Research Article

Received: 22 November 2009, Revised: 25 December 2009, Accepted: 4 January 2010, Published online in Wiley Online Library: 29 April 2010

(wileyonlinelibrary.com) DOI: 10.1002/pat.1670

The synthesis and properties of a reactive

flame-retardant unsaturated polyester resin
from a phosphorus-containing diacid
Chen Zhang a,b, Jun Yi Huang a,c, Shu Mei Liua * and Jian Qing Zhaoa
A transparent flame-retardant unsaturated polyester resin (FR-UPR) was obtained by reacting propylene glycol (PG) with
maleic anhydride (MA), phthalic anhydride (PA), and 9,10-dihydro-10[2,3-di(hydroxy carbonyl)propyl]-10-
phosphaphenanthrene-10-oxide (DDP) synthesized from 9,10-dihydro-9-oxa-10-phosphaphenanthrene 10-oxide (DOPO)
and itaconic acid (ITA). The chemical structure of the resulting FR-UPR was confirmed by FTIR, 1H-NMR and 31P-NMR. The
average molecular weight and viscosity of the FR-UPR were determined by gel permeation chromatography (GPC) and
viscometer, respectively. Thermal stability was studied by thermogravimetric analysis (TGA) both in air and nitrogen to
determine the thermal decomposition mechanism, and the apparent activation energy (Ea) was calculated by both the
Kissinger and Ozawa methods. Compared to unsaturated polyester resin (UPR), the higher Ea of FR-UPR3 implied an
improved thermal stability. According to variations of the limited oxygen index (LOI) values, the UL 94 rating of vertical
burning test and scanning electron microscopy (SEM) photographs of char residues, the flame retardance of cured FR-UPR
was enhanced with increasing DDP content. The study of fire reaction tests, using a cone calorimeter, suggested that
there was a significant reduction of flammability in the FR-UPR. Copyright ß 2010 John Wiley & Sons, Ltd.

Keywords: unsaturated polyester resin; synthesis; thermal stability; flame retardance

INTRODUCTION phase, and become an important replacement for halogen flame

retardants.[9] Since Endo[10] firstly reported 9,10-dihydro-10[2,3-
Unsaturated polyester resin (UPR) is one of the most important di(hydroxycarbonyl)propyl]-10-phosphaphenanthrene-10-oxide
thermosetting materials, and is used alone or in fiber-reinforced (DDP), which is a phosphorus-containing diacid monomer, many
composites in the field of naval construction, offshore appli- researchers have focused on its particular properties and
cations, water pipes, coatings, building construction, automo- application.[11–16] Chang et al.[11,16] synthesized poly(ethylene
tives, etc.[1,2] The general grade UPR is a blend of styrene with the terephthalate)-co-poly(ethylene DDP)s (PET-co-PEDDP)s with
condensation product of 1,2-propylene glycol (PG) with maleic high thermal stability and a high limit oxygen index (LOI) value
anhydride (MA) and phthalic anhydride (PA). >33. Lin et al.[12] prepared epoxies with DDP, which showed
The major drawback of UPR is its flammability. For example, a good thermal stability and achieved flame-retardant UL 94 V0
general grade UPR containing styrene has a low limiting oxygen specification grade, when 22.3% DDP was incorporated.
index (LOI) value of 19.[3] The high flammability of any polymeric
system brings out a considerable hazard. Two methods have been
used to make UPR flame-retardant and safe: the introduction of
flame-retardant elements into the polyester structure using a reactive * Correspondence to: S. M. Liu, College of Materials Science and Engineering,
flame retardant[4]; the blending of non-reactive flame retardants with South China University of Technology, Guangzhou 510640, China.
UPR.[5] However, the use of non-reactive flame retardants has several E-mail: [email protected]
disadvantages such as poor compatibility, migration, reduction of
a C. Zhang, J. Y. Huang, S. M. Liu, J. Q. Zhao
the mechanical properties, transparency, and electrical properties.[6] College of Materials Science and Engineering, South China University of
The goal of current research is to develop and employ reactive flame Technology, Guangzhou 510640, China
retardants to enhance the flame retardance of polymers. Mansour
b C. Zhang
et al.[7] synthesized two kinds of UPR with high LOI values containing
School of Chemistry and Environment, South China Normal University,
halogen (bromine) in the backbone of the polymer chain. Guangzhou 510006, China
Although halogen atoms (such as bromine and chlorine) form
the most widely applied flame retardant materials, the European c J. Y. Huang
The Key Laboratory of Polymer Processing Engineering, Ministry of Education,
Community (EC) has proposed to restrict the use of halogen flame
South China University of Technology, Guangzhou 510640, China
retardants because highly toxic and potentially carcinogenic
halogenated furans and dioxins may form during combustion.[8] Contract/grant sponsor: Specialized Research Fund for the Doctoral Program of
Higher Education
For this reason, halogen-free flame retardants for polymeric
contract/grant number: 200805611098.
materials have attracted increasing attention in recent years. Contract/grant sponsor: Cooperation Project of Industry-University-Research of
Phosphorus-containing flame retardants give off less toxic Guangdong Province and Ministry of Education of P.R.China
combustion products, influence the reaction in the condensed contract/grant number: 2008B090500007.
1768

