1557
V O L U M E 26, NO. 10, O C T O B E R 1 9 5 4
< serine was observed, a complete inversion of the order of these three compounds relative to phenol. Other inversions relative to the phenol sequence includcd methionine < valine with npropyl alcohol-water (ratio by volume i o to 30) and isoleucineleucine < methionine Ir-ith collidine-lutidine-water (ratio by volume 3 to 3 to 4j. Relative to phenol, the latter solvent improved the separat,ion of v:tline and methionine without sequence inversion. Both solvent and filter paper characCYCLIC~ I N ACIDS. O teristics affected the sequences of the cyclic amino acids. The order of the cyclic amino acids was independent of the type of filter paper. With the phenolic solvents the sequence of the cyclic amino acids was as follows: tyrosine < tryptophan < phenylalanine < proline. K i t h n-propyl alcohol on Whatman No. 1 filter paper and with all tert-butyl alcohol solvents, the sequence proline < tyrosine < tryptophan < phenylalanine was observed, differing from the order obtained with phenolic solvents only in respect to the complete inversion of proline. On S & S 507 and S & S 589 Blue Ribbon filter papers. proline was inverted only in respect to tryptophan and phenylalanine by n-propyl alcohol.
The sequence of the cyclic amino acids with collidine-lutidine was essentially the same as with the aliphatic alcohols except for the possible inversion of tryptophan and tyrosine, the complete sequence for collidine-lutidine being as follows: proline < tryptophan 5 tyrosine < phenylalanine. Tryptophan was not as stable as the other amino acids during chromatography (Table I). Fresh solutions of tryptophan UPUally gave a purple spot with ninhydrin. However, solutioris of tryptophan which were stored for several \!-eeks often gnve both purple a n d yellow spots when chromatographed in various solvents including collidine-lutidine and phenol. The fastermoving, yellow spot frequently appeared when freshly prepared tryptophan solutions were employed and the chromatograms developed in phenol. The brilliant yellow fipot developed very slowly, several days a t room temperature being required to bring out its maximum brightness. A third tan spot falling hetween the purple and the yellow spots was observed when filter paper strips, spotted a t the origin with freshly prepared tryptophan solution, were stored a t room temperature for 3 months before they were developed in 70 volume % tert-butyl alcohol.
(SMALL-SCALE FILTER PAPER CHROMATOGRAPHY)
A Rapid Two-Dimensional Procedure LOUIS B. ROCKLAND and J. C. UNDERWOOD Fruit and Vegetable Chemistry Laboratory, Agricultural Research Service,
'
A small-scale two-dimensional filter paper chromatography procedure has been developed for the separation of amino acids and related compounds. .4s little as 0.1 y of amino acid contained in 0.1 to 1.0 rl. of solution can be separated and identified on a 5inch-square sheet of filter paper irrigated in a No. 4 American Medical museum jar. A tert-butyl alcoholformic acid-water mixture is employed as the first solvent and a single phase phenol-ammonia-water mixture as the second solvent. Variations in R,.values as a function of irrigation temperature have been studied for both solvents over the range 5' to 40" C. -4 chart has been prepared showing the loci of 60 amino acids, peptides, carbohydrates, and organic acids on a diagrammatic small-scale, two-dimensional chromatogram. Chromatograms of test mixtures containing 18 amino acids, a hydrochloric acid hydrolyzate of casein, and an enzymatic hydrolyzate of lactalbumin contained 16, 12, and 16 distinguishable spots, respectively.
D
ATTA et al. (IO),hla and Fontaine (I?'), and Boissonnas ( 3 ) have reported procedures for two-dimensional filter paper cliromatography of amino acids which employ filter paper sheets approximately one half of the dimensions suggested by Consden et nl. (9). For the past 3 >-ears the present authors have employed further reduced scale procedures of one- and two-dimensional filter paper chromatography for the identification of the nitrogenous constituents in citrus juices (20, 25, 29) and other plant extracts. The reduced scale procedures afford greater speed and convenience, increased capacity, lower equipment, material, and labor costs, and conservation of space. Using three or more replicate chromatograms per sample and careful control of important variables, it has been possible to improve the precision of two-dimensional paper chromatograms. In the laboratory of the authors, one person has prepared an average of over 100 two-dimensional filter paper chromatograms per week.
U.S.
