)
20H at -20C
2H at +70C
c)

Fig.V.2-2: Rupture stress of wet regenerated B407 cellulose for different solutions, treatments, regenerating baths
Initial solutions: a) 5cellulose/7.6NaOH/water; b) 5cellulose/7.6NaOH/6urea/water; c) 5cellulose/7.6NaOH/0.7ZnO/water;
Thermal treatment: 20 hours at -20C; 20 hours at +25C and 2 hours at +70C;
Regeneration bath: water at +5C, +25C, +70C and 10%acetic acid at +25C
_____________________________________________________________________________________________________
- 205 -
V.2.1.4. Influence of additives
Although the addition of zinc oxide improves cellulose dissolution and stability of solutions, it
does not increase the mechanical properties of regenerated cellulose samples (see Fig.V.2-2 a and
c). This can be explained by the presence of zinc oxide crystals that cannot be either dissolved in
water during regeneration or washed out from the regenerated cellulose. The presence of ZnO
crystals is seen in SEM images (Fig.V.2-3b). These crystals may generate rupture points in the
sample.
As far as ZnO can be dissolved either in acids or in strong basis (see Chapter III, Fig. III.2.12), it
could be possible to regenerate cellulose from cellulose/NaOH/ZnO/water solution in relatively
strong acids. As mentioned previously, this could be for a way for the film manufacture but not
for the sponges.

a) b)
Fig.V.2-3: SEM observations of regenerated B407 cellulose from
a) 5cellulose/7.6NaOH/water and b) 5cellulose/7.6NaOH/0.7ZnO/water,
after 2H gelation at 70C and regeneration in water at 70C

From the results on cellulose dissolved in the presence of urea, we can see that gelation at 70C is
not interesting to produce sponges: the rupture stress is low and even inferior to the one
obtained in NaOH without additives. Cellulose/NaOH/urea/water solutions are better than
without urea whatever the regeneration bath is.

The tensile tests performed for cellulose/7.6NaOH/water solutions with or without additives
and without reinforcing fibres and porophores showed that, first, regeneration in a hot bath
decreases the mechanical properties of wet regenerated cellulose whatever are solution treatments
and additives, and second, that gelation at room temperature gives the highest rupture stress of
the material.
As far as the presence of Na
2
SO
4
,10H
2
O crystals requires a high temperature to melt to be
washed from regenerated cellulose, one of the ways to reach higher rupture stress is to perform
solution gelation at room temperature, to regenerate cellulose in water also at room temperature
and finally to heat the regenerated cellulose blocks for the Na
2
SO
4
,10H
2
O crystals to melt.
_____________________________________________________________________________________________________
- 206 -
The presence of ZnO does not increase mechanical properties and will not be studied anymore in
the following. The presence of urea seems to increase the rupture resistance of samples and thus
can be a promising way to be tested for the preparation of sponges.
The best conditions at this stage are solutions with urea, gelled at +25C and regenerated in water
at +25C.

solutions with reinforcing fibres without porophores
V.2.2.1. Influence of added reinforcing fibres
The addition of fibres into a matrix usually leads to the increase in the mechanical properties of
the obtained composite material. In the viscose process, cellulose fibres are added to cellulose
solution in order to improve the mechanical properties of sponges.

Two types of reinforcing fibres were used: cotton and fluff with DP1100 (more details are given
in part II.1.2). Due to their high DP, these cellulose fibres cannot be dissolved in
7.6NaOH/water, even at -6C (see Chapter II, Fig.II.1.1).
Before being added to the cellulose solution, the reinforcing fibres were pre-impregnated by
2.7%NaOH/water solution for some minutes. Then they were added into cellulose solution and
mixed for 1.5hour. To mix impregnated fibres with cellulose/NaOH solutions, a blade mixer was
used (see part II.3.5). Long cotton fibres coiled around the blades and the mixture was not
homogeneous. We mainly used fluff fibres which was not having this difficulty.
The samples for tensile measurements were prepared in the same way as in the previous section.
The comparison of the stress at rupture for wet regenerated cellulose with and without the
reinforcing fibres is shown in Fig.V.2-4. Samples were gelled at +25C for 20 hours and
regenerated in water at +25C and +70C.

0
1
2
3
4
5
in water at +25C in water at +70C
regeneration bath
rupture stress
(daN/cm)
cellulose matrix
without fibres
+2% fluff

Fig.V.2-4: Comparison of the rupture stress of wet regenerated cellulose with and without reinforcing fibres.
Matrix: 5B389 cellulose/7.6NaOH/water
All-cellulose composite: Matrix + 2g of fluff fibres pre-impregnated in 2.7%NaOH
Treatment: gelation at +25C for 20 hours; regeneration in water at +25C and +70C
_____________________________________________________________________________________________________
- 207 -
Two observations can be made:
As expected, the addition of reinforcing cellulose fibres increases the value of the rupture
stress (Fig.V.2-4).
The regeneration bath of water at +25C shows better results, regarding the resistance to
rupture, as compared with regeneration at +70C, as already shown.

V.2.2.2. Influence of fibres treatment of the rupture stress
As far as the addition of fluff fibres improves the mechanical properties of regenerated cellulose,
a special attention was paid to the adhesion between the fibres and the matrix. The goal is to
increase the adhesion that should lead to the increase of the mechanical properties.

Fluff fibres were treated with different adhesion promoters (see their chemical formulations in
part II.1.3). Instead of being pre-impregnated by 2.7%NaOH/water, the 2.3g of fibres were pre-
impregnated by a solution containing 2.7%NaOH/water+the adhesion agent. A reference was
prepared without adhesion agent. Wet treated fibres were added to cellulose solution in the same
way as described above. The results of tensile tests are shown on Fig.V.2-5.

0
1
2
3
4
5
w
i
t
h
o
u
t

r
e
i
n
f
o
r
c
i
n
g

f
i
b
r
e
s
f
i
b
r
e
s

t
r
e
a
t
e
d

w
i
t
h

2
7
g
/
l

N
a
O
H
w
i
t
h

N
a
O
H
+
0
.
5
%
G
l
y
m
o
w
i
t
h

N
a
O
H
+
1
%
G
l
y
m
o
w
i
t
h

N
a
O
H
+
1
%
A
M
E
O
Rupture stress
(daN/cm)
regeneration in
water at +25C
regeneration in
water at +70C

Fig.V.2-5: Rupture stress as a function of fibres treatment for 5B389 cellulose/7.6NaOH/water + 2.3 treated fibres
solutions gelled for 20hours at +25C and regenerated in water at +25C or +70C.
Data for regenerated cellulose without reinforcing fibres are also shown.
Treatments of fibres are directly indicated on the graph.

First of all, whatever is the treatment of fibres, regeneration in water at room temperature
showed better mechanical results as compared with regeneration at 70C.
_____________________________________________________________________________________________________
- 208 -
Secondly, a comparison of results obtained for regenerated cellulose, without reinforcing fibres
(on the left side of the graph Fig.V.2-5), with the ones obtained for fibres pre-impregnated with
NaOH without adhesion agent, shows that the rupture stress is considerably increased.
Finally, the treatment of fibres by the adhesion agents (GLYMO, AMEO, CPTMO) does not
improve the rupture stress.

V.2.2.3. Influence of freezing on rupture stress and adhesion between fibres and matrix
When making real sponges, porophores (Na
2
SO
4
,10H
2
O crystals) are added to create large pores.
Because it is not possible to form thin films with porophores, thick samples have to be made in
order to obtain reproducible rupture stress measurements. To anticipate the forthcoming
experiments with crystals, parallelepiped blocks of regenerated cellulose with reinforcing fibres
and without porophores were prepared and studied. The goal is to investigate the influence of
freezing on the mechanical properties of wet samples and on the adhesion between the
reinforcing fibres (cotton or fluff) and cellulose matrix.

