O r g u n i c Reugents

A Group of Twelve Papers

Presented at the Second Annual Summer Symposium sponsored by sthe
Division of Analytical and Miero Chemistry and Analytical Chemistry,
Wesleyan University, Middletown, Conn., dune 24 and 25, 1949

The Role of Organic Reagents

In the Chemistry of Specijic, Selective, and Sensitive React ions
FRITZ FEIGL
Ministerio da Agricultura, Laboratorio da Produ$io Mineral, Rio d e Janeiro, Brazil

Translated by RALPH E. OESPER, University of Cincinnati, Cincinnati, Ohio

The chemistry of specific, selective, and sensitive re- ganic reagents; it points to problems still to be ex-
actions has the following two tasks as its main ob- amined, especially to the discovery of new organic
,jectives: (1)the gathering and classification of, and reagents, and the improvement of analytical meth-
commenting on, experimental findings which are ods based on the use of organic reagents; and finally,
connected, either directly or indirectly, with the it shows the necessity of correcting erroneous gen-
specificity, selectivity, and sensitivity of analytical eralizations. To demonstrate the significance of the
methods. This material is to be found both in chemistry of specific, selective, and sensitive reac-
analytical papers and in papers on other branches of tions there are discussed here instances taken from
chemistry; (2) the selection from analytical papers the action of atomic groupings in organic reagents,
of experimental findings t h a t are of interest to other new complex compounds formed by organic reagents,
fields of chemistry; these are primarily to be sought real and false equilibria by formation of metallo-
in original papers and not in abstracts. This leads organic compounds, masking of reactions, and inner
to a better understanding of the effectiveness of or- complexes and adsorption compounds and lakes.

0 RGASIC reagents play an important role in modern meth-

ods of analysis, which are based on chemical reactions of the
materials that are to be detected or determined. This is evi-
organic reagents, another generalization must also be taken into
consideration-that the employment of organic reagents in
chemical analysis is as old as analysis itself. If this were true,
denced by numerous comprehensive publications and periodical the tremendous development of preparative organic chemistry
surveys in the journals of practically all countries (26), and by during the second half of the 19th century would certainly have
the space which the test- and handbooks of analytical chemistry stimulated an intensive use of organic reagents. This was not
devote to discussion of the employment of organic reagents. The the case; in fact, many organic compounds, whose convenient
first essay in English on the use of organic reagents in inorganic preparation or even reactivity with inorganic substances has been
analysis was published by Feigl ($6). The acquired knowledge in well known for decades, have been recognized as reagents only
this field has now become so great that three escellent special within a comparatively fe\v years.
works on organic analytical reagents have been published re- An inspection of the literature reveals that up to about 1905
cently (50,98,108). organic reagents were recommended and employed in inorganic
Some authors include under this heading all organic compounds analysis to a relatively limited extent. Grateful tribute should
that are used in the course of analytical operations. This, how- be paid to Cazeneuve, Denigks, Griess, Illinsky, Knorre, Pozzi-
ever, does not conform to the requirement that a compound is Escot, and Raschig, pioneers in the utilization of organic reagents.
designated as a reagent only if it reacts with a material which is Hence the stimulus to make estensive use of organic reagents did
to be detected or determined. Among organic compounds that not ceme from organic chemistry, which had long possessed an
function as solvents, extraction media, flotation agents, wash ma- arsenal of analytically usable organic compounds. For a very
terials, protective colloids, etc., it is a question not of reacting long time analytical chemistry felt no urge to pay attention to
materials, but of analytical aids, 7%hich admittedly are sometimes this storehouse. The merit of having done this belongs to the
of the highest importance in carrying out analytical processes. few Forkers who first appreciated the prime analytical impor-
Such auxiliary materials also include organic materials that are tance of the Werner complex chemistry to the formation of
employed as indicators of all varieties or as primary standards in metallo-organic compounds. The introduction of dimethyl-
titrimetry. If the requisite limitations are enforced regarding glyovime and dicyandiamidinc as nickel reagents (Tschugaeff,
1298
V O L U M E 21, NO. 11, N O V E M B E R 1 9 4 9 !299

1905; Grossmann, 1907), of nitrosophenylhydroxylamine as a reagents are not treated exciusively because of their analytical
copper and iron reagent (Baudisch, 1911), and of alizarin as an usefulness. Their behavior can contribute greatly to extending
aluminum reagent (Atack, 1915) ushered in, beginning around the knowledge of specificity, selectivity. and sensitivity. and their
1920, a new tendency in analytical chemistry. This movement characteristics can aid in securing insight into other provinces of
is characterized by a far-reaching consideration of coordination chemistry.
chemistry, and above all by the employment of organic reagents
for the production of soluble and insoluble metallo-organic com- GROUP ACTIOS Ih-O R G . i S I C REAGENTS
ples compounds ( 2 1 ) . The reaction ability of organic compounds is always due to the
Organic reagents can be regarded from t x o viewpoints. The activity of certain groups, which can be termed reactive or affine
first is purely practical, in that the sole consideration is whether (binding) groups. These may involve (1) acidic groups, whose
and how an organic reagent makes it possible to accomplish an hydrogen atoms are replaceable by metals; (2) atoms or groups
analytical task. In this case, important factors are sensitivity, that can be bonded to metal atoms through auxiliary valences;
certainty, speed, and accuracy of tests or determinations. In the (3) the simultaneous action of principal and ausiliary valences
foreground is the prescribed succession of operations-the “pro- localized on different parts of the organic molecule; and (4)con-
cedure” of an analytical method. I n this regard, an analytical densation reactions, usually those that proceed through elimina-
incthod may be characterized by an organic reagent, and both tion of n-ater. The products of chemical changes 1 to 4 are not
t h r niethod and the reagents will retain their interesting features necessarily stoichiometrically defined and isolable compounds.
as long as they we not displaced by something better. The same types of reaction and combining forms can also result
The second viewpoint takes account not only of the practical from surface reactions of solids with no production of stoichio-
solution of an analytical problem, hut all the details of an analyti- metrically definable compounds. Such products are known as
cal method and their scientific bases. In general, this is taken to adsorption compounds, and interesting insights into their forma-
mean the application of physical-chemical laws to analytical pro- tion have recently been given by organic reagents.
ceduies. Especial emphasis is placed here on a numerical con- The principle of group actions in organic reactions, now gener-
ception of the conditions under which chemical changes can be ally accepted, contributes greatly to an understanding of the
carried out to reach analytical objectives, taking due account of effect of organic reagents on the composition and constitution of
the stoichiometry, equilibria, and kinetics of the reactions in- the reaction product. It provides an initial orientation as to the
volved. This niethod of scientifically treating analytical proce- classes of organic compounds that are likely to offer success in
duIes was originated by Wilhelm Ostwald, continued by I\-.Bott- the search for new reagents. The presence of certain groups is a
ger, and has become a fundamental of modern analytical chemis- necessity for the reactivity of an organic compound, but by itself
try, particularly through the work of I. 11.Kolthoff. Its essence does not ensure the occurrence of an analytically usable reaction.
is that the qualitative phenomena of chemical changes are, as a The latter requires that the reaction products possess certain
rule, taken as a priori facts, and attention is directed primarily to characteristics. Among these are insolubility in water, solubility
their quantitative aspects. However, of no less importance to a in organic liquids, and color in the solid or dissolved state under
scientific continuation of the classical analytical method of work- normal or ultraviolet light. In addition, the reaction products
ing, which rests on the utilization of chemical reactions, is the should be produced rapidly and by even small amounts of the
scientific treatment of the following questions, through which the material being sought or determined. The reaction should be
qualitative phenomena are brought to the front: How is it pos- confined to a particular material or to a highly restricted number
sible to arrive a t new analytically useful reactions and reagents? of substances (specific and selective action).
How can known reactions be used in analysis with the highest ilt present no definite predictions can be made as to which
efficiency? How can improvements of analytical procedures be conditions satisfy the requirements. At first glance it would
discovered? appear hopeless to seek new organic reagents by any means except
The present state of chemical knowledge will not provide pre- pure empiricism. Fortunately, this is not the case. Previous
cise ansn-ers to all parts of these questions, especially if nothing experiences concerning the activity of organic reagents, \-,-hose
beyond physical-chemical laws is taken into consideration. number is steadily increasing, point to definite regularities whose
Severtheless, pure empiricism is not the only way out of this observance is very helpful. Thus, there is no absolute independ-
difficulty; points of reference and guiding principles are available ence of group action. The production of metallo-organic com-
for handling these problems. I n the first place, it is necessary to pounds is dependent on pH, and this relationship is often so pro-
consider in detail the chemism of known unequivocal and sensi- nounced that it constitutes an important factor for the attain-
tive reactions. Efforts must be made to secure a clear under- ment of selective and sometimes even specific actions of an or-
standing of the relations between the composition and constitu- ganic reagent. Furthermore, it is sometimes possible by the addi-
tion of reactants and reaction products, as well as of the relation tion of masking agents to lower the concentration of some ionic
of the particular type of reaction to the phenomena which are the species to the point a t which certain precipitations or color reac-
bases of an analytical employment: solubility, volatility, color, tions fail. The action of such masking agents, which frequently
fluorescence, etc. Furtherrequirements are an extensive acquaint- are organic compounds, may be remarkably selective. Finally,
ance n-ith and intimate consideration of all the factors that may the action of organic reagents may be annulled by the presence of
influence the certainty and sensitivity of reactions in a positive reducing or oxidizing agents. Hence, the analytically useful
or negative direction. In such studies, especial attention must reactivity of a certain group in organic compounds should never
be given to many provinces of chemistry, particularly when the be considered apart from the entire reaction theater.
study includes phenomena which-though not of direct analytical The external effect of the reaction milieu on the reactivity of a
use-often are instances of events that may take place in analyti- given group in organic compounds is accompanied by the impor
cal processes. tant influence that may be exerted by the rest of the molecule of
There is a t hand a great fund of factual knowledge which can which the group is a part. Inasmuch as it is often possible to
be d r a m on for solving these problems. The assembling, ar- modify this residue very materially by introducing other atoms
rangment, and appraisal of this material are among the most im- or groups, it becomes experimentally possible to test the activity
portant tasks of the chemistry of specific, selective. and sensitive of the reactive groups with respect to their reliability or vari-
reactions ( 2 1 ) . A further task is to retain, from analytical stud- ability. Such studies furnish information of practical as well as
ies, observations that are of significance with respect to other theoretical interest.
fields of chemistry. rlccordingly, within the framen-ork of the The earliest studies of group actions were made by Baudisch
chemistry of specific. selective, and sensitive reactions, organic (3, 4). After demonstrating that the ammonium salt of nitroso-
 1300 ANALYTICAL CHEMISTRY

phenylhydroxylamine (cupferron I ) precipitates copper and iron Froiii the standpoint of group action, special int,erest is at.-
from acid solutions, he tested the behavior of derivatives of cup- tkiched to dioximes in which the active groups are bound to nitro-
ferron and found that the introduction of groups into the phenyl gen rather than to carbon atoms. Such compounds are repre-
ring of cupferron does not impair the precipitating action of the sented by oxalenediamide oxime (12) and osalenediuramidr
active XO and N O H groups, The salts of cupferron, to which oxime (33). The former, VIT, forms an inner complex nickel sal1
coordination formula I a can be ascribed, and its derivatives are of the same constitution as the nickel salts of other 1,2-dioximes.
inner complexes. Seocupferron, 11, which v a s iecomniended by I n contrast, VI11 behaves as a dibasic acid and, in ammoniacal
Baudisch and Holmes ( 5 ) ,has a greater precipitating sensitivit) solution, produces an insoluble diammine, whose coordinatiou
than the parent compound. This reveals the so-called size or formula is shown by VIIIa. Hence, it must be assumed that the
weighting effect exerted by the part of the molecule attached to ---COKHt group exerts a distinct acidifying action. This effect
the affine groups. I t is probable that the increased size which obviously is related t o the configuration of the t,\w hydroxy
would result from the attachment of a third ring, ai: shown in TI1 croups in the amphi position (67).
could lead to a still more sensitive precipitating reagent

