V O L U M E 2 4 , NO. 4, A P R I L 1 9 5 2
643 LITERATURE CITED
C !I20 cin
-1
range are associated with the C-C-0
group rather
C than nith the - G O link. Nevertheless, the modifications imposed upon this vibrational mode (whatever it may be) by the rest of the molecule in alcohols, peroxides, and hydroperoxides alter its frequency sufficiently to differentiate among the three types in oxidation mixtures nhen the molecular weight is low. A t higher niolecular weights it is difficult to distinguish between the alcohol and peroxide. It is preferable to identify the individual coniponents of oxidation mixtures by fingerprinting with the spectra of the isolated compounds \Thenever possible. ACK\OW LEU t,aken. 2-Butanol. The li uid was refluxed 4 hours overopotassiun; hvdroxide pellets. T%e fraction boiling between 99 and 101 ! l o v e d by Developer, Ctn. 1 7 8.1 12.0 6.4
Effect of Temperature on RF. Tubes containing developer weie placed in the ice storage room overnight to come to tempeiature 1" t o 4". Spots of amino acid were placed a t 0.6 cm. u p standard strips from the developer-surface line. The developers used were made u p from collidine-lutidine, 1to 1by volume, M hich \vas diluted n-ith water (on a weight basis) so as to contain 61.1, 65.0, 70.0, 75.0, and 80.0% organic liquid. These xere compared with a similar set run a t 252-5.5'. The data are shown in Figure 4. The effect of temperature in this system is not nearly so large
ANALYTICAL CHEMISTRY
646 (where it exists a t the lower concentrations of organic phase, close to the critical solubility a t 25") as is t,he effect of concentration. Xcholas and Rimington (21 j found that t,he same R p was obtained (with free porphyrins) as 5' as a t 21" with lutidinr containing 40% water (volume/volume) even though the tvio liquids are completely miscible a t 5" while a t 21" the solution of this concentration is saturated m-ith water. Effect of "Equilibration." Strips were run in duplicate at 24" after equilibration for 1.5 hours and 6 days, respectively, with the 'vapor of the developer (collidine I1 saturated with water). The standard strips contained spots of valine a t 0.6, 4.0, 7.0, and 10.0 em up the strip from the solvent source. The data are Ph. shown in Table 11; siniilar resultsare found in the last Pr. two columns of Table I. It, n.ould appear that the A. I?, value as determined in 60 70 00 the standard manner is not % Collidinr- lutidinr greatly affected by long Figure 4. Effect of Temperequilibration, but that the ature on RF of Amino Acids change in RF with position in Aqueous Collidine-Lutiof the spot is lessened. dine hIuller and Clegg ( Z U ) RF plotted against per cent ( b y showed that the rateof rise weight) of collidine-lutidine (1 to 1 mixture b y volume) in water. X of liquid in filter paper values a t 1-4' values at 2525.5'. V, valink; OG, glycine: Ph. equilibrated with water Dhenylalanine; P r , proline: A . asvapor for various times departicacid pended 011 the time of exDosure to the saturated vapor up to about 30 minutes. Thereafter, although the paper continued to take up water slowl~for several days, the rate of rise of liquid was not much affected.
'
Table 11.
Effect of Eauilibration
Iliatance of Spots from Solvent Source. Cm 0.6 4.0 7.0 10.0
RF Values 33 0.38 29 0.26 24 0.17 17 0.11 Time of Equilibration -~ 6 days 1 . 5 hours Temperature, O C. 24 24 Time to Rise 11.4 Cm.. M i n , 222 203 0 0 0 0
Effect of Prewashing of Paper and Prefiltration of Developer. It has been shown ( 2 ) that when pure liquids, such as ordinary distilled water, are filtered through filter paper, the rate of flow in continuous filtration usually slows with time, often markedly. Prefiltration of the water gave the most consistent improvement in constancy of rate of filtration. It seemed desirable to determine whether such effects might be of importance here. I t also seemed desirable to determine the possible effects of the impurities, present in the bron-n discoloration which accumulates a t the solvent front in collidine development, on Rp values. Standard strips of paper were washed by development with collidine I saturated with y t e r , or n-ith water alone. These Other strips were not washed. were dried in the oven a t 80
.