The cellulose/NaOH/water + reinforcing fibres solutions were prepared as in the section
V.2.2.1:
2g of fibres were pre-impregnated in 8%NaOH/water for some minutes, at room
temperature
the mixing of fibres with cellulose solution was performed at -6C, for 1.5 hour
a thermal treatment was applied
only the shape of specimens for tensile tests was changed (parallelepiped blocks versus
film)

It should be noted that when studying parallelepiped blocks, the value of the rupture stress is up
to ten times smaller than the ones obtained for films. This dramatic decrease can be explained by
the fact that there is no skin since the specimens were cut in a block. Moreover, as the cutting of
samples was performed by hand, specimens are more irregular. Even by change of the razor
blade each time, cellulose fibres are difficult to cut and we probably induced cracks in the sample.

In order to be as close as possible to the Spontex conditions of sponge preparation, the
regeneration of samples was performed in water at +70C. The rupture stress of wet
parallelepiped samples is shown in Tab.V.2-1.
Reinforcing fibres
Cellulose solution thermal
treatment
Rupture stress,
daN/cm
Gelation at 25C for 20 hours 0.13
Cotton
Freezing at -20C for 20 hours 0.26
Gelation at 25C for 20 hours 0.47
Fluff
Freezing at -20C for 20 hours 0.43
Tab. V.2-1: Rupture stress of wet regenerated parallelepiped cellulose samples from 5B389 cellulose/7.6NaOH/water
reinforced by 2g of cotton or fluff fibres pre-impregnated with 8%NaOH/water at room temperature,
regeneration in water bath at 70C.
_____________________________________________________________________________________________________
- 209 -
First of all, Tab.V.2-1 shows that the cotton fibres give very low rupture stresses in comparison
with fluff fibres. This is explained by the fact that the cotton fibres enrolled around the blades of
the mixer and thus that the solutions were not homogeneous.
Freezing either improves or does not change mechanical properties depending on the nature of
the reinforcing fibres, cotton or fluff respectively. In order to better understand what happened
during freezing, we observed by optical and electron microscopy the samples containing cotton
fibres, since there is a difference between the two thermal treatments. The pictures obtained with
optical microscope and SEM are shown on Fig.V.2-6.

If we compare the observations with optical microscope a and c of the Fig.V.2-6, it seems that
cotton fibres are almost unchanged before and after their introduction into the cellulose solution
after a gelation at +25C for 20 hours. The photo of sponges prepared with the freezing step
(Fig.V.2-6e) reveals that the fibres are cylindrical whereas natural cotton (Fig.V.2-6a) has a flat
ribbon shape: cotton fibres have been modified during the step at -20C.
On the other hand, after freezing and after gelation (Fig.V.2-6 pictures d and f in comparison
with picture b), we can see the matrix is attached to the cotton fibres, as seen by SEM. This is
compatible with the fact that both cases (freezing and gelation) give similar results in term of
rupture stress. The variation seen in Tab.V.2-1 for cotton is most probably due to the
inhomogeneous structure of the mixture (bad mixing). In both cases, some matrix material is
attaching to the cellulose reinforcing fibres. It is known that the external layer of cotton fibres is
not easily dissolved by NaOH/water [CUI2006]. Nevertheless, a treatment in NaOH/water in
dissolving conditions (-6C or -20C) seems to help the cellulose from the matrix to interact with
the surface of the fibre, so that it will stick to it after regeneration.

_____________________________________________________________________________________________________
- 210 -

OBSERVATION WITH AN OPTICAL MICROSCOPE
The bar scale represents 100m
OBSERVATION WITH A SEM

a)
b)
Initial dried cotton fibres

c) d)
Cotton fibres inside a regenerated cellulose matrix
Thermal treatment: gelation for 20 hours at +25C

e) f)
Cotton fibres inside a regenerated cellulose matrix
Thermal treatment: freezing for 20 hours at -20C

Fig.V.2-6: Observation of cotton fibres in the initial dry state (a, b) and after different thermal treatment and regenerated
in water at +70C: c) and d): gelation for 20 hours at +25C; e) and f): freezing for 20 hours at -20C.

_____________________________________________________________________________________________________
- 211 -
V.2.2.4. Influence of the thermal history of fibres treatment with 8%NaOH aqueous solution on
the mechanical properties of wet regenerated samples
The goal of this section is to investigate if the mechanical properties of wet regenerated samples
(regeneration bath: water at 70C) are influenced by the following:
a) Temperature of fibre treatment before they are mixed with cellulose/NaOH solution and
b) Thermal history (gelation, freezing) of the mixture composed of cellulose/NaOH/water
solution and pre-impregnated fibres.

a) We know that at least two conditions have to be fulfilled to dissolve cellulose in sodium
hydroxide: (i) NaOH concentrations should be in the interval of 7-10% at -6C 0C
temperatures according to the phase diagram [SOB1939] and (ii) to have cellulose with a DP
inferior to about 400-600 [YAM1990]
The DP of the reinforcing fibres is at least 1100. These fibres cannot be fully dissolved in sodium
hydroxide, even in 8%NaOH aqueous solution, at -6C. But in these conditions of low
temperature, the fibres are swelling.

The fibres were placed in a large amount of 8%NaOH aqueous solution for 20 hours at different
temperatures: +25C, -5C and -20C. At the end of this impregnation, they are removed from
the alkaline solution, spin-dried and observed with an optical microscope (Fig.V.2-7).

a)

b)

c)

Fig.V.2-7: Microscopic observations of cotton fibres pre-impregnated with 8%NaOH/water for 20 hours at
a) +25C; b) -5C; c) -20C
Scale bars represent 100m

It is clear that after 20 hours in 8%NaOH/water, at +25C (Fig.V.2-7a), the aspect of the fibres
did not change much (Fig.V.2-6a). Keeping fibres at -6C or freezing them at -20C in
8%NaOH/water solutions clearly changes the aspect of fibres (pictures b and c respectively):
fibres are swelling. After freezing at -20C, the swollen part of the fibre is rather homogeneous
and smooth, while after being kept at -6C the diameter of swollen zones varies from place to
place. Are these treatments influencing the mechanical properties of wet regenerated samples?

b) Two types of fibres treatment were chosen: 20 hours at 25C and 20 hours at -6C, with
8%NaOH aqueous solution. After one of these two treatments, fibres were spin-dried and then
mixed with cellulose/7.6NaOH aqueous solution following the procedure described in part
II.3.5. The mixture was then either:
_____________________________________________________________________________________________________
- 212 -
aged for 2 hours at +5C and then gelled at +25C for 20 hours or
aged for 2 hours at +5C and then frozen at -20C for 20 hours or
not aged at all and gelled at +25C for 20 hours or
not aged at all and frozen at -20C for 20 hours.

The results of rupture stress for these cases are presented in Fig.V.2-8.

0
0.2
0.4
+5&-20 -20 +5&+25 +25
Thermal treatment
Rupture stress
(daN/cm)
Pre-impregnation at -6C
Pre-impregnation at +25C

Fig.V.2-8: Rupture stress for wet regenerated samples (in water at +70C) as a function of thermal treatment for
5B389 cellulose/7.6NaOH/water+2 fluff pre-impregnated in 8%NaOH at +25C ( , one value at +5&-20) and at -6C ( )
+5&-20: 2 hours of ageing at +5C then 20 hours of freezing at -20C;
-20: 20 hours of freezing at -20C (without ageing);
+5&+25: 2 hours of ageing at +5C then 20 hours of gelation at +25C;
+25: 20 hours of gelation at +25C (without ageing).

The following conclusions can be made from this graph.
Regarding the results two by two, freezing at -20C and gelation at +25C, with or
without ageing (2 hours at +5C), it is clear that the ageing step does not increase
mechanical properties.
Secondly, pre-impregnating the fibres at -6C does not improve the resistance to the
rupture compared with a pre-impregnation at room temperature.
Freezing (with or without aging) does not lead to a considerable increase of the rupture
stress.