Polym. Adv. Technol. 2011, 22 1768–1777 Copyright ß 2010 John Wiley & Sons, Ltd.
FLAME-RETARDANT UNSATURATED POLYESTER RESIN

Since UPR is like PET and epoxy with high oxygen content in
Table 1. Synthesis recipes
the molecular structure, DDP will probably be effective on the
enhancement of flame retardance for UPR. And the structure of
DDP allows for the facile incorporation of different phosphorus Sample ID
contents as an integral part of the molecular chain of UPR by melt
polycondensation. It was therefore chosen as a reactive flame Substrate (g) UPR FR-UPR1 FR-UPR2 FR-UPR3
retardant for the synthesis of phosphorus-containing UPR to
improve flame retardance and to avoid the migration problems MA 25.0 25.0 25.0 25.0
of blending flame retardants. flame-retardant unsaturated PA 36.7 28.7 20.6 12.6
polyester resin (FR-UPR) was based on MA, PA, DDP, and PG. DDP 0 10.0 20.0 30.0
The synthesized DDP and FR-UPR were identified by FTIR, PG 49.8 47.2 44.7 42.2
1
H-NMR, and 31P-NMR. Average molecular weight, apparent St 60.7 60.9 61.2 61.5
viscosity, thermal stability, and flame retardance were deter- DCP 1.5 1.5 1.5 1.5
mined, and fire reaction tests were examined. HQ 0.015 0.015 0.015 0.015

of xylene were added and heated to 1508C. To the stirring mixture,

EXPERIMENTAL itaconic acid (130 g) was added incrementally over 3 hr. The crude
product was purified by crystallizing from acetone. The melting point
Materials of the purified product is 193.68C, and the conversion rate is 82.8%.
9,10-dihydro-9-oxa-10-phosphaphenanthrene 10-oxide (DOPO, Anal. Cal.: C, 58.97%; H, 4.37%; Found: C, 59.21%; H, 4.35%.
trade name of HCA) was supplied by Sanko Co., Ltd, Japan.
Itaconic acid was obtained from Shanghai Experiment Reagent Polymerization procedures of UPR
Co., Ltd, China. 1, 2-propanediol(PG), MA, PA, itaconic acid (ITA),
styrene(St), hydroquinone(HQ), and Dicumyl peroxide (DCP) were Reactions were carried out in a four-necked round-bottom flask
of analytical grade and used without further purification. (500 ml) fitted with a condenser, mechanical stirrer operated at
approximately 300 rpm, and a nitrogen gas inlet to facilitate the
removal of water and oxygen from system. The temperature of
Synthesis of DDP
the reactants was continuously monitored and controlled using a
The DDP were synthesized and purified according to the method mantle heater. A detailed recipe used in this study is given in
used by Chang et al.,[16] as well as Wang et al.[13] In a four-necked Table 1. The preparation rout of FR-UPR is shown in Fig. 1.
round-bottom 1000 ml flask equipped with a mechanical stirrer, The polyesterification reaction was conducted by the non-
condenser, nitrogen inlet, and thermowell, 216 g of DOPO and 500 ml catalyzed fusion method.[17] The reactions were carried out by

Figure 1. Preparation route of DDP and FR-UPR.