To all of t,hese strips valine was a p lied a t 0.6, 5.0, 8.0, and 10.0 cm. above the developer line, and t l e strips were developed with collidine I saturated with water. Table 111 shows the results. In another experiment, standard st,rips were washed by development with collidine I saturated wit.h wat,er. Spots of valine were applied t o this paper which was developed with coliidine I eiituratrd with water, and the results compared with standard strips doveloped wit,h the same mixture and with the mixture d been filtered just before use. S o significant difference ilur.$ \vas obtained. The developer rose faster in t.he do as lied paper, an observation reported by Muller and Clegg (go), but prefiltering of the developer did not affect rate of rise. I n another experiment, a sheet of Whatman So. 1 filter paper was developed for 16 hours with collidine I saturated with water, by descending chromatography (6, 9) in a cabinet. The sheet was then nir dried and heated 20 minutes at 80" to 85". ' T h e solvent front had advanced 37 cm. in the machine direction of the paper. At, the solvent front were found the brown colorations commonll- referred to as "brown front." St.andard strips were cut from thia sheet' in the machine direction. One set from the brown front region with the stains at the narrow (lower) end of the strip; another set with the stains a t the wider end. Another set \vas cut from well behind t,he stains, where the paper had been washed by the develo er, and a fourth set from well in front of the browi region 14hel.e &e developer had not yet reached. Spots of valine were applied along the strips at 0.6, 4.0, 7.0, 9.0, and 10.5 cni. from the develo er surface. The strips, equilibrated 2.5 hours, were develope% with collidine I saturated with water, at 23-24". The strips were run in triplicate. There was no evidence that the washing or the accumulated discoloration affected the Rp values found for the valine spots applied a t the different distances from the developer surface.
Table 111. Distance of Spots from Developer Surface, Cm. 0.6 5.0 8.0 10.0
Effect of Prewashing of Paper
Collidine-washed 0.40 0 41 0.26 0.24 0.15 0.13 0.10 0.10
RF Value Water-washed 0 41 0.26 0.li 0.10
S o t prewashed 0.40 0 25 0.16 0.10
20
Time of Equilibration. LIiu. 0 20
144
Time for Sol\-ent t o Rire 1 2 C'rn.. Min. 140 143 170
0
Sorption by Paper. Standard strips of Whatman 30.1 paper were dried overnight a t 80" to 85". Each strip was transferred rapidly to a weighing bottle, which was stoppered, and weighed to 0.1 mg. Test tubes had been prepared as in the standard procedure. Each contained 0.5 ml. of water-saturated collidine 11, and had stood for 16 hours, so t h a t the atmosphere would be saturated with vapor. The strips were suspended in the tubes so as not to touch the walls of the tube, except a t the top edge of the strip xhere it was suspended, and not to touch the solution. A t intervals t,he strips were removed, quickly replaced in the weighing b o t t k , and neighed. The temperature lay between 23.0" and 23.8 Strips mere removed a t int,ervals of 1 to 46 hours.
.
The results are shown in Figure 5 , Each point is the average from duplicate strips, except the &hour point, where only one strip was usable. These results indicate that the sorption process continues for a fairly long time.. Muller and Clegg (2U) found that filter paper talres up \eater for several days, when suspended in the saturated vapor. I t is the general observation that the process of sorption of n-ater vapor and other liquids by cellulose is a slow one. In a preliminary experiment 0.5000-gram filter paper cuttings were immersed in 5.000 grains of mixtures of collidine I with water. The mixtures covered the concentration range: 0.188 to 0.793 "mole" fraction, calculated on the basis that the average molecular weight of collidine I Iyould be that for CsHIIS. The experiments were run a t 25" and a t 9.5-10". Analysis by refractive index measurement showed that water was preferentially sorbed over t,he entire range of mixtures, and that relatively more water {vas sorbed a t the lower concentrations of collidine.
