ANALYTICAL CHEMISTRY
1358 shown that the apparent reduction potential corresponds to the decomposition potential, ED, on the polarographic wave of a compound. Table I11 shows that a t a p H of 7 the apparent oxidation potential is close to ED. At a p H of 5 the agreement is not so good, but the apparent oxidation potential corresponds to a potential on the v o l t a m e t r i c aave. ANALYTICAL kPPLICATION S
Qualitative. The half-wave potential of a phenol at a known pH value can be used for qualitative identification. In cases where more than one phenol have the same half-wave potential, further characterization ~ o u l dbe necessary. Qualitative identification of several phenols in a mixture would be possible only if the individual half-wave potentials were a t least 0.2 volt apart. A wave observed to cover a potential range of more than 0.2 volt would indicate a mixture of phenols whose half-wave potentials R ere too close to yield separate waves. Quantitative. In cases where it is necessary t o know the approximate concentration of a phenol, voltammetry would find application. This might be the case with drinking Jvater or industrial waste products, where concentrations of phenols must be held below some maximum value. While the accuracy of this method is only within &14%, it offers a great advantage in its speed and simplicity. -4total of 6 minutee is required to run a sample. Samples containing as little as 0.1 p.p.m. of phenol could be run directly, while more dilute samples could be run if first subjected to processes designed to concentrate the phenol. Antioxidants. Egloff and coworkers (11) have found that effective inhibitors have critical oxidation potentials within the region of 0.6 to 0.9 volt. The half-aave potential of a new
phenol could be determined in much less time than the critical oxidation potential and would show whether more detailed testing for use as an antioxidant was warranted. LITERATURE CITED
Bowman, Ji. H., Ph.D. theais, Vniversitg of Pittsburgh, 1950. Campbell, T., and Coppinger, G., J . ii?n. CRem. SOC.,74, 1469 (1952). Conant, J. R.,and Pratt, AI. F., C‘hem. Recs., 3, 1 (1926). Cosgrove, S. L., and Waters, W. A , , J. Chem. Soc., 1949, 3189. Fieser, L. F., J . A m . C’henr. Soc., 52, 5204 (1930). Fletcher, W. E., Ihid., 68, 2726 (1946). Gibbs, H. D., Philipp. .I. Sci., 4, 133 (1910). Hedenburg, J. F., M.S. t,hesis,t-nirersity of Pitkburgh, 1952. Julian, D. B., and Ruby, W. R., J . A m . Ckeni. Soc., 72, 4719 (1950). Laitinen, H. A , , s n d Rolthoff, I. 3f., J . Phus. Chnn., 40, 1061 11941). Loury, C. D., Jr., Egloff, G., hlorreli, J. C., and Dryer, C. G., Ind. Eng. Chem., 25, 804 (1933). Muller, 0. H., a n d Raumberger, J. P., ,J. A m . Chein. SOC.,61,590 (1939). Rogers, L. B., and h i d , S 8 , Ji., Pittsburgh Conference on Analytical Chemistry and Applied Spectroscopy, March 5, 1952. Scudder, H., “Electricvtl Conductimty and Ionization Constants of Organic C‘omponrrrls,” p 246, S e \ \ York, D Van Nostrand Co.. 1914.
