Extraction - Analytical Chemistry (ACS Publications)

---

66

ANALYTICAL CHEMISTRY

(60) St. John, C. V., Fleck D., and Tepe, J., AXAL.C H E ~ I 23, . , 1289 (1951). (61) Samuelson, O., Svensk Keni. Tids., 62, 221 (1950) (62) Samuelson, O., and Gaitner, F., Acta Chem. Srand.. 5 , 596 (1951). (63) Schubert, J., AN.~L.CHEM.,22, 1359 (1950). (64) Schubert, J . , J . Am. Chem. Soc., 73, 4488 (1950). (65) Schubert, J., and Lindenbaum, A,, iYature, 166,913 (1950). (66) Schuler, R. H., Boyd, A. C., and Kay, D. J., J . Chem. Education, 28, 192 (1951). (67) Shimizu, H., Chem. High Polymers ( J a p a n ) , 7, 108 (1950). (68) Sillen, L. G., Arkiv K e m i , 2, 477 (1950). (69) Ibid., p. 499. (70) Sillen, L. G., A-ature, 166, 722 (1950). (71) Sinsheimer, R. L., and Koerner, J. F., Science, 114, 42 (1951). (72) Stein, W. H., and Moore, S., Sci. American, 184, 35 (1951).

(73) (74) (75) (76)

Spedding, F., and Dye, J., J . Am. Chem. Soc., 72, 5350 (1950). Teicher, H., and Gordon, L., ANAL.CHEM.,23, 930 (1951). Tompkins, E. R , Ibid., 22, 1352 (1950). Totter, J. R., Volkin, E., and Carter, C., J . Am. Chem. Soc., 73, 1521 (1951).

(77) Van der Heijde, H., and Aten, A. H. IT.,J . Phys. and Colloid Chem., 55,750 (1951). (78) Volkin, E., and Carter, C. E., J. Am Cheni. Sac., 73, 1519 (1951). (79) Kickbold. R.. Anaew. Chem.. 62. 448 (1950). (80j Wickbold, R., Z. anal. Chem., 132, 2 4 i (1951). (81) Ihid., p. 321. (82) Ibnd., p. 401. (83) Wilson, J. B., J . Assoc. O ~ CA .g r . Chemists, 33, 995 (1950). (84) Yoshino, Y., and Kojima, RI. Bull. Chem. Soc. J a p a n . 47, 7 (1950). RECEIVED Xovember 12. 1951.

EXTRACTION LYMAN C. CRAIG T h e Rockefeller Institute f o r Medical Research, iVew York 21, AW. F.

T

HE increased number of reviews dealing with extraction

which have appeared during the past year are indicative of the growth of interest in the subject (22, 58, 47,62,61, 85, 92). I n spite of the greatly increased interest in analytical extraction as a laboratory tool, the contributions on the subject for the past year have not brought forth startling results. They have rather indicated only a steady improvement in apparatus and the general understanding of the subject as a whole. Each year there is a steady improvement in the quality and availability of the solvents required. On t h e other hand, there has not been a general improvement in the understanding of the factors required in order t o choose or compound a highly selective system for any given separation. Unfortunately, with our present understanding, the choosing of a system for difficult separations still requires much experimentation and effort. As literally hundreds of solvent combinations are possible, weeks of careful experimentation may be required t o find a system giving sufficient selectivity for the purpose. The problem would now appear t o overshadow by far any other which concerns extraction. I n last year's review ( 2 2 ) the most important factors t o be considered in choosing a satisfactory system were discussed. These still hold for the nen er apparatus discussed in this revieir, eycept possibll- for the completely automatic countercurrent distribution machine ( 2 7 ) which is capable of applying thousands of transfers t o a given separation. Here a high 0 value ie not so important as a spstem adhering closely t o the law of Kernst, one in which the phases separate quickly, and certain other aspects of a technical nature. Inability t o choose a selective system on theoretical grounds reflects lack of understanding of the phenomena which produce a given partition ratio. T h e ratio is broadly a measure of the relative attraction or repulsion the competing phases show for the solute. T h e relative tendency the solute has for associating or forming hydrogen bonds with solvent molecules t o form complexes may be a governing factor. At least reasoning along this line helps t o bring the absolute K of a given solute %tithinthe proper range to be useful, but it does not throw much light on the more subtle requirements t h a t produce a high @ value for a given binary mixture. I n general, if the components of a given binary mixture differ considerably in structure and the first suitable solvent tried gives by chance a poor 6 value, a change to a different solvent combination will give a n improved @ value On the other hand, if the two components are similar structurally, such as leucine and isoleucine, the 0 value is not so easily improved. I n the letter case the search for a system need n o t be entirely a