641
V O i L U M E 24, NO. 4, A P R I L 1 9 5 2 ,41so, relatively less water was sorbed a t the lower than at the higher temperature. These results are consistent with the interpretation that the paper and the collidine compete for the wat,er. The more collidine is present, the more effective the competition. The solubility of water in collidine increases with decrease in temperatwe in these ranges (collidine evolves heat when mixed with water). Although adsorption tends to increase with derrease in temperature, the effect, if present in the interaction of the paper with the water, is masked by competition from the greater solubility of the water in the collidine a t the lower t,emperature. The differences were not very giwrt. Effect of Concentration in Binary Developers on RF Values. Six amino a(>idswere chosen to give a range of RF values, and to represent, acidic, neutral, alkaline, aliphat'ic, and aromatic types. These were: aspartic acid, glycine, histidine hydrochloride, phenylalanine, proline, and valine. The folloTTing binary mistures were studied: one component was always water; the others being collitline-lutidine (1 to 1, volume /volume); methyl, ethyl, n- and isopropyl and n-, iso, sec-, and tert-butyl alcohols; and diosane. ;Iqueous solutions were prepared in which the concentrations covered the range of mole fraction 0 t o 1. \Vith the Ixinially miscible liquids the range was restricted. K i t h colli-
dine-lutidine the concentr:itions were recorded in per c w t . Duplicate standard strips were run, without preliminary equilih a t i o n , and the temperatures of all the experiments fell l~etweeii 24.0 and 26.5, though in any given experiment the teniptmtuw did not change more than 1". On each strip a single spot (0.15 inicroliter) of solutio11 was placed. The amino acids were run i n piirs: glycine and valine, histidine hydrochloride and phenylal:triincL, and aspartic. : i d anti proline, except in the n-butyl :il(*ohol mixture, whew t t?e amino acids were run in three's: histidine hydrochlol~idr~,I)henylalanine, and valine: aspartic acid, glycine, and prolinr. The results are shown in Figures 6 and T but the points for aqiartir acid are omitted hecaause they ivould unduly congest tlici fiyui,c*s. 5
t
.7
9 IO
IQ
%
1 Figure 7 . Effect of Concentration in Binary Developers o n RrValues
Developers. 7, n-butyl: 8,,isoh,ityl: 4, sec-butyl alcohol; 10, collidinelutidine 1 t o 1 h y v o l u m e .%mino acida 0 , glycine: X , 1-aline: Ophistidine; 0,phenylalanine; A proline
Y
R i t h mkcible pairs ol liquids, in gt~iic~al all Rp values increased \yit,h increasing water content,, tending t,oward 1.0 in pure \yater. The esceptions are phcnylalalline (0.96) and histidine (ca. 0.65), TIME OF EXPOSURE IN H R S . the lower RF values of which iiitlicat,e probable adsorption on the Figure 5. Sorption by Paper from Vapors of WaterSaturated Collidine I1 filter paper. I n general, histidine shows the lowest RF under 7 gain in weiaht, G . ralculated on basis of d r y paper, plotted against nln~ostall conditions, but i t s the zone of this amino acid tends time of exposure, in hours to sti.t.nl; badly with high \\-ater content it was difficult to determine the RF value. The R p values decreased with decreme in water cont,ent, and in all ewept I. 0 methyl and ethyl alcohols. the Rp values tended 8 to converge and hecome 0.0 in the pure organic liquid. The KI.values of aspartic acid are not .6 .4 shown. Usually they were lowest at low n-ater content, and highest a t high water contrnt, thus .2 coming closr to the l i p of each of the other amino RF acids a t sonic intcrmetiiate concentration of devel.8 oper. I n gc~ncr:tl,the 1,elittive order of the amino acids rein:+incd the saine throughout, though there .6 seemed t o h r :i few raws of crossing of the RF zs. .4 concentrnt~ioncurves. Thew curves became niore .2 curved the greater the number of carbon atoms in the homologous alcohols. I n general, for a given amino acid and R given 111ole fraction of wat,cr, say 0.5, the magnitudes of the RF values fall in the order: methyl>ethgl>n-propyl = isopropyl (escept for phenylalanine) >tert-butyl alcohols. I n the mixtures of partially miscible liquids there was no change in order of R F values with concentration. I n all these nlistures the order of RF values at any given concentration was phenylalanine > Figure 6. Effect of Concentration i n Binary Developers on RF Values valine >proline >glycine>histidine>aspartic acid, RF values are plotted against mole fraction of organic coinponent of aqueou8 developer, except in 10, plotted i n weight 70 except in the collidine-lutidine system, lvhere Developers. 1 , methanol: 2, e t h y l alcohol: 3 , n-propyl; 4, isopropyl; 5,, tert-hutyl; 6, histidine>glycine over part of the rangc. In dioxane. Amino acids 0 , glycine: X, 7.aline; 0, histidine: 2 phenylalanine; A proline