Shaw, J. A., private coniinunication. Taube, H., J . Am. Chem. Soc., 63, 2453 (1941) for review June 17. 1952. -4ccepted June 26, 1953. Presented at the Pittsburgh Conference on Analytical Chemistry aud Applied Spectroscopy, Pittsburgh, Pa., March 6. 1952. Contribution 862 from the Department of Chemistry, University of Pittfiburgh. Abstracted from the thesis of John F. Hedenburg presented to the Graduate School of the University of Pittsburgh as partial fulfillinent of the requirements for the M.S. degree. RECEIVED
Separation of Phenols by Partition Chromatography T. R . SREENEY AND J. D. HULTRIAN Naval Reseurch Laboratory, Wunhington 25, D. C;
T
H E so-called tar acids coiist,itute B mitjor and, to the woodpreserving industry, a possibly important fraction of high temperature coal tar creosote. In considering the separation and identification of the const,ituents of a certain fraction of t8hetar acids, one that consisted almost entirely of n misture of phenols, the technique of partition chromatography appeared to be applicable. The method of partition chromatography, originally developed by Martin and Syngr ( 2 ) for the separation of t.hc acetyl derivatives of amino acids Tvhich were obtained from protein hydroIj-zates, consists in partitioning a solutr, between two inimisc.it)le solvents, one of which is immoililized by adaorhiiig it on an iiiert solid. In practice, t,he inert solid holding the immobile solvent is packed in a suitable chromatographic tube and, after the solute is placed on the column, the mobile solvent, is percolated through the column, whereupon t,he solut,eis partitioned between the two solvents according t o its partit,ion coefficient.. Under proper conditions the solute will pass through the column in a small zone and eventually emerge from the bottom. When the solute consists of a mixture of compounds, their respective ratcs of passage through the column are a function of their partition coefficients as well as the relative amounts of mobile and immobile phases and hence a separation may often be achieved. For the present invedtigation the use of a system such as thr water (on silicic acid)-cyclohexane system described by Zahner and Swann (6) for the separa.tion of phenol from petroleum cresylic acid seemed particularly applicable, since cresylic acid is a
trade nAmP u v d to tlrqlgnstr a C I ude mixture consisting largely of phenolic ( onipourids but contaminated with neutral oils, sulfur conipounds, and nitrogen compounds. Furthermore, these workers indicated that their system could also be used for the separation of homologs of phenol. Actually, becauw the solubilities of the materials studied appeared to be about the same in iso-octane as in cyclohexane and because pure iso-octane was more readily available, the present study x-as made on the system water (on silicic acid)-iso-octane. In order to employ this system for the separation and identification of phenols in unknown mixtures it was necessary t o study first the behavior of a number of known representative phenols. This paper reports the effect of changes in the r a t e r content of the silicic acid support on the rate of movement of phenolic compounds thiough the column and on their rerolution. EXPERIBZENTA L
The chromatograms in this study were obtained using 12 grams of silicic acid as the support, varying quantities of water as the immobile phase, and isc-octane as the mobile phase. The sample solutions of the individual compounds used contained 1 micromole of solute per ml. (solvent ieo-octane); solutions of mixtures of compounds contained 1 micromole of each per ml. Materials. Phenol, Merck reagent grade. 1-Naphthol, 2-naphthol, p-phenylphenol, o-phenylphenol, and p-tert-butylphenol were recrystallized from water; melting point,s agree with literature. 2,6-Xyle~ol was purified by steam distilling twice: melting point 141.5-143’ C.
, NO. 9, S E P T E M B E R 1 9 5 3
__-r _ ..-.eseareh program on the eheinieal nature of creosote as it is related to the protective action against destruction of wooden piling by marine horers, it was deemed important to develop a method for the resolution of phenolic mixtures obtained from tar acid fractions of creosote, since the latter may be of importance in the process of preservation. The technique of partition chromatographyemploying the system water (on silicic acid)-iso-ootane Eoupled with the spectrophotometric examination of the effluent is shown to he effeetivefor the resolution of certain mixtures of phenols. The proportion of water to silicic acid in the system is critical with respect to the rate of movement of the different phenols through the column and to the diffusiveness of their respective zones in the column. By running more than one chromatogram with different pmportions of water, the resolution obtainable with a given mixture may he enhanced. The size and position of the ring substituents have a m a r k e d effect on the resolution obtainable. A phenolic extract of meosote was chromatographed to yield nine distinct fractions.
1359
tight rubber sleeves. Twelve grams of silicic acid were weighed out and transferred without delay to a mortar in the dry box. The desired quantity of water wan added drnnwise bv mema of a buret and the mixture was eronnd v addition of several drops. Grinding by means-of a mechanical
de& &uear&e.