random trial affair. Early in searches carried out at the, Rockefeller Institute for selective systems for the separation of antimalarial bases by countercurrent distribution, buffers (24)were found t o contribute greatly t o selectivity. Furthermore, the optimum buffer concentration. pH, etc., could be quickly established by constructing a plot of concentration or p H against the partition ratio. A system in which the K vias sensitive t o pK change was usually highly selective. The same approach waa used t o advantage in the penicillin studies. Golumbic et al. ( 4 0 ) used a similar approach in their study with phenols. ,4graphical approach similar in principle has been employed by Engel et al. ( 3 2 )in their search for selective systems to be used in separating estrogens. A four-component system containing water, methanol, chloroform, and carbon tetrachloride seemed of greatest interest. The proportions of the first two were held constant and a plot of K against per cent composition of chloroformcarbon tetrachloride was then enlightening. Certain theoretical deductions from the data were given, which showed the error of considering partition ratios in terms of relative solubilities of the aolute in the two phases. This misconception had previously been pointed out by Craig and Craig ( 2 3 ) on general grounds. d large amount of data on liquid-liquid equilibrium for systems of three or more Components has now become available and has been listed in the excellent review (86) of Treybal. However, i t is given in the form most useful for industrial application and not from the analytical viewpoint. Sonetheless, valuable suggestions may be obtained from these data. Saunders ( 7 3 )has made an extensive study of different nitriles as selective solvents for the separation of aromatics from aliphatic petroleum hydrocarbons. H e developed several generalizations regarding the most effective type of nitriles. The mere determination of single partition ratios for solutes in a given system does not' alone give sufficient data for choice of the most useful system (see 26) for separations, but i t is one of the first necessary steps. Those interested in a broad survey of part,ition ratios will find much of intereEt in the papers of Collander (19) in which the effect of changing the structures of the solutes is discussed. I n one study the partition ratios of 200 organic compounds in ethyl ether-water are measured and in the other 145 substances in the system isobutyl alcohol-water have been considered. The second system was usually the least selective. This is in line with the generalization t h a t the highest selectivity is t o be found in systems whose phases differ most in polarity (25). Other considerations, however, make the butanol system a very useful one for separating the more polar solutes, particularly when a n automatic countercurrent distribution

V O L U M E 2 4 , N O . 1, J A N U A R Y 1 9 5 2 apparatus ( 2 7 ) with many tubes is available and a certain degree of selectivity can be sacrificed. I n homologous series such as the normal fatty acids a p value in the range of 2 t o 4 for a CH2 difference was usually found by Collander. This is of the order shown by the systems developed by Barry et al. ( 7 ) for separating the homologous normal fatty acids. Simple systems of the type studied by Collander were of little use because of nonlinear partition isotherms. In general, any change in a solute molecule which affects its polarity will strongly affect the partition ratio. This fortunate circumstance together with a countercurrent distribution pattern provides the information required in order t o permit unambiguous determination of the molecular weights of larger peptides by assay for the substituting group (8). -4reagent attacking the free NH2 groups of the peptide such as the Sanger reagent (2,4-dinitrofluorobenzene) is permitted t o react only partially with the peptide under controlled conditions, so that a considerable fraction of unreacted peptide remains. Then a countercurrent distribution pattern determined directly on the reaction mixture will give a series of bands. The monosubstituted band will naturally occur nearest the band of unsubstituted peptide, the disubstituted derivative further removed, and the trisubstituted still further refor the radical introduced then moved. rl quantitative anal permits a calculation of molecular weight. The method clearly showed that gramicidin-S is a derapeptide. Attempts t o determine its molecular weight hy other methods ( 8 2 ) had always led t o a n ambiguous result. When the molecular weight is derived in this manner the final conclusion is not based on a single measurement but on several fractions throughout the monosubstituted band. These data must also be in agreement with the figure derived from the disubstituted band. Moreover, if the several bands all point to a single molecular weight and the bands are of such shape that impurity is not indicated, the data as a whole give considerable added confidence in the purity of the starting material. Golumbic and Weller ( 4 1 ) have developed a method for accurately measuring large partition ratios and have called it “interchange extraction.” Plattner et al. (68) have devised a scheme for separating azulenes which is based on differences in their extractability from sulfuric or phosphoric acid by a n organic solvent such as carbon tetrachloride, as the acid is progressively diluted with water. It amounts t o a n elution curve in which a constant volume of water is added t o the acid each time an extraction is made. Mathematics for calculating a theoretical elution curve are proposed. The separation of binary mixtures is given. APPARATUS

To the experimentalist one of the charms of extraction is offered by the great opportunity for the construction and design of new and novel apparatus. Each year brings its new crop. This year several extractors for removal purposes have been reported (18, 48,80). Holmes (50) described a simple extractor t o avoid formation of emulsions. Use of syringes in small scale extraction experiments has been discussed (10, 69). Smith et al. (79) described a constant temperature separatory funnel for determining liquid-liquid equilibria. A number of new designs for accomplishing countercurrent distributions have appeared (27, 4S, 93). These operate on the principle of decantation. Tschesche and Konige (86) describe a glass spiral type for larger scale work. Lathe and Ruthven (66) have described an apparatus which can be rather easily constructed and which possesses a number of advantages; has the disadvantage t h a t the solutions come into contact with polythene tubing. I t does not separate the phases by decantation but by permitting the heavier phase t o flow slowly through narrow polythene tubing, where a quick tip will cause it t o move t o the next chamber rat’her than return t o the one from which i t came. I t has given precise distribution curves.