Department o f Agriculture, Pasadena 5, Calif,
EXPERIMENTAL
Materials. These materials are the same as those given in iLFactorsAffecting Separation and Sequence of Amino iicids." General Procedure. The general procedure was essentially the same as that described ( 2 3 ) for small-scale, ascending, onedimensional chromatography with the exceptions and additions indicated below. Five-inch squares of filter paper were spotted 5 / * inch from the sides of the lower left' corner with approximately 1 X 10-4 ml. of 0.06M solut,ions of test mat,erials using selffilling micropipets ( 2 4 ) or a Gilmont ultramicroburet (0.01 ml. in capacity). The filter papers were mounted on t h r wooden dowels so that the machine direction was parallel t'o the solvent flow and suspended in hmerican bledical museum jars containing approximately 75 ml. of solvent I. The solvent was allowed to migrate to within l / 4 inch of the upper edge of the paper. Spproximately 7, 8.5 ,
Gn BI EB
3(i 64 .8 i 02
78 73 7R
A0
20
42 84
17 32 81 24 4fi 11 62
38 15 22 64 6R
4 B
4I
.? R
-4 E .4 t
AC CC CE CI I?'
DJ
E 7'
I' I'
u sr GE G4 GO NR GI, Ga
GLl
0I . 111
IIE H
c
€IS TIP
IS
LA
LE I4 LY
-21E .If s MD NL NV
OR
PS PR a R f values are not necessarily identical u i r h thost. listed in other scctions j trcer-because of difference: in ternlrcratcre a t wI.ic1. cl.roniarogra1i.s uere prepared.
-
+olvent I
Bi Ba
I'H
uf
I-
S4
SE
su T .4
TH
TR
1' y 11
59 66 77 69 71)
69
34
DI.-AIanine 3- Alanine DL-Alanylalanine ut,-Alanylglycine ~~-m-.itnino-n-butyric acid y-.ioiino-n-biityric acid a-Ailiinoisobiitsric acid d-.~iiiinoiiobiityrjcacid o-.imino-a-etliylbutyric arid r)i.-o-Ariiinooctanoic acid i.-.irxinine nionohydrcrliloride Dr.-A.;parapine ni.--ispartic acid Ascorbic acid DL-Citriilline L-Cysteine liydrocliioride
58
,
82 50 4-1 27 42 10 4 .j 37 47 33 31 15
1-aIII e
38 80 50
60 57
a
25 SI 40 42 75 57 34 72
so
31 15 41 39 84 9 84 82 15
43 54 6Y 84 26 82 83 70
68 36 31 84 72 22
80 66 81 82 78 69
75 51 50 41 33 42 41 38 54 48 65
87 71 43 78 34 34 39 50 78 59 78
I3,ienkolic acid 1:tliionine Friirtose Gliiroiaiuine L-GIiita (nine r,-GIiitarnic acid Gliitatllione ioxidizedj Glutathione (reduced) Glycine Glycyl-DL-alani ne Glyeylylycine Glucose 1.-Histidine iuonohydrocliloride, inonoliydrate I.- Homocysteine I,- Ilouiocystine ~~-1loiri0serine r.-HydroxyproIine nL-Isoleucine Lantliionine r.-Leiicine nL-Lmicglglycine Dr.-Lysine monohydroFlilorirlr .... _.
DL-Afethionine DL- Afethionine sulfone DL-hlethionine sulfoxide DL-Norleucine DL-h-orvaline DL-Ornithine monohydrochloride DL-Phenylalanine @-Phenylserine L-Proline Sarcosine hydrochloride DL-Serine Sl,Cr"*e -.~ . .
Taurine DL-Threonine DL-Tryptophan L-Tyrosine nL-Valine
Color with Ninhydrin a t 25' C. Purple Purple b Purple Purple Purple Purple e Purple b Purple b Purple Purple Purple Tan Blue-purple Light purple8 Purple Pink-purple Pink-purple Light purple Purple Purple Tand Purple Purple Purple Purple Purple Pink-purple Purple e Purple Tand Gray-piirule 6 Pink-purple Pink-purple Purple Yellow Pink-purple Purple Purple Purple Purple Purple Purple Purple Purple Purple Purple Pink-purple Purple Yellow PurpleC Purple Tan$ Purple Purple Pink-purple Purple Purple
1562 Table 111. Variation of R / X 100 Values of Seven Amino acids (With different lots of filter paper a t two temperatures) R f X 100 t'aluesa At 21.50 c. At 24.0' C. .Imino Acid Lot l b Lot 2 h Lot l b Lot 2 b Aspartic acid 29 22 29 23 Glutamic acid 34 28 36 28 Arginine 14 7 13 8 Serine 25 20 26 21 A1ani n e 33 26 33 27 Valine 55 44 50 43 Proline 36 28 34 30 a Each figure represents average value obtained from three replicate chromatograms, using solvent I , one-dimensional chromatograms. b 9 & S 589 Blue Ribbon filter paper.