1769

Polym. Adv. Technol. 2011, 22 1768–1777 Copyright ß 2010 John Wiley & Sons, Ltd. wileyonlinelibrary.com/journal/pat
 C. ZHANG ET AL.

heating the reactants from room temperature to 1708C, under a N2

atmosphere for 1.5 hr. Then the temperature was held at 1808C for
4 hr, raised to 1908C, and maintained at this temperature. The
progress of the reaction was determined by taking small (relative
to the reaction mass), accurately weighed samples (approximately
0.5 g) at appropriate time intervals and measuring the acid
numbers according to the method described by Curtis et al.[18]
Each sample was dissolved in approximate 50 ml of a toluene–
methanol mixture (1:1) and titrated with 0.1 N methanolic
potassium hydroxide using phenolphthalein as indicator. The
reaction was finished until the acid number decreased under
20 mg KOH/g. The reaction product was then allowed to cool, and
dissolved in 40% styrene (by weight of styrene with respect to the
overall resins), which acted as a solvent and a crosslinker for the
unsaturated polyesters. DCP (as initiator) and HQ (as free radical
inhibitor) were subsequently added.
After pouring into a steel mold, the UPR solution was heated in
stages: 1008C (2 hr); 1208C (2 hr); and 1408C (1 hr). Transparent Figure 2. FTIR spectrum of DDP.
samples were obtained by cooling slowly in air.

Characterization Fig. 2, the strong absorption around 1180 cm1 corresponds to

the vibration of P –– O. DDP also shows strong absorptions at
The 31P-NMR and 1H-NMR spectra were determined by a Bruker 920 cm1 corresponding to P–O–C (aromatic) stretching, P–C
AVANCE Digital 400 MHz NMR. Element analyses (EAs) were stretching absorption around 1431 cm1, C – – O absorption
performed by a Thermal Flasher Scientific Flash EA112 element around 1710 cm1, and –COOH absorption around
analyzer. FTIR spectra were recorded on a Bruker VERTEX 70 FTIR 3200–3400 cm1.
spectrophotometer after the samples were prepared as potassium The 1H-NMR spectrum of DDP is presented in Fig. 3 with
bromide pellets. The average molecular weight and the molecular dimethyl sulfoxide (DMSO) as solvent and TMS (tetramethylsi-
weight distributions of the polyesters were measured by Waters lane) as internal standard. The chemical shift data of DDP are:
2414 gel permeation chromatography (GPC). The thermogravi- d(H1, H4) ¼ 12.40 ppm, d(H2) ¼ 2.63 ppm, d(H3) ¼ 3.31 ppm,
metric analysis (TGA) was performed on a NETZSCH TG 209 F1 d(H5) ¼ 2.36 ppm. The d of 7.24–8.18 ppm accounts for chemical
thermal analyzer under either flowing nitrogen or an air shifts of H on the diphenyl. This is in good agreement with the
1
atmosphere at a heating rate of 108C/min (or 5, 20, 408C/min). H-NMR spectra reported by Chang et al.[11,13]
The TGA results were corrected for blank experiments. The
reproducibility of the results in the blank measurements was better
Characteration of FR-UPR
than
0.01 mg. All samples were measured in one time. The
limited oxygen index (LOI) was tested by an oxygen index/FTA The synthesized UPR was purified by dissolving in an
instrument from Fire Testing Technology Co. Ltd according to appropriate amount of non-polar organic solvent (benzene,
ASTM 2863-98. The vertical burning test was carried out by using a toluene, CCl4, or CH2Cl2) and washing thoroughly several times
UL 94 flammability meter from Fire Testing Technology Co. Ltd with distilled water.[19] The purified UPR was characterized
according to UL 94 classification. 3.2 mm thickness sample bars following evaporation of the solvent by rotary evaporator, and
were made in a standard steel mold. Observation of the char drying in a vacuum oven[20].
surface morphology was performed by a NOVA NANOSEM In Fig. 4, both FR-UPR3 and UPR have the absorption around
430 scanning electron microscopy (SEM). The voltage used was 1290 cm1 indicating an ester group. Compared to DDP, the
typically in the range of 1–30 kV. Energy dispersive spectrometer
(EDS) analysis was performed using an ADD350 þ HKL Fast EBSD
system attached to a SEM. The apparent viscosity of UPR was
measured with a Brookfield DV-1t viscometer at 258C. A cone
calorimeter (FTT—Fire Testing Technology Ltd, UK) was used with
an incident heat flux of 50 kW/m2 using the cone shaped heater.
According to ISO 5660, the heat release rate, mass loss rate and
specific extinction area were obtained from the cone calorimeter
experiment using online software. The cone data reported here
were obtained by burning samples in the horizontal position with a
100  100 mm2 exposed surface.