67 The investigator who has not had experience with countercurrent distribution will usually approach it as a n experiment and try t o set up a n apparatus for multiple extractions with as little expense and effort as possible. The literature now describes a number of designs with different advantages. However, assembling a n extraction train with high numbers of units requires a t best a considerable expenditure of time and money. A makeshift arrangement may be the most costly in the end. When the research deals with the separation of unknown and often costly materials, all possible uncertainties should be eliminated a t the start. I t n-ould therefore seem a mistake to build an apparatus that will not meet the widest possible requirements. The design of cell reported by Craig et a!. ( 2 7 ) incorporates nearly all the features required by countercurrent distribution in its present stage of development. I t is more expensive to build than certain other types reported, but well may prove t o be the cheapest from the standpoint of the research. This viewpoint becomes much more important when the train is extended to several hundred units and automatic equipment t o operate it is added. For several years it has been obvious that countercurrent distribution would compete more favorably with chromatography for difficult separations were it not for the labor involved in applying high numbers of transfers. The fully automatic countercurrent distribution machine ( 2 7 ) goes a long way toward eliminating this shortcoming. It has a n automatic filling device for the upper phases and a mechanical robot for putting the apparatus through any choice of movements. A fraction collector for collecting the effluent phases during the single withdrawal procedure is controlled by the robot. A small cocurrent device is attached t o the train, which can be used t o counteract the effects of phase distortion by higher concentrations of the solute. When thousands of equilibrium stages are involved, large volumes of solvent are required. This can be overcome in large part by the “recycling” procedure ( 2 7 ) . Here the effluent tube from the last cell is connected t o the 0 tube, so that the upper phases pass over the lower more than once. This permits the application of several thousand transfers t o a mixture of two or three closely related bands with no more solvent required than that needed t o fill the machine once. A procedure for making rapid and precise weight analyses (28) was developed t o meet the need posed by the much larger number of fractions presented by the automatic equipment. A simple rotating evaporator for recovering the residues rapidly and under mild conditions has been described (26). Kies and Davis (55)have developed a stage continuous extraction train of simple design. It accomplishes essentially the same effect as the Johnson and Talbot extractor (51)or any of the stage continuous extractors when one of the liquid phases is held stationary and the solute is introduced batchwise with the other moving phase [for a review of these extractors see Craig and Craig (25, p. 224)l. Kies and Davis suggest the term “cascade dietribution process,” a very descriptive term. The process is analogous in principle t o the “single withdrawal” procedure ( 2 7 ) of countercurrent distribution and therefore comparisons should be made on this basis rather than t o the less efficient “fundamental” procedure tvhich Kies and Davis chose for comparison. The novel feature of their extractor lies in the absence of a stirring device t o promote the achievement of equilibrium. Instead a diffuser plate of the type familiar in the Kutscher-Steudel type of estractor is used. A wad of glass wool n-as found t o be as effective a diffuser as sintered glass. The extractor should prove very effective for separations involving systems in which solute interchange is very rapid. The experimental curves showed a degree of deviation from the calculated which could be due to lack of equilibrium or nonlinear partition isotherms. The deviations fortunately favored separation of the binary mixture used as test substances. Eisenlohr (31) has described a new centrifugal countercurrent extractor.