Thus it is not necessary to decrease the temperature of pre-impregnation and to apply an ageing
step.

The addition of the reinforcing fibres (fluff) leads to the increase in the rupture stress of wet
regenerated samples. But the effect is rather small and we tried to find a method to increase the
mechanical properties by improving the adhesion matrix/fibres. It appears that the addition of an
adhesion promoter (GLYMO, AMEO or CPTMO) is not efficient.
_____________________________________________________________________________________________________
- 213 -
The treatment of the reinforcing fibres at -20C does not improve the mechanical properties of
the samples.
Finally, as previously in the case of wet regenerated cellulose, without reinforcing fibres,
regeneration in water at +25C gives better mechanical properties than regeneration in water at
+70C.

solutions with reinforcing fibres and porophores
In the last step of our investigation, we added Na
2
SO
4
,10H
2
O crystals to create a porosity and
thus to obtain a sponge from NaOH process. Fig.V.2-9 shows a wet regenerated sample, seen by
optical microscopy.

Fig.V.2-9: Wet sponge made in Cemef from the NaOH process.

Tensile tests were performed on wet sponge sample after two different thermal treatments of
cellulose solution + reinforcing fibres + porophores: 20 hours of freezing at -20C or 20 hours
of gelation at +25C. No ageing step was applied as we saw that it does not improve the results.
The fibres of fluff were pre-impregnated with 8%NaOH/water at -6C. Samples were
regenerated at 70C. The values of the rupture stress and other characteristics of sponges are
reported in Tab.V.2-2 and their calculation is given below (details are given in part II.3.6).

Property Calculation Meaning
Rupture stress
FR
S
F
F
R
=
In our case FR is directly
determined from the graph
F is the tensile force when rupture appears (daN)
S is the initial specimen section (cm)
FR is the rupture stress (daN/cm)

Density
d
w
d
V
m
d =
d represents the density of the sponge (g/dm
3
)
md is the weight of the dry sample (g)
Vw is the volume of the wet sample (dm
3
)
Water absorption
Abs.
d
d w
m
m m
Abs
) (
.

=

Abs. is the absorption capacity
mw represents the weight of wet sample (g)
md is the weight of dry sample (g)
Shrinkage
Shr.
w
dw w
A
A A
Shr

= .
Shr. is the shrinkage of the sponge sample
Aw is the surface of a never-dried-wet sample (cm)
Adw is the surface of a dried-and-swollen again sample (cm)

15mm
_____________________________________________________________________________________________________
- 214 -
All the properties described above depend on the conditions of preparation of sponges. But they
are also closely connected to the density. In other words, when two samples have very different
densities, the comparison of other characteristics does not mean much. In our case (Tab.V.2-2),
the density values are similar, around 25g/dm
3
. It is thus possible to compare the mechanical and
physical properties as a function of the thermal treatment applied to the mixtures before
regeneration, i.e. if the mixture was frozen or not.

With Na2SO4,10H2O crystals

Without
crystals
Impregnation Treatment FR d Abs. Shr.

FR
+20C -20C 0.17 24.9 19.6 37

0.31
-6C -20C 0.20 25.8 20.6 27

0.40
-6C +25C 0.09 24.6 9.7 58

0.35
+20C +25C 0.05

Sponges from Viscose process

FR d Abs. Shr

CV 1.6-1.7 35 24-25 6

CE 1.5 30 29 1.5

Tab.V.2-2: Properties of sponges prepared from NaOH and viscose processes,
For the viscose process, the properties are given by Spontex
CV: Vapour coagulation; CE: Electrical coagulation part V.1.1-

From the results reported in Tab.V.2-2, it is clear that the properties of sponges prepared in
Cemef with the NaOH process are very low in comparison with the ones of standard sponges,
prepared industrially with the viscose process.

Now, let us compare the frozen and gelled sponges, prepared from the NaOH process.

V.2.3.1. Rupture stress
As seen in Tab.V.2-2, the presence of crystals forming the pores dramatically decreases the
resistance of sponges. This is an expected result since the amount of solid materials is decreased.
But the difference between Cemef and Spontex sponges is extremely high. Sponges made with
the NaOH process have very poor resistance to rupture.
From the results of the rupture stress (F
R
) obtained for sponges containing regenerated cellulose
+ reinforcing fibres + porophores (Tab.V.2-2 column 3), we can see that the freezing step allows
to double the resistance to rupture. A pre-impregnation of reinforcing fibres with
8%NaOH/water at -6C instead of +20C slightly improves the rupture stress, but this increase
is within the experimental error.
These tensile tests on final sponges (system containing regenerated cellulose + reinforcing fibres
+ porophores) reveal that the freezing step enables to improve the mechanical resistance of
sponges. This result is in complete contradiction with the values of the rupture stress obtained
_____________________________________________________________________________________________________
- 215 -
when the mixture does not contain the porophores, see Tab.V.2-1. Without porophore, the
freezing step (-20C) has no effect on the tensile stress of the wet regenerated material.
Pre-impregnation of the reinforcing fibres at -6C does not seem to be, in our experiments, a
necessary step for improving the rupture stress.
The most important finding of this measurement is the absence of suitable mechanical property
of the NaOH processed sponges, made in Cemef.

V.2.3.2. Absorption
The comparison of the absorption values (Abs.) between NaOH and viscose sponges shows that
the viscose ones have better properties, but that the NaOH ones are not very far from the
viscose sponges. Absorption is linked primarily to the amount of porosity and to the way pores
are structured. Porosity of viscose and NaOH processed sponges is compared using SEM and
shown on Fig.V.2-10.

NaOH sponges contained a cellulose matrix from 5B389 cellulose/7.6NaOH/water solution,
reinforcing fibres (fluff) and a macroscopic porosity due to the presence of Na
2
SO
4
,10H
2
O

crystals. Regeneration in water at +70C, in several baths, was performed to obtain the final wet
product. Viscose sponges contained a Raywhite DP1044 cellulose matrix, reinforcing fibres
(cotton, flax and fluff) and a macroscopic porosity due to the presence of Na
2
SO
4
,10H
2
O

crystals.
The coagulation/regeneration step was performed by electrical coagulation and the blocks were
washed in water at +70C.
In both cases, the sponges were left uncovered to dry naturally and then they were observed with
the SEM. Without drying, sponges have to be observed with the SEM in Environmental mode
(ESEM), and it is not possible to obtain high enough magnification. This is why we dried them,
to be able to use classical SEM.

Fig.V.2-10 a and b reveal an important and fine porosity in viscose sponges (on the top) which
explains the high water absorption value (25-30%). Small pores have diameter inferior to 5m. As
the Na
2
SO
4
,10H
2
O crystals have diameter between 100m and 3mm, it is obvious that the small
pores observed on this picture do not come from the melting of crystals. In the viscose
process, they are due to the gases (CS
2
, H
2
S) that are released during the
coagulation/regeneration step, when the structure of cellulose is recovered. In sponges prepared
from cellulose/NaOH solutions with a freezing step (Fig.V.2-10 c and d), the quantity of micron-
size pores is low and they are not regularly distributed. Indeed, in the NaOH process, no
chemical reaction occurs and thus no gas leaves the sample that could lead to the formation of
the micro pores. This explains the low absorption value. Finally, sponges prepared from
cellulose/NaOH solutions with gelation at +25C (Fig.V.2-10 e and f) contain no micron-size
pores in the regenerated cellulose. Freezing step also plays a role in the formation of some small
pores at all, thus decreasing water absorption (see Tab.V.2-2).