0
1 Ia

That group actions III urg:,tnic precipitants are not alwajs b u

so rl.0 iridependent of the rest of the organic molecule as in the examplet
IT I11 cited is revealed by a comparison of 8-hvdrouyquinoline (8-
Practically all water-insoluble metal a d t a of the nitroauuyl iluinolinol) IX, and 8-hvtiroxyquinaldine, X The former has
hydroxylamines are soluble in organic liquids, especially chloro- tound ektensive use as a precipitant ( 7 , 9 ) for numerow mcJtals in
form. This is in conformity with their character as inner coni- dcetate-buffered solutions or in ammoniacal solutions masked
UI Ith tartrate. Inner complexes with the general coordinatioii
plexes. The color of the solid copper and iron salts is not. identi-
cal with that of t,he molecularly dissolved materials. The water- tnrmiila at e formed.
insoluble salts show a striking instability toward hydroxyl ions.
and hence are convert,ed by alkalies into the corresponding metal
hydroxides and the soluble alkali salts of the organic compound
The metal cupferronate is readily decomposed by heat, which
indicates that in general the decomposability of the organic COKJ-
ponents is retained in the salts. Thus, noteworthy information
concerning the properties of metallo-organic compounds can hr
secured from studies of group actions of organic reagents.
Interesting examples of the tracing out of group actions are d natural to Assume that compounds IX and X n.ould
I t ~ o u l be
provided by the 1,2-dioximes (16). Tschugaeff’s (92) discover! ,,\hibit analogous precipitating action, as they differ merely b?
that nickel can bc quantitatively precipitated by diniethylgly- one methyl group. This is the case, with the notable exception
oxime and benxildiosime was subsequently substantiated nith uf their behavior toward aluminum. Rferritt and Walker 179
numerous 1,2-dioximes (78). The nickel inner complexes of 1,2- lound that aluminum is not precipitated by X, and made analyti-
dioximes, IT’, in which the two S O H groups are in t,he anti posi- cal use of their otxervation. They assume that the increase itl
tion, are assigned the coordination structure, IF’a. Soule (87) volume due to the methyl group prcvcnts R stable binding vritti
recommended furildioxime as a nickel reagent, even though it is the aluminum atom.
only slightly soluble in n-ater. Wallach (95) found that cyclo- The fact that the slight increment in volunie due to the intro-
hesanedionedioxime, rr1, which he prepared and which is readily iluction of a methyl group into an organic moleculc can have
water-soluble, precipitates nickel quantitatively, and in fact has :i marhed effect may tie seen from studies of bipyridine derivative-
higher precipitating sensitivity than dimethylglyoxime. This hy Sniirnoff (85) and X l l i n k and I\-ihaut (106). A s early a-
remarkable finding demonstrates that the precipitating action 1895, Blau (10) found that ferrous salts combine with three inolr-
due t,o a given grouping may he independent of the solubility or cules ot a,a’-bipyridine, X I , to produce soluble red salts, which
insolubility in water of the organic reagent. Diehl and associates contain the complex cation [Fe(a,a’-dip),J-+. The foimation oi
194) have recently described a convenient method for preparing this colored cation in acidic solution was iecommended 30 years
c:yc.lohexanedionedioxime and have pointed out the advantages of later tor the detection and colorimetric determination of iron
this water-soluble reagent for the detection and quantitative (R6, 58). Thc authors mentioned found that derivatives XI1 t o
~lt~tf~rminationof nickel and palladium. XVT of bipyridine do not give a color reaction with ferrous salts
Perhaps these compounds are not capable of forming
., red heuammine ions of iron because of their ewcssivc
0 ypatial requirements.
R-C=h’OH R-C=R‘ Lo)-?=NOH HZC’ ‘?=NOH
I I
R-C=NOH R-C=W’
\Yl
h &=NOH
OH
‘d
Hz
IV I\’a v VI X I (a,a’-dip) XI1
V O L U M E 21, N O . 11, N O V E M B E R 1 9 4 9 1301

the isomer of l-nitroso-2-naphthol-i.e., 2-nitroso-1-naphthol-

must also be a precipitant for cobalt. Because, however, the
recognition of the significance of group actions is rather recent,
the pertinent trials were not made until 1919. Bellucci (6) found
that the anticipated precipitating action is about eight times a*
XI11 XIF'
sensitive as that of the isomeric 1-nitroso compound. The fad
that 2-nitroso-1-naphthol-4-sulfonic acid gives a color reaction
A A A P\ with cobalt salts was observed as early as 1885 by Hoffmann (69),
but it was not until 1938 that Sarver ( S I ) made a thorough study
of this reaction and applied it to the colorimetric determination of
xv XVI cobalt.
I t is not a mere coincidence that effects of sulfonic groups,
Tartarini (91) was the earliest to observe that cuprous salt's are similar to those just discussed, are again observed with 8-hydroxy-
likewise capable of coordinating with a,a'-bipyridine to produce quinoline, I X , which reacts with numerous metal ions to produce
colored compounds. I t is interesting that a,a'-diquinolyl, insoluble inner complexes of the coordination structure IXa.
XIII, forms a purple-red complex ion, [Cu(a,o-quin)p] +, with The sulfonated derivatives, 8-hydroxyquinoline-5-sulfonic acid,
cuprous ion (11). Hoste (61) recently described a very senskive, XIX, and 7-iodo-&hydrouyquinoline-5-sulfonic acid, XX, are not
strictly specific test for copper and also a colorimetric method of precipitants under the conditions prevailing in the oxine pre-
determining this metal, which is based on the production of this cipitations from dilute solutions. However, this does not signif)
tetrammine ion. I t is probable that not only diquinolyl but also that the inner complex-forming groups in the two sulfonic acids
all other bipyridine derivatives, which do not form hexammine have lost their activity. This is shown clearly by their action
ions with ferrous ion, will be found t o react with cuprous ions, for toward solutions of ferric salts. Oxine gives a brown-green pre-
this latter reaction occurs with the less space-demanding tetram- cipitate, whereas X I X and X X yield brown and brown-red solu-
mine ions. If so, it will have been demonstrated that the selec- tions, respectively. Yoe (107) found X X to be a sensitive and
tive action of a given group is raised to specificity by the volume extremely selective iron reagent, which makes possitde a colori-
increase derived from the attached organic molecular remainder, metric determination of iron. He named this reagent ferron.
In any case, it may be concluded that steric factors can play sig- Inasmuch as the intensity maximum of the color reaction is at
nificant parts in determining the reactivity of certain groups in the ratio 1 iron to 3 ferron, there can be little doubt that coordina-
the organic molecule. tion formula XXa can be assigned to the soluble colored ferric,
Because water-solubility is a very desirable characteristic in an product, even though the compound has not yet hem kolated
organic reagent,, and organic compounds can be made water- So3H SOaH
soluble by the introduction of sulfonic groups, the behavior of I I
su1fonat)edderivatives of organic precipitating agents raises inter-
esting questions. This matter was experimentally explored for
the first time in 1928, when van Klooster (65) studied l-nitroso-
2-hydroxy-3:6-naphthalenedisulfonic acid, XS'III, commonly
known as nitroso 12 salt. This compound is a derivative of
l-riitroso-2-naphtho1, XVII, whose action as precipitant for co- XIX xcx XXa
bal t was discovered as early as 1885 by Illinsky and Knorre (62).
The cobalt inner complex, XVIIa, is produced in this instance. Here again, a precipitant has been modified into a color reagtint
van Klooster found that a red coloration develops in acetate- by the introduction of sulfonic groups. Compounds analoguus
buffered solution, is stable against subsequent addition of mineral to XXa have not yet been detected in the case of other metal?
acid, and makes possible a sensitive colorimetric determination of that can be precipitated as oxinates. However, an inner complcnu
colmlt. The color reaction probably is due to thr production of anionic binding of the mrtal is very probable because when tllr
the wvntrr-soluble cohalt compound, XVIIIa metal salt solutions are tre:itctl rTith an eycess of X I S or X X , n o
precipitation iollows if the system is then
made basic with ammonia or caustic alkali

0"" (y):'.
,A,''\
SO,H SOaH
1 ,chi
r

so:,
ON-

SO8
1
( 2 2 ) . Thissignifies that a masking agent
has been formed bv the inclusion of a sulfonic
group in the oxine molecule. I t appears thus
that the introduction of sulfonic groups
into organic precipitants does not destroy
theactivity of inner complex-forming groups,
but, produces soluble electrolytes with inner
SS'II XVIIa XVIII NS'IIIa complex anions instead of iiisoluble inner
complexes. This indicates an interesting
way by which new color reagents and mask-
Comparison of coordination formulas XVIIa and XVIIIa ing agents may be obtained.
shows that the inner complex-forming group has not lost its reac- Because hydroxy and carboxy groups are known to confer
tivity through the introduction of sulfonic groups. However, a water-solubility on organic compounds, it seems logical to inquire
water-insoluble precipitant has been converted to a water-soluble whether the introduction of these groups will affect organic inner
color reagent, which leads to EL water-soluble compound with a complex formers in the same ways as the inclusion of sulfonic
colored inncr complex anion. In addition, nitroso R salt ex- groups. Ephraim found that salicylaldoxime, X X I , is a selective
hibits a greater selectivity than the parent compound. precipitant for copper, with which it forms a yellow-green inner
l-l\'itroso-2-naphthol precipitates not only cobaltic, but also complex, XXIa. A gravimetric method has been based on this
cupric, nickel, and ferric salts; of these only the cobalt compound finding (19). His investigations of the behavior of derivatives of
is acid-stable. In the light of present experiences with group ac- aalicylaldoxime showed (20) that compounds X X I I to XXV have
tion in organic compounds, it seems to be a matter of course that t'he same precipitating action as XXI.
1302 A N A L Y T I C A L CHEMISTRY

is the basis of a sensitive test and a colorimetric procedure for

determining zirconium (39,66).
The establishment of the fact that particular group actions are
retained after certain other groups are put into organic com-
pounds is of great importance in improving organic reagents or
XXI XXIa XYII discovering new ones. However, the fact that group actions of
reagents can also be nullified by introducing other groups is of
CH=SOH no less importance, for sometimes the selectivity and sensitivity
CH3 CH3 I
of organic reagents can be raised through such effects. Llodern
studies of the analytical application of organic reagents are giving
ever-increasing consideration to group actions. Experimental
studies which broaden and deepen knowledge of group reactions
bH are of significance to the chemistry of specific, selective, and sensi-
tive reactions, even when they do not lead to results of immediate
XXIII XXIV XXI. analytical utility.
It appears, therefore, that the introduction of solubilizing hy- DISCOVERY OF NEW COMPLEX COMPOUNDS
droxy groups into the molecule of an organic precipitant does not THROUGH USE OF ORGANIC REAGENTS
annul the precipitating action of its characteristic groups. This
The activity of certain atoms and groups in organic compounds
finding can be utilized in the analytical development of group is not merely of analytical interest with respect to the specificity,
actions.
selectivity, and sensitivity of a reagent. Often the composition,
Fichter and Goldach (46) described a test for hydrazine based constitution, and properties of metallo-organic compounds are of
on the formation of a water-insoluble aldazine, resulting from great importance to complex chemistry. Sometimes during
the condensation of hydrazine with benzaldehyde: analytical studies, new metallo-organic compounds have been iso-
lated which can be fitted into certain classes of complex com-
2D C H O + H??;--SHp + pounds, and thus the systematics of complex compounds have