68

ANALYTICAL CHEMISTRY MATHEMATICAL CALCULATIONS AND THEORETICAL CONSIDERATIONS

I n extraction the great variations in procedure present just as much of a challenge to the mathematically inclined as the physical operations involved do t o the experimentalist. Interest along this line keeps pace with the technical progress and with the general interest in extraction. Sandell ( 7 2 ) has discussed the rationale of analytical extractions. Morrison (61) has given a good discussion of the subject. As regards countercurrent distribution, most of the recent effort has been directed ton-ard more rapid calculation of theoretical curves (4,42, 76, 90). A formula for rapid calculation of single withdrawal curves has been given (21). The relationship of continuous and discontinuous extraction has been treated by Scheibel ( 7 4 ) . Excellent graphical calculations have also been given by Compere and Ryland (20). Kies and Davis (65) have presented formulas for calculating curves obtained with their extractor. The question of the achievement of true equilibrium and the factors which cont,rol the rate has altvak-s been a n interesting one. Solutes and systems vary widely in their rate of approach t o equilibrium. The majority of apparatus for both continuous operation and countercurrent distribution has been designed t o promote the most rapid approach. I n general, amino acids reach equilibrium rapidly, as Felix et al. ( 3 3 ) found. This may be due t o the systems used. The highly charged amino acids exhibit a solubility sufficient for the purpose only in the organic phases of systems in which much v a t e r is also present. Here the two phases approach mut.ual solubility and there apparently is little resistance to the passage of the solute across the interface. However, the fatty acids also were found t o approach equilibrium (7) very rapidly in isopropyl ether-buffer systems, in spite of the greater differences in polarity of the phases. The penicillins in the same type of system reached equilibrium much more slotvly (6). Polypeptides with molecular weights of the order of a feiv thousand apparently reach equilibrium ( 2 5 ) in no more time, during a countercurrent distribution, than many solutes of one tenth the size. Even more surprising is the finding t h a t the protein insulin can be fractionated (44)almost ideally in the proper system by countercurrent distribution with no ext r a time allo~ved for the equilibration period. With these observations in hand it is difficult t o believe t h a t rates of diffusion (75) can play a very large role in controlling the rates of approach to equilibrium in liquid-liquid systems, a t least where there is sufficient agitation for momentary emulsification of the phases. Most countercurrent distribution runs could probably be made with very much shorter equilibration periods. The larger molecules of interest t o biochemistry usually show a marked tendency to associate either with themselves or with other molecules in the solution. They do not behave as ideal solutes. K h e n the association is with like molecules t o form dimers, trimers, etc., a nonlinear isotherm results. This type of isotherm is to be avoided if possible when separations are t o be made. On the other hand, if the solvent molecules are the ones surrounding the solute molecules to form the complex, the isotherm will be linear. Thus incorporation of sufficient acetic acid into the system ( P I ) gives a good approach t o linear isotherms with the higher fatty acids, the bile acids ( I ) , and certain peptides. At the same time emulsion tendencies are inhibit’ed. Many heterogeneous complexes show sufficient stability ( 6 6 ) in t h e appropriate systems t o permit their distribution as ideal solutes. Here there is no dissociation. A nonlinear partition isotherm produces usually a skewed countercurrent distribution pattern. If several such solutes are present and the objective of the run is separation of all the solutes, there are disadvantages. Either the front or last band, depending on the direction of skewing, can be obtained free of the others very easily with a high percentage recovery but a t the expense of the desired result for the other bands.

In the continuous or stage continuous process operated by contiiiuous center feed, such deviations are more serious. These processes are designed usually to separate the mixture into tlvo fractions in a single run. One solute (or family of solutes) move8 toward one end of the column and may be called the extract, while the other (or different family), the raffinate, moves in the opposite direction. However, because of the curvature of the nonlinear partition isotherm the solute may a t a lower concentration favor the heaLier solvent and move in the direction of its flow, but as the concentration builds up the K may shift to favor the lighter solvent. An unpredictable quandary thus arises. When there is such a change of preference of the solute from one phase t o the other, the substance has been called a “solutrope” (78). A good example of the phenomenon was found by Smith et a l . (79) in the system benzene-pyridine-water. APPLlCATIONS

In the inorganic field separations thus far have involved only simple extraction, even though certain distribution data in the literature indicate t h a t countercurrent distribution would definitely extend the usefulnePs of the extraction approach, Many separations do not require the countercurrent process. Weinhardt and Hixson (91) have found methyl isobutyl ketone to be very selective in preferentially extracting sodium dichromate from a n aqueous hydrochloric acid solution containing dichromate and vanadic acid. Separation factors well over 4000 were obtained. They give a short but excellent review of the use of extraction for inorganic separations. I n a n excellent discussion of the role of extraction in analytical chemistry Morrison (61) also gives a review of inorganic separations. Separation of the rare earths should challenge those interested in extraction. .4sselin et a/. ( 3 ) have studied the liquid-liquid partition of the rare earth salts from the standpoint of separations. Thorium nitrate could be separated easily from the other nitrates in the system 1-pentanol-aqueous ammonium thiocyanate. The system I-butanol-water also gave differences. Moeller and Jackson (60) have studied the separative extraction of t,he rare earths in the form of 5,7-dichloro-8-quinolinol chelates. Only fractional separations were achieved by single extractions and obviously a countercurrent process is indicated. Lithium chloride has been separated from sodium chloride by extraction with butanol (53). An observation of interest has been made by Morton (62). He found extraction by butanol of enzyme solutions useful in removing extraneous materials. Such a procedure did not cause denaturation and made further purification much easier. Countercurrent distribution should prove to be a n excellent method of purification of alkaloids. It has already led to the isolation of a number of new alkaloids. Applications have dealt ergot (45,46),wilforine, and wilfordine with Veratrum v i d e (M), (9, 11). The Brodie methyl orange distribution technique has been applied t o the colorimetric determination of alkaloids in tissues ( 3 7 ) . i n lipide chemistry t,here is a great need for some tool t h a t will separate the intact materials or slightly modified derivatives. A promising start on fat oxidation products has been made by Zilch and Dutton (95) and by Fugger et al. (36‘). Introduction of a hydroxyl group into a fat shifts its partition ratio considerably. Improved systems for fatty acid separations have been reported (7). The most striking contribution of countercurrent distribution during the past year has probably been made in the polypeptide field, It has been used by Brink et al. in purifying subtilin (I$), by Brockmann et al. in purifying polypeptide antibiotics from Actinomycetes (13-15), by Payne et al. in purification of ACTH ( 6 7 ) , by du Vigneaud and collaborators on vasopressin (87) and oxytocin (63), by Felix et al. on clupein (SS), by Velick and Udenfriend in their isotope dilution method (88) for amino acid