RESULTS AND DISCUSSION

Characteration of DDP
The synthesis route of DDP is shown in Fig. 1. DDP was
characterized by elemental analysis, FTIR and NMR. As shown in Figure 3. 1H-NMR spectrum of DDP.
1770

wileyonlinelibrary.com/journal/pat Copyright ß 2010 John Wiley & Sons, Ltd. Polym. Adv. Technol. 2011, 22 1768–1777
FLAME-RETARDANT UNSATURATED POLYESTER RESIN

Table 2. Assignment of the resonance peaks of FR-UPR3

Proton Chemical shift (ppm)

a 4.33, 4.26, 4.15, 4.06, 3.64

b 1.33, 1.23
c 3.28, 2.93, 2.61
d 2.37
e 5.28, 5.05, 4.78
f 7.93,7.71
g 7.52–6.87

that MA participated in the reaction. The peaks at around

3.64–4.33 ppm (Ha) are the chemical shifts of the hydrogen on
Figure 4. FTIR spectra of FR-UPR3 and UPR. the ester group, and chemical shifts at 2.61–3.28 ppm(Hc)
account for the hydrogen on the carbons connected to the
carbonyl groups. The hydrogen on –CH2 adjacent to
carbonyl absorption band shifts to a considerably higher phosphorus shows chemical shifts at 2.37 ppm(Hd) and the
wavenumber (from 1711 to 1727 cm1), because the ester chemical shifts at around 1.33 ppm(Hb) is assigned to the
carbonyl absorbs at higher frequency than the corresponding hydrogen on methyl groups.
carboxyl one. FR-UPR3 gives the absorption at 918 cm1 The difference of 31P-NMR spectra between DDP and FR-UPR3
corresponding to P–O–C (aromatic) stretching, which is the is small as shown in Fig. 6 since the ring containing phosphorus in
characteristic absorption of FR-UPR or DDP, but does not exist in DDP is on the side chain and the variation of chemical shift
UPR. influenced by the structure of the main chain is not obvious.
The chemical structure and the 1H-NMR spectrum of According to the characterization above, DDP has been
FR-UPR3 are illustrated in Fig. 5, and resonance peak incorporated into the UPR chain completely.
assignments of 1H-NMR are shown in Table 2. For DDP, the
chemical shift at d ¼ 12.40 ppm is assigned to the two
Average molecular weight and apparent viscosity
hydrogens of carboxylic acid. The disappearance of the
chemical shift in FR-UPR3 indicates that DDP takes part in As shown in Table 3, the average molecular weight (Mn and
the polycondensation completely. At the same time, the Mw) of FR-UPR is lower than that of UPR and decreases with an
appearance of the peaks at 5.28, 5.05, 4.78 ppm (He), which increase in DDP content. Under the same polyesterification
account for chemical shifts of H on the C –– C bond, indicates condition, DDP reduces the molecular weight of the product

Figure 5. 1H-NMR spectrum of FR-UPR3.

1771

Polym. Adv. Technol. 2011, 22 1768–1777 Copyright ß 2010 John Wiley & Sons, Ltd. wileyonlinelibrary.com/journal/pat
 C. ZHANG ET AL.

31
Figure 6. P-NMR spectra of DDP and FR-UPR3.

Table 3. Average molecular weight and properties of UPR and FR-UPR

Sample ID Content of phosphorus (%) Mn Mw Mz d(Mw =Mn ) Acid number h (mPas)

UPR 0 3688 6149 10,198 1.67 15 932

FR-UPR1 0.54 3656 6076 10,330 1.66 18 869
FR-UPR2 1.08 3560 5940 9656 1.67 16 848
FR-UPR3 1.62 3379 5173 8029 1.53 15 832
Mn : number-average molecular weight; Mw : weight-average molecular weight; Mz : Z-average molecular weight.

because the biphenyl configuration of DDP has a large steric TGA shows the differences in thermal behavior of different
effect. All of FR-UPR have a middle apparent viscosity, which is UPRs. All values of the initial decomposed temperature (T5) of
low enough to assure proper impregnation and easy to FR-UPR are higher than UPR in N2, and the TG curves shift toward
process. The viscosity gradually decreases with increasing higher temperature with increasing DDP content in both N2 and
DDP content. air, indicating that FR-UPR is characterized by slightly higher
thermal stability than UPR. TGA confirms that the chemical
structure of polyesters influences the thermal stability of cured
Thermal stability
UPR. And, as more DDP is incorporated in the chain of UPR, better
The thermal degradation of cured FR-UPR has been studied and thermal stability is shown by the chemical modification of
experimentally compared to cured UPR. Their thermograms in N2 polyesters. This result differs from the reports of other
and in air are shown in Fig 7 and 8, respectively. The characteristic researchers’,[21,22] where a phosphorus-containing monomer
temperature obtained from TGA is summarized in Table 4. introduced into epoxy system reduced its thermal stability.

