V O L U M E 25, NO. 7, J U L Y 1 9 5 3 compound in a distillate fraction boiling below the boiling point of the pure compound is not conclusive evidence of azeotrope formation, but it may be regarded as suggestive. The emphasis is here on the identification of perylene in a concentrate of an Edeleanu fraction of a lubricating oil. On the strength of absorption spectra obtained with synthesized samples of anthracene and pyrene in various nonpolar solvents, the presence of anthracene and possibly also of pyrene in the concentrate of perylene is claimed. It is well to remember that Wanless, Eby, and Rehner’s observations were made on solutions in benzene, whereas the spectra referred t o here were measured in isooctane as the solvent. There is a red shift of the order of 50 A. for the relevant bands on changing from iso-octane to benzene.
1013 at4030A., and the 1,2,5,6-dibenzanthracene bandat 3950A. There is no evidence in support of Zerbe and Folkens’ (9) conclusion of the predominance of polynuclear aromatic hydrocarbons in the Edeleanu extracts of mineral oils. ACKNOWLEDGMENT
The authors express their thank3 to Max Sulzbacher, The Grosvenor Laboratory, London, for the synthesis of the specimens of pyrene and perylene which were used for reference purposes, and to H. S. Boyd-Barrett and Frank Long, formerly of Petrorarbon, Ltd., for the synthesized anthracene. LITERATURE CITED
CONCLUSION
The quaiititative estimates of anthracene, pyrene, and perylene were based on spectrophotometric determinations obtained with synthesized specimens of these compounds. There is little tloubt that a detailed analysis of the content of polynuclear aromatic hydrocarbons in high boiling petroleum fractions can be made by spectrophotometry with the aid of chromatographic separation of the constituents of narrow distillation cuts. Although at the present stage the possibility cannot be excluded of the. presence of very small percentages of 1,2-benzanthracene, it is certain that .3,4-benzopyrene, and 1,2,5,6-dibenzanthracene, there cannot be present more than 1, 0.03, and O.S%, respectively, of these compounds. These estimates were made from the 1,2-henzanthrarene band at 3830 A the 5,4benzopyrene band ~
(1) -%dams,K. G., and Richardson, E. (2)
(3) (4)
(5) (6) (7)
(8) (9)
N., ANAL. CHEM.,23, 129 (1951). Clar, E., “Aromatische Kohlenwasserstoffe,” 2nd ed., RerlinGottingen-Heidelberg, Springer-Verlag, 1952. Cleaves, A. P., Carver, hl. S., and Hibbard, R. R., Natl. Advisory Comm. Aeronautics, Tech. Note 1608 (1948). Coggeshall, N., and Posefsky, J., J . Chem. Phys., 19, 1980 (1951). Friedel, R. A , and Orchin, hI., “Cltraviolet Spectra of Aromatic Compounds,” Spectrum No. 569, New York, John Wiley & Sons, London, Chapman & Hall, 1951. Jones, R. N., Chem. Revs., 32, 1 (1943). Orchin, hl., Reggel, L., Friedel, R. h.,and Woolfolk, E. O., Bur. Mines, Tech. Paper 708 (1948). Wanless, G. G., Eby, L. T., and Rehner, J., Jr., ANAL.CHEY. 23, 563 (1951). Zerbe, C., and Folkens, K., Brennstof-Chern., 16, 161 (1935).
RECEIVED for review December 24, 1961. Accepted May 1 , 1953.
Production, Particle Characteristics, and Spectra of Tellurium Hydrosols for Spectrophotometry RALPH A. JOHNSON, Department of Chemistry, University of Illinois, Urbana, Ill.
To form a basis for the development of the spectrophotometric deterniination of tellurium as the elementary tellurium hydrosol, fundamental studies on the production, particle size, and spectra of the sols have been carried out. The effects of hypophosphite, gum arabic, and acid concentrations have been studied and statistically analyzed. Electron microscopic investigations revealed the particles as elongated spindles or rhombs, the ratio of length to breadth being about 15 to 1 and with
T
ELLURIUM may be determined in trace concentrations by photometric methods involving light transmission through elementary tellurium sols. The spectrophotometry of the method is complicated because both the optical density characteristics and the wave length characteristics of the sol spectra vary continuously with the size of sol particles. Since particle size is determined by the conditions of sol formation, the spectral characteristics are an implicit function of those conditions. The investigation reported here is intended as a fundamental study of the interrelationships indicated above. It is preliminary to development of a quantitative analytical method described in detail in another paper ( 7 ) . ELECTRON MICROSCOPY OF TELLURIUM PARTICLES
Tellurium particles were mounted for electron micrographic studies on collodion film supported on 20-mesh copper Electromesh screen. . 4 droplet of the sol Fas placed on the film and
lengths and breadths distributed normally with a standard deviation of about 10% relative. For blue sols, the length is about 1800 A,; for amber sols, length is nearly 300 A. The color variation is due to a nonspecific, conservative absorption band, which undergoes a bathochroniic shift with increase in particle length. Two specific, consumptive absorption bands are found at 7400 A. and at 2900 A. For the ultraviolet band, the optical density increases regularly with decrease in particle size.