V O L U M E 24, NO. 1, J A N U A R Y 1 9 5 2 analysis, and by Kenner (54)in purifying synthetic peptide derivatives. Where isolation of amino acids from hydrolyzates is desired, i t is a useful tool (25, 27, 33). It has settled the question of the purity and molecular weight of gramicidin-S (8). When the problem of. the investigation of the metabolic fate of a drug is encountered, countercurrent distribution is rapidly becoming the method of choice. Recent studies along this line include the isolation of a degradation product of folic acid (OS?),of quinoline ( G j ) , of chloroguanide (ZQ),of methadone (go), of salicylic acid ( Z ) , of p-aminobenzoic acid (83), of tocopherol compounds in feces (71),and of the metabolic products formed from the cinchona alkaloids when given t o man ( 1 6 ) . Flavin and Graff (34)have used countercurrent distribution t o good advantage in studying the utilization of guanine for nucleic acid biosynthesis by protozoa. The conversion of guanine t o adenine was clearly shown. Engel and collaborators (36,7 7 ) continue their excellent work on the separation of estrogens. Separation of sterols should become more interesting with the greatly increased numbers of transfers ( 6 7 ) now possible. I n spite of the sticky nature of bile, countercurrent distribution has been shown t o be an effective tool for separating bile acids both free and conjugated (1). Here the higher numbers of transfers were required. Countercurrent distribution has been of considerable interest in the streptomycin type of antibiotic ever since Titus and Fried ( 8 4 )first showed t h a t the original streptomycin was a mixture of two streptomycins. Such interest continues (57, 58, 81). Rauen and Waldmann ( 7 0 ) have studied the separation and characterization of pterins. The most effective system was 1-butanol-0.02 S hydrochloric acid. Analysis could be made by the intense fluorescence, which permitted a good run t o be made on 30 to 60 micrograms of material. The photo-oxidative degradation was studied with the same technique. Miscellaneous applications of countercurrent distribution have included interesting studies on the auxins in cabbage (49),the methyl substituted anilines ( 3 9 ) )the purification of sarmentogenin (17 ) , the dephosphorylation of nucleotides (j), the fractionation of polythene (SO), and the study of heparin ( 9 4 ) . Neuivorth et al. ( G 4 ) found the Schiebel column t o be very effective for removal of t a r acids from tar distillates. LITERATURE CITED (1) Ahrens, E. H., Jr., and Craig, L. C., Federation Proc., 10, 154

(1951). and Smith, P. K., (2) Alpen, E. L., Ivlandel. H. G., Rodrvell, V. W., J . Pharm. Exptl. Therap., 102, 150 (1951). J . Phys. & (3) Asselin, G. F., Audrieth, L. F., and Comings, E. W., Colloid Chem., 54, 640 (1950). (4) Bacher, J. E., J . Am. Chem. SOC.,73, 1023 (1951). ( 5 ) Bacher, J. E . , and Allen, F. IT., J . B i d . Chem., 188, 59 (1951). (6) Barry, G. T., Sato, P., and Craig, L. C., Ibid., 174, 209 (1948). (7) Ibid., 188, 299 (1951). (8) Battersby, A. R., and Craig, L. C., J’. Am. Chem. Soc., 73, 1887 (1951). (9) Beroaa, Lf., ANAL.CHEX, 22, 1507 :1950). (10) Ibid., 23, 1055 (1951). (11) Beroza, &I., J . Am. Chem. Soc., 73, 3656 (1951). (12) Brink, N. G., Mayfield, J., and Folkers. K., Ibid., 73, 330 (1951). (13) Brockmann, H., Bauer, K., and Borchers, I., Chem. Ber., 84, 700 (1951). (14) Brockmann, H., and Grubhofer. N., ~ ; a ~ u r ~ i s s e n s c h a ~37,494 ten, (1950). and Kalbe, H., Chem. (15) Brockmann, H., Grubhofer, N., Kass, W., Ber., 84, 260 (1951). (16) Brodie. B., Baer, J. E., and Craig, L. C., J. B i d . Chem., 188, 567 (1951). (17) Callow, R. K., hleikle, R. D., and Taylor, D. A. H., Chemistry & Industry, 1951,336. (18) Cohen, S. L., J . Lab. Clin. Med., 36, 769 (1950). (19) Collander, R.. Acta Chem. Scand., 3, 717 (1949); 4, 1085 (1950). (20) Compere, E. L., and Ryland, A, I n d . Eng. Chem., 43,239 (1951). (21) Craig, L. C., ANAL. CHEM.,22, 1346 (1950). (22) Ihid., 23, 41 (1951). (23) Craig, L. C., and Craig, D. C., in Weissberger, A., “Technique of Organic Chemistry,” Vol. 111, pp. 171-311, Kew York, Interscience Publishers, 1950.