Figure 7. TG curves of UPR with varying DDP contents in N2. Figure 8. TG curves of UPR with varying DDP contents in air.
1772

wileyonlinelibrary.com/journal/pat Copyright ß 2010 John Wiley & Sons, Ltd. Polym. Adv. Technol. 2011, 22 1768–1777
FLAME-RETARDANT UNSATURATED POLYESTER RESIN

Table 4. TGA results of UPR and FR-UPR

Thermal stability data Char yield (%)

In N2 In air 6508C

Sample ID T5 T10 Tmax T5 T10 Tmax1 Tmax2 In N2 In air

UPR 247 293 365 259 291 362 574 3.4 2.5
FR-UPR1 256 306 373 251 295 380 562 7.2 3.5
FR-UPR2 252 307 378 251 300 382 561 8.5 5.9
FR-UPR3 264 315 382 265 312 377 589 10.3 12.0
T5: the initial decomposed temperature; T10: a temperature at 10% weight loss; Tmax: the maximum decomposition temperature in the
thermal degradation in N2; Tmax1: the maximum decomposition temperature in the first stage of thermal degradation in air; Tmax2: the
maximum decomposition temperature in the second stage of thermal degradation in air.

The thermal degradation pattern[23] of all copolymers exhibits Apparent activation energy of thermal degradation
one degradation step in N2, but shows the trend of two separate
degradation steps in air, with two maximum decomposition To estimate the kinetic parameters of the degradation process,
temperatures (Tmax1 and Tmax2) in the DTG curves calculated from dynamic thermogravimetric experiments were carried out in
Fig. 9, as noted in Table 4. This implies that the existing oxygen nitrogen at heating rates (b) of 5, 10, 20, and 408C/min (Fig. 9),
changed the thermal degradation process of UPR. respectively. The apparent activation energy (Ea) of the different

Figure 9. TG curves of UPR and FR-UPR3 at different values of heating rate in N2.

Figure 10. Application of the Kissinger method to the experimental data of UPR and FR-UPR3.
1773

Polym. Adv. Technol. 2011, 22 1768–1777 Copyright ß 2010 John Wiley & Sons, Ltd. wileyonlinelibrary.com/journal/pat
 C. ZHANG ET AL.

Figure 11. Plots of lg b versus 1000/T at different conversions of Figure 12. Activation energy (Ea) as a function of the degree of con-
FR-UPR3, calculated by the Ozawa method. version for the decomposition process of UPR and FR-UPR3 calculated by
the Ozawa method

degradation stages of UPR and FR-UPR3 can be evaluated by a result of the incorporation of DDP into the UPR chain.
the Kissinger[24] differential method and the Ozawa[25] integral This is consistent the activation energy calculated with the
method, respectively. Kissinger method.
With the Kissinger method and the temperature correspond-
ing to the maximum weight-loss rate (Tp) from the DTG curves, Ea
Flame retardance
for the decomposition of the UPR and FR-UPR3 systems was
estimated from the slope of a straight line of lg(b/T2p ) versus 1000/ It is clear from the LOI data in Table 5 that the flame retardance of
Tp, as shown in Fig. 10. Ea was calculated to be 150.4 kJ/mol for cured FR-UPR increases with phosphorus content in the
UPR and 174.5 kJ/mol for FR-UPR3. copolyesters. The enhancement of the flame retardance of
The Ozawa calculation method is applicable to all points of FR-UPR can also be observed from the vertical burning test
the TG curves, thus they are capable of providing reasonably results. FR-UPR becomes more difficult to flame with increasing of
reliable data, and are often used in analysis of thermal phosphorus content. 1.62% phosphorus in the resin allows
degradation mechanism. Ea at different degrees of conversion FR-UPR3 to achieve a V-0 rating in the UL 94 test.
has been also evaluated with the Ozawa method. Figure 11 To further investigate how a phosphorus moiety in FR-UPR
shows nearly parallel fitted straight lines, indicating the improves flame retardance, a NOVA NANOSEM with an EDS
applicability of this method to our system in the conversion analysis system was used to examine the variations of char
range studied. On the basis of isoconversional analysis, Ea of residues on the surface of different burned resins. Char residues
the thermal decomposition process at different conversion for SEM scan were prepared as described previously.[27] In Fig. 13,
calculated by the Ozawa method from TG curves is provided clear differences in the morphology of three char residues have
in Fig. 12. The calculated Ea values of UPR are similar to been observed. The residue particles of UPR are piled loosely.
those reported by other researchers.[26] Figure 12 illustrates FR-UPR2 forms a layer of continuously and smoothly consolidated
the Ea values of FR-UPR3 are consistently higher than those char with a few small pores (Fig. 13b). In FR-UPR3, the formation
of UPR. In the thermal decomposition process, it is necessary of a compact whole carbon layer demonstrates that phosphorus
for FR-UPR3 to overcome a higher energy barrier. This plays an important role in flame retardation.[28] During burning,
implies that the degradation mechanism was changed as phosphorus in polymers is converted to phosphoric acid and