of R/ values. A second, identical experiment was conducted using a second lot of the same type of filter paper obtained from the manufacturer a t a later time. The shapes and slopes of the curves obtained with the second lot of filter paper were similar but not identical to those shown in Figures 6 and 7. In general, R/ values were highly reproducible within one lot of filter paper but an average of 5 to 7 units lower with the second lot of filter paper with solverit I as shown in Table 111. In Figure 7 the R, values obtained from chromatograms irrigated with solvent I1 tended to increase slightly with increasing temperature. The relatively small temperature coefficients of the Kj values obtained with solvent 11, a phenolic solvent of fixed coniposition, are in contrast to the large R/ variations with temperature found Rith similar, water-saturated solvents because of composition changes in the latter two-phase system a t critical solution temperatures (5, 16). Burma (6) has studied the effect of temperature on the R j values of four amino acids over a more limited temperature range (10 to 30" C.) with four water-miscible solvents. The R, tis. temperature curves obtained by the present authors for solvent I are similar to those obtained by Burma for isopropyl alcoholwater (7 to 3) and pyridine-water (T to 3 ) over the same temperature range. The Rj minima found for all amino acids a t about' 30' C. with solvent I were not observed by Burma who studied a more limit,ed temperature range. However, Burma did find a minimum for aspartic acid a t 25" C. with the pyridine-water solvent. I n the present study secondary maxima were found for arginine with solvent I. and valine and proline with solvent 11. The R f variations with temperatiire are greatest for compounds which migrate most rapidly. Inversions with temperature of the arginine-proline sequelice with solvent I1 and the aspartic acid-serine and asparaKitie-argiiiine sequences with solvent I1 suggest the possibilit,y of improving chromatographic separations through temperature regulation. Reproducibility of R, Values of Amino +ids. Table IT' shows the average R, values, range of Rj values, and st'andard deviation for each of eight representative amino acids obtained from 60 replicate two-dimensional chromatograms prepared on S & S 589 Blue Ribbon filter paper with solvents I and 11. Calculations of the standard error of the means of replicate R, values indicate that the Rj values of the amino acids will have a standard error of lese than 2% a h e n three to five replicate values are employed to compute the average. Two-Dimensional Chromatographic Map of Amino Acids and Related Compounds. Figure 8 shows a diagrammatic map of the positions occupied by various amino acids, peptides, carbohydrates, and related compounds on small-scale filt'er paper chromatograms prepared with solvents I and I1 on S & S 589 Blue Ribbon filter paper a t about 19" C. The R/ values used in the construction of the diagram were obtained by averaging the values obtained from three or more replicate two-dimensional chromatograms. The sulfhydryl compounds, cysteine, homocysteine and glutathione, could not be detected on Whatman No. 1 filter paper under the same conditions, probably because of
ANALYTICAL CHEMISTRY their decomposition to unknown oxidation products on the filter paper. With S & S 589 Blue Ribbon filter paper these same compounds were detected in the normal manner with ninhydrin. However, the presence of two additional weak spots was usually observed on the same chromatogram of each of the resoective compounds. One of these spots corresponded, in each case, with cystine, homocystine, or oxidized glutathione, indicating oxidation of the compounds on the filter paper before irrigation with solvent I. I n each case the third spot \vas located a t the apes of a right triangle formed by the three spots obtained from chromatograms of each of the pure reduced compounds. The R / values of the third spot corresponded in each case to the Rj value of the pure, reduced compound obtained with solvent I and the R j value of the disulfide (cystine, etc.) obtained with solvent 11. Therefore, it would appear reasonable to suggest that a fraction of each sulfhydryl compound was oxidized to the corresponding disulfide form on the filter paper both before and after irrigation with solvent I, but before irrigation with solvent 11. Histidine could be detected on two-dimensional chromatograms only when applied a t levels five or more times the amount required for satisfactory spots with most other compounds.
Table IV.