_____________________________________________________________________________________________________
- 216 -
Scale bar corresponds to 200m Scale bar corresponds to 50m

a)

b)
Sponges made from the viscose process: cellulose Rayonier DP1044 with a mixture of cotton, flax and fluff
regeneration in water at +70C

c)

d)
Sponges made from the NaOH process: 5B389 cellulose/7.6NaOH/water with 2 reinforcing fluff fibres
pre-impregnated with 8%NaOH/water at -6C, porophores added,
20 hours of freezing at -20C, regeneration in water at +70C

e)

f)
Sponges made from the NaOH process: 5B389 cellulose/7.6NaOH/water with 2 reinforcing fluff fibres
pre-impregnated with 8%NaOH/water at -6C, porophores added,
20 hours of gelation at +25C, regeneration in water at +70C.

Fig.V.2-10: SEM observation of dried sponges.

_____________________________________________________________________________________________________
- 217 -
As it could be anticipated, knowing the gaseous origin of the good water absorbance properties
of the viscose sponges, NaOH sponges cannot be as good as viscose ones. The absence of
chemical reaction is not bringing any micron-size holes. To form them can only occur through
the phase separation process that is occurring during regeneration. It seems that there is a large
difference in the production of holes between samples that went or not at -20C. The freezing is
changing something in the structuration of the solution that is leading to a phase separation
during regeneration that is producing tiny holes.

V.2.3.3. Shrinkage
The shrinkage (Shr.) corresponds to the decrease of surface area of a wet sponge as compared
with a wet/dried/re-wet sponge. In high performance sponges the shrinkage should be as low as
possible, like in viscose sponges where it reaches 1.5% and 6% for sponges prepared from
electrical and vapour coagulation methods respectively. A high value means that a large quantity
of pores has been closed.

The shrinkage of the NaOH sponges is very high, up to 58%, as compared with less than 10%
for viscose sponges. This is a rather puzzling result since it should be in coherence with the
absorption properties. We showed that the absorption is low in NaOH sponges because of the
absence or the low quantity of micron-scale holes. These should be the first candidates for
closing after drying.

In order to understand what happens during the natural drying of sponges, we dried them with
supercritical CO
2
technique.
This specific drying is used in order to replace solvent by air. This technique allows avoiding the
pores collapse due to the liquid surface tension and keeping intact the initial micro- or
nanostructure of 3D-objects. The supercritical fluid used in our study to dry samples is
supercritical CO
2
(sc CO
2
) mainly because its critical point is easily obtainable (P=73.8bar and
31.1C). As water was not compatible with supercritical CO
2
, the swollen cellulose was washed
five times with acetone, solvent miscible with this supercritical fluid. Following this exchange of
solvents the cellulose was in a swollen state in acetone and the supercritical CO
2
drying can be
applied.
CO
2
supercritical extraction was performed at laboratory scale in CEP, at Sophia-Antipolis,
France.

First of all, the regenerated cellulose + reinforcing fibres + porophores system was prepared as
described in Chapter II: with the freezing step and the regeneration in water at +70C. It results a
wet regenerated sponge. From this sponge, two samples were observed:

_____________________________________________________________________________________________________
- 218 -
The wet regenerated sponge is directly placed in an acetone bath. There, an exchange
between water, present in the cellulose sample, and acetone, present in the bath, is
occurring. When the equilibrium is reached, the sample is dried in supercritical CO
2
.

The wet regenerated sponge is first dried at ambient air, and then re-wet in water to
obtain a wet/dried/re-wet sample. This latter is then placed in an acetone bath where
both solvents water and acetone are exchanged. The super-critical CO
2
drying is
applied.

Fig.V.2-11 a and b show the microscopic observations of the wet and wet-dried-rewet samples,
respectively, seen by SEM.

a)

b)
Fig.V.2-11: SEM observation of a wet (a) and a wet/dried/re-wet (b) NaOH sponge prepared from 5B389
cellulose/7.6NaOH/water solution, reinforcing fibres pre-impregnated with 8%NaOH/water at -6C, porophores added.
20 hours of freezing at -20C, regeneration in water at +70C, dried using super-critical CO2 technique

From these pictures we can note that the micro porosity of a wet/naturally dried/re-wet NaOH
sponge (Fig.V.2-11b) has dramatically decreased as compared with the micro porosity of a wet
NaOH sponge (Fig.V.2-11a). These observations can explain the high value (>30%) obtained for
the shrinkage (Tab.V.2-2). This means that drying the frozen sponges prepared with NaOH
process at ambient air is closing must of the micro-pores of the sponges.

For non-frozen NaOH sponges, it is clearly not the case because of the absence of micro
porosity (Fig.V.2-10f). What are closing in the NaOH sponges are the large pore made from the
traces of crystal blocks of porophore. And these some large pores are not closing in the case of
viscose. One reason that should be investigated is the way OH- groups are organised at the
extreme surface of the solid cellulose parts, and why they are prevented to form a strong
hydrogen network with neighbouring cellulose surfaces in the case of viscose while they are
doing it for NaOH sponges.

Moreover, if we look at the Fig.V.2-11a, we can see that the porosity is lower than the one
observed for viscose sponge (Fig.V.2-10b) but not very far. This can explain a not too bad
absorption value (20%) for a frozen NaOH sponge.
_____________________________________________________________________________________________________
- 219 -
V.2.4- Conclusion on the role of the freezing step
The results concerning the frozen NaOH sponges (system with regenerated cellulose +
reinforcing fibres + porosity left after the melting of Na
2
SO
4
,10H
2
O crystals) are:
The freezing step leads to a swelling of reinforcing fibres but does not improve the
adhesion between the regenerated cellulose matrix and the reinforcing cellulose fibres.
During the freezing step, certain porosity is created which is not the case when sample
are gelled at room temperature before regeneration.
During the freezing step, the free water present in cellulose solution crystallised and
formed needle crystals (Fig.V.2-12). Tensile tests on such a sample should give a break
along the imprint of ice crystals. But it is not the case (unfortunately, we do not have
photos), the breaking occurs across these imprints and does not take into account them.
Moreover, we saw in the section V.2.2.3 that thin specimens are ten times more resistant
than thick one, probably because of the skin effect. Thus we can think that a skin is
formed around the ice crystals improving the mechanical properties of frozen sponges.

Fig.V.2-12: Imprints of ice crystals in a regenerated cellulose prepared from 5B389 cellulose/7.6NaOH/water,
frozen for 20 hours at -20C and regenerated in water at +70C.

The influence of freezing step on the rupture stress is very pronounced when the samples contain
Na
2
SO
4
,10H
2
O crystals and not important without.

The results of Cemef sponges confirm the findings of Spontex, i.e. that a freezing step at -20C is
needed to obtain a material that will not shrink and will have better mechanical properties.
Nevertheless, at least with the Cemef preparation, the mechanical properties of the NaOH
sponges are very poor. An interesting question lies in the difference of results between the wet
material obtained by the NaOH process with and without porophores, aside the large difference
in mechanical properties between these two cases. Without porophore, a freezing step has no
effect. With porophores, it dramatically increases the mechanical properties. Two hypotheses can
be made taking into account that the effect of freezing on the structure of the solution is not very
clear (DSC results seem to show that a small amount of water is crystallised):
With porophores, freezing allows a long time in the best dissolution range, leading to a better
solution in the small thin walls that are between porophore crystals. Upon regeneration, they
_____________________________________________________________________________________________________
- 220 -
give thin cellulose wall with good properties. Without freezing, these walls are not continuous
and are breaking easily.
There is a specific interfacial interaction between the porophore crystals and the cellulose
solution that is wetting the crystal better at low temperature, leading to the formation of
continuous walls upon regeneration.

A change in the porophore system should bring some clues whether it is due to the chemical
nature of the porophores.

_____________________________________________________________________________________________________
- 221 -
V.3- CONCLUSIONS

Mechanical properties of the regenerated cellulose matrix

Testing only the cellulose matrix reveals not only that the freezing step does not improve
the rupture stress but also that the water regeneration bath at +70C is not favourable to
obtain good mechanical properties. Gelation and regeneration in water at room
temperature give the highest rupture stress for a cellulose/NaOH/water solution
We saw that the presence of urea or ZnO in NaOH/water improves the cellulose
dissolution and delays the gelation of solutions. But concerning the mechanical tests, the
two additives behave differently: urea increases the rupture stress whereas ZnO decreases
it, probably because of the presence of ZnO particles in the regenerated cellulose sample.