evaporated before being placed in the microscope. The instrument used was an RCA Model B electron microscope, operating at 55 kv. The electron micrographs in Figures 1 and 4 are photographs of the beam image at 9000 X magnification, photographically enlarged to 70,OOOX. Some sols m r e shadow cast with platinum and palladium, but no additional information was obtained from them. The primary particles appear to be elongated rhombs or spindles. Also the primary particles appear to be agglomerated in a remarkable way-viz., they form pairs. Rarely is a primary particle found alone or are more than two particles found together. Because of the occasional asymmetry of orientation of particles with respect to each other and because, in some cases, there appears to be a continuous line of demarcation between two primary particles, it is tentatively concluded that the doublets are agglomerates and not crystallographic twins. It is noted that pairing is found throughout the size range of particles studied.
1014
ANALYTICAL CHEMISTRY Table I.
Dimensions of Sol Particles
Dimensions are obtained from sols repreaented in Figures 1 t o 4. Ail dime?sions &re in Angstrom units. Sol? %reformed according t o promdure given m text. For sols of the same tellunum Oonoentration. variations m sol p r o ~ e h ties are produced by w r y i n s hYpophosphite concentration. Width of Pairs Length Sol Te. Standard Stand+ Color Figure P.P.M. Mean deviation Mean deviation 1788 147 14 233 20.3 1A Blue 1233 135 17.3 14 179 1R-4A PurDile 1063 93 10 187 1 3 . 6 Purple 4R 986 140 14.6 4 188 4c Purple 965 107 19.8 4D 2 2U7 Purple 518 113 14 141 16.6 IC Red 68 15.1 310 14 116 ID Amber
OL
600
400
09-
1 600
400 WAVE LENGTH IN m,u
Figure 2.
i
graphs were optically magnified and measured with an ocular micrometer oalibrated against electron micrographs of tobacco mosaic virus. Dimensions of particles shown in Figures 1 and 4 are given in Table I. Discussion is given below. PARTICLE SIZE AND OPTICAL PROPERTIES
The possibility of producing sols of continuously varying colors by proper control of reaction conditions has interested a number of investigators. Systematic studies by Gutbier ( 6 ) and by Ostwald (fa)led to Ostwald's rule that the color of sols goes up in spectral order as particle size decreases. In the highly colored metal and metalloid sols, the particle dimensions are less than the wave length of light. In this size range, electromagnetic scattering is considerable and accounte. for the continuous variation in color. Following the lead of Lord Rayleigh (f4), Mie (fd) made a theoretical and mathematical treatment of the interaction of light with very small, spherical, absorbing and conducting particles (metals). Mie's theory applied to gold sols correctly predicts the observed color variations-~., the color shifts hathochromically with increase in particle size (6). Gana ( 4 ) extended the Mie treatment to nonllpheriod particles. He
2m
Spectra of Tellurium Sols
showed that for gold sols a small deviation from the spherical wan sufficient to displace the absorption maximum toward longer wave lengths-i.e., sol color toward the blue (10). Similar theoretical and experimental studies were made by Gribnau ( 6 ) on selenium as an example of a metalloid sol. Good agreement between theory and experiment was found in the visible region of the spectrum. A brief study and comparison of the spectra of selenium and tellurium sols w m made by Auerbrtch (1). Krishnan (8, 9)used the depolariaation of scattered light in an optical study of tellurium sols. Spectra of various tellurium sols obtained in the present investigation are shown in Figure 2. Comparison of these spectra with corresponding electron micrographs in Figure 1 ~ervesto illustrate the following discussion. The principal absorption band in the visible is nonspeoific with respect to wave length and dependent upon particle size. A hypsochromic shift parrallels a decrease in particle size and a tcndency toward equilinearity. Correspondingly, the sol color shifts in the order: blue purple red amber a B the particles become smaller and less elliptical. For the spectra in Figure 3, a decrease in wave length is associated with 8 decrease in the length of particles; but, for this case, there is an increase in the width of the particles (cf. Table I), This behavior indicates two possible controlling factors: the length of particles and the axial ratio. The opticd density of the visible band shows little variation for blue, purple, and red sols. I n amber sols, the visible and ultraviolet bands overlap. In the resulting mutual re-enforcement, a sharp increase of optical density is effected. The behavior of the ultraviolet band stands in contrast to that of the visible band, Not only is the wave length shift of the former relatively slight, but the optical density is regularly increased with the bathcchromic shift of sol color and decrease in particle size. The ultraviolet band and the small band appearing around 740 mp appear to he specific and due to consumptive absorption.