~~~
~~~
wm transferred oracticallv cluantitati;eiv to
.-..~ ~~
~
~~
~
~~~
~
~~~~
~
into the tube. During this pIocess enough iso&tane wan add& from time t o time to maintain a supernatant layer. The rod was h d l y washed down vith iso-octane before removal.
3,5-Xylenol and 4-methyl-2-terGbuty I pure from the National Research Council 2,6-Di-t~tbutyl-4-methylphenolwas I hol; melting point 69-69.5' C. c-Ethylphenol (Eastman Kodak Co. p. yl.\l Y., .I hefore use. 0-Cresol, Fisher C.P. Thymol (4methyl-2-isopropyIphenol), J. T. Baker, U.S.P. Silicic acid, Mallinckrodt analytical reagent, lOC-mesh, "specially prepared for chromatographic analysis by the method of Rsmsey and Patterson." Iso-octane (2,2,4tnmethylpentane), Phillips Petroleum Co. uure grade. This was sufficientlvtransoarent a t 230 mu without burifi&tion. The iso-octmc fro& the experiments was recovered by shaking several times with 10% sodium hydroxide solution and twice with distilled water. It was then dried over anhydrous calcium chloride and 6 n d v distilled over metallic sodium. The iso-oct,ane so rccovered u.& satisfactory for the chramato&~hie experiments.
.._"
Although in practice i t is usus1 to saturate the mobile phase with respect to tho immobile, this prooeedure was avoided in the present work. Becausc thc cffcct of changes in the water content of the silicic acid support was being studied and because, in many cases, the support was far from saturated with water and very hygroscopic, i t was felt that to pass a water-saturated iso-octane solution through the column would alter the water content of the support to an unknown degree and hence endangrr control of the experiment.
Apparatus. Spectrophotometric measurements were made in square 1.0-cm. quartz cells with a Beckman Model D spectrophotometer modified so as t o be usable in the ultraviolet region of the spectrum. The chromatographic apparatus (Figure 1) consisted of a glass column, 1.9 em. in inside diameter and 20 em. long. I n the bob tom waa sealed a 10/22 female joint into which fitted a sinteredglass disk sealed t o the end of B male joint; the latter wa8 sealed onto the end of a small piece of glass tubing (12 X 0.8 em.). On the top of the oolumn was sealed a female 29/42 joint in which was placed a Y-tube. One end of the Y-tube was connected by a. 24/40 joint t o a solvent reservoir and the other, by a similar joint, to a pressure-equalizing tube leading to a T-tube connected to the top of the solvent reservoir through another Y-tube. One opening of the latter Y-tube was used to fill the reservoir. The T-tube was connected throueh a needle valve oresmre r e d a t o r
Figure 1. Chromatographic Apparatus
quired. A h e nickel &en, which fitted snugly into the tube, w m then fitted carefully to the top of the adsorbent and pressure was again applied. If the adsorbent sank below the screen, the latter was again adjusted. Enough pressure was then applied to cause the solvent to emerge from the column at the rate of about 1 ml. per minute, and the solvent level wyas allowed to fall t o the level of the screen, a t which time the column was ready to receive the sample. A t no time was the solvent allowed to fall below the top of the packing. During the preparation of the packing a total of 60 to 75 ml. of solvent will have been collected. The mole per ml. bas added carefully to the top of the packing and, at the same time, the first of s series of calibrated test tubes WBB -D
vkbt was allowed to flow ontd the pGking; as'this dimppdeared an additional 2-ml. portion of solvent was used t o fix the sample, after which the pressure and flow from the reservoir were regulated to give a condant rate of effluent. Absarbmeies %,erethen obtained on each fraction a t an appropriate wave length.