69 (24) Craig, L. C., Golumbic, C., Mighton, H., and Titus, E., J . Bid. Chem., 161,321 (1945). (25) Craig, L. C., Gregory, J. D., and Barry, G. T., Cold Spring Harbor Symposia Quant. B i d , 14, 24 (1949). (26) Craig, L. C., Gregory, J. D., and Hausmann, R., ANAL.CHEM., 22, 1462 (1950). (27) Craig, L. C., Hausmann, W., rihrens, E. H., Jr., and Harfenist, E. J., Ibid., 23, 1236 (1951). (28) Ibid., p. 1326. (29) Crounse, N. N., J . Org. Chem., 16, 492 (1951). (30) Desreux, V., and Spiegels, XI. C.. Bull. soc. chim. Belg., 59, 476 (1950). (31) Eisenlohr, H., Chem. Ing. Tech., 23, 12 (1951). (32) Engel, L. L., Slaunwhite, W.R., Jr., Carter, P., and Olmsted, P. C., J . Bid. Chem., 191, 621 (1951). (33) Felix, K., Rauen, H. M . , Stamm, W.,and Zimmer, G., 2. physiol. Chem., 286, 199 (1950). (34) Flavin, hl., and Graff, S., J . Bid. Chem., 191, 55 (1951); 192, 485 (1951). (35) Fried, J. F., White, H. L., and Wintersteiner, O., J . Am. Chem. Soc., 72, 4621 (1950). (36) Fugger, J., Cannon, J. A., Zilch, K. T., and Dutton, H. J., J . Am. Oil Chemists’ SOC.,28, 285 (1951). (37) Gettler, A. O., and Sunshine, Irving, ASAL. CHEM.,23, 779 (1951). (38) Golumhic, C., Ibid., 23, 1210 (1951). (39) Golumbic, C., and Goldbach, G., J . Am. Chem. SOC.,73, 3966 (19511. (40) Golumhic, C., Orchin, M., and TVeller, S., Ibid., 71, 2624 (1949). ’ (41) Golumbic, C., and Weller, S., ANAL.CHEM.,22, 1418 (1950). (42) Gregory, J. D., and Craig, L. C., Ann. N . Y . Acad. Sci., 53, 1015 (1951). (43) Grubhofer, N., Chem. Ing. Tech., 22, 209 (1950). (44) Harfenist, E. J., and Craig, L. C., J . Am. Chem. SOC.,73, 877 ( l 9 X ). (45) Hashimoto, T., J. Pharm. SOC.J a p a n , 66, 22 (1946). (46) Hellberg, H., Farm. Reay, 50, 17, 33 (1951). (47) Ihid., pp. 301-9. (48) Hemmings, A. TI’., Analyst, 76, 117 (1951). (49) Holley, R. W., Boyle, F. P., Durfee, H. K., and Holley, A., Arch. Biochem., 32, 192 (1951). (50) Holmes, F. E., ANAL.CHEM.,23, 935 (1951). J . Chem. Soc., 1950, 1068. (51) Johnson, J. D. A., and Talbot, -4., (52) Kalopissis, G., Chimie & Industrie, 64, 563 (1950). (53) Kato, T., and Hagim-ara,Z., Technol. R e p f s . Tdhoku I m p . C’niv., 14,lO (1950). (54) Kenner, G. IT., Chemistry & Industry, 1951, 15. (-55) Kies, X f . W., and Davis, P. L., J . Biol. Chem., 189, 637 (1951). (56) Lathe, G. H., and Ruthven, C. R. J., Biochem. J., 49,540 (1951). (57) Leach, B. E., De Vries, IT. H., Nelson, H. A , Jackson, W.G., and Evans, J. S., J . Am. Chem. SOC.,73, 2797 (1951). (58) Leach, B. E., and Teeters, C. XI.,Ibid., 73, 2794 (1951). (59) hledalia, A. I., and Stoenner, R. IT,, ANAL. CHEM.,23, 545 (1951). (60) hloeller, T., and Jackson, D. E., Ihid., 22, 1393 (1950). (61) Morrison, G. H., Ibid., 22, 1388 (1950). (62) Morton, R. K., Xature, 166, 1092 (1950). (63) LIueller, J. XI., Pierce, J. G., Davoll, H., and du Vigneaud, V., J . Biol. Chem.. 191. 309 11951). (64) Neuworth, M. 6., Hofmann, Ti., and Kelly, T. E., I n d . Eng. Chem., 43, 1689 (1951). (65) Novack, L., and Brodie, B. B., J . Biol.Chem., 187, 787 (1950). (66) O’Keeffe, A. E., Dolliver, M.A,, and Stiller, E. T., J . Am. Chem. SOC.,71, 2452 (1949). (67) Payne, R. IT., Raben, M. S., and bstwood, E. B., J Biol.Chem., 187, 719 (1950). (68) Plattner, P1. A., Heilbronner, E., and Keber, S.,Helu. Chim. Acta, 33, 1663 (1950). (69) Rauen, €1. &and I., Waldmann, H., Experzentia, 6, 387 (1950). (70) Rauen, H. M.,and Valdmann, H., Z. phusiol. Chem., 286, 180 (1950). (71) Rosenkranta, H., hlilhorat, A. T., and Farber, M., J . Bid. Chem., 192, 9 (1951). 172) Sandell. E. B.. A n a l . Chim. Acta. 4. 504 (1950). (73) Saunders, K. W., I n d . Eng. Chem., 43, 121 (1951). (74) Scheibel, E. G., Ibid., 43, 242 (1951). (75) Sherwood, T. K., Evans, J. E., and Longcor, J. V. A., Ibid., 31, 1144 (1939). (76) Slaunwhite, W.R., Jr., ANAL.CHEM.,23, 687 (1951). (77) Slaunwhite, W. R., Jr., Engel, L. L., Olmsted, P. C., and Carter, P., J . Bid. Chem., 191, 627 (1951). (78) Smith, A. S., I n d . Eng. Chem., 42, 1206 (1950). (79) Smith, J. C., Stilbolt, V. D., and Day, R. W.,Ibid., 43, 190 (1951). (80) Stepanov, F. N., Vul’fson, N. S., and Mikova, I. A., Zaaodskaya Lab., 16, 1131 (1950).