Table 5. Cone calorimeter results and flame retardance for cured UPR and FR-UPR

Mass
residue TTI PHRR TTP THR Hc SEA average
Sample (%) (sec) (kW/m2) (sec) (kJ) (MJ/kg) (m2/kg) LOI UL 94

UPR 5.1 6 655 170 146 28.5 787 19 No-distinguished

FR-UPR1 5.5 8 495 170 112 17.9 979 22 V2
FR-UPR2 6.1 8 380 215 89 13.9 1099 26 V1
FR-UPR3 9.0 12 375 155 83 11.6 1124 29 V0
TTI: time to ignition; PHRR: peak heat release rate; TTP: time to peak (PHRR); THR: total heat released; SEA: specific extinction area; Hc:
effective heat of combustion.
1774

wileyonlinelibrary.com/journal/pat Copyright ß 2010 John Wiley & Sons, Ltd. Polym. Adv. Technol. 2011, 22 1768–1777
FLAME-RETARDANT UNSATURATED POLYESTER RESIN

Figure 13. SEM photographs of char residue of (a) UPR, (b) FR-UPR2, and (c) FR-UPR3. 2k.

further thermal decomposition leads to the formation of dots in Fig. 14) have been taken from EDS observations at 10k
polyphosphoric acid.[9] The polyphosphoric acid esterifies and magnification before and after burning. In Fig. 14, the
dehydrates the pyrolyzing polymer and a simultaneously forming distribution density of phosphorus element in the residue is
phosphorus-rich carbonaceous layer further inhibits pyrolysis larger than that before burning and the phosphorus content on
reactions. This protective layer is resistant to even higher the surface increases from 1.8 to 9.6% after burning. These
temperatures and shields the underlying polymer from attack by results along with the variation of Ea of degradation
oxygen and radiant heat. mentioned-above reveal that the phosphorus moiety confers
On the other hand, the distribution and amounts of flammability resistancy mainly by the condensed phase
phosphorus on the surface of FR-UPR3 (shown as the bright flame-retardant mechanism.

Figure 14. SEM photographs of P-mapping and EDS analysis of FR-UPR3 (a) before burning and (b) after burning. 10k.
1775

Polym. Adv. Technol. 2011, 22 1768–1777 Copyright ß 2010 John Wiley & Sons, Ltd. wileyonlinelibrary.com/journal/pat
 C. ZHANG ET AL.

Fire reaction tests different phosphorus contents. UPR burns rapidly and the HRR
peak is much larger than those of the FR-UPRs. The difference in
The cone calorimeter is a bench-scale fire test apparatus which THR curves is presented in Fig. 15(b), and indicates that an
provides a wealth of information on combustion behavior.[29] increase of phosphorus content reduces the total amount of fuel
Some cone calorimeter results have been found to correlate well available for combustion. At the end of burning, all FR-UPRs
with those obtained from large scale fire tests, so that it can be produce a greater of amount of residue than UPR, particularly
used to predict the behavior of materials in a real fire.[30] It FR-UPR3 for which the residue amount is 1.8 times that of UPR as
quantifies fire parameters such as peak heat release rate (PHRR), shown in Fig. 15(d), indicating an improvement in the thermal
heat release rate (HRR), total heat release (THR) and smoke stability of FR-UPR. The increasing amount of residue with
emission (SEA—specific extinction area). phosphorus content is in agreement with the results obtained
In the present study, sample thickness was maintained at from thermogravimetric analysis.
5
0.5 mm to minimize any thickness effects.[31] Table 5 shows Smoke extinction area (SEA), which represents the relation-
some data obtained from the cone calorimeter. All FR-UPR ship between the volatile property and smoke emission, is
samples have higher TTI (time to ignition) values than UPR, and another very valuable parameter.[32] It can be observed from
FR-UPR3 has the highest value. An increase in phosphorus Fig. 15(c) and Table 5 that the FR-UPRs released higher
content from 0 to 1.62% decreases PHRR values from 655 to quantities of smoke from volatiles than UPR during all burning
375 kW/m2 (a 42.7% reduction) and THR values from 146 to tests. This may be attributed to the contribution of an aromatic
83 kJ (a 43.2% reduction). Hc is the quantity of heat produced structure from DDP; aromatic compounds produce more smoke
by the combustion of unit quantity of a material and is in a fire. The increase of released smoke together with char
measured in the cone throughout the burn period from the residues during burning results is not sufficient flaming of
heat release rate divided by the mass loss at each time. In FR-UPR. This is the main reason why THR, PHRR, and Hc are
FR-UPR3 sample, Hc is reduced to 40.7% of the control sample reduced greatly. However, the increase of SEA can be avoided by
(UPR) value. introducing modified layered double hydroxides (LDH) into UPR
Selected results are graphically shown in Fig. 15(a–d). Figuire to improve the strength of the char residue layer formed during
15(a) shows curves of HRR against time for UPR and FR-UPR with the burning process[33].