If operating conditions are maintained relatively constant and the s a n e batch of silicic acid (stored in a dry box) is used, the
A N A L Y T I C A L CHEMISTRY
1360 position of the peak tube will not vary by more than 1 in the higher tubes and will usually remain constant in the tubes below about 15. Changes in temperature, which were not controlled in these experiments, probably account for slight peak position variations in the higher numbered tubes when larger total volumes were collected. DISCUSSION
The phenolics chosen for this study are representative of some of the types found in coal tar-namely, phenol, alkylphenols, phenylphenols, and naphthols. Isomers were chosen with the view of studying the effect of the position and size of the substituent group on the behavior of the phenolic in the column.
I-WAPHTHOL
P *3,5-XYLEWOL PIL.p-PHENYLPHENOL
V, VI, and VI1 as a group; other such group separations are apparent a t other water levels. There are certain concentrations of water below which some phenols will, for all practical purposes, remain on the column permanently-for example, below about 5.5 ml. of water for VI, VII, and VI11 and below about 5 ml. of water for all except I, 11,and 111.
It has generally been confirmed that the movement of zones through a column is sensitive to small changes in composition of the developer solution and many workers, taking cognizance of this fact, have extended the poqsible resolution of certain mixtures by adding a small quantity of polar solvent to the developer a t certain stages in the development of the chromatogram. In employing this method the finding of suitable solvent combinations is sometimes a tedious task. The fact that the water content of the silicic acid column is so critical in the water-iso-octane system for the separation of phenols indicates that as an alternative to changing the mobile phase during a chromatogram of an unknonm mixture it may be advantageous to run a few short chromatograms using different quantities of water on the column. In this way one group of compounds in a mixture may be made to come through the column relatively unresolved in the first several fractions collected, while another group in the same chromatogram may be resolved in the subsequent fractions. Then, by decreasing the watex content of the column and running another chromatogram of the same mixture, the previously unresolved compounds may be resolved while the previously resolved compound may now remab on the column. 14
d L w & I
I3 1.2
1
2
3
5
4 NL
OF
6
.'
1
+
u-
--.I
1
12 g.SlLlClC ACID 6 5 Y L WATER
E
7
WATER
Figure 2. Change in Position of Peak Tube with Variation in Volume of Water Added to Silicic Acid
The effect produced on the chromatograms of several compounds by varying the water content of the column, while maintaining the quantity of support constant, is shown graphically in Figure 2, where the position of the peak tube (tube of greatest concentration) is plotted against the water content of the column. Because the position of the peak tube is a measure of the rate of movement of the compound through the column, these curves show the change in the rate of the movement with a change in the water added to the column. As required by the theory of partition chromatography (Z), the movement of a band through the column increases in velocity as the water content of the column is increased. The slopes of the curves increase sharply as the water content of the column is decreased, although the rate of change of the slopes is less for those phenols having lower water solubility (smaller partition coefficient) because of the greater blocking of the hydroxyl function by substituents in the ortho position. It can be seen from Figure 2 that no one water concentration is optimum for the separation of all these phenols; separations which can be effected a t one water concentration cannot be effected at another.
For example, using 7 ml. of water, phenol can be easily separated from the other homologs studied. However, using, say, 5.5 ml. of water, it is seen that phenol could not be separated from 2-naphthol and p-phenylphenol, although good separations could still be obtained from the other homologs. Compounds I, 11,and 111(Figure 2) can easily be separated from each other below about 5 ml. of water but not above this amount; V and VI can be s e p arated at 5.5 or 6 ml. of water but not a t 7 ml., and the same holds true, for example, for the pairs V, VI1 and IV, VI. At 7 ml. of water compounds I, 11, and I11 can be separated as a group from
0
Figure 3.