CZ)--CH=N-K=CH a
Because alcoholic solutions of benzaldehyde must be used, thr
been enriched; members of totally new classes of complex com-
pounds have been discovered in the course of analytical research,
Willard and Smith (86, 105) found that tetraphenylarsonium
aldehyde may precipitate in the aqueous solution. Reasoning chloride, XXX, is an excellent precipitant for large-volume and
from the knowledge that the precipitation action of the aldehyde complex anions. Here is shown clearly the influence of the ionic
group is preserved even after the introduction of hydroxy groups volume on the occurrence of precipitation reactions, an effect
into the benzaldehyde molecule, Feigl and Schwarz (43) tested that is met in numerous precipitations of alkaloid bases by means
the action of water-soluble salicylaldehyde and found that it
likewise produces an insoluble aldasine. This reagent permits a of complex acids. The findings of Wllard and Smith have stimu-
more reliable test for hydrazine, and its use provides a suitable lated researches on the behavior of other cations of large volume.
weighing form in the gravimetric determination of hydrazine. I t is now known that the ferrous bipyridyl sulfate, XXXI (eo),
and the cation of the dye rhodamine, XXXII(68), are selective
Interesting applications of group actions are based on the fact
precipitants for complev ions and can be used analytically.
that organic reagents, which produce colorless precipitates with
metal ions, can often be easily modified, by preparative proce- (CsHj)ih +C1- [Fe(a,a’-dip)a ]++SO,--
dures, to produce derivatives that give colored precipitates. XXX XXXI
Thus, a study of the reactivity of rhodanine, XXVI, showed that
the ability to produce insoluble silver salts (through action of the Hayes and Chandlee (54), while
C6H,--S( CBHL)~
imino group) is retained in all rhodanine derivatives with an in- working out a new specific meth-
tact imino group. Such derivatives are easily obtained by means CsH4.C/, ‘0 od for the gravimetric determi-
of condensation reactions of the carbonyl or methylene group of I I\ / nation of palladium, discovered
CO -0 CcHz-S( CGHS)?
rhodanine. Thus it was possible to obtain dimethylamino- interesting complex compounds
benzylidene rhodanine, XXVII, the first organic reagent that XXXII of p-furfuraldoxime, XXXIII.
forms a deep colored (red) silver salt (69). Because of the replaceability of
the hydrogen atom of its NOH group, this compound is an
(Ag) HN-CO HN-CO acid possessing the radical XXXIIIa. -4s it is known from the
SA AH2 SCI J=CH--C)--Pi(CH3!,
coordination formula of palladium dimethylglyoximate that one
‘d ‘d oxime group can be coordinated on palladium through its nitro-
gen atom, the question arises as to how the S O H group func-
XXVI SXVII tions in furfuraldoxime. Hayes and Chandlee found that the
yellow compound, A, is precipitated from acidic palladium solu-
The arsonic acids, XXVIII, demonstrate a similar color effect tions that contain chloride ions. This compound corresponds
following the introduction of new groups. Their precipitating to the diammine type that is so frequently encountered among
action is analogous to that of arsenic acid. In particular, the complex palladium salts (38). This type of compound might
precipitation of zirconium as the white insoluble salt of alkyl be expected to be produced with all soluble palladium salts, no
arsonic acids has been recommended for the gravimetric deter- matter what the corresponding anion, but such is not the case.
mination of this metal (66). The precipitating action of the The insoluble intercalation compound, B, was obtained from solu-
AsO(OH2) group is preserved in p-dimethylaminoazophenyl- tions of palladium sulfate, while solutions of palladium nitrate
arsonic acid, XXIX. However, the action is improved because produced compound C.
of the intense red color of the zirconium salt. The use of XXIX

XXIX SXXIII (F) XXXIIIa ( f )

V O L U M E 21, NO. 11, NOVEMBER 1 9 4 9 1303
PdClz.2F PdS04.4F Pdfz.2F
A B C
H~~--CI)-O
‘\\ / 1 t
Me -Me- -Me-
Accordingly, furfuraldoxime presents an unusual case in which
the composition of the precipitate is affected by the anion which
was originally combined with the cation. Among the palladium
compounds of furfuraldoxime, A has the lowest solubility. so that XXXV XXXVI
its production is always assured by the addition of chloride ions. These examples make it plain that metallo-organic compounds,
The interesting observations by Hayes and Chandlee on the discovered in the course of analytical studies, may be of considera-
divergent behavior of furfuraldoxime toward palladium sulfate ble importance to complex chemistry. The compounds isolated
and nitrate are not cited in the abstracts of their paper because by West and Spacu lead to the expectation that polynuclear sta-
these results are not of direct analytical significance. This omis- ble chelate compounds with considerably greater numbers of
sion demonstrates plainly the danger that findings of importance rings are much more possible than was hitherto assumed. Reac-
to other fields may not receive proper attention if they are made tion possibilities, which had previously passed unnoticed, were
in the course of analytical investigations. The chemistry of
thus indicated.
specific, selective, and sensitive reactions performs an important Kot only may reactions involving organic reagents result in the
task in that it lays hold of such findings and indicates their sig- production of new metallo-organic complex compounds of sur-
nificance. prising structure, but surprising reaction paths may be traversed.
Other instances in which organic reagents led to new complex
Pertusi (76) proved that a white crystalline compound, 2Bzd.-
compounds were offered through the behavior of masked ammoni-
4HF.HgF2, precipitates from fluoride solutions on the addition
acal thallic salt solutions toward 8-hydrovyquinoline, thionalide,
of an acetic acid solution of benzidine and mercuric succinimide.
and other organic precipitants. Precipitates of formula-pure,
The product is so stable that it provides a sensitive and specific
stable inner compleses of trivalent thallium, hitherto unknown,
test for fluoride and is the basis of a gravimetric determination
were produced in this manner (27). Thallic oxinate and dibrom-
(73). No coordination formula for this remarkable mercury
osinate are very suitable for the selective detection and gravi- compound has been given. It has been suggested that it is an
metric determination of thallium (32). I t is probable that other
addition compound of benzi line hydrofluoride with mercuric
thallic inner complexes, which invariably contain a low percent-
fluoride:
age of thallium, will also be of analytical use. I n view of the
fact that thallium salts have pharmacological action, the inner F H . H ~ x - ~ - - - s H ~ . H [F.~~.;]
romplexes of this metal, which are soluble in organic liquids, may
possess an interest of their own. FH.HLV- -SHZ.H F.
The elucidation of analytically valuable effects, or of those that
are observed as by-products of analytical researches, is a funda- The complex compound discovered by Pertusi is unique with
mental objective of the chemistry of specific, selective, and sensi- respect to both its composition and the unsymmetrical binding of
tive reactions. West and Amis (104) discovered a sensitive reac- mercuric fluoride. There is the further singularity that the use
tion of palladium with fuchsin hydrochloride (Fs), which de- of mercuric succinimide is essential to its production. Accord-
pends on the production of the brov-n compound 3PdC12.2Fs. ingly, this presents another instance in which the composition of
From his studies \Vest (103) has concluded this palladium com- the precipitate produced by the action of an organic reagent with
pound can bc represented by the coordination formula XXXIV: a cation is influenced by the anion originally associated with the
cation. Such effects should be taken
fully into account in the produc-

=\C===g -
J
1++
1[PdCla]--
tion of molecular compounds.
Feigl and Suter (44)made some
interesting observations in studies
designed to elucidate Cooper’s dis-
covery (14) that palladium di-
X S S I V (3PdC12.2Fs) methylglyosimc, XXXT’II, is solu-
ble in alkali hydroxide. This solu-
bilitv results from the fact that
No complex compounds of this typc had been obaerved previ- this inner complex has the character of an acid because of the
ouslv. two unsalified and merely coordinatively bound S O H groups.
These conclusions from \Test’s studies make it necessary to These oxime groups are salified by strong alkali to produce water-
re-examine the complex compounds of benzidine (Bzd). I t has soluble alkali salts possessing inner complex palladium-bearing
been recommended that Cu(CSS)2.Bzd be employed in quantita- anions, XXXVIII.
tive analysis for the precipitation of copper (89) and Hg(CS),.Bzd 0 0
as a wpighing form in the determination of mercury (71). These H,C-C=S Y=C--CH,
materials belong to a class of complex compounds of the general
formula MeX?.Bzd, that were studied by Herzog ( M ) ,Spacu (88), ~ ‘Pd/
,, \ ,
+ 20H- -e-
and Barcel6 ( 2 ) . Spacu views these compleves as interpenetra- H,C-C=X’ K;=C--CHI
tion compounds, [MeBzd]Xz. This formulation has been ac- OH HO
cepted in the analytical literature (99) but cogent arguments can
be brought against it. The occupation of two coordination posi-
tions on one metal atom by one molecule of benzidine, as shown in
XXXY, is impossible on steric grounds because of the too great
distance between the two S H , groups. The formulation as a
bimolar chelate compound with the coordination formula XXXVI
is much more nrobahlr. This is analogous to the complex pxlla-
diuni fuchsin compound.
1304 ANALYTICAL CHEMISTRY
or Pd(DH)z + 2KOH +KzIPdDzI + 2Hz0 Sandell (79) in the reaction of gallium salts with &hydroxy-
quinoline (oxine). The pH range for the quantitative precipita-
hccordiiigly, palladium dimethylglqoxime should be regarded
tion of gallium oxinate lies between 6 and 8. Consequently, no
aa an amphoteric compound, which functions as an inner complex
visible formation of this salt occurs at pH 2.6 to :3. Sandell
in neutral or acid media and as an acid in alkaline media, The found that all the gallium passes, as oxinate, into the organic sol-
justification for speaking of a dimethylglyoxime palladium acid
vent if a solution of oxine and the metal salt is shaken M ith chloro-
was supplied by the isolation of its potassium, barium, manga-
form a t this pII. Gallium can be detected and determiried by
nese, palladium, and leadsalts. The lead salt is particularly inter-
this procedure even though other oxinate-forming metals are
esting because of its scarlet color and the brown palladium salt
present.
for being a dimer of the hitherto unknown yellow palladium di-
methylglyoxime. The findings relative to palladium dimethyl- Even though no perceptible precipitation of gallium oxinate
glyoxime make it very likely that other inner complexes of organic ucrurs a t pH -3, total absence of reaction a t this acidity is not
necessarily indicated. Every precipitation reaction, and there-
compounds, which contain two acid groups, one salified, the other fore also the reaction
coordinated, may dissolve in caustic alkalies to form alkali salts
of acids with inner complex metal-bearing anions. (Salicylal- Ga ++ + 3HOx ----L Ga(Ox)r + 3H +

doxime, isatin-8-oxime, phenylglyoxilic acid oxime, etc., are cases begins and proceeds to a slight extent even prior to the visible
in point.) The production of such water-soluble alkali salts has separation of a precipitate-Le., before its solubility product is
a certain similarity to the formation of alkali salts of sulfonated exceeded. In the case of the gallium-oxine reaction a t pH -3,
the solution contains a slight quantity of molecularly dissolved
inner complev formers. In both cases, the production of inner gallium ovinate in equilibrium with its dissociation products
complex metal-bearing anions indicates masking possibilities. However, gallium oxinate is readily soluble in chloroform, and
The palladium in the alkaline solution of palladium dimethyl- when the solution is shaken with the solvent, the oxinate distrib-
glyoxime is masked toward practically all reactions. However, utes itself between the two liquids. Because the partition equi-
librium greatly favors the ciiloroform, the gallium oxinate is ex-
the dimethylglyoaime is also masked against the familiar action tracted from the water solution, and is then replenished by the
of nickel ions. Hence, organic compounds can be masked as a renewed formation of the salt. If the extraction with chloroform
consequence of their inner complex binding, and this opens inter- is repeated, this process continues until the gallium is entirely
converted into oxinate and collected in the chloroform.
esting prospects. The behavior of alkaline solutions of palla-
dium climethylglyoxime toward cyanide ions is remarkable (36) This extraction of gallium a t pH values below those a t which it
The reaction is: is precipitated from water solutions is the only known instance of
[PdDz]-- + 4 C S - + [Pd(CN)r]-- + 2D-- the analytical utilization of an equilibrium shift of this kind
However, an analogous behavior may be anticipated with such
This corresponds to a demasking of the dimethylglyoxime, inner complexes as dissociate to only a slight extent in aqueous
which in this way can be brought to reactivity with nickel to pro- solution and are soluble in organic liquids. In fact (b8), acid
duce the familiar red compound. A specific and selective test solutions of alkali vanadate, which exhibit no change when oxine
tor cyanide in alkaline solution was developed from this behavior, IS added, form a green lager when they are shaken with chloro-