Mechanical properties of the regenerated cellulose matrix + reinforcing fibres

The addition of reinforcing fibres improves the mechanical properties but the improvement is
relatively low. So we tested two ways to improve it.

The addition of adhesion agent (GLYMO, AMEO and CPTMO) does not increase the
rupture stress of samples.
The pre-impregnation of reinforcing fibres with 8%NaOH/water, at -6C, induces the
swelling of fibres but does not improve the rupture stress.

Mechanical properties of the regenerated cellulose matrix + reinforcing fibres + porophores

First of all, we can note that the NaOH sponges have lower properties (rupture stress, absorption
and shrinkage) in comparison with the viscose sponges. Moreover, the influence and the
necessity of freezing in NaOH sponges clearly appear.

Despite the low values of the rupture stress of these final sponges, prepared in Cemef, we
can see that the freezing step improves the mechanical properties of samples.
The very bad absorption value of a non-frozen NaOH sponge can be explained by an
absence of micro-porosity. In a frozen NaOH sponge, we can observe (SEM) some
micro-pores but their amount is low in comparison with a viscose sponge, this result is in
agreement with the absorption.
The drying at the ambient air of NaOH sponges is closing micro-pores leading to a high
value of shrinkage.
_____________________________________________________________________________________________________
- 222 -
[CUI2006] CUISSINAT C., NAVARD P. Swelling and dissolution of cellulose Part II: free

[ISO1998] ISOGAI A., ATALLA R.H. Dissolution of cellulose in aqueous NaOH solutions.
Cellulose (1998), 5, p.309-319


[SOB1939] SOBUE H., KIESSIG H., HESS K. The cellulose-sodium hydroxyde-water system as
a function of the temperature. Z. Physik. Chem. B (1939), 43 (3), p.309-328

[YAM1990] YAMASHIKI T., MATSUI T., SAITOH M., OKAJIMA K., KAMIDE K.,
SAWADA T. Characterisation of cellulose treated by the steam explosion method. Part 1:
Influence of cellulose resources on changes in morphology, degree of polymerisation, solubility
and solid structure. British Polymer Journal (1990), 22, p.73-83

Conclusions
_____________________________________________________________________________________________________
- 223 -
C
C
O
O
N
N
C
C
L
L
U
U
S
S
I
I
O
O
N
N
S
S

The work performed in the present thesis tackles both scientific questions and industrial
problems. This study is focused on the structure of solvents, 7.6NaOH/water,
7.6NaOH/6urea/water and 7.6NaOH/0.7ZnO/water, and on the properties of solutions of
cellulose dissolved in these solvents, in particular, the rheology, the conditions of gelation
(temperature and kinetics) and the influence of freezing the solutions on their structures. These
results were used to evaluate the possibility of making films and sponges.

Structure of solvents: NaOH/water, NaOH/ZnO/water and NaOH/urea/water

The phase diagrams of cellulose solvents, NaOH/water, NaOH/urea/water and
NaOH/ZnO/water, were studied using the DSC technique.

The NaOH/water binary phase diagram was studied in details for 0-30%NaOH/water solutions.
The results are in agreement with the NaOH/water phase diagram reported in literature
[PIC1893] [ROL1964]. The possibility to calculate at all compositions and temperatures the
composition of the mixture and the existence of a eutectic peak (NaOH,5H
2
O+4H
2
O at -33.4C)
allowed studying the influence of the incorporation of cellulose.

The urea/water binary system forms a eutectic mixture, at -11.4C, composed of urea+8H
2
O. In
the ternary NaOH/urea/water system, both NaOH and urea eutectic mixtures are formed
independently when enough water is present in solution. When water is missing, NaOH is
attracting water at the expense of urea. NaOH and urea do not interact.

Mixing ZnO with NaOH/water leads to the formation of ZnO-NaOH-water complexes which
change the NaOH/water phase diagram by decreasing the melting temperature and enthalpy of
NaOH eutectic. ZnO reacts with 3NaOH molecules and can be surrounded by up to 25 water
molecules.
Conclusions
_____________________________________________________________________________________________________
- 224 -
Structure and properties of cellulose/NaOH/water solutions

Cellulose was dissolved in NaOH/water with NaOH concentrations from 7 to 9%. The addition
of cellulose in NaOH/water does not change the melting temperatures of the NaOH eutectic
and of free ice. This means that the presence of cellulose does not disrupt the phase diagram of
the solvent.

The melting enthalpy of the NaOH eutectic peak significantly decreases when cellulose content
increases. This allows to calculate the amount of NaOH linked to each anhydroglucose unit
(AGU) and thus to determine the maximal amount of cellulose that can be dissolved in a given
NaOH/water solution. The dissolution of Avicel cellulose in NaOH/water is occurring if at least
four NaOH are bound to one AGU which corresponds to a weight ratio m
cell
/m
NaOH
1.
The constant enthalpy of free ice (no dependence on cellulose content) means that water
molecules linked to cellulose are the only ones that are involved in the eutectic through NaOH.
(i.e. 9 water molecules per NaOH).
As the upper limit of NaOH concentration that can dissolve cellulose is 10% [SOB1939], the
maximal Avicel cellulose concentration that can be dissolved in a sodium hydroxide aqueous
solution is 10%, which is a relatively low value for an industrial process.

Other than Avicel cellulose samples were studied: steam-exploded Borregaard (DP = 389) and
Solucell (DP = 198 and 307). The enthalpy of the eutectic peak is also decreasing in the presence
of cellulose as in the case of Avicel, but the amount of the decrease depends not only on cellulose
concentration but also on its origin. This is due to the fact that the cellulose solutions other than
pure cellulose materials are difficult to dissolve properly and completely.
The viscosity versus shear rate measurements showed that the steam-exploded
cellulose/NaOH/water solutions are not molecularly dispersed and behave as a suspension. This
is compatible with the fact that cellulose chains are coated with several NaOH and water
molecules.

The intrinsic viscosity of cellulose/NaOH/water solutions is decreasing with temperature
increase. The quality of the solvent thus decreases and the cellulose-cellulose interactions are
favoured leading either to compactisation of macromolecules in the case of dilute solutions and
to gelation in the case of semi-dilute solutions.

Structure and properties of cellulose/NaOH/urea/water and cellulose/NaOH/ZnO/water
solutions

Although some parts of cellulose fibres remain non-dissolved, the presence of urea or zinc oxide
in cellulose/NaOH/water solutions improves the cellulose dissolution (microscopic observation).
The intrinsic viscosity of cellulose/NaOH/water solutions in the presence of both additives is
slightly higher than without them demonstrating better solvent (NaOH/water + additives)
quality, but it decreases with temperature increase.
Conclusions
_____________________________________________________________________________________________________
- 225 -
The limit of cellulose dissolution in NaOH/urea/water determined with DSC is the same as in
NaOH/water. All these solutions have the same structure: cellulose chains coated with NaOH
and water molecules. As previously, this explains the flow behaviour under shear being close to
the one of a suspension.

It seems that the role of urea is to trap a part of water which is not favourable for the cellulose
dissolution. Higher is the urea content up to a certain value, higher is the gelation temperature,
which means that solvent quality increases. Gelation occurs at low temperatures if there is too
much urea in the solution probably because the proportions between NaOH and water are such
that the NaOH hydrates are not able to dissolve cellulose anymore. The maximal urea content to
delay gelation depends on the cellulose origin, but the reasons of this difference stay an open
question.