- - -
V O L U M E 25, NO, 7$ J U L Y 1 9 5 3
1015
In this paper and in the following paper, the visible band is treated in detail. Although considerable data have been collected on the ultraviolet band, further work is necessary before evaluation of its analytical usefulness czn be made. PARTICLE SIZE AND SOL FORMATION
The process of tellurium sol formation which was studied in this investigation appears to he similar in m m y respects to the process described by LaMer and Dinegar (1l)for the formationof uniformly dispersed sulfur sols by a chemical reaction. There is an induction period which follows the mixing of the reagents, during which a chemical reaction builds up the concentration of precipitable material (tellurium), thus leading to nucleation. The nucleation process usually occurs rapidly and brings with i t a rather sudden depletion in the concentration of precipitatable material in the original phase. Since nucleation cau take place only a t a relatively high concentration+.e., high relative supersaturation-the nucleation process man chokes its own progress by causing tellurium depletion to concentrations below the critical level necessary foi nucleation. As the concentration is lowered, the growth process takes over and concentration levels permitting nucleation are not again reached. The process outlined has elements which lead to remarkably uniform and reproducible particles. A relatively long induction period permits variations in reagent addition and mixing to equalize themselves and makes possible nucleation from a homogeneous medium. Because the nucleation period is rela, tively so brief, the conditions and period of growth are very nearly the Same for all particles; hence the final particles are very much alikei.e., they approach monodispersion. The data in Table I indicate that particles formed under the conditions of this investigation vary shout +IO% relative standard deviation.
PHYSICAL CONOITlONS OF SOL
A limited number of experiments were casnea out on m e pnysical conditions for sol formation by reduotion of tellurium(1V) with hypophosphite. From consideration of the results obtained, a set of conditions was chosen and maintained constant throughout the remainder of the investigation. It may be added that this phase of the subject invites further study.
2P.P.M.
I ~
An exception to the general rule is provided by the increase in particle length with increase of tellurium concentration. The extension of the particles doubtleas reflects iw1 extra growth capscity because of higher tellurium concentration, The effect of "foreign nuclei" is not great if a good grade of distilled water and reagents is used throughout. Although traces of micraparticulate matter are alwaya present, supersaturations are used which produce an overwhelmingly large proportion of tellurium nuclei; induced nucleation is thus negligible. Foreign particles, such as lint, serve as foci for agglomeration of so1 particles. The induced agglomeration is probably more troublesome than induced nucleation.
0.2
k VI z
g
OA
2 I?
:a6 0
0.8
eo0
350
500
WAVE LENGTH
Figure 3.
450
400
I N M,U
Effeot of Tellurium Concentration on Visible Band of Purple Sols
Figure 4.
?n,wox c. lnp.p.m.
~ ~ i . ~e~eotroq ~ e d miomgrsphs at A. B.
The pmticle size of a sol is largely controlled by the rate a t which reduction proceeds. If the rate of reduction is great, many nuclei are formed before the growth process supersedes nucleation. If the rate of reduction is small, the frequency of nucleation will be correspondingly less. Consequently, the growth process o m become the primary one when fewer nuclei are in existence and develop them into particles relatively large in comparimn to those formed simultaneously with a myriad of other nuclei from greater supersaturation. This effect is illustrated by the series in Table I and Figure l. I n progressing from A to D, the particles become smaller in length and width as a result of increased rate of formation brought about by succeasively greater hypophosphite concentrations. The width of the psrticles decreases With respective increases in tellurium concentration and reaction rate, as shown in Table I and Figure 4.
Effeot of Tellurium Concentration on Particles from Purple Sols 2p.p.m. 4P P . ~ .