and presumably a quantitative procedure can be worked out. 811 form; this indicates that chloroform-soluble vanadium oxinate
the previous methods of detecting cyanide involved the liberation has formed. It is obvious that a selective extraction of inner
of volatile hydrocyanic acid and its subsequent identification. complexes combined with the establishment of definite pH values
The new test is thus a distinct step forward. It can be used to can be of great significance in enhancing the selectivity of organic
reveal cyanide in illuminating gas. reagents.
The literature of analytical chemistry contains numerous other Generally, organic reagents react directly with the substancpk
statements concerning metallo-organic compounds, which have to be detected or determined. But in certain cases a substance
analytical usefulness, but their method of formation and constitu- reacts only by shifting the equilibrium of another reaction in
tion have not been satisfactorily explained or have been incor- which an organic reagent is a partner. This equilibrium reaction
rectly interpreted. Certain reactions have been described, even may depend on pH or a redox reaction. Feigl and da Silva (34)
though the empirical formulas of the reaction products are not showed that the filtrate obtained from a solution of a nickel salt
known. Certainly, a laboratory study of such statements would that had been treated with an excess of dimethylglyoxime is a
lead to many new compounds and reaction possibilities. very sensitive reagent for basic materials of all kinds. As a re-
sult of the incomplete precipitation of the nickel-dimethylglg-
oxime from a neutral nickel solution, there is established the
TRl’E AND FALSE EQUILIBRIA IN REACTIONS WITH equilibrium :
ORGANIC REAGENTS
Every chemical reaction leads to an equilibrium, characterized
Xi’ + + 2DH2 F= Xi(DH)Z + 2H+
hy the equilibrium constant, which represents the ratio of some The solution is therefore saturated with nickel dimethylglq-
power of the concentrations of the starting and final materials. oxime and contains all the reactants for producing this inner com-
Special analytical interest attaches to the establishment of equi- plex, and a130 free hydrogen ions a t a concentration just sufficient
librium in reactions in which the hydrogen ion concentration to prevent formation of the precipitate. Every material that
plays a role, as in all reactions of organic reagents that are acidic withdraws hydrogen ions from the equilibrium will consequentl?
in character. Independence of the acidity is unusual in such in- occasion the precipitation of red nickel-dimethylglyoxime in an
stances. Usually, the establishment of certain pH values is amount that is equivalent to the hydrogen ions consumed. Basi(
necessary to secure practically quantitative reactions. Careful materials which dissolve so slightly in water that they have ncl
studies have demonstrated that sometimes the selectivity of effect on acid-base indicators-e.g., calcium carbonate, magnp-
organic precipitants can be substantially increased by establish- sium ammonium phosphate, calcium arsenate-react on their sur-
ing certain pH values and maintaining certain pH ranges. faces with hydrogen ions of the equilibrium solution, and thus
Particularly noteworthy are the effects of an extraction in the cause deposition of red nickel dimethylglyoxime. Many practical
case of pH-dependent precipitation reactions which lead to metallo- applications can be made. For instance, it is easy to distinguish
organic compounds that are soluble in organic liquids. I t is quickly between “fusible and infusible white precipitate” (Hg-
then possible a t certain pH values to secure complete reaction, a C12.2SH3and HgKH2C1), a problem that ordinarily requires a
result that could not have been expected from the equilibrium of quantitative determination of the mercury-nitrogen ratio. Cal-
the precipitation reaction. This was proved for the first timp by rite and dolomite ran be readily distinguished hv this means, I ~ P
V O L U M E 2 1 , NO. 11, N O V E M B E R 1 9 4 9 1305

well as alumina either anhydrous or hydrated. It is clear that first condition is accepted, the dissolution is delayed, whereas if
equilibrium solutions can be prepared from other acid-soluble the second is held to be valid, the precipitation is delayed.
inner complexes. The nickel-thionalide equilibrium solution The abnormal resistance of preci itates, which are considered
has proved very suitable. from the standpoint of a false equiligrium, must be demonstrated
A counterpart of the acid equilibrium solution of nickel di- immediately following the production of the compound. Conse-
rnethylglyoxime is found in a basic equilibrium solution of silvc~i quently, it has nothing to do with “aging” of precipitates. The
chromate. This is prepared by shaking an excess of silver chro- latter is a process that can be followed with respect to the elaped
time, especially in the case of preci&tation of hydrated oxides.
mate with a solution of ethylenediamine (en) and then filtering The precipitate, in such cases, exhibits progressive resistance to
fdn). The bright yellow solution houses the equilibrium’ attack by acids or bases.
[-4g en]&r04 ehg2CrOa + 2en The following instances of false equilibria in the action of or-
ganic precipitants possessing acid character are taken from the
\L hen solids or dissolved materials which bind ethylenediamine literature (98).
react with this equilibrium solution, red-brown silver chromate Cobaltic nitrosonaphthol can be precipitated from acetic acid
precipitates as a consequence of the disturbance of the equilib- solution, but once formed it is stable against mineral acids.
rium. Water-insoluble materials with acid characteristics can be Rubeanic acid precipitates copper from acetic acid, and cobalt or
nickel from ammoniacal solution. The metal rubeanatea are
thus revealed topochemically in a manner that is completely resistant to dilute mineral or acetic acid. Cobaltic thioglycollic
analogous to the detection of basic materials by means of the acid anilide has to be brought down from ammoniacal solution,
acid equilibrium solution. Positive results with acid and basic but it then is stable against acetic acid. Palladium can be p r e
equilibrium solutions are given also by solid substances which cipitated by 6-nitroquinoline from neutral solution; the re-
cipitate is, however resistant to strong hydrochloric acid. &ly
withdraw hydrogen ions or ethylenediamine molecules from the sodium salt of dipicrylamine is able to precipitate potassium,
equilibrium solutions through adsorption. Instances are alumina cesium, and rubidium, whereas dipicrylamine itself is ineffective.
and hydrated silica. On the other hand, the precipitated alkali salts are resistant to
An example of the application of an equilibrium disturbance dilute acids.
was given by Velluz and Pesez (98). Their method of detecting All the preceding organic compounds are inner complexes, and
ammonia and volatile organic bases is based on the fact that ‘2 it is probable that a study of the behavior of other inner complexes
solution containing zinc and oxinate ions: will produce fluoresciria would reveal further instances of acid resistance, which should not
zinc ovinate on contact with bases. The reaction is claimed to be exist according t,o the precipitation conditions. Severtheless,
superior to the litmus reaction for ammonia. such anomalies should not be viewed as a characteristic of inner
Equilibrium shifts, which can be put to analytical use, are d S 0 complexes. The examples of acidic equilibrium solutions of
possible with redox reactions that involve organic compounds. nickel dimethylglyoxime and nickel thionalide showed that doubt-
less there are typical inner complexes, whose formation and be-
The earliest known example is the sensitive test for zinc de-
scribed by Cone and Cady (IS). I t is based on the finding that havior toward hydrogen ions conform exactly to a true equilib-
in the presence of zinc ions diethylaniline is oxidized to colored rium. I t is remarkable that all the organic precipitants referred
p i d a l compounds by potassium fwricyanide in strongly acidi- to previously have one property in common: Their solutions (in
ed solution. This action is due to the formation of acid-insolu- water or organic liquids) present an equilibrium between the
hle zinc ferrocyanide, which thus removes ferrocyanide ions de-
rived from the redox equilibrium. Later, Szebelledy and Tanay tautomeric forms of inactive pseudo-aci form and an active aci
(90)recommended p-phenetidine as the reducing compound. iwm. When such equilibria exist, the concentration of the acl
Lapin’s (69) test for ferrocyanide by means of tetrabase (ptetra- form is invariably increased by a rise in pH. This provides a
methyldiaminebiphenylmethane) depends on the removal of plausible explanation of the fact, that these reagents are to be
ferrocyanide ions by zinc or mercury ions. Other pertinent ev-
amplqs are the sensitive tests for copper and cyanide ions (86, 76, used at pH values that guarantee a sufficiently high concentratioo
8.4) in the system: copper-benzidine(0-tolidine)-cyanide- of precipitating organic anions, but it does not explain why the
(thiocyanatc), and the Eegriwe ( 1 8 ) test for magnesium in the formed precipitat,es are resistant to far lower pH values. The
system: p-aminophenol-ammonia-magnesium, which iq based equilibria between pseudo-aci and aci forms of a precipitant imme-
on the production of a colored magnesium lake.
diately set themselves up, a fact that i s readily discerned from the
I n all these cases, the tests are dependent on the formation ot ilissolution of the particular organic compouiitl i n alkali and ita
colored oxidation products of organic reagents, which do not form r~eprecipitationwhen t,his solut,ion is made aoi(1.
these products in the absence of the materials to be detected. The question arises as to whether it is possilde to have precipi-
The term “induced oxidation” is sometimes incorrectly applied ration or solution anonislies of metallo-organic compounds with-
in such instances. According to the physicochemical nomencla- out the participation of hydrogen ions. This has thus far re-
ture an induced reaction requires the simultaneous occurrence of ceived no attention, alt’hough when organic reagents are used the
an inducing reaction with a common reaction partner, the influence of pH on the formation and stability of metallo-organic
“actor ” The evamples just cited do not involve this type of compounds is of particular pructical importance. However, two
inducing reaction or the action of an acto1 ; rather they involve instances show that false eqiiilil>i~iacan occur even with reactions
exclusivelv the shifting of redox equilibria through the removal of that are indrpenrlent of pH.
a reaction product. Hammett and Sottery ( 6 3 ) proved that aluminum can be pre-
The examples stressed concern true equilibria; there exist also cipitated by aurintricarboxylic acid only from acetic acid-acetate
“false equilibria” in precipitation reactions of inorganic ions by buffered solution, and ihat an ammoniacal solution of the reagent
organic reagents. The term false equilibrium is applied when a brings down nothing but hydrated alumina. Consequently, it
should be possible to convert this precipitated inner complex into
precipitate is acid-resistant within a pH range where it is not hydrat,ed alumina by mean9 of ammonia. This change does not
formed. This phenomenon is outwaidly analogous to the famil- take place. In contrast’, thc chromium salt of aurintricarboxylic
iar behavior of cobalt and nickel sulfides, which cannot be brought acid, which can be precipit’ated under the same conditions as the
down from acid solutions, but once precipitated from alkaline aluminum salt, is decomposed by ammonia. Recently it was
found ( S 6 ) that palladium solutions .containing cyanide ions, or
media are resistant to dilute acids. Strictly speaking, it is not more correctly complex Pd(CS)I--- ions, produce no precipitate
possible to state what the true equilibrium is in the case of a false on treatment with oxine. The establishment of an equilibrium
equilibrium; the question remains undecided as t o whether the would therefore demand that precipitated palladium oxinate be
pH value, which limits the precipitability, determines the true soluble in alkali cyanide. However, palladium oxinat,e remains
unaltered even when warmed with alkali cyanide. Other palla-
equilibrium, or whether this is governed by the much lower pH dium inner complexes with 1,2-dioximes and salicylaldosirne
value, which makes possible a solution of a precipitate. If the show a normal behavior in that they are not precipitable from
1306 ANALYTICAL CHEMISTRY
;solutions of complex palladium cyanide, and once they are shown by dissolved materials. In many cases, this hindering is
.brought down, they dissolve immediately in alkali cyanide solu- due to the formation, through the aid of suitable masking agents,
tions.
of soluble complex compounds to such an extent that the concen-
1 There is, as yet, no completely plausible and experimentally tiation of the ionic or molecular species originally present falls
confirmed explanation as to why false equilibria arise in the pre- below the concentration necessary for the occurrence of a given
cipitation and solution of metallo-organic compounds (25). I t is reaction. However, this signifies that the stability of the result-
not known whether an abnormally delayed precipitation or dis- ing complex compound, and not the mere formation of the com-
solution is involveJ. The writer believes the latter to be more plex (24, 42), is a factor of prime importance.
probable, and that this anomaly corresponds to a more or less The masking of reactions of materials in true solution is one of
persistent passivity. Hence the term “passive precipitates” the most effective means of improving the selectivity of precipita-
seems appropriate for characterizing this anomaly. Such pas- tion and color reactions. Hence the search for organic com-
sivity may be the result of an abnormal low tendency by the par- pounds that can function as masking agents should be just as
ticles of the precipitate to add and retain water on their surface vigorous as attempts to find organic compounds that enter into
(hydration). I n other words, hydrophobia is exhibited, and this precipitation or color reactions. Organic compounds that can
stands in the way of a direct contact with hydrogen ions, and accomplish maskings could be called “organic masking agents”
also impedes a surface dissociation of the precipitate particles. in order to stress their special function. Up to the present, little
Possibly the reason for the hydrophobia of inner complexes is the attention has been directed to the fact that certain groups are
fact that in their molecules the metal atoms may be enclosed on doubtless involved when organic masking agents are active. I t
all sides by organic radicals that of themselves have no hydro- may be expected that due regard for group actions will facilitate
philic groups and hence possess no centers for hydration. There the search for new selective masking agents, an endeavor that has
is a relation to the fact that so many inner complexes are not thus far aroused little interest. The possibility of converting
readily wetted by water, and also to their frequent solubility in selective precipitants into masking reagents by introducing sul-
organic liquids. fonic groups has been discussed above. The solubility of certain
The assumption that hydration is connected with the normal inner complexes in alkali indicates that alkaline solutions of inner
behavior of inner complexes, and that the lack of hydration is re- complex formers may be capable of serving as masking agents.
lated to a passivity, is supported by observations of metal oui-
nates toward dilute hydrochloric acid (3‘7). CHARACTERIZATIOS OF IYRER COMPLEXES
As no metal oxinate can be precipitated unless the pH of the Erroneous generalizations are encountered in the characteriza-
solution exceeds a certain value (with a few exceptions the lower tion of inner complexes-Le.,
limit is pH 4 5 ) , all freshly precipitated metal oxinates should dis- metallo-organic compounds in
solve inimediately in 3 N hydrochloric acid. This is true of the which a metal atom is linked with certain atoms of the organic
majority of the oxinates, including those that are known to carry molecule through principal and auxiliary valences to form a ring.
water of crystallization; their anhydrous Droducts obtained after It is frequently stated that inner complexes are distinguished by
drying at 120” C. display no passivity. In contrast, the freshly
precipitated oxinates of aluminum, iron, chromium, and gallium exceptional insolubility in water, stability against reagents, color,
dissolve just as slug ‘shly or just as little in this acid as the dried participation of functional groups responsible for ring formation
products. Especialt impressive is the coniparison of aluminum in all reactions which result in opening the ring, inability to co-
oxinate and thallium oxinate trihydrate, which are precipitable ordinate with further molecular species, and solubility in organic
in the same pH range. The thallium salt cannot be completely
dehydrated even by prolonged drying a t 120” C. (32). I t dis- liquids. There are inner Complexes to which this characterization
solves immediately in acid. I n contrast, freshly precipitated or applies, NOT, however, when the number of inner complexes
dried aluminum oxinate in 3 N hydrochloric acid remains un- has increased so greatly, it is imperative to set matters right.
altered at first and noticeable solution occurs only gradually. -4 great number of normal salts possess considerable water-
The fact that metaloxinates which carry water of crystallization insolubility. Insoluble sulfides, sulfates, chromates, oxyhy-
show no passivity, even after they are dehydrated, is probably due drates, etc., are examples. Other complex compounds are no
to their being immediately rehydrated on contact with dilute hy- more soluble in water than inner complexes. Probably the notion
drochloric acid, whereas the passive metal oxinates dissolve only that inner complexes have an outstanding insolubility originated
to the extent corresponding to their lesser hydration capability. from the knowledge that the precipitation sensitivity of colored
Experimental studies will be needed to determine the soundness inner complexes as a rule is very high. However, the precipita-
of this tentative assumption of a passivity of metal oxinates due tion sensitivity, which is determined by the ease of perceiving
to hydrophobia, and to determine whether it must be taken into small amounts of precipitate, is by no means a function solely of
account in all instances of an abnormally sluggish dissolution of the insolubility of a precipitate. I t depends also on the rapid
inner complex salts. Passivity can arise from other causes, and overcoming of supersaturation, and the form species, color, and
can occur with other organo-metallic compounds. The assem- diffractive ability of the particles that make up the precipitate.
bling of experimental material for researches on false equilibria and The influence of such factors becomes readily apparent on com-
for testing their possible analytical employment is one duty paring the precipitation of nickel by hydroxide and dimethyl-
of the chemistry of specific, selective, and sensitive reactions. glyoxime. The production of red crystalline nickel dimethyl-
glyoxime is perceptible in solutions which are so dilute that the
MASKING OF REACTIONS AND BEHAVIOR OF INNER light green amorphous nickel hydroxide precipitate can no longer
COMPLEXES be seen. Nevertheless, if the filtrate from a nickel hydroxide
When assembling and arranging material on chemistry of precipitation, or water saturated with nickel hydroxide, is tested
specific, selective, and sensitive reactions, it was found that the with dimethylglyoxime, no trace of a nickel reaction can be ob-
literature of analytical chemistry contains well-entrenched errone- served.
ous generalizations concerning organic reagents. Misstatements Sormal, insoluble salts, such as sulfides and sulfates, frequently
with respect to the terms “organic reagent” and “induced oxida- prove much more resistant to attack by acids and bases than
tion” have been pointed out above. There are similar unjustified inner complexes. The latter, for the most part, are resistant to
generalizations concerning the formation of metallo-organic com- either acids or bases, but seldom t o both, as is frequently true of
pounds. insoluble normal compounds. In fact, there are k n o m a t present
I n the first place, “masking” is often regarded as identical with only two inner complexes whose resistance actually is abnormally
the production of complexes. Masking is the appropriate term great-cobaltic nitrosonaphthol and copper phthalocyanine (15)-
for the prevention of characteristic reactions that are normally which as yet has found no use in analytical xvork. These two,
V O L U M E 2 1 , NO. 11, N O V E M B E R 1 9 4 9 1307