Reproducibility of R j X 100 Values
(Obtained from sixty replicate two-dimensional chromatogramsa) Solvent I Solvent I1 Standard Standard AverdeviAverdeviAmino .icid age Range ation, ob age Range ation, ab Aspartic acid 27 22-32 2.3 12 9-15 1.3 Glntamic acid 38 31-43 2.6 2% 18-28 1.2 Arginine 91 87-94 2.0 17 6-24 4.2 2.0 Asparagine 17 13-22 37 32-42 2.4 Serine 28 22-32 2.5 38 28-42 2.6 3.7 57 53-63 2.6 Alanine 39 32-47 it: 70-82 2.1 Valine 62 51-70 4.1 Proline
4.7 91 87-94 1.4 3.3 2.0 a 6 S: S 589 Blue Ribbon paper lot 1. Irrigation witti solvent I a t 20 to 21' C. Irrigation with solvent I1 a t room temperature (22" to 25' C.). 44
33-57
AV.
b r =
dgl, S d2
ACKYOW LEDG\IE\T
The authors airh t o thank 11.S.Dunn, R . B. Henderson, arid >I. J. Horn for generous gifts of various compounds and Patricia C. SlodowsLi for technical aqsistance during the course of this ivork. LITER4TURE CITED
Bentley, €I. R., and Whitehead, J. K., Biochem J . , 46, 341 (1950). Block, R. J., LeStrange, R . , and Zweig, G., "Paper Chromatography," New York, Academic Press, 1952. Boissonnas, R. h.,Helz. Chiin. Acta, 33, 1966 (1950). Briiih. 31. K.. Routwell. R . B.. Barton. A . D.. and Heidelbereer. C . , Science, 113,4 (1951). Bull, H. B.. Hahn. J. W., and Baptist, V. H.. .J. .4m. Chrni. Soc., 71,550 (1949). Burma, D. P., . t h t u r e , 168,565 (1951). Cain. L.. and Berry, H. K., l-nic. of Texas PubI., No. 5109, 77 I
il95l).
Cassidy, H. G., "Adsorption and Chromatography," Technique of Organic Chemistry, Vol. V, pp. 312, 325-6, Sew York, Interscience Publishers. Inc., 1951. Consden. R., Gordon, A . H., and llartin, A. J. P., Biochenz. J . . 38,224 (1944). Datta, S. P., Dent, C. E.. and Harris, H., Scierice, 112, 621 (1950). Dent, C. E., Biochem. J . , 43,169 (1948). DeVay, J. E., Chang, W. H., and Hossfeld, R. L., J . Ani. Chem. Soc;, 73,4977 (1951). Draper, J., and Pollard, A. L., Science, 109, 448 (1949). Jirgensons, B., L'niv. of Texas Publ., No.5109, 56 (1951). Kowkabany, G. N., and Cassidy, H. G., ANAL. CHEM.,22,817 (1950). Ibid.,24, 643 (1952). Ma, R. M., and Fontaine, T. D., Science. 110, 232 (1949).
V O L U M E 26, NO. 10, O C T O B E R 1 9 5 4 (18) hlcFarren, E. F., ANAL. CHEM.,23, 168 (1951). (19) McFarren, E. F., and Mills, J. A., Ibid., 24, 650 (1952). (20) Miller, J. hl., and Rockland, L. R., Arch. Biochem. and Biophys., 40,416 (1952). (21) Muller, R. H., and Clegg, D. L., . b . i L . CHEM.,23, 403 (1951). (22) Pratt. J. J., and Auclair. J. L., Science, 108, 213 (1948). (23) Rockland, L. E., Blatt. J. L., and Dunn, &I.S., -bar.. CHEM., 23, 1142 (1951). (24) Rockland, L. B., and Dunn, M. S., Science, 109, 539 (1949). (25) Rockland, L. B., Underwood, J. C . , and Bearens, E. A, Calif. Citrograph, 35,490 (1950). (26) Shibatani, A., and Kukuda. 3I.,Bioche?n. J . ( J a p a n ) , 38, 181 (1951). (27) Strain, H. H., A i v . i L . CHEM.,22,41 (1950). (28) Toennies, G.. and Iiolb, J. L., Jbid., 23, 823 (1951).
I563 (29) Underwood. J. C., and Rockland, L. B., Food Research, 18, 17 (1953). (30) Wernimont, G., and Hopkinson, F. J., IND. END. CHEM.,d N A L . ED.,15,272 (1943). (31) White, J. W., and Maher, J., Arch. Biochenz. and Biophys., 42, 360 (1953).