D.
14p.p.m.
The rate of precipitation is increased with increase of temperature. Also increase of temperature produces smaller sol particles and a shift of the minimum in the trsnsmittancy curve toward shorter wave lengths. To obtain the necessary control of this variable without resort to a thermostat, the precipitations are carried out a t the boilingpoint. The reducing agent is added rapidly by blowing it out of a pipet into the rapidly swirling tellurium mixture. The induotion period preceding nucleation is of sufficient length to permit complete mixing before nucleation is begun. Hence, nucleation and growth are assumed to take place in a uniform medium. precipitations are complete in 15 minutes after addition of reagent, and this period is the digestion time used in all experiments. After digestion, the sols are cooled in tap water for 15
ANALYTICAL CHEMISTRY
1016
nunutes, then transferred to volumetric flasks and diluted to volume. CHEMICAL EFFECTS IN SOL FORMATION
The satisfactory range for gum arabic concentration is 0.05 t o 0.5%. Stabilizing ower decreases rapidly as the concentration drops below 0.058 Within the useful ran e, stabilization. is best at 0.3 to 0.5%. Concentrations greater t i a n 0 . 5 7 give rise to excessive foaming which makes accurate dilution very difficult.
The choice of reagents in this study was made after consideration of certain theoretical requirements and the results of some preliminary experiments. More critical and extensive studies were then made on the reagents chosen. Doubtless other reagents could be used and deserve similar study. The results of the present investigation are probably general enough to permit inferences concernin the behavior of analogous reagents.
g
Reducing Agent. The reagent usually prescribed for reducing tellurite for photometric tellurium determination is *tannous chloride (2, 3, 16, 1 7 ) . This reagent has certain objectionable features: It nearly always contains colloidal stannic acid which may induce nucleation or agglomeration of tellurium particles. I t is difficult to prepare the reagent free of such material and, once prepared, there is a tendency for stannic acid to c-ontinue to form. Certain precautions must be taken with the .;tannous solutions to preserve their reducing capacity because the reagent is so easily hydrolyzed and so easily oxidized by oxygen in the air. Hydroxylamine and hydrazine salts have been used for tellurium sol formation, but are not suitable for the low te1luriu:ir concentrations of interest in spectrophotometry.
OF V I S I B L E B A N D . 0.05 Yo CUM ARABIC 0.5 70
0
1'
' 1
---0.05 0.5 M 300 0
I
HY POPHOSPH I T E II
I' I
0.2
I
0.4
Figure 6. Effect of Gum Arabic Concentration on Wave Length of Visible Band
Increase of gum arabic causes a small shift in the visible minimum transmission band to shorter wave lengths. The band i n t tie ultraviolet' is not significantly shifted by similar changes. T h ~ i effect is illustrated in Figure 6. The effect of increasing gum arabic on optical densities is qualitatively similar to that of increasing hypophosphite; it is much less pronounced, however. With respect to optical properties, hypophosphite and gum arabic appear to have additive effects-Le., no significant interaction. Acid. Tellurium(1V) solutions must be maintained acidic to prevent their slow hydrolysis to insoluble tellurous acid. Because of the ability of chloride to complex tellurium, hydrochloric acid is particularly effective as a stabilizer. The stock telluriuni solution used in this investigation was stabilized with 0.2 Y hydrochloric acid.
0.2 0.4 CONCENTRATION OF REAGENT CM)
Figure 5. Effect of Hypophosphite Concentration on Wave Length of Visible Band
A more suitable reagent is hypophosphorous acid, which is vomniercially available and may be used without further purification. I t is free of particulate matter and, unlike other reagents of similar reducing power, is not oxidized on standing in the air. .ipplication of hypophoephite in photometric tellurium determination has been reported by Southern (16). In order to test independently the effects of hypophosphite and hydrogen ion concentration, the former was added as a hypophosphite-hypophosphorous acid buffer (1 : 1 mole ratio). It was prepared from C.P. sodium hypophosphite monohydrate and ITSP purified hypophosphorous acid (30 or 50%). The practical lower limit of hypophosphite concentration during tellurite reduction is 0.05 M . Below this limit, the slo\vly formed particles are too large for good stability. Sol properties are much more dependent upon conditions of precipitation with resulting loss in p y x i o n . Increase of hypophosphite concentration increases t e rate of sol formation, correspondingly decreasing particle size and shifting the visible band toward shorter wave lengths. The overlapping of the visible band and ultraviolet band imposes an upper limit on the hypophosphite concentration at approximately 0.4M . The wave length shifts of the visible band with hypophosphite concentration are summarized in Figure 5.