(a) 700 (b) 160

UPR 140
600
FR-UPR1
2
Heat Release Rate kW/m

FR-UPR2
2

120
MJ/m

500 FR-UPR3
100
400
Total Heat Released

80
300
60
UPR
200
40 FR-UPR1
FR-UPR2
100 20 FR-UPR3

0 0
0 100 200 300 400 500 600 0 100 200 300 400 500

Time s Time s

(c) (d) 100

2100

UPR UPR
1800 FR-UPR1 FR-UPR1
m /kg

80 FR-UPR2
FR-UPR2
2

1500 FR-UPR3 FR-UPR3

Mass Residue (%)
Specific Extinction Area

60
1200

900
40

600

20
300

0 0
0 100 200 300 400 500 600 0 100 200 300 400 500 600

Time s Time s

Figure 15. Cone calorimetric results for UPR and FR-UPR with different phosphorus contents: (a) heat release rate, (b) total heat released, (c) specific
extinction area, and (d) mass residue versus time curves.
1776

wileyonlinelibrary.com/journal/pat Copyright ß 2010 John Wiley & Sons, Ltd. Polym. Adv. Technol. 2011, 22 1768–1777
FLAME-RETARDANT UNSATURATED POLYESTER RESIN

CONCLUSIONS [5] S. Nazaré, B. K. Kandola, A. R. Horrocks, Polym. Adv. Technol. 2006, 17,
294–303.
In this work, in order to improve the flame retardance of [6] R. M. Perze, J. K. W. Sandler, V. Altstädt, T. Hoffmann, D. Pospiech, M.
Ciesielski, M. Döring, J. Mater. Sci. 2006, 41, 341–353.
polymeric materials, a modified unsaturated polyester resin [7] S. H. Mansour, J. N. Asaad, S. L. Abd-El-Messieh, J. Appl. Polym. Sci.
(FR-UPR) with DDP as the reactive flame retardant was 2005, 102, 1356–1365.
successfully synthesized and studied. FR-UPR had relatively [8] S. Y. Lu, I. Hamerton, Prog. Polym. Sci. 2002, 27, 1661–1712.
lower average molecular weight and viscosity than UPR. [9] S. Hörold, Polym. Degrad. Stab. 1999, 64, 427–431.
TG curves of cured FR-UPR showed higher thermal stability [10] S. Endo, T. Kashihara, A. Osako, T. Shizuki, T. Ikegami, U.S. Patent
1978-11-28.
than UPR through the comparison of data obtained from TG [11] S. J. Chang, F. C. Chang, J. Appl. Polym. Sci. 2000, 72, 109–122.
curves. The Ea results calculated by the Ozawa integral method [12] C. H. Lin, C. Y. Wu, C. S. Wang, J. Appl. Polyml. Sci. 2000, 78, 228–235.
and the Kissinger differential method also revealed that the [13] L. S. Wang, H. B. Kang, S. B. Wang, Y. Liu, R. Wang, Fluid Phase Equilib.
thermal stability of FR-UPR3 was improved as the degradation 2007, 258, 99–107.
[14] D. Y. Wang, Y. P. Song, J. S. Wang, X. G. Ge, Y. Z. Wang, A. A. Stec, T. R.
mechanism was changed by the incorporation of DDP. Hull, Nanoscale Res. Lett. 2009, 4, 303–306.
The flame retardance of cured FR-UPR increased with [15] D. Y. Wang, X. Q. Liu, J. S. Wang, Y. Z. Wang, A. A. Stec, T. R. Hull, Polym.
increasing DDP content. Incorporating 1.62% phosphorus into Degrad. Stab. 2009, 94, 544–549.
FR-UPR allowed it to meet the UL 94 V0 rating and achieved an [16] S. J. Chang, F. C. Chang, Polymer 1998, 39, 3233–3240.