0
I 5
I
IO
I
15
I
20
I
42
I
46
J
1
2
5
ANALYTICAL CHEMISTRY
648 soine cases the RF values of the last three amino acids were very close together. There was no apparent correlation observed between the shapes of the RF cs. concentration (mole fraction) curves and the curves for refractive index w. mole fraction, for surface tension 2's. mole fraction, viscosity us. niole fraction, or solubility of amino acid us. mole fraction where these could be compared. If the miscible alcohols were compared as developers at mole fraction 0.5, say, the order of magnitude of K p value! was niethyl>ethyl>npropyl = isopropyl > tt.rt-l,utyl aclohols. The order of magnitudes of the dielectric constants of the aqueous mixtures of this mole fraction is methyl >ethj.l >n-propyl = isopropyl > t e r t butyl alcohols. Howevei,, at 0.5 mole fraction, diorane-water shows higher RF values than tert-butyl alcohol, but has a lower dielectric constant. \Then the ratios of RF values of glycine to valine are plotted against concentration of developer, almostd itlentical curves \\-ere obtained for ethyl, n- and isopropyl, ant1 trrt-butyl alcoholr and dioxane. The curve for methanol \vas different. JVhen the ratio: of K F values of valine to phenylalanine \\ ei'c plotted, different curves were obtained with each developer. If the difference in K p values between valine and glj.cine was plotted against concentration of developer, the curves for the group of developers ahove were ve1.y similar i'l,oni 0 t o about 40% of solveiit, when the curves diverged. The methanol curve was different over the entiw i'ange of concentration. The shape of the aspartic acid zone changed with the witor content of the developer. \t-ith a low water content there \vas :I tendency to streak in the dirction of flow of developer. JVith increased water content! the zone became rounded, and a t high water content, it hecame eloilgated in a direction perpendicular t,o the direction of flow. The cause of this is not known. It ma). depend on changes in degree of ionization ( I S , 15jand soluliil- , it)- (aspartic acid i.s relatively insoluble in water). It, may also have some relation t o the ability of filter paper to eschangr ca t ioiis. DIscL~ssIo~
The course oi development in paper chroniat'ography may lie visualized a t least provisionally in the following terms: A zone of the misture t o h a separat.ed is placed as a spot or streak on t h e paper in a position which is t o be some (relat,ively small) distance a m y from the surface of the developer. The edge of the paper i 5 dipped into the developer, or, alternatively, t h e pa er is first exposed t o the vapors of the developer (equilibrat,edf, and then the edge near tmhezone is dipped into the developer. T h e developer passes by capillary action along t,he paper past the zone of mixture. T h e rate of movement of t h e liquid is governed at least by its viscosity, surface tension, anti tleiisity, and by the porous structure of the paper. h sinall elenient of liquid, say :it the advancing edge of the de\-eloper, conies into contact with the zone of niisture :titer rising froiii the surface of the developer to the region of application of the zone. T h e liquid picks up the sulistanre(x) in its passage over the zone, possibly ~,eachingsaturation, if the zone lie a large oiie. When the elenieiit of solution reaches the: : f edge of the zone it conies into c,ont:ict \vith a region of paper empty" n 3 h respect to yolute. Here, in conformity with Le C'hatelier's principle, solute passes out of t h e developer. The small element, of developer moving i'or\vaid progressively loses solute until after a very small distance kjeyond the edge of t h e initial zone it becomes virtually empty of solute. Thus t h e developer operates as it moving vehicle whivh traiis1inrts solute from the rear to the front of a zone. T h e nest folloa-ing