Protective Colloid. The protective colloid usually specified for Starch, dextrin, and gelatin are inferior to gum arabic. Also for this purpose, Crossley (2) found gum tragacanth to be without) advantage. The gum arabic solution should be made fresh before uAe because it slowly becomes cloudy on standing. I t should be centrifuged before use. the tellurium hydrosol determination is gum arabic.
'Table 11. Analysis of Variance of Effects of Reagents on Spectral Characteristics of Visible Band 1 Concentrations during sol development: tellurium, 12.5 p.p.m.; hypophoiphite, pH 2 buffer (four levels), ,0.05, 0.1, 0.2, 0.5 1'6: gum arabic (thrw levels), 0.05, 0.2, 0.5%: sulfuric acid (two levels), 0.01, 0.2 N. Data analyzed: wave lengths, in mp. of optical density maxima] Sum of Degrees of Mean Reagents Squares Freedom Square F 99,534 3 3 3 , 1 7 8 691" Hypophosphite 10,914 2 5 , 1 8 3 10Sa Gum arabic 1,734 36" 1,734 1 .4cid Hypophosphite plus gum arabic 1,428 6 238 3 Ob Hypophosphite plus acid 424 3 108 2 3 G u m arabic plus acid 244 2 122 2 6 Hypophosphite plus gum arabio plus acid (error) 287 6 48 Totals 114,565 23 * High1 significant, 99% level. b Signi&ant, 9570 level.
In the low tellurium concentrations in which analytical sols are formed, the tendency toward hydrolysis is not so great, and lower acidities are permissible. Satisfactory sols are formed a t p H 2 ; with increase of p H above this value, the sol development is slow and incomplete. Increase of acid concentration above 0.2 A' brings an increase in agglomeration and should be avoided. Hydrogen ions promote the reduction of tellurite. Accordingly, an increase of acidity brings about smaller particle size and a hypsochromic shift of the visible band. Tellurium. Also important in quantitative applications is the effect of tellurium concentration on the particles and spectra. The effects are illustrated in Figures 3 and 4. Increase of tellurium concentration yields an increase in particle length. Cor-
V O L U M E 2 5 , NO. 7, J U L Y 1 9 5 3 respondinglg, a bathochromic shift in the visible band is brought about. Also with increasing tellurium concentration, the band becomes sharper and the maximum optical density per unit concentration becomes greater. The spectral effects resulting from variation in tellurium concentration are discussed in greater detail ( 7 ) . STATISTICAL SU’\IMARY O F REAGENT EFFECTS
For the study of reductmt, gum arabic, and acid effects, H t’tictorial design was used. -4n analysis of variance derived therefrom for the visible band is given in Table I1 for wave length characteristics. -411 three variables show a highly significant effect,, with the reductant decidedly showing the greatest deet tmd acid the least effect. Only one significant interaction is observed-namely, reagent-gum arabic. Because first-order interactions are so small, it is assumed that the second-order interact,ion is a valid estimate of error n.ithin replications. ACKNOWLEDG.MENT
The author wishes to acknowledge the assistance of Albert E. \-nt,ter, Jr., who prepared the electron micrographs and made measurenient.~on them.
1017 LITERATURE CITED
(1) .Auerhach, R., Kulloid-Z., 38, 343 (1926), ( 2 ) C‘rossley, P. B., Analyst, 69, 206 (1944).
(3) DeAIeio, R. H., . k . i L . C H E Y . , 20, 488 (1948). (4) Gans, R., Ann. Physik, 62, 331 (1920). (5) Gribnau, B., Kolloid-Z., 77, 289 (1936); 82, 15 (1938). (6) Gutbier. A.. 2. unoru. Chem.. 32. 91 11902): 42. 174 (1904). , , (7j Johnson’, R: A , , Kwan, F. P:,and Westlake, D.’, ANAL. CHEM.. 25, l O l i (1953). (8) Krishnan. R.S.,Kolloid-Z., 84, 2 (1938). (9) Krishnan, R. S., Proc. I n d i a n Acad. Sci., A5, 407 (193i). (10) Kruyt, H. R.. “Colloid Science,” Vol. I. D. 34. .InistcrcIam. Elsevier Press, 1952. (11) LaRler, V. K., and Dinegar, R. H., J . Am. Cheni. Soc., 72, 4847 (1950). (12) Mie, G., Ann. Physik,25, 377 (1908). (13) Ostwald, Wo., Kolloidchem. Beih., 2, 409 (1911). (14) Rayleigh, Phil.J f u g . , 47, 375 (1899). (15) So:thern, H. K., in Report BR-606, May 31, 1945; cited in Analytical Chemistry of the Manhattan Project,” S e w Tork. AIcGraw-Hill Book Co., 1950. (16) Volkov, S. T., Zacodskaya Lab., 5, 1429 (1936). (17) Zemel, V. S., Ibid., 1433 (1936). RECEIVED for review January 9, 1953. hrcepted .4pril 18, 1953. Prpiented before the Division of Colloid Chemistry at the 123rd Xeeting of t h e A M E M C A Y CHEMIC.AL SOCIETY, Los Angeles. Calif.