5
10
I5
25 30 T U B E NO
20
35
40
45
50
Chromatogram of Mixture of Phenols
To illustrate the separation of phenols obtainable in this way the following mixture was chromatographed: phenol, 3,5-xylenol, 2,6-xylenol, 1-naphthol, o-phenylphenol, p-tert-butylphenol, and 4methyl-2,6-di-tert-butylphenol. When the immobile phase consisted of 6.5 ml. of water, only incomplete separation of 2,6di-tert-butyl-4-rnethylpheno1, 2,6-xylenol, and o-phenylphenol was obtained but good separations of the other compounds with the exception of p-tert-butylphenol (Figure 3). When the mixture was chromatographed using 4.5 ml. of water (Figure 4), excellent separation of the same compounds was obtained, which were inseparable using 6.5 ml. of water. I n this way all the compounds were separated by means of two short chromatograms without changing the mobile solvent. Extending the second chromatogram by collecting a greater number of fractions in order to separate a larger number of compounds is impractical, using only iso-octane as the mobile solvent, because of the increasing widths of the peaks obtained in the higher numbered fractions.
V O L U M E 25, NO. 9, S E P T E M B E R 1 9 5 3 The change in the diffusiveness of the zone of the compound in the column (the shape of the peak in the effluent diagram) with a change in the water content of the column is shown for the case of o-phenylphenol in Figure 5 and for the case of 1naphthol in Figure 6, where the peak tubes are assigned a value of zero, the tubes approaching the peak’s negative numbers, and the tubes follon~ing the peak’s positive numbers. The diagrams for these compounds illustrate the general o5 trend: As the water content of the column is increased, the peaks become sharper and the spread becomes correspondingly less. In separating mixtures of unknown composition and concentration the effect of concen0 5 IO 15 20 2 5 tration on the position of the r u e € NO. peak tube may be important and may be different for each Figure 4* Chromatogram of Mixture of c o m p o u n d . The effect of Phenols changes in concentration on the partition coefficient has been discussed by Craig and Craig (1) and is probably small for dilute solutions. The effect of concentration on the position of the peak in the effluent diagram was not studied extensively. One compound, 2-naphthol, was run a t three different concentrations using 12 grams of silicic acid and 6.5 ml. of water. The results, shown
1361 in Table I, indicate the the peak did not vary appreciably with small changes in concentration but that the increased spread at higher concentrations may interfere with good resolution. Of the compounds studied, the ortho isomers, in every case, moved through the column with greater velocity than did the meta or para and, consequently, the former will have the smallest R value (2) a t any given water content of the column. As the water content of the column is decreased it can be seen from Figure 2 that, for any given compound or isomer, the R value will increase,
1-NAPHTHOL 12 g S I L I C I C A G I O 1 = 7 N L WATER
1.3
- 3 - 2 - 1
0
I
2
3
4
5
6
7
T U B E NO. Figure 6. Effect of Variation in Volume of Water Added to Silicic Acid on Height and Spread of Peak in Effluent Diagram
- 2 - 1
0
I 2 T U B E NO.
3
4
5
Figure 5. Effect of Variation in Volume of Water Added to Silicic Acid on Height and Spread of Peak in Effluent Diagram
Table I.
-0
Effect of Concentration of 2-Naphthol on Peak Position
(12 grams of silicic acid, 6.5 ml. of water) Concentration, Mg. per 3 M1. Peak 1.5 14 0.5
0.1
16
17
The order in which a group of compounds emerge from a partition column follows, in general, their partition coefficients. Phenols with a substituent in the ortho position have a distinctly higher partition coefficient in water-cyclohexane systems than do their meta or para isomers (4, because the proximity of the ortho substituent to the hydroxyl group hinders ( 3 ) ,or masks, the function of the latter. This property is reflected in the 1.0 relative ease of separation of ortho substituted phenols 08 from their meta and para 08 isomers as shown in the c a s e s of t h e i s o m e r i c 07 xylenols, p h e n y 1 p h e n o Is, 06 and naphthols (I-naphthol u < r may be considered to be m o r e h i n d e r e d t h a n 2os u3 naphthol) in Figure 2. Be* 04 cause their partition coeffi03 cients lie so close together, the separation of meta from 02 para substituted isomeric 01 phenols is much more diffia0 cult. This effect is shown 0 5 10 I5 20 25 30 in Figure 7 for a mixture of T U B E NO the three isomeric cresols. Figure 7 . Chromatogram The mixture was chromatoof Isomeric Cresols graphed using 6.5 ml. of
Spread 14 11 7
ANALYTICAL CHEMISTRY
1362 water to yield a clean separation of the ortho isomer but only one peak for the combined meta and para isomers. That partial separation of the meta and para isomers was obtained was determined by examining spectrophotometrically a tube on each side of the peak-namely, tubes 16 and 21. The increase in resolution obtainable by increasing the length of the column or by employing other mobile solvents was not investigated.