The addition of zinc oxide in NaOH/water is also efficient to better dissolve cellulose and delay
gelation up to the limit of ZnO solubility, which is reached at the molar ratio NaOH:ZnO=10.
Further increase of ZnO content does not lead to the improvement in dissolution or gelation. We
concluded that the improvement of solvent quality is due to the formation of ZnO-NaOH-water
complexes. These complexes are preventing the cellulose-cellulose interactions but do not modify
the gelation mechanism of cellulose/NaOH/water solutions.

Influence of freezing step on thermal properties of cellulose/NaOH/water solutions

To keep cellulose/NaOH/water solutions at -20C for several hours seems to improve the
mechanical properties of regenerated cellulose products. That is why the influence of the freezing
step on the thermal properties of cellulose solutions was investigated. Using DSC, we studied the
melting behaviour of cellulose/NaOH/water and cellulose/NaOH/ZnO/water solutions frozen
to -28C and kept at this temperature for several hours. Cellulose does not change the phase
diagram of the NaOH/water binary system, so keeping the solution at -28C leads to the
crystallisation of a part of free water. This free water melts at a certain temperature that depends
on NaOH concentration. During the isothermal step at -28C of several hours, some water is
detached from cellulose, about 1H
2
O per AGU, and is released. This detached water
crystallises as it was alone, giving the melting peak at 0C. Several hours of freezing are needed
for detached water to appear. The same occurs for cellulose/NaOH/ZnO/water solutions.

Mechanical properties of regenerated objects prepared from cellulose/NaOH/water solutions.
Influence of freezing.

Tensile tests on regenerated cellulose with and without reinforcing fibres and porophores were
performed.

Conclusions
_____________________________________________________________________________________________________
- 226 -
Without reinforcing fibres and without porophores, the best properties are obtained when the
cellulose/NaOH/water solution is gelled at room temperature for about 20 hours and
regenerated in water at room temperature. Regeneration in water at temperatures around 70-80C
leads to a strong decrease in the mechanical properties of regenerated samples as compared with
regeneration at room temperature. The freezing step (keeping solution at -20C for 20 hours)
does not improve the mechanical properties of regenerated cellulose.

All-cellulose composite, consisting of a matrix of regenerated cellulose and reinforcing fibres of
cotton or fluff (wood pulp with high DP), was then prepared. Although the addition of fibres
improves the rupture stress, this improvement is not high enough because of a bad matrix/fibre
adhesion.
During the freezing step, the reinforcing fibres are swelling but this does not induce better
adhesion matrix/fibres and thus does not explain better properties of frozen sponges as
compared with non-frozen ones.

The Na
2
SO
4
,10H
2
O crystals used as porophores were then added to the cellulose/NaOH/water
+ reinforcing fibres system to obtain the final product. Tensile properties of frozen and non-
frozen NaOH sponges are lower as compared with the ones of viscose sponges. Nevertheless,
the freezing step improves properties of NaOH sponges like the rupture stress, the shrinkage and
the absorption. Indeed, a SEM observation of the final product reveals that frozen sponges have
a relatively good micro-porosity but the pores are closing after drying. On the contrary, non-
frozen sponges have low micro-porosity leading to weaker properties.

P PE ER RS SP PE EC CT TI IV VE ES S
The proper knowledge of the phase diagrams and the physical properties of cellulose dissolved in
NaOH-water-additive mixtures is opening the way to a deeper study of the structure of the
cellulose mixtures by techniques able to probe the organisation of the molecules, like NMR, high
resolution X-ray scattering and dynamic light scattering. Such results, combined to our work,
could bring more information about the way hydrogen bond networks are organised and on
which compounds are formed. This should open the way to find either new additives better
dissolving cellulose at the molecular level (decreasing the amount and size of micro-gels) or either
treatments (activation) of cellulose.

Appendixes
_________________________________________________________________________________________________________
-227-
A
A
P
P
P
P
E
E
N
N
D
D
I
I
X
X
E
E
S
S

-228-
Appendixes
_________________________________________________________________________________________________________
[CIO2005] CIOLACU D., POPA V.I.; Cellulose chemistry and technology (2005), 39 (3-4), 179-188

-229-
A AP PP PE EN ND DI IX X 1 1
0
2000
4000
6000
8000
10000
12000
0 10 20 30 40 50 60
2Theta ()
I (cts)
B530
B407
L307
L198

X-Ray diffraction patterns for initial dry cellulose samples (described in the part II.1.1.3)
B530: non-steam exploded Borregaard cellulose (DP530)
B407: steam-exploded Borregaard cellulose (DP407)
L307: electron-beam irradiated Lenzing cellulose (DP307)
L198: regenerated Lenzing cellulose (DP198)

The ref [CIO2005] reports that the characteristic peaks for:
Cellulose I appear at
o 14.8 for (101)
o 16.6 for (10-1)
o 22.6 for (002)
Cellulose II appear at
o 12.1 for (101)
o 19.8 for (10-1)
o 22.0 for (002)

Appendixes
_________________________________________________________________________________________________________
-230-
0
0,04
0,08
0,12
0,16
0,2
0,001 0,01 0,1 1
stress (Pa)
G
'

(
P
a
)
0.3Hz
0.08Hz
0.1Hz
0.16Hz
0.02Hz
0.05Hz

Storage modulus G as a function of the applied stress for various frequencies, at 20C
for 2.3% B500 cellulose/7.6%NaOH aqueous solutions
Measurements on the StressTech rheometer

0
0,04
0,08
0,12
0,16
0,2
0,001 0,01 0,1 1
stress (Pa)
G
'
'

(
P
a
)
0.02Hz
0.16Hz
0.1Hz
0.08Hz
0.05Hz
0.3Hz

Loss modulus G as a function of the applied stress for various frequencies, at 20C
for 2.3% B500 cellulose/7.6%NaOH aqueous solutions
Measurements on the StressTech rheometer

Appendixes
_________________________________________________________________________________________________________
-231-
0
0,2
0,4
0,6
0,8
1
1,2
1,4
-25 -20 -15 -10 -5 0 5 10
TC
H
e
a
t

f
l
o
w

(
W
/
g
)
experimental points
eutectic peak
ice peak
sum

0
0,2
0,4
0,6
0,8
1
1,2
-25 -20 -15 -10 -5 0 5 10
TC
H
e
a
t

f
l
o
w

(
W
/
g
)
experimental points
eutectic peak
ice peak
sum

Appendixes
_________________________________________________________________________________________________________
-232-
0
0,2
0,4
0,6
0,8
1
1,2
-25 -20 -15 -10 -5 0 5 10
TC
H
e
a
t

f
l
o
w

(
W
/
g
)
experimental points
eutectic peak
ice peak
sum

0
0,2
0,4
0,6
0,8
1
1,2
1,4
-25 -20 -15 -10 -5 0 5 10
TC
H
e
a
t

f
l
o
w

(
W
/
g
)
experimental points
eutectic peak
ice peak
sum

0
0,2
0,4
0,6
0,8
1
1,2
1,4
-25 -20 -15 -10 -5 0 5 10
TC
H
e
a
t

f
l
o
w

(
W
/
g
)
experimental points
eutectic peak
ice peak
sum

7.6%NaOH + 6% urea aqueous solution
0
0,2
0,4
0,6
0,8
1
-50 -40 -30 -20 -10 0
TC
H
e
a
t

f
l
o
w

(
W
/
g
)
experimental points
eutectic peak
ice peak
sum
urea peak

7.6%NaOH + 12% urea aqueous solution
0
0,2
0,4
0,6
0,8
-50 -40 -30 -20 -10 0
TC
H
e
a
t

f
l
o
w

(
W
/
g
)
experimental points
eutectic peak
urea peak
ice peak
sum

7.6%NaOH + 14%urea aqueous solution
0
0,2
0,4
0,6
0,8
-50 -40 -30 -20 -10 0
TC
H
e
a
t

f
l
o
w

(
W
/
g
)
experimental points
eutectic peak
ice peak
sum
urea peak
stable eutectic

7.6%NaOH + 18%urea aqueous solution
0
0,2
0,4
0,6
0,8
-50 -40 -30 -20 -10 0
TC
H
e
a
t

f
l
o
w

(
W
/
g
)
experimental points
eutectic peak
stable eutectic
sum
urea peak

NMR investigations of NaOH-water-samples
I- OBJECTIVE
The current report contains the results of NMR spectroscopic experiments
with NaOH water and NaOH water ZnO solutions. By means of the nuclei
1
H and
23
Na is to be examined whether in the concentration interval of 6%
to 10% NaOH discrete NaOH hydrates (NaOH water complexes) are
detectable. In addition the temperature influence is to be examined within
the range of 20C to 20C.