which are resistant to concentrated hydrochloric acid, dissolve in group actions. Duke (17) showed that accurate gravimetric
concentrated sulfuric acid, from which they are reprecipitated methods of determining copper and nickel can be based on pre-
unaltered on the addition of water. The copper compound is not cipitation by means of ammoniacal solutions of salicylaldehyde
attacked by fused caustic, and can be sublimed in vacuo. How- and weighing of the resulting inner complexes, Duke also inves-
ever, no other inner complexes are even remotely so resistant as tigated the reactions of solutions of salicylaldehyde in methyl-
the naturally occurring silicates, sulfides, and certain ignited amine and several derivatives of this primary amine, and obtained
metal oxides. findings that are of great interest to the study of group actions.
Inner complexes do not possess abnormal stability, although in Current generalizations concerning the color qualities of inner
some cases components of inner complexes exhibit a stability that complexes need to be corrected. It is true that the first inner
can legitimately be regarded as abnormal. Thus, trivalent cobalt complexes used in analytical practice were distinguished by par-
and trivalent thallium form no known soluble or insoluble normal ticularly intense color qualities. However, numerous inner com-
salts vith organic or inorganic acids. Inorganic cobaltic or thal- plexes are only slightly colored, and others, such as zirconium
lic compounds invariably contain the metal as a constituent of mandelate (67) and thallic thionalide ( 2 7 ) are colorless. .4s yet,
complex anions or complex cations, such as [TICla]-, [Co- only a few cases have been found in which colorless components
(XH3)b]-++,and [Co(XOz)a]---, or they are nonelectrolytic com- produce colored inner complexes. (An example is bismuth thion-
plex compounds. All these compounds are made up of three com- alide.) I n the great majority of instances, inner complexes are
ponents. On the other hand, cobaltic and thallic inner complexes colored when the metal ion, the inner complex-forming organic
consist always of two components ( 2 3 ) . This represents the reagent, or both are colored. This rule also applies to normal
simultaneous principal and auxiliary valence binding of the salts. A peculiarity of the color quality of inner complexes is the
metal. Cobaltic nitrosonaphthol is a familiar instance of an fact that frequently with colored starting materials there is not
inner complex trivalent cobalt salt. Further examples are the only a deepening of color in the product, but frequently a char-
stable, water-insoluble compounds cobaltic thioglycollic acid acteristic color change. The color of inner complexes doubtless is;
anilide, X X X I X (9),and thallic thionalide, X L (27). connected n-ith coordinative binding within the molecule, but CO-
H ordinative binding is also found in other types of complex com-
D-=i--)Co,$ 2 -
pounds. Accordingly, color and color intensification are also
observed in addition and intercalation compounds. The color
conferred by coordination of a,a’-bipyridine on ferrous and 00p-
XXXIX per ions is a pertinent example.
The assertion that both functional groups attached to the metal
atom react concurrently when inner complexes enter into chemical
reactions is based on an extremely interesting study by Karrer
and collaborators (64). They found that inner complexes of
silver and o-amino and o-hydroxy acids react with alkyl halides
XL in such manner that, in addition to silver halide, there is produced
not only an ester of the carboxylic acid, but, a t the same time, also
In these inner complexes, there is stabilization not only of the S-:dkyI and 0-esters of the o-situated imino and hydroxy groups.
higher valence of the metal, but also of the organic portion of the Holvever, this reaction course is not realized with all inner com-
compound. Thioglycollic acid anilide and thionalide are mercap plru silver salts of this type nor with all alkyl halides; an analo-
tans (thiols) and therefore, when not in combination, are rapidly gous mode of reaction has, as yet, never been observed with other
oxidized even by air to the corresponding disulfides. As constitu- inner complexes. Hence, a simultaneous reaction of functional
ents of an inner complex, they are stable, despite their being com- groups in inner complexes may not be taken as characteristic of
bined Tvith trivalent cobalt and thallium, \Those reductions by a this class of compounds.
mercaptan would be a logical expectation. \Vhen examining the question as to whether inner complexes
Pfeiffer and his associates (77) discovered a very impressive are coordinativdy saturated compounds, i t is necessary to take
stabilization of an organic compound resulting from its incor- account of their physical state. The great majority of solid!
poration into an inner complex. They found that compounds of inner complexes appear to be coordinatively saturated com-
very diffwent kinds are produced by the reaction of metal salts pounds, but there are notable exceptions. Inner complex metal
with ammoniacal solutions of salicylaldehyde, XLI. Calcium oxinates with definite mole numbers of water of crystallization
and magnesium form fairly soluble nitrogen-free compounds, are known, in which the binding of water molecules cannot be
corresponding to XLIa. Zinc, copper, and nickel-i.e., am- brought into conformity with the maximum coordination num-
mine-producing metals-precipitate in the form of nitrogenous bers of the metal atoms contained in the inner complex. There
compounds, to which can be assigned the constitution of inner is, for instance, a trihydrate of thallic oxinate, which can be de-
complexes of an imide of salicylaldehyde, XLII. The completely hydrated a t 120‘ C. to a stable monohydrate ( 3 2 ) . The inner
surprising thing about this divergent behavior is the fact that an complex thorium and uranyl oxinates hold an additional molecule
imide of salicylaldehyde cannot be isolated in the free state. of oxine with great tenacity (8). Cobaltic nitrosonaphthol, in
Therefore, its production in considerable amounts is occasioned nhich the cobalt has the coordination number 6, contains two
solely through the formation of inner complexes, Accordingly, niolecules of water of crystallization. At present i t is not pos-
the zinc, nickel, and copper salts of salicylaldehyde, which have sible to make precise statements concerning the localization-i.e.,
the structure XLIa, can be converted into inner complexes of the points a t which the coordinated water or ovine molecules are
salicylimide by treatment with ammonia. attached to the inner complexes.
The solubility of inner complexes in organic solvents also can
presumably be connected with their coordinative saturation. If
it is assumed that dissolution represents a chemical process which
leads to the production of a solvate, then solutions of inner com-
plexes in organic liquids must be assumed to contain solvates of
inner complexes. These solvates can be regarded as coordination
compounds of molecules of the inner complex with molecules of
Other o-hydroxy aldehydes and o-hydroxy ketones behave the solvent.
similarly to salicylaldehyde and provide excellent instances of The solubility of inner complexes in organic liquids has an
1308 ANALYTICAL CHEMISTRY