RECKIVED for review June I;, 1953. Accellted June 7 , 1954, Presented before the Division of Analytical Cheinistry a t they123rd Meeting ofjthe AMERICAX CHEMICAL SOCIETY, Los Angeles. Calif.. March 1953. Mention of special instruments or materials throughout this paper does not imply that they are endorsed or recommended by the Department of Agriculture o r e r others of a similar nature not mentioned.
Study of Solvents Used in Adsorption Chromatography PATTERSON B. M O S E L E Y ' and A R T H U R L. L E R O S E N ~ Coates Chemical Laboratories, Louisiana State University, Baton Rouge, La.
J A C K K. C A R L T O N lnstitute of Science and Technology, University o f Arkansas, Fayetteville, Ark,
Several pure solvents, as wTell as binary and ternary solvent mixtures, were investigated as developers in adsorption chromatography. The R values of a number of organic compounds developed with the pure solvents and adsorbed on silicic acid and calcium hydroxide make available useful data regarding the developing power of these solvents. Two- and three-component solvent mixtures were used in developing o-nitroaniline on silicic acid. These studies offer data on solvent mixtures covering a wide range of developing power. Observations of previous workers on the lack of correlation between certain properties of the solvent and its developing power are substantiated.
N
LXEROUS studies have been reported in which an effort was
made to determine the role of solvent in adsorption processes. Patrick and Jones ( 6 ) in an attempt, t o correlate solubilit'y of the absorptive in the Folvent with adsorption affinity, found that as the absorption on silica gel from nitrobenzene, toluene. carbon tetrachloride, carbon disulfide. and kerosine was higher., the less soluble wap the adsorptive iii the solvent. Another investigator ( ? ) concluded that no valid correlation could be made between adsorptive solubility and atlsorption affinity. Considering the pclarity of the solvent, He>-maiin and Boye ( 2 ) suggested that, if adsorption depends 011 t,he dipole moment of the adsorptive, solvents having high dipole moments should compete more strongly with the adsorptive in adsorption. However, those authors ( 2 ) found no corrclatioii between dipole moment of the solvent and adsorption. After a s h d y of the adeorption of a number of adsorptivities from a variety of solvents on alumina. Jacques and Mathieu ( 4 j observed t,hat in a binary mixture the component having the higher dielectric constant was more strongly adsorbed; in binary mixtures, where one ?omponent was kept the same for all runs, the adsorption of the coniponent of higher dielectric constant increased as the difference in dielectric constant between the two increased; in ternarv mixtures, the component with the highest dielectric constant was most strongly adsorbed. In addition to dipolar attractions among solvent, adsorptive and adsorbent hydrogen bonding must be given consideration. 1
2
Present address. Hercules Powder Co., U'ilnlington. Del Deceased
Elder and Springer ( 1 ) . a~suniingthe structure of silica gel to yield the grouping, 0
--st
// \\
OH picture adsorption of fatty acids to occur i n this nlrtnner 0
\
HO
//
OH... . .O
Schroetler ( 8 ) has shown that certain derivatives of diphenylamiile and A-ethylaniline exhibit K value8 which can be explained on the basis of intramolecular hydrogel1 bond formation, or the lack of it, depending on the position of substitution of a nitro group. For example, 2-nitrotlipheiiylnmine is much less strongly adsorbed than ~-nitrodiphenylaiiiiiie,the difference being attributed essentially to the freeing of the nitro group in the 4position as compared to its limited availability in the 2-position where it is effectively bound through a hydrogen bridge t o the amino nitrogen. LeRosen and coworkers ( 5 ) have studied the effect of position in ortho and para substituted benzenes and have also found that decreased adsorption is observed where intramolecular hydrogen bonds are possible. Hoyer (3) has also investigated the significance of hydrogen bond formation in chromatographic processes. Each of the examples cited involves the study of the adsorption of certain organic molecules from an inert solvent. It should be remembered that, given a solvent such as methanol which is itself capable of hydrogen bonding to the adsorbent, competition offered by the solvent may be strong enough to prevent the ntlsorption of the organic compound being chromatographed. In this case an R value of 1.00 is measured. I n this ptudy a preliminary survej- was made of a number of pure solvents. This was followed by an extensive study of several two- and t,hree-component solvent mixtures of varying compoFiition employing o-nitroaniline as the substance to be chromatographed. It is hoped t>hatthe data made available by this study will aid ot'her researchers in selecting the appropriate solvent or solvent mixture for the separation in which they are interested.