1.2 I .I
1.0 0.9
2.2
0.8
2.0
0.7
1.8
0.6
1.6
0.5
14
04
0.3 0.2
0.1
0 0.7 0.6 0
5
10
15
20 25 TUBE NO
30
35
40
Figure 8. Chromatogram of Mixture of Phenols with Ortho Substituents of Increasing Bulk
The pronounced effect of the size of an ortho substituent on the resolution of a mixture of phenols and the effect of a small change in the water content of the partitioning column were demonstrated by chromatographing a mixture of phenols with ortho substituents of increasing bulk-namely, o-cresol, o-ethylphenol, 2-isopropyl-4-methylphenol (thymol), o-phenylphenol,
12 g. SILICIC ACID 5.8 ML. WATER
0.5
> 0
0A 0.3
2
a rn
0.2
0
5",
OJ
I\ L P -
T U B E NO
Figure 9. Chromatogram of Mixture of Phenols with Ortho Substituents of Increasing Bulk
and 2-tertbutyl-4-methylphenol. The use of 6.5 ml. of water effected the separation of o-cresol only (Figure 8), whereas the use of 4.5 ml. gave almost complete separation of the five compounds (Figure 9). Another example of the type of changes that occur in the effluent diagram with changes in water content of the column and how this may be used in analyzing an unknown mixture is shown in Figure 10. These curves were obtained from a certain phenolic fraction of creosote which, unfortunately, appears to be of such complexity that it would undoubtedly be more amenable t o
TUBE NO. Figure 10. Chromatograms of Phenolic Fraction of Creosote Using columns of silicic acid with various quantities of water added
V O L U M E 2 5 , N O . 9, S E P T E M B E R 1 9 5 3 chromatographic analysis after preliminary fractionation by other techniques. However, it serves well the purpose of illustration. Chromatographing an iso-octane solution of this fraction through a column prepared with 7.25 ml. of water yielded four distinct zones (Figure 10, A ) . When a column prepared with 5.8 ml. of water was used (Figure 10, B ) , it was found that the zones with peaks a t tubes 17, 26, and 29 in Figure 10, A , have been pushed out of the picture (peak tube 17, Fipurp 10, A , is altogether different spectrophotometrically from peak tube 23, Figure 10, B ) and the zone of tubes 4 to 7 has nox been separated into four new zones. The identity of the zones was determined by running their ultraviolet spectrum. When the column w a s prepared with only 2 ml. of water (Figure 10, C), it was found that the peak a t tube 12 (Figure 10, B ) had shifted to tube 48 and that their spectra were practically identical. Furthermore, the peak a t tube 4 (Figure 10, B ) had shifted only to tube 5 , while the peaks at tubes 8 and 23 did not materialize. When the sample was chromatographed with no m-ater added to the silicic acid (Figure 10, D),it was found that the band with the peak of tube 5 (Figure 10, C ) was separated into three distinct bands which represented about 73% of the material in the original band. An additive spectral curve obtained from the three zones (Figure 10, D ) was
1363
very similar but not identical with that of the m:ttr.ri:il t,ube 5 (Figure 10, C).