Table 1 Composition of the samples and measuring temperatures

sample 1 2 3 4 5 6 7 8 9
% NaOH
% ZnO
6
0
7
0
8
0
8
0,2
8
0,5
8
1
8
1,5
9
0
10
0
-20C
-10C
0C
10C
20C
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X

II- EXPERIMENTAL
II.1- Sodium lyes
From water-free sodium hydroxide and/or zinc oxide and water without ions
the caustic soda solutions and/or zinc containing caustic soda solutions
were made by weighing the components in polyethylene containers, which
were sealed.

II.2- NMR measurements
The NMR investigations has been carried out with a spectrometer INOVA
500 (Varian Inc.) at measuring frequencies of 499.84 MHz for the
1
H nuclei
and 132.22 MHz for the
23
Na nuclei. The samples were measured in 2 mm
OD capillaries in a 5 mm OD sample tube. The gap was filled with
deuterated chloroform. That served as stabilizer (NMR lock) and also as
reference substance for the determination of the chemical shifts of the
1
H
spectra. The chemical shifts of the
23
Na spectra were referred to the
sample with the smallest NaOH concentration at the highest temperature
(6%, 20 C), so that all shift values are positive.
The spin-lattice relaxation times T
1
were measured with the inversion
recovery method. Control measurements of the transversal relaxation
times T
2
with the CPMG-method resulted in an agreement with T
1
at all
concentrations and temperatures in the limits of the measuring errors, so
that the evaluation and discussion without T
2
could be done.

III- RESULTS AND DISCUSSION
III.1-
23
Na measurements
In liquids and in diluted solutions, which were measured here, the Na nuclei
have a large mobility. An extreme movement-conditioned line narrowing
arises. Therefore the chemical shift for
23
Na can be measured exactly. It
reacts very sensitively to the environment of the Na nucleus and is suitable
therefore very well for the indication of ion water interactions, which are the
subject of the investigations here. [1] By the movement-conditioned line
narrowing specified above the spin-lattice and spin-spin relaxation times
(T
1
and T
2
) are about alike.
The measured relaxation time is determined by the mobility of the Na
nuclei and by the electrical field gradients at the position of the nucleus.
The environment of the nucleus yields a contribution to gradients of the
electrical field at the position of the nucleus. In the here investigated

system this contribution comes from field gradients by the water molecules
located near the Na nucleus. Highly symmetrical arrangements of the water
molecules (e.g. tetrahedral or octahedronal) do not yield a contribution to
the field gradients.
In accordance with these considerations, changing or preferred
hydratations of the Na ions should identifiable by a nonlinear behaviour of
the relaxation times or the chemical shifts as a function of the NaOH
concentration.
In the concentration range between 10% and 40% NaOH expected minima
and maxima could be observed. [2]
As the Figures 1 to 4 show, only linear relations exist in the interval from
6% to 10% NaOH and also with additive ZnO from 0% to 1.5% for all
examined temperatures.
Figure 1 Results of
23
Na relaxation times for the lyes
Figure 2 Results of
23
Na relaxation times for the lyes with ZnO

6 7 8 9 10
5
10
15
20
25
30
35
40
45
20 C
10 C
0 C
-10 C
-20 C
2
3
N
a

r
e
l
a
x
a
t
i
o
n

t
i
m
e

T
1

[
m
s
]
0,0 0,4 0,8 1,2 1,6
5
10
15
20
25
30
35
40
45
2
3
N
a

r
e
l
a
x
a
t
i
o
n

t
i
m
e

T
1

[
m
s
]
20 C
10 C
0 C
-10 C
-20 C

Figure 3 Results of the
23
Na chemical shifts for the lyes

23
Na chemical shifts for the lyes with ZnO

III.2-
1
H measurements
The relaxation behaviour of the protons in the water molecule is determined
predominantly by the direct dipolar interaction of the two protons. That is a
difference to the quadrupole relaxation of the
23
Na-nuclei. In the case of
extreme narrowing (
c
1) the relaxation time is indirectly proportional to the
middle thermal correlation time
c
.
The relaxation behaviour of the water protons in the caustic solutions is a
single-component exponential one at all examined concentrations and
temperatures. This means that either all protons have the same interaction
and mobility or (compared to the aquisition time and the relaxation time) a very
fast exchange between possible different states is present. This results in a
6 7 8 9 10
0,0
0,2
0,4
0,6
0,8
1,0
1,2
20 C
10 C
0 C
-10 C
-20 C
2
3
N
a

c
h
e
m
i
c
a
l

s
h
i
f
t

[
p
p
m
]
0,0 0,4 0,8 1,2 1,6
0,0
0,2
0,4
0,6
0,8
1,0
1,2
2
3
N
a

c
h
e
m
i
c
a
l

s
h
i
f
t

[
p
p
m
]
20 C
10 C
0 C
-10 C
-20 C

middle relaxation time, weighted whit the relative fractions of the water
molecules into the corresponding states.
The measured relaxation times T
1
are represented in Figure 5 as a function of
NaOH concentration and temperature. An extrapolation of the T
1
values (at 10
and 20 C) to the NaOH concentration 0% results nearly exactly in the
relaxation times of pure water. The reduction of the relaxation times with
increasing NaOH concentration can be explained qualitatively by an increase
of the correlation times (dicrease of the mobility, viscosity increase) and/or a
shift of the fractions of the water in different states (free, bound). We could not
be observe a dependency on the ZnO concentration (Figure 6).
The chemical shifts of the protons (Figures 7 and 8) show likewise only very
small changes as a function of the concentrations. That permits no reference
to any changing of water structures within the examined concentration range.

Figure 5 Results of
1
H relaxation times for the lyes

Figure 6 Results of
1
H relaxation times for the lyes with ZnO

6 7 8 9 10
0,2
0,4
0,6
0,8
1,0
1,2
1,4
1,6
1,8
2,0
2,2
2,4
20 C
10 C
0 C
-10 C
-20 C
1
H

r
e
l
a
x
a
t
i
o
n

t
i
m
e

T
1

[
s
]
0,0 0,4 0,8 1,2 1,6
0,2
0,4
0,6
0,8
1,0
1,2
1,4
1,6
1,8
2,0
2,2
2,4
20 C
10 C
0 C
-10 C
-20 C
1
H

r
e
l
a
x
a
t
i
o
n

t
i
m
e

T
1

[
s
]

1
H chemical shifts for the lyes

1
H chemical shifts for the lyes with ZnO

6 7 8 9 10
4,8
4,9
5,0
5,1
5,2
5,3
5,4
5,5
20 C
10 C
0 C
-10 C
-20 C
1
H

c
h
e
m
i
c
a
l

s
h
i
f
t

[
p
p
m
]
0,0 0,4 0,8 1,2 1,6
4,9
5,0
5,1
5,2
5,3
5,4
5,5
20 C
10 C
0 C
-10 C
-20 C
1
H

c
h
e
m
i
c
a
l

s
h
i
f
t

[
p
p
m
]

IV- SUMMARY
Changing or preferred hydratations of the Na ions should to be identifiable
by a nonlinear characteristic of the relaxation times or the chemical shifts
as a function of the NaOH concentration.
As the Figures 1 to 4 show, only linear relations exist in the interval from
6% to 10% NaOH and also with the additive ZnO for all examined
temperatures.
In the concentration range between 10% and 40% NaOH expected minima
and maxima could be observed for
23
Na earlier.
The relaxation behaviour of the water protons in the caustic solutions is a
single-component one at all examined concentrations and temperatures
(Figures 5 and 6).
The chemical shifts of the protons (Figures 7 and 8) show likewise only
very small changes, which do not permit a reference to any changing of
water structures within the examined concentration range.