eminently practical analytical Importance. The fact that they chloroform; its solvent capability for inner complexes seems to be
dissolve in liquids that are immiscible with water provides the exceptionally great. Table I, which gives orientation relative to
possibility that inner complexes formed in water and suspended the chloroform solubility of various inner complexes, shows that
in it can be transported into the organic solvent by extraction. there is a definite connection with group actions.
It, therefore, would be extremely advantageous if soluhility in Studies of the solubility of inner complexes in organic liquids
organic solvents actually were a characteristic of all inner com- should take due account of dioxane (diethylene oxide). This
plexes, but unfortunately this property does not prevail in such cyclic ether, which is miscible with water and many organic
generalized form. Solubility in organic liquids to notable propor- liquids in all proportions, apparently is an excellent solvent for
tions is not general, nor have organic solvents been found to certain inner complexes. This property may have analytical
possess equal solution capabilities for inner complexes. Further- importance, for dioxane can also serve as solvent for organic re-
more, different metal salts of the same inner complex former may agents that are insoluble in water. Feigl and Heisig (38)de-
show a thoroughly divergent behavior toward a particular sol- scribed the preparation of a fluorescing hydrosol of aluminum
vent. The quantitative data concerning the solubility of inner oxinate by adding a dioxane solution of this inner complex drop
complexes in organic liquids are still scanty, and there have been by drop to water. The literature of colloid chemistry contains
no studies of the action of mixtures of solvents. Consequently, no record of the preparation and study of hydrosols of inner
there are no values available for comparing the solubilities of the complexes. It may be assumed that hydrosols of these salts and
salts with those of organic inner complex formers. Such data other metallo-organic compounds, and also of pure organic com-
would have great theoretical and practical value in deepening the pounds, can be prepared with the aid of dioxane as solvent.
knowledge of inner complexes and solution processes. Should this prove true, it will again demonstrate that findings
Among organir solvents. a preferred position is occupied by derived from the chemistry of specific, selective, and sensitive

Table 1. Solubility of Inner Complexes in Chloroform

Inner Complex Forming Compound - Coordination Formula Solubility
NO. Name Formula of Inner Complex in CHCL

CH
CHI--C / \C-CHs
1 Acetylacetonr (!Ha.C'O CHz.CO.CH3 Very sol.
0I !
\ /'

&fejz
Hz
S---Ni/*

2 Dicyandiamidine HN=C
/
\
x
/
Insoluble
NH-C=NH
OH
C H --C-NOH
3 Dimethy lglyoxime Slightly 801.

4 Dimethylglyoxime mono-
methyl ether Very sol.

0
NH----S-C&
/ / 1
5 Diphenylthiocarbazone 5=C C=S---- Ale/, Very sol.
\ \
N=K-CeHj

6 1,gHydroxyquinoline

7 Insoluble
7-Iodo-&hydroxyquinoline5-
sulfonic acid
05: I
OH
V O L U M E 21, NO. 11, N O V E M B E R 1 9 4 9 1309

Very sol.

Insoluble

II i
10 Sitrosonaphthol-3,6-disul- Insoluble
fonic acid

HOSS

11 Sitrosophenol Very sol.

12 Nitrosophenylhydrosylamine &-N-OH Very sol.

s
13 4-Nitrosoresorcinol Insoluble

H0

14 Quinaldinic acid Iiixoluble

c o - C O O H

13 II ulmmic acid Insoluble

16 Salicylaldoxime
Slightly sol.

@p""
17 Salicylaldehyde Very sol.

H
18 Salicylaldoxime methyl ether O-:;KOCHSII c
Very sol.

19 Salicylaldoximine Insoluble

o('c
C
H u/2
 1310 ANALYTICAL CHEMISTRY

reactions can be very useful in fields with which analytical chem- been quantitatively precipitated. This explanation cannot be
istry has no direct connection. accepted as applying to all precipitations brought about by tannin.
Metals can be quantitatively precipitated by tannin from am-
Inner complexes are soluble without decomposition in melts of moniacal or slightly acid solutions which are masked with tar-
the particular inner complex former (41). S o tests have yet trate. Such solutions contain practically no oxyhydrate particles
been made to determine whether solubility in melts is a property produced by hydrolysis, but complex metal-tartrate ions. Fur-
exhibited solely by inner complexes or whether normal salts of thermore] weakly acid solutions of alkali salts of tungstic and
molybdic acids, and acid solutions of complex niobium and
low-melting organic acids likewise show this behavior. It has tantalum compounds, when treated with tannin yield quantita-
been established that metal hydroxides and oxides and metal tive precipitations, in the form of adsorption compounds of the
salts of inorganic and organic acids react with fused %hydroxy- respective metal acids. Acid solutions of molybdates and tung-
quinoline to form metal oxinates (51). The formation of the states contain negatively charged particles of the acid anhydrides,
which cannot be flocculated by negatively charged tannin parti-
dark colored oxinates of iron and vanadium in this way makes cles. Finally, most metal precipitates produced by the action of
possible rapid, sensitive, and selective tests for ferric oxide and tannin are highly colored. This contradicts the evperience that
vanadium pentoxide even in presence of greater quantities of no considerable color changes accompany the mutual flocculation
other basic and acid metal oxides. of oppositely charged hydrosols. Thus tannin precipitations
cannot be due exclusively to mutual coagulation of colloid parti-
The experiences accumulated thus far indicate that inner com- cles, as has been assumed, and consequently other evplanations
plexes cannot be characterized with certainty by means of special must be sought.
chemical properties or peculiarities of solubility and color. All
such generalizations have proved to be in error. However, this A clearer idea of the action of tannin is obtained if considera-
error doubtless stimulated the search for organic reagents which tion is given to (1) surface reactions of colloidal particles sus-
because of their constitution are capable of forming inner com- pended in water, and (2) the reaction capability of tannin due to
plexes. In numerous cases, reactions which lead to inner com- the groups it contains.
plexes have proved to be selective, sometimes even specific, when Surface reactions of colloid particles should be regarded from
conducted under appropriate conditions. Practically all such the general standpoint of the reaction of solids with dissolved
reactions possess high sensitivity and have therefore found appli- reactants. In this case, a reaction need not proceed solely via
cation in qualitative and quantitative microanalysis. If, in addi- dissolved portions of the solid; ions or atoms of the surface of the
solid can also act directly. The products of such surface reac-
tion, consideration is given to the significance of inner complexes tions may form an independent new soluble or insoluble phase,
for complex chemistry, for questions of color, solubility, and but reaction products may also remain in the phase association
passivity, etc., it is understandable that the treatment of these of the solid. The latter condition is possible only when reaction-
salts within the bounds of the chemistry of specific, selective] and capable ions or atoms of the lattice surface do not leave the origi-
nal phase association even after the reaction has taken place,
sensitive reactions is of the greatest importance. This corresponds to an incomplete, nonuniversal reaction of sur-
face atoms or ions, which is restricted to the surface layers of the
ADSORPTION COMPOUNDS AND LAKES solid and leaves its interior unaltered. This type of reaction
A rather considerable number of organic reagents, a t suitable leads to an accumulation of an initially dissolved reactant in the
pH values, react with metal salt solutions to precipitate products surface of a solid, and corresponds to the concept of an adsorption
(45). A model example of such a surface reaction is provided
whose composition is not constant, but depends on the reaction when pure alumina is treated with dilute acid or alkali in prepara-
conditions. As a rule, the molecular ratio between precipitated tion for chromatographic studies. No separate phase made u p
material and reagent is much in favor of the former. Similar of the aluminum salt of the particular acid or of alkali aluminate
combinations of materials, whose composition shows still more ex- is formed; rather acid radicals or alkali are held on the surface of
the alumina-Le., basic aluminum salt or aluminate is produced
treme departures from stoichiometrically definable compounds, there. Consequently, in both cases, the surface reacts and fixes
are known in colloid chemistry as the result of adsorptions. the reaction product. The mode of reaction of alumina powder
Reasoning correctly that such precipitations had some sort of holds also for alumina particles in sol form, and analogous surface
reactions can be readily conceived with colloidal particles of other
connection with adsorption, the term “adsorption compound” basic and acid metal oxides.
was a logical choice, Flagg (51) has introduced the appropriate In order to understand the reactivity of tannin and to bring it
term “adsorption precipitant” for organic reagents that bring into accord with the adsorption model on alumina just cited, it is
about the precipitation of adsorption compounds. There is as necessary to consider the composition and constitution of tannin.
The latter is a collective name given to water-soluble mixtures
yet no clear understanding of the details of the action of such found in manv plants and consisting of esterlike compounds of
adsorption precipitants or reagents, and Flagg is justified in stat- various sugars with phenolcarboxylic acids. Pentagalloylglucose
ing “that the term is largely a disguise for ignorance of basic (48, 63) is surprisingly similar to Chinese tannin, and perhaps is
facts.” Tannin precipitations and the formation of metal lakes its chief constituent. This compound, which has been prepared
synthetically, has the structure:
which are of analytical importance fall into this class. An
approach from the standpoint of the chemistry of specific, selec- CHO.digal1 OH
tive, and sensitive reactions offers new and interesting aspects.
(CHO.digall)l H o b - C O . 0
TANNIN PRECIPITATES

The classic studies of Schoeller and his school have demon-

strated that tannin functions as a precipitant for numerous
metals, and that this action can be used as the basis of a series of
S‘ I
L CH20.digall
AH
06 Ho+-cooH