ANALYTICAL CHEMISTRY

70 (81) Swart, E. A., Lachevalier, H. S . , and Waksman, S. A.. J . Am. Chem. Soc., 73, 3253 (1951). (82) Synge, R. L. M., Biochem. J., 39, 363 (1945). (83) Tabor, C. W., Freeman, M.V., Baily, J., and Smith, P. K., J . Pharm. Exptl. Therap., 102, 98 (1951). (84) Titus, E., and Fried, J., J . Bid. Chem., 168, 393 (1947). (85) Treybal, R. E., Ind. Eng. Chem., 43,79 (1951). (86) Tschesche, R., and Konige, H. B., Chem. Ing. Tech., 22, 214 (1950). (87) Turner. R. A.. Pierce. J. G.. and du Vieneaud.. V... J . Bid. Chem.. 191, 21 (1951). (88) T’elick, 5. F., and Udenfriend, S., Ibid., 190, 721 (1951). I

,

I

(89) Way, E. L., and Bennet, B. M., Ibid., 192, 335 (1951). (90) Way, E. L., Signorotti, B. T., March, C. H., and Peng, C. T., J . Pharm. Erptl. Therap., 101, 249 (1951). (91) Weinhardt, -4.E., and Hixson, A. S . ,Ind. Eng. Chem., 43, 1676 (1951). (92) Weisz, E., Magyar Kem. Lapja, 6, 59761 (1951). (93) Weygand, F., Chem. Ing. Tech., 22, 213 (1950). (94) Wolfrom, M. L., Montgomery, R., Karabinos, J. V., and Rathgeb, P.. J . Am. Chem. SOC.,72, 5796 (1950). (95) Zilch, K. T., and Dutton, J., ASAL. CHEW.,23, 775 (1951). RECEIVED October 31, 1951.

ORGANIC MICROCHEMISTRY C. 0. WILLITS AND C. L. OGG Eastern Regional Research Laboratory, Philadelphia 18, Pa.