LOI value of 29. SEM photographs indicated that, for char [17] W. R. Sorenson, T. W. Campbell, Preparative Method of Polymer
Chemistry. Interscience, New York, 1968, 65–78.
residues, the carbon layer became more compact and whole with [18] L. G. Curtis, D. L. Edwards, R. M. Simons, P. J. Trent, P. T. V. Bramer, Ind.
increasing DDP content. EDS results of FR-UPR3 determined that Eng. Chem. Prod. Res. Dev. 1964, 3, 218–221.
before and after burning the phosphorus content on the surface [19] J. W. Chen, L. W. Chen, J. Appl. Polym. Sci. 1999, 73, 35–40.
increased from 1.8 to 9.6%, confirming the condensed phase [20] M. S. Farahat, A. A. Azim, M. A. Raowf, Macromol. Mater. Eng. 2000,
283, 1–6.
mechanism of flame retardation. The results of the fire reaction [21] Y. L. Liu, C. S. Wu, Y. S. Chiu, W. H. Ho, J. Polym. Sci. A Polym. Chem.
tests indicated that relative to the UPR sample, the PHRR value of 2003, 41, 2354–2367.
FR-UPR3 decreased by 42.7% and THR by 43.2%, whereas the char [22] W. E. Khatib, B. Youssef, C. Bunel, B. Mortaigne, Polym. Int. 2003, 52,
residue amount and SEA values increased with increasing 146–152.
phophorus content. The enhanced thermal stability and excellent [23] K. L. Erickson, J. Therm. Anal. Cal. 2007, 89, 427–440.
[24] H. E. Kissinger, Anal. Chem. 1957, 29, 1702–1706.
flame retardance of FR-UPR could bring some benefits for resin in [25] T. Ozawa, Bull. Chem. Soc. Jpn 1965, 38, 1881–1886.
application, such as a basic ingredient used in halogen-free [26] A. Baudry, J. Dufay, N. Regnie, B. Mortaigne, Polym. Degrad. Stab.
flame-retarded insulating coatings. 1998, 61, 441–452.
[27] D. Price, L. K. Cunliffe, K. J. Bullett, T. R. Hull, G. J. Milnes, J. R. Ebdon, B.
J. Hunt, P. Joseph, Polym. Degrad. Stab. 2007, 92, 1101–1114.
[28] Y. L. Liu, G. H. Hsiue, Y. S. Chiu, R. J. Jeng, C. J. Ma, J. Appl. Polym. Sci.
1996, 59, 1619–1624.
REFERENCE [29] M. Zanetti, T. Kashiwagi, L. Falqui, G. Camino, Chem. Mater. 2002, 14,
881–888.
[1] P. Penczek, P. Czub, J. Pielichowski, Adv. Polym. Sci. 2005, 184, 1–95. [30] V. Brabauskas, Fire Mater. 1984, 8, 81–87.
[2] S. Vlad, S. Oprea, A. Stanciu, C. Ciobanu, V. Bulacovschi, Eur. Polym. J. [31] B. K. Kandola, A. R. Horrocks, P. Myler, D. Blair, Compos. Part A 2003, 34,
2000, 36, 1495–1501. 863–873.
[3] P. A. Cusack, A. W. Monk, J. A. Pearce, S. J. Reynolds, Fire Mater. 1989, [32] L. B. Manfredi, E. S. Rodriguez, M. Wlayka-Przybylak, A. Vázquez,
14, 23–29. Polym. Degrad. Stab. 2006, 91, 255–261.
[4] M. P. Luda, A. I. Balabanovich, G. Camino, A. Hornung, Polym. Adv. [33] C. M. C. Pereira, M. Herrero, F. M. Labajos, A. T. Marques, V. Rives,
Technol. 2003, 14, 741–748. Polym. Degrad. Stab. 2009, 94, 939–946.

1777

Polym. Adv. Technol. 2011, 22 1768–1777 Copyright ß 2010 John Wiley & Sons, Ltd. wileyonlinelibrary.com/journal/pat