Spectrophotometric Determination of Tellurium as Hydrosol IZiLl’II i.JOHNSON, FIZiNCIS P. ICWiN’, AND DON WESTLAKE Department of C h e m i s t r y , C7niz.ersitj of Illinois, Urbana, I l l .
‘l’elluriuni(I\) m a ) he deterniined spectrophotometrically i n trace concentrations upon formation of t h e colored hjdrosol b) reduction to TeO. Sols so produced may exhibit a variet) of colors, according to t h e conditions of their formation. T h e color variation arises because t h e absorption band i n the tisible is nonspecific!; i n particular, i t is d u e to conservative absorption, simple scattering, which is dependent upon particle dimensions a n d geometry. Choices of conditions of sol formation a n d \*ate length for spectrophotometric measurement are iniplicitlj related. Because of variations i n t h e spectra with changing tellurium concentration, special consideration is demanded for t h e selection of wate length for Beer’s law measurements. T h e sols are formed b? reduction with hypophosphite a n d stabilized with g u m arabic. By varying t h e a m o u n t of hypophosphite, prototype sols of blue, purple, and red areobtained. Red sols prove superior i n linearity of Beer’s law relationship, reproducibilitj, and stabilit) against agglomeration. For t h e blue sols there are marked deviationb from Beer’s law, a n d t h e results are more susceptible t o small variations i n conditions of formation. Purple sols are intermediate i n behavior. Blue or purple sols may be indicated for spectrophotometrj i n t h e presence of interferences absorbing i n the blue.
T”
LLURIL-31 may be determined spertrophotometrically as the highly colored, soluble iodotellurite complex ( 5 )or as the rlementary tellurium hydrosol. The latter is formed by reduction of tellurite and the method has been applied to tellurium in i
1 Piesent addresa, Department of Biocheniiitr\ l l r d i c a l School Chicago 11, TI1
Sorthwwtern Tnivei,it>
ores (10, 1 2 ). in steels ( b ) , in industrial atmospheres ( 3 , 6, 9 ) , and in biological materials (3, 6). The sol methods have also been described by Shakhov ( 7 ) and Southern (8). The optical properties of tellurium sols vary greatly according to the size anti geometry of the sol particles; thus they are dependent upon the conditions of formation of the hydrosol. The conditions for forming a variety of tellurium hydrosols have been presented and the relationships between particle characteristics and optical properties discussed ( 4 ) . It is the purpose of the prePent investigation to produce sols of different spectral t1q)es ant! to study the suitability of each type for spectrophotomrtry, with particular reference to the Beer’s law relationship, stability, and interferences. The reducing agent used is hypophosphorous arid and the protective colloid is gum arabic. FORMATION O F HYDROSOLS
T o simplify and systematize the study, conditions are chosen to form prototypes with the following colors: blue, purple, and red. Amber sols may also be produced, but they are not suitaljle for quantitative work because their visible and ultraviolet ban(l. are not separate and exhibit anomalous optical density behavior. A s many conditions as possible arc’ held constant for the production of all sols: Preliminary dilution is made so that the volume in which the The reaction mixture containing all reagents but the reducing agent is heated to the boiling point. The reagent is then added rapidly from a pipet with vigorous stirring and the mixture is allowed to digest for 15 minutes at or very near the boiling point, then cooled in a water bath at room temperature for 15 minutes before being diluted to final volume (50 ml.). In all cases, the reaction mixture is made to contain 0,3y0gum arabic. This amount achieves a high degree of stability of the sol, but avoids excessive foam production found when larger nmounts of gum arabic are used. sol develops is always 40 ml.