T

H E scope of the present review is essentially the same as that of other years. Judged from the number of papers written during the past year, there is no apparent lessening of interest on the part of microchemists in their efforts to improve existing methods of analysis. Noticeable is the increased attention being given both here and abroad to the subject of standardization. APPARATUS

The accuracy, precision, and utility of the Cornwall syringe, Chaney pipet, and modified tuberculin syringe for delivering from 0.2 to 7.0 ml. of liquid were compared by Dern et al. (29). They concluded that syringes held the advantage in speed and accuracy and also in adaptability, because other than integral F olumes could be measured and viscosity, surface tension, and temperature variations had less effect on the constancy of volumes delivered from syringes than from pipets. British Standards (16) specifies the design and dimensions of washout pipets for volumes from 0.1 to 1.0 ml. with tolerance and detailed methods for using the pipets. Flaschka ( 4 1 ) has described a combination microburet and Jones reductor which greatly simplifies titrations u-ith solutions that must be kept in an inert atmosphere. Two new melting point apparatus have been described, one by Hilbck (61), the other by Hippenmeyer (62). The apparatus of Hilbck uses exchangeable thermometers which increase the accuracy of reading \Then used over a wide temperature range. The distillation and sublimation apparatus described by Erdos ( 3 5 ) is designed for use under reduced pressure and with small amounts of material. Conolly and Oldham (27) also described a distillation apparatus for quantities of 1 ml. or less. A turntable for carrying out six simultaneous saponifications with reflux condensation was described by Griitzner and Hintermaier (62). The modified Conway cell designed by Leurquin and Delville ( 8 7 ) has three compartments, one for sample, one for reagent, and another for absorbent solution, all covered by a glass plate. The sample and reagent can be mixed without contaminating the absorbent solution. A light, simple bomb for sodium peroxide fusions of micro, semimicro, or macro amounts of starting material was described by Wurzschmitt (162). WYth this bomb ethylene glycol was used in place of potassium nitrate and sugar, as the fusion temperature was lower and the metal of the bomb less affected. Electrical apparatus described during the past year include a a small high temperature, high vacuum furnace by Alberman (I), simple, precke temperature regulator for electric furnaces by LBvy and Schick (91), an apparatus for microconductometric titrations of volumes less than 1 ml. by Stock (135), and a photoelectric micronephelometer designed for use with 0.1 to 0.5 ml. of solution by hIason (98). Other apparatus included a drying tube described by Bradlow and Vander TVerf (IS), containers and manometers for micro-

manometric methods by Mohle (101I. and a device for making pressure prpcipitation of sulfides by Stock and Heath (136). Balances. Two articles published by Hodsman, one a report on a symposium on microchemical balances ( 6 7 ) and the other a discussion of developments in microohemical balance design (66), give a comprehensive picture of balance design, materials, and manipulation, as well as the sensitivity, precision, and practical aDdication of results. Brown (IS)described an imDroved 5-me. aluminum foil rider for Ainsworth microbalances which reduced the rider seating errors. KomArek (80) discussed the origin of microbalances and Muller (103) presented an unpublished paper by Eigenberger in which proposals were made for improving the microlift balance for determining liquid densities. Improvements in the radioactive electronic microbalance were discussed by Feuer (37) and several modifications of torsion ultramicrobalances were reported by Korenman et al. (86). The magnetically controlled quartz fiber microbalance described by Edu-ards and Baldwin ( 3 4 ) had load capacities of 1.2 mg. unbalanced and 100 to 150 mg. balanced with a sensitivity of 0.1 microgram for 6 minutes of arc.

-_

-

ELEMENTAL AX 4 LYSIS

Carbon and Hydrogen. The determination of carbon and hydrogen in organic compounds continues to be the popular subject of analytical papers; achieving accuracy of analysis n-ith simpler apparatus and technique is usually the goal. Kainz (76) has described an apparatus in which nitrous oxides are eliminated in the gravimetric procedure through the use of metallic copper. The advantage claimed for the copper absorhrnt is that it not only removes the oxides of nitrogen but eliminates the conditioning runs, as it is nonhygroscopic. Belcher and Ingram ( 7 ) advocate the use of manganese dioxide for the removal of oxides of nitrogen, and using their rapid oxygen flow technique this absorbent can be used outside the combustion tube and at room temperatures. For the slow or conventional combustion method they have described a modified combustion tube packing. The rapid Combustion procedure continues to gain in popularity and Ingram ( 7 0 ) has described advantages of the rapid combustion process for carbon and hydrogen and outlined methods that have been successfully used in routine analysis. Wagner (146) has described a somewhat modified combustion tube which uses readily exchangeable extensions, the one without lead dioxide filling for compounds that do not contain nitrogen and the other with an aged lead dioxide filling for use when nitrogenous organic compounds are burned, He advocates not wiping absorption tubes prior to weighing, a point in accord with methods being adopted in the United States. With increasing interest in organofluorine compounds, i t is but natural that studies have been made of methods for the successful analysis of these compounds. Belcher and Goulden (6) have proposed a method in