OH
excellent precipitation and separation procedures (82). Schoel- Pentogalloylglucose m-Digallic acid
ler’s explanation (83)of tannin precipitations is very enlightening,
a t first glance, and has been incorporated in the analytical litera- Their polyphenol components are characteristic of all tannins.
ture. Despite their high molecular weights, tannins are water-soluble
because they contain so many phenolic hydroxy groups. Their
He points nut that the solutions of salts of iron, aluminum, water solutions contain both true and colloidally dissolved tannin.
chromium, etc., which are precipitable by tannin, are without There is little doubt that the content of phenol groups confers
exception solutions of hydrolyzable metal salts and consequently reactivity on colloidal particles of tannin as well as on dissolved
contain hydro1 sis products (metal ouyhydrates) in hydrosol tannin molecules. Therefore, solutions of hrdrolpzable metal
form. Such soy particles are positively charged. Aqueous tan- salts can react in different ways with colloidallv dispersed and
nin solutions are colloidal dispersions m which the tannin parti- genuinely dissolved tannin. Surface reactions of hydrolysis prod-
cles carry negative charges. Therefore, the colloidal particles ucts may occur with either dissolved or colloidal tannin, surface
can undergo mutual discharge and flocculation with continuous reactions of colloidal tannin particles with metal ions are possible,
disturbance of the hydrolysis equilibrium until the metal has and finally there can be reactions between metal ions and tannin
V O L U M E 21, N O . 11, N O V E M B E R 1 9 4 9 1311

in true solution. 411 these possibilities lead to metal-tannin

bindings, and the particular conditions will determine which reac-
tion will prevail. However, under no circumstances will the re-
sulting precipitate be homogeneous nor will it conform to a
stoichiometric formula. I t is also apparent that complex metal
ions in ammoniacal solution will react ryith colloidal or truly dis-
solved tannin in so far as the products are not soluble in water.
The assumption that the reactivity of tannin is due to its phe-
nol groups provides a good explanation as to why metal acids are
precipitable by tannin. Phenols can react with metal acids to ci 0
t’orm phenol esters, as shown by the precipitation of formula-pure
esters of tungstic, molybdic, and vanadic acids by S-hydroxy- SLIII SLIV
quinoline (97). An esterlike binding betmen tannin and metal IIoehlau’s findings appear to have been completely forgotten,
acid can obviously occur as well on the surface of colloidal acid or their significance with respect to the aluminum-alizarin lake
anhydride particles as on the surface of colloidal tannin particles. was not appreciated. They demonstrate that the red aluminum-
The intense color of metal-bearing tannin adsorption compounds alizarin lake cannot be either an inner complex aluminum alizari-
is likewise in harmony with the idea that the tannin is reacting nate or a misture of this salt with unchanged alumina. The red
chemically. As shown by its formula, tannin contains many lake, produced by adding ammoniacal alizarin solution to alumi-
ortho- situated hydroxy groups. When a metal salifies a group num solutions and then acidifying with acetic acid, has distinctly
uf this kind it is possible for the metal to be coordinated u i t h the different properties from the compound isolated by Moehlau.
oxygen of a neighboring hydroxy group; inner complex bindings The lake is resistant to ammonia and is decomposed by hrdro-
result, with the color effects that so frequently accompany them. chloric acid much more dilute than 1-V.
In fact, the precipitation products of‘the noncolloidal gallic and
digallic acid have approsimately the same color as tannin pre- Discussions of the formation, composition, and constitution of
cipitates.
color lakes must take due account of the rule that lake reactions
Hence tannin precipitates can be regarded as products of sur- employed in analysis occur in the pH range of the respective mctal
face reactions of colloid particles, in which phenolic hydrosy hydroxides (osyhydrates) which are then in the sol or gel form.
groups of the tannin participate. However, the products of such I t is therefore probably the metal hydroxide, and not the metal
surface reactions are not metal tannates or tannin esters of metal ions, which reacts in the aqueous solution with the dyes. This
acids in an independent phase, but the tannin binding takes place provides a tie in with the metal tannin precipitations discussed
on the surface of colloidal particles, whose interior remains un- above, 1% hich nere viewed as surface reaction products of sol par-
affected. The designation “adsorption compound” thus ac- ticles. I t is plausible that alizarin reacts with alumina:
quires real meaning because tannin precipitations involve effects
that can be viewed from the standpoint of adsorption as well as
[a41(OH)3], + H-lliz --+ [ ~ 4 l ( O H ) ~ ] , - ~ . ( O H ) ~ . r+l l ~H20
~liz
chemical binding. This schematic representation indicates that the aluminum
atoms of the alumina surface do not all react with these alizarin
FOR\I4TION O F METAL LAKES molecules, which would result in the production of aluminum
The term “lake” (color lake or metal color lake) comes from alizarinate as an independent phase. The reaction proceeds only
the chemical technology of textile dyeing. I t signifies the fising partially with principal and auxiliary valence binding of the
of dyes by inorganic metal compounds, usually oxides or oxyhy- alizarin on the surface of the alumina without forming a new
drates. There is no uniform concept as to the nature of the fix- phase. This corresponds to a chemical adsorption of alizarin.
ing. In a fundamental study, Werner (102) showed that lake- I n the purely formal sense, it signifies a formation of basic alumi-
forming dyes are inner complex formers, and that colorless inner num alizarinate in the surface of the alumina. I t might also be
comples-producing organic compounds, which form colored stated that an “alizarinisation” of the alumina surface occurs.
metal salts, are irreversibly bound by the corresponding oxides in The reaction picture presented for the aluminum-alizarin lake
a way analogous to that by which dyes are held by metal mor- can be appropriately applied to metal lakes of dyes, which possess
dants. On the other hand, Weiser (96), on the basis of careful inner complex-forming groups. Therefore, such lakes are to be
experimental investigations, champions the view that the taking regarded as adsorption compounds, in which dyestuffs are held
up of dyestuffs by the so-called mordants is an adsorption process by principal and auxiliary valences to metal atoms on the surface
which does not lead to chemical compounds. Both concepts are of gel or sol particles of the respective hydrosides (osyhydrates).
discussed in connection with metal color lakes in the literature of Because of this binding, color lakes have a similarity to inner
organic, colloid, and analytical chemistry (47, 49,60, 80, 100). complexes. This is shown by the fact that lakes frequently have
Reactions that lead to lakes play an important role in analytical a more intense or different color from that of the free dye. This
chemistry. After Atack (1) showed that the aluminum-alizarin divergence is the basis of the analytical usefulness of many lake
reaction could be utilized for the sensitive detection and colori- reactions.
metric determination of aluminum, many lake-producing reac- In the color lakes there is realized an inner complex binding
tions were described for the detection and determination of vari- which cannot be accomplished by a reaction of metal and dye-
ous metals (101). Most of the lake-forming dyes are acidic com- stuff ions with production of inner complexes. As shonn by
pounds, whose constitution is such that they are fundamentally Moehlau, it is not permissible to rule out the possibility of a pro-
capable of producing inner complexes. Although this fact seems duction of inner complex metal salts of acid dyes under special
to support Werner’s assumption that lakes are inner complexes, circumstances. I n isolated cases, including the analytical forma-
such inner complexes have been actually isolated in surprisingly tion of lakes, products may closely approach the composition of
few instances, and then only under conditions that do not corre- inner complexes. This has been recently shown by Liebhafsky
spond to those prevailing during the production of lakes in and Winslow (70) within the system hafnium-alizarin. However,
analytical procedures. all experience indicates that the production of adsorption com-
pounds is the preferred mode of reaction.
Aluminum-alizarin lake is often cited as a model of lake pro-
duction. Moehlau ( 7 4 ) prepared a formula-pure inner complex Feigl and Zocher (45)discovered interesting examples of the
aluminum alizarinate, XLIII, and described its properties. This behavior of organic inner complex formers in surface reactions
salt, which was obtained by the reaction of monopotassium with metal oxides.
alizarinate with aluminum chloride, is insoluble in 1 N hydro-
chloric acid and dissolves in ammonium hvdroxide. Moehlau If alumina is shaken with a concentrated chloroform solution of
correctly assumed that an ammonium salt of an aluminum 8-hydroxyquinoline, the filtrate exhibits a yellow fluorescence.
alizarinic acid. S L I V , is produced and he substantiated this idea This shows that a chloroform-soluble fluorescent compound has
by the isolation of the corresponding alkali earth salts. been formed. However, the alumina, which appears to be un-
1312 A N A L Y T I C A L CHEMISTRY
changed, likewise fluoresces and this fluorescence persists despite lead also to results that are interesting from the standpoint of
prolonged washing with chloroform. Inasmuch as the fluores- new analytical methods.
cence of metal oxinates is connected with the inner complex bind-
ing of the metal, it must be supposed that the surface of the alu- The adsorption of oxine on the surface of alumina, forming a
mina is “oxinated,” analogous to the “alizarination” of the sur- fluorescent product, can be used as a highly sensitive test for the
face of the alumina in the aluminum-alizarin lakes. If the ex- detection of oxine (46). When 0.5 liter of oxine dissolved in wa-
periment is repeated with increasingly diluted oxine solutions, a ter or organic liquid is shaken with alumina and filtered, a
region is reached in which no aluminum oxinate is formed as a fluorescent product is obtained, even a t a dilution of 1 to 10,000,-
separate phase, but there is only a holding of the oxine on the 000,000. This is a good example of the sensitivity of a test
surface. This latter attachment is clearly revealed by the fluores- based on detection of traces.
cence of the alumina. It appears, therefore, that the inner com- Small quantities of calcium, magnesium, or aluminum in paper
plex binding of the oxine on the surface of the alumina-i.e., the can be detected by spotting the paper with a solution of oxine or
chemical adsorption of oxine-is a preliminary stage in the for- sulfonated oxine, or by exposing the paper to the vapors of oxine
mation of aluminum oxinate. at room temperature. Fluorescent flecks are obtained through
the contact of oxinate-forming materials (37).
The sulfonic acids of oxine give no precipitate with water solu- Palladium cyanide can be specifically identified by spotting
tions of aluminum salts, and as yet no aluminum compounds 01 wit,h an appropriate indicator solution; colored addition lakes are
this type have been isolated in solid form. If, however, alumina fwmed (31).
is treated with aqueous solutions of these sulfonic acids, fluorescing
LITERATURE CITED
products are obtained. This indicates that inner complex bonds
have been formed through surface reactions, even though thi? (1) S t a r k , F. W.. J . Soc. $‘hem. I n d . . 34,936 (1915).
(2) Barcel6, J., Anales SOC. espafi. f l s . z/ q u h , 31,852-60 (1933,.
binding cannot be accomplished in the form of solid inner com- (3) Baudisch, O., Chem.-Ztg., 33, 1298 (1909).
plexes. The behavior of oxine sulfonic acids toward alumina is (4) Bnudisch, O., et al., Be?., 45,1164 (1912): 48,1665 (1915); 49,
further evidence that the reactivity of inner complex-forming 180, 191, 203 (1916).
groups in organic compounds is not annulled by the introduction (5) Baudisch, O., and Holmes, S.,2.anal. Chem., 119,241 (1940)
(61 Bellucci, I., Gam. chim.ital., 49,284 (1919).
of sulfonic groups. The chemical reaction of oxine and its sul- (7) Berg, Richard. “Die analytische Verwendung von o-Oxychinoliri
fonic acids with alumina, which can be followed through the for- (Oxin) und seiner Derivate,” Stuttgart, F. Enke, 1938.
mation of fluorescing products, demonstrates the reaction possi- (8) Berg, R . , and Becker, E., 2. anal. Cheni., 119, 1 (1940).
bilities of dyes, whose constitution makes them potential inner (9) Bersin, T h . , Ibid., 85, 428 (1931).
( I O ) Blau, F., Monatsh., 19, 647 (1898).
complex formers. The inner complex binding of oxine on the (11) Brrckenridge, J. G.. Lewis, R. W. J., and Quick, L. A., Can. J .
surface of alumina corresponds to the inner complex binding of Research, 17B, 258 (1939).
dyes on the surface of alumina. In one case, fluorescence results: (12) Chatterjee, R. J., J . I n d i a n Chem. Soc., 15, 608 (1938).
in the other, the color characteristic of lakes appears. (13) Cone, N. H . , and Cady, L. C., J . Am. Chem. Soc., 49, 2214
(1927).
When considering these color lakes from the standpoint of com- (14) Cooper, A., J . C h e m , M e t . M i n i n g SOC.S . A f r i c a , 25, 296
plex chemistry, it is essential to keep in mind that the deepening (1925).
of a color or a color change is not unique with inner complex bind- (15) Dahlen, M. A., I n d . E n g . C‘hem., 31,839 (1939).
(16) Diehl, H., “Application of the Dioximes in Analytical Chemis-
ing; such color phenomena can also arise through purely coor- t r y , ” Columbus, Ohio, G. Frederick Smith Chemical Co..
dinative binding. Accordingly, it may be expected that “addi- 1940.
tion lakes” must exist as well as “inner complex lakes.” The (17) Duke, F. R., ISD. ENG.CHEM.,ANAL.ED., 16,750 (1944)
former would be colored adsorption compounds in which the per- (18) Eegriwe, E . , Z . anal. Chem., 118,98 (1939).
tinent surface reaction consists of an addition reaction-Le., (19) Ephraim, E . , Be?., 63,1928 (1930).
( 2 0 ) Ibid., 64, 1210 (1931).
coordination of dyestuff molecules. The production of such lakes (21) Feigl, Fritz, ”Chemistry of Specific, Selective, and Sensitive
has now been indubitably demonstrated (70) in studies dealing Ileactions,” S e w York. Academic Press, 1940.
tl-ith the action of organic bases on palladium cyanide. Strong (22) Ibid., p. 187.
(23) Ihid., p. 240.
diacid bases (B), such as ethylenediamine and a,a’-bipyridine, (24) Ibid., Chap. 11’.
form addition compounds of the type Pd((3S)a.B. These com- ( 2 6 ) Ibid., Chap. XI.
pounds correspond to the long-known diammine Pd(CN)2.2”7, (26) Feigl, Fritz, I m . ENG.CHEM.,ANAL.ED.. 8, 404 (l!C3tii.
R hich is obtained by treating palladium cyanide with ammonia. f.27) Feigl, Fritz, Nature ( L o n d o n ) , 161,435 (1948).
( 2 8 ) Feigl, Fritz. unpublished studies.
\Teak organic bases are not capable of producing diammines as a (29) Feigl, Fritz, Z . anal. Chem., 74, 380 (1925).
separate phase. However, their coordination capability is re- (30) Feigl, Fritz, and Barbosa. P. E., ISD. Esc,. C H E Y . , AXAI,.h..
vealed by an attachment on the palladium atoms of the surface of 14. 519 119421.
palladiuin cyanide. Adsorption compounds of palladium cya- (31) Feigi, Fritz, and Baurnfeld, L., Anal. Chim. Acta, 3, 15 (1949).
(32) Ibid., 3,83 (1949).
nide with organic bases are obtained. Colored adsorption com- (33) Feigl, Fritz, and Christiani, -1. v , , Z. anal. Chem., 65, 341
pounds-color lakes-are obtained when colored R eak organic (1925).
bases are used. In the majority of cases, the color of these addi- 134) Feigl, Fritz, and da Silva, C.E’. d., ISD. Ex).CHmr., AN.?.L
tion lakes is more intense, and often different from the rolor of ED.,14, 316 (1942).
( 3 5 ) Feigl, Fritz, and Feigl, H . E., A n a l . Chim.Acta, 3,300 (1949’
the corresponding basic dyes. (36) Feigl, Fritz, and Hamburg, H.. 2 . anal. Chem., 86, 1 (1931).
The behavior of acid-base indicators, which contain NHz or (37) Feigl. Fritz, and Heisig, G. B., A n a l . Chim. Acta, 3, (1949).
dkylated SITz groups, is especially interesting. They form in press.
lakes nith palladium cyanide, and the color of the product is (38) Feigl, Fritz, and Heisig, G. B., Ibid., in press.
(39) Feigl, Fritz, Krumholz, P., and Rayinarm, E . , NZlii;rochemie, 4,
essentially that of the particular indicator in acid solution. This 395 (1931).
effect probably presents a contribution to the resonance theory of 4 0 ) Feigl, Fritz, and Miranda. L. I . , IKD. ESG.CHELI.,As.LI.. ED..
acid-base indicatoi s. The addition lakes of palladium cyanide 16,141 (1944).
are examples of not only a new rlass of color lakes but also a new 4 1 ) Feigl, Fritz, Miranda, L. I.. mid Spinelli, H. A , , Ministerio agr.,
Dept. nacl. produqao miiitral, Lab. producao lnineral ( B r a -
class of metallo-organic adsorption compounds. Hitherto. the d ) ,11, 81 (1943).
known examples of this class of adsorption compounds invariably 142) Feigl, Fritz, and Raacke, I. D., A n a l . Chim. Acta, 1, 317 (1947):
liad basic or acid metal oxides as their inorganic components. 2, 397 (1948).
(43) Feig!, Fritz, and Schwarz, It., Rrc. trav. chim., 58, 474 (1939).
(44) Feigl, Fritz, and Suter, H. .A,, J . Chem. Soc., 1948, 378.
CONCLUSIONS (46) Feigl, Fritz, and Zocher, H. W., J . Colloid Sci., in press.
(46) Fichter, F., and Goldarh, rl., Helw. Chim. A d a , 15, 1513 (1932).
The foregoing considerations and experiments, which constitute (47) Fieser, L. F., and Fieser, M.,“Organic Chemistry,” pp. 837,
an attempt to elucidate the formation and constitution of lakes, 867, Boston, D. C. Heath, 1944.
V O L U M E 21, NO. 1 1 , N O V E M B E R 1 9 4 9 1313
Fischer, E., and Bergmann, M., Ber., 51,1760 (1918). ucccld. sci. Torino, 57,384 (1922); and subsequent publica-
Fischer, H., Unruh, A. v., and Kurt, F., “Handbuch der ana- tions.
lytische Chemie,” p. 305,Berlin, 1942. Sandell, E. B., IND.ENG.CHEM.,S SAL. ED., 13,844 (1941).
Flagg, J. F., “Organic Reagents Used in Gravimetric and Volu- Sandell, E. B., “Polarimetric Determination of Traces of
metric Analysis.” New York, Intersrience Publishers, 1948. Metrtls,” p. 70,New York, Interscience Publishers, 1944.
Zbid., p. 30. Sarver, L. A., IND.ENG.CHEM.,ANAL.ED., 10,378 (1938).
Zbid., pp. 107-119. Schoeller, W. R., and Powell, A. R., “Analysis of Minerals and
Hammett, L. T., and Sottery, C. T., J . Chem. SOC., 47, 142 Ores of the Rare Elements,” 2nd ed., London, Charles Grif-
(1925). fin & Co., 1940.
Haves, J. H . , and Chandlee, G. C., ISD. ESG. CHEM..ANAL. Zbid., p. 6.
ED., 14,491 (1942). Sievertz, A., and Hernsdorf, A., Z . angew. Chem., 34,3 (1921)
Hayes. W. G., and Jones, G. W., Zbid., 13,603 (1941) Smirnoff, A., Helv. C h i m . Acta, 4,802 (1921).
Herzog, W., Chem.-Ztg., 50, 642 (1926). Smith, J.,ANAL.CHEM.,20,937 (1948).
Hieber, W., and Leutert, F., Ber., 62,1839 (1929). Soule, B. A,, J . Am. Chem. SOC.,47,981 (1925).
Hill, R., Proc. R o y . SOC.London, 107,205 (1930). Spacu, G., B u l . SOC.$tiinfe Cluj, 3,285-307 (1927).
Hoffrnann,O., Ber., 18,46 (1885). Spacu, G., and Makarovici, C. G., 2. anal. Chem., 102,350
Holleman, A. F., “Lehrbuch der organische Chemie,” Chap. (1935).
20.p. 837,Berlin, W. de Grupter & Co., 1935. Szebelledy, L., and Tanay, I., Zbid., 106,342 (1936).
Hoste, J., Research, 1,713 (1948). Tartarini, G., Gazz. chim. ital., 63,597 (1933).
Illinsky, M., and Knorre, G. v., Ber., 18,699 (1885). Tschugaeff, L., Ber., 38,3520 (1905); 2 . anorg. Chem., 46,144
Karrer, P., Helv. C h i m . A c t a , 6,3 (1923). (1905).
Karrer, P., and Naegli, C., and Weidmann, H., Zbid., 2,242 Vellur, Leon, and Peser, Maurice, Ann. pharm.fi.anc., 4,10-12
(1919). (1946).
Klooster, H. S. van, J . Am. Chem. SOC.,43,746 (1921). Voter, R. C., Banks, C. V., and Diehl, H., ANAL.CHEM.,20,
Kulberg, L., Mikrochemie, 20,153 (1936). 458,652 (1948).
Kumins, C. A., AN.<L.CHEM.,19,376 (1947). W‘allach, 0.. Ann., 437,175 (1942).
Kuznetsov, V. I.,J . A n a l . Chem. ( R u s s i a ) , 2,179 (1949). Weiser, H. B., “Inorganic Colloid Chemistry,” Vol. 2,Chap.
Lapin, L. N., T r u d y Uzbekskooo Gosudarat. U n i v . Sbornik 16,New York. John Wilev & Sons, 1938.
Rabot K h i m . , 15,150-68 (1939). Welcher, F. J., “Organic Analytical Reagents,” New York.
Liebhafsky, H. A., and Window, E. H., J . Am. Chem. Sot., 69, D. Van Nostrand Co., 1947-48.
1130 (1947). Ibid., Vol. 1, 1947.
Mandal, K. L., Science and Culture, 5 , 719 (1940). Ibid., Vol. 2,p. 277.
Merritt, L. L., and Walker, I. K., IND.ENG.CHEM.,ANAL.ED.. (100) Ibid.; Vol. 4,p. 359.
16,357 (1944). (101)Zbid., Vol. 4,Chaps. 14,15,1948.
Miller, C. F., Chemist-Analyst, 26,35 (1947). ! 102) W’erner, A., Ber., 41,1062 (1908).
Moehlau, A,, Ber., 46,4143 (1913). (103)West, P. W., Anal. C h i m . Acta, 2,133 (1948).
Pertusi, C., B t t i ZZZ congress0 naz. chim. pura applicata, 1930, (104)West, P. W., and Arnis, E. S., IND.ENG.CHEM.,ANAL.ED., 18,
573-5. 400 (1946).
Pertusi, C., and Gsstaldi, E., Chem.-Ztg.. 37,609 (1913). ilO5) Willard, H. H., and Smith, G. >Ibid., I., 11, 186,269 11939).
Pfeiffer, P., and Buchholz, E., and Bauer, O., J . prakt. Chem., (106)Willink, H. T., and Wibaut, J., Rec. trav. chim., 54,275 (1935).
129,163 (1931). (107)Yoe, J. H., J . Am. Chem. SOC.,54,4139 (1932).
(108)Yoe, J. H., and Sarver, L. A., “Organic Analytical Reagents,”
Ponzio, G., et al.. Gazz. chim. ital., 51,11,213-25 (1921);52,I, Kiew York, J o h n Wiley & Snns. 1941.
289-301 (1922); 52, 11, 145-60 (1922); 53, 15-19,20-4,
25-35, 297--305,305-11. 311-18, 318-27 (1923); A t t i R E C * E I V E D July 5 , 1949.