ill
peak
From the separations discussed it is seen that the sample has been srparated into nine distinct parts in thc, c~ffluc~nt rractiona. ACUNOWLEDGMENT
The authors wish to thank Harold J. E. Segrave for his assistance and interest in this work and (‘arl J. Wessel, National Reseal c:h C‘ounoil,for furnishing sonic of the phenols. LITERATURE CITED (1) Craig, L. C . , a n d Craig, D., “Technique of Organic Chemistry,” Vol. 111,p. 171, New York, Interscience Publishers, 19.50. (2) Martin, A. J. P., a n d Synge, R. I.. hl., Biochem. J., 35, 1358 (1941). (3) Stillson, G. H., Sawyer, D. IT.,arid H u n t , c‘. K., J . A m . Chem. Soc., 67, 303 (1945). (4) Koolfolk, E. O., et al., Bur. Mines, Bull. 487, 7 (1950). (5) Zahncr, R. J., and Swann, W. B., As.tr.. CHEN.,23, 1093 (1951). Aprils, 1953. Acccptcd
.I\iiie %,
1853.
Colorimetric Determination of Molybdenum with Mercaptoacetic AcidFRITZ WILL 1111, AND JOIlN EX. YOE Pratt Trace .inulysis Lmboratory, D e p a r t m e n t os C h e m i s t r y , Cinirersity of Virginia, CharlottesvilIe, Va.
The purpose of this investigation was to study the reactivity of mercaptoacetic acid with inorganic ions, in particular the molybdate ion, and to develop a colorimetric method for determining molybdenum, especially in materials containing large amounts of iron. Procedures have been developed for the colorimetric determination of molybdenum in the presence or absence of iron and also for the determination of both molybdenum and iron in the same aliquot. The method is simple and accurate and should be useful for the determination of molybdenum in steels, molybdates, and other materials, especially those containing large amounts of iron.
S
POT-PLATE studies showed that acetylmercaptoacetic acid gives a yellow color with molybdenum in acid solution; iron praduces a blue color in acid solution but the color fades immediately to colorless. In alkaline solution the iron complex is purplish-red. These color reactions were found to be revereible both in acid and in alkaline media. Mercaptoacetic acid (HS.CHI.COOH, also called thioglycolic acid, thioethanoic acid, and thiolacetic acid) gives color reactions similar to those of the acetyl derivative. However, since mercaptoacetic acid has a sensitivity of about 0.1 p.p.m. of molybdenum and is readily available in the pure state, it was preferable to investigate it rather than its acetyl derivative, Because the molybdenum complex is yellow and the iron complex colorless in acid solution, it appeared that mercaptoacetic acid might be used for the determination of molybdenum, even in the presence of iron. In fact, investigation has shown that both molybdenum and iron may be determined, the molybdenum being determined in an acid solution, where the iron complex is colorless, and the iron in an alkaline solution, where the molybdenum complex is colorless. Lyons ( 3 ) and Swank and Mellon ( 7 ) reported mercaptoacetic acid as a colorimetric reagent for total iron; the latter mentioned that molybdenum gives a yellow or orange color a t high con-
’ Present address, dluminum Co. of America, New Kensington, Pa.
centrations. Hamence ( 1 ) showed that mercaptoacetic acid could be used for the detection of molybdenum by the production of a yellow eolor in acid solution. Richter (6) and Xleyer ( 4 ) reported the use of the reagent for the determination of molybdenum but also used hydrazine sulfate in the process. Miller and Lowe (6) employed mercaptoacetic acid, but only as a reducing agent for molybdenum(V1) in the classical thiocyanate determination of the latter. In the investigation, detailed studies were made to determine the effect of a number of variables on the molybdenum-mercaptoacetic colored complex and to develop procedures for the colorimetric determination of molybdenum, in either the presence or absence of iron, and for the determination of both molybdenum and iron in the same aliquot. APPARATUS .4NI) REAGENTS
Instruments. Absorbancy measurements were made with a Beckman spectrophotometer, Model DU, using 1.00-cm. cells. All pH measurements were made with a Beckman glass electrode pH meter, Model G. Visual color comparisons were made in 50-ml. Nessler cylinders (tall-form). Reagent Solutions. Mercaptoacetic acid is a water-soluble, colorless liquid. A 5% solution (volume) of mercaptoacetic acid, neutralized with ammonium hydroxide, was used for most of the work. If impurities are present in the reagent, a slightly turbid