V- LITERATURE
[1] G. J. Templeman et al.: J. Amer. Chem. Soc. 94, 5578 (1972)
[2] J. Kunze, A. Ebert et al.: Z. phys. Chemie, Leipzig 266, 49 (1985)

Appendix
Tables: Values of the relaxation times and chemical shifts

1
H-measurements of leys

1
H - T1 [s]
concentraton;
NaOH [%] 20 C 10 C 0 C -10 C -20 C

6 2,09 1,56 1,23 0,89 0,59
7 1,98 1,49 1,18 0,83 0,56
8 1,86 1,38 1,08 0,78 0,52
9 1,73 1,32 1,04 0,74 0,49
10 1,61 1,24 0,96 0,68 0,46

1
H chemical shift [ppm]
concentraton;
NaOH [%] 20 C 10 C 0 C -10 C -20 C

6 4,928 5,037 5,144 5,257 5,379
7 4,934 5,042 5,142 5,254 5,367
8 4,943 5,045 5,146 5,251 5,360
9 4,952 5,048 5,148 5,250 5,356
10 4,961 5,058 5,153 5,253 5,353

1
H-measurements of leys with ZnO (8% NaOH)

1
H - T1 [s]
concentraton;
ZnO [%] 20 C 10 C 0 C -10 C -20 C

0 1,86 1,38 1,08 0,78 0,52
0,2 1,86 1,37 1,04 0,8 0,55
0,5 1,89 1,43 1,05 0,78 0,53
1 1,9 1,45 1,08 0,77 0,53
1,5 1,87 1,4 1,04 0,76 0,52

1
H chemical shift [ppm]
concentraton;
ZnO [%] 20 C 10 C 0 C -10 C -20 C

0 4,943 5,045 5,146 5,264 5,36
0,2 4,936 5,044 5,15 5,254 5,363
0,5 4,936 5,043 5,143 5,251 5,359
1 4,925 5,028 5,137 5,244 5,349
1,5 4,915 5,019 5,126 5,231 5,341

23
Na- measurements of leys

23
Na - T1 [ms]
concentraton;
NaOH [%] 20 C 10 C 0 C -10 C -20 C

6 39,3 31,5 23,9 17,1 11,3
7 37,5 29,9 22,5 16,1 10,6
8 35,4 28,1 21,1 15 9,8
9 33,7 26,5 19,9 14 9
10 31,8 24,9 18,6 12,9 8,3

23
Na chemical shift [ppm]
concentraton;
NaOH [%] 20 C 10 C 0 C -10 C -20 C

6 0 0,203 0,409 0,622 0,836
7 0,063 0,271 0,476 0,686 0,896
8 0,143 0,349 0,555 0,765 0,972
9 0,213 0,417 0,625 0,831 1,042
10 0,29 0,496 0,697 0,906 1,118

23
Na- measurements of leys with ZnO (8% NaOH)

23
Na - T1 [ms]
concentraton;
ZnO [%] 20 C 10 C 0 C -10 C -20 C

0 35,4 28,1 21,1 15 9,8
0,2 33,7 27 20,7 15 9,77
0,5 33,8 26,8 20,3 14,5 9,55
1 32,6 25,8 19,5 13,9 9,12
1,5 31,4 24,9 18,8 13,4 8,81

23
Na chemical shift [ppm]
concentraton;
ZnO [%] 20 C 10 C 0 C -10 C -20 C

0 0,143 0,349 0,555 0,765 0,972
0,2 0,156 0,361 0,564 0,762 0,974
0,5 0,16 0,366 0,571 0,778 0,983
1 0,166 0,38 0,59 0,797 1,002
1,5 0,183 0,388 0,596 0,803 1,009

Dr. H.-P. Fink Dr. A. Ebert
Division Director Project Manager

RESUME
Il est connu que la cellulose peut-tre dissoute dans des solutions aqueuses dhydroxyde de
sodium dans un petit domaine de concentrations (6-10%NaOH) et de tempratures (-10C/-
4C). Les solutions de cellulose ainsi obtenues ne sont pas stables et glifient en fonction du
temps et de la temprature. Il a galement t montr que la prsence dadditifs tels quoxyde
de zinc et ure amliore la dissolution des fibres de cellulose et la stabilit des solutions.
Afin de comprendre la structure des solutions de cellulose dans NaOH/eau avec et sans additifs,
nous avons tudi les diagrammes de phase des solvants dune part et linfluence de la cellulose
sur les thermogrammes de DSC dautre part. Nous avons ainsi pu dterminer (i) la limite de
dissolution de la cellulose dans des solutions de NaOH/eau et NaOH/ure/eau et (ii) la nature
des interactions entre les diffrentes espces en solution : chanes de cellulose, NaOH, ure, ZnO
et eau.
Nous avons galement tudi les proprits rhologiques des solutions de cellulose dans
NaOH/eau, NaOH/ure/eau et NaOH/ZnO/eau. La glification des solutions de cellulose est
dune part retarde par la prsence dadditifs et correspond une diminution de la qualit du
solvant quand la temprature augmente. Cependant le mcanisme de glification reste le mme
pour les trois solvants tudis.
Pour finir, ltude des proprits mcaniques, et notamment la rsistance rupture, sur des objets
de cellulose rgnre a montr limportance des paramtres de glification et de rgnration.

ABSTRACT
It has been reported that cellulose can be dissolved in aqueous sodium hydroxide solutions in a
narrow range of NaOH concentrations (6-10%) and temperatures (-10C/-4C). Cellulose
solutions obtained are not stable and gelation occurs with time and temperature increase. It has
been also showed that the presence of additives, such as urea or zinc oxide, in solvent system
improves the dissolution of cellulose fibres and the solution stability.
In order to understand the structure of cellulose solutions in NaOH/water without or with
additives, we first studied phase diagrams of solvents and then the influence of the addition of
cellulose in these solvents on the DSC thermograms. We determined (i) the limit of cellulose
dissolution in NaOH/water and in NaOH/urea/water and (ii) the nature of interactions
occurring between cellulose chains, NaOH, urea, ZnO and water molecules.
We also studied rheological properties of cellulose solutions in NaOH/water, NaOH/urea/water
and NaOH/ZnO/water. The presence of additives delays gelation (in time and in temperature)
but does not change its mechanism. Gelation of cellulose solutions is due to a decrease of solvent
quality which favours the cellulose-cellulose interactions with the temperature increase.
Finally, the study of mechanical properties, and especially the measurements of the stress at
rupture of regenerated cellulose objects showed the great influence of gelation and regeneration
parameters on the properties of the final products.

NaOH Alkali

Transféré par

Droits d'auteur :

Formats disponibles

NaOH Alkali

Transféré par

Informations du document

Copyright

Formats disponibles

Partager ce document

Partager ou intégrer le document

Options de partage

Avez-vous trouvé ce document utile ?

Ce contenu est-il inapproprié ?

Droits d'auteur :

Formats disponibles

NaOH Alkali

Transféré par

Droits d'auteur :

Formats disponibles

ECOLE DES MINES

[ATA1984], the latter being thermodynamically more stable.

, which has a triclinic unit with one chain, and of I

, and (b) O3-HO5 and

depends on the cellulose origin [ATA1999].

is the cellulose molecular weight given in the article and 162 is

adhesion promoters, from Hls France, were used:

= is the fraction of water released and m

: the shear rate (s

Vous aimerez peut-être aussi