2nd Annual Summer Swmpoeium - Organic Reagents

Polysubstituted 1,lO=Phenanthrolines and

Bipyridines as Multiple Range Redox Indicators
Further Applications as Specifi Organic Analytical Reagents
B’4RREN R’. BR4NDT’ 4ND G. FREDERICK S;ZIITH, r’niversity of Illinois, Urbana, I l l .

dine, 1,lO-phenanthroline, and terpyridine for use as redox indica-

T HE ‘*ferroin!’ reaction involves chelate ring formations
having as their basis the specific group configuration
I, I
tors.
Use of the above formulations in the photomet,ricdetermination
of iron, copper, molybdenum, and cobalt, and the detection of
(=S--C---C-X=l. This sperific group occurs in the hetero- other metals.
cyclic ring nitrogen compounds. of which the following are the As anion precipitants for separations and determinations such
mo-t familiar examples: 2,2’-hipyridine: 1.10-phenanthroline: :ts perchlorate, periodate, persulfate, and thiocyanate from chlo-
and 2,2‘,2”-terpyridine. rate, iodate, sulfate, and cyanide, respectively.
As masking reagents (in t8hequantitative separation of traces of
This type of compound is capable of beiiig altered over a wide aluminum in the presence of large amounts of iron, as a typical
r:trige of modifications through the suhstitution of various t>-pe mample).
replacement groups, such as meth)-l, phenyl, csrbosy, nitro, and
halogen, for one or more of the hydrogens of the psrent substance. HISTORICAL
These derivatives of the parent substances can be shon-n to affect
the propertirs of the resulting reagents under the following cate- General. The preparation, properties, and analytical applica-
gories : tions in the use of the organic reagents having the ferroin specific
grouping thus far investigated are described by Smith and Richter
Formation of the complex divalent ferrous cations with bipyri- (58), Walden and Edmonds ( 4 2 ) , and Oesper (SI). The most
1 Present address, Purdue University. Lafayette, Ind. recently published source of reference and the most complete