175
ON SOME ASPECTS OF ADSORPTION PHENOMENA, WITH ESPECIAL REFERENCE TO THE ACTION OF ELECTROLYTES AND TO THE ASH-CONSTITUENTS OF PROTEINS By W. M. BAYLISS, D.Sc., F.R.S., Assistant Professor of Physiolog-y i University College, London
From the Physiological Laboratory o} University College, London
(Received March 7th, I9o6) CONTENTS PAGE
I. The General Law of Adsorption and the Nature of the Union between . Protein and Ash . . II. The Washing-out of Electrolytes from Proteinis . . . III. The Kinetics of the Adsorption Process . . . . . IV. The Action of Electrolytes . V. Conditions under which Adsorption Compoundls become Dissociated . . . . . VI. Specific Adsorption . VII. The Application of the Facts of Adsorption to Various Processes VIII. Summary of Results . Table of Properties of Dyes
I.
.
ADSORPTION UNION BETWEEN PROTEIN
THE GENERAL LAW
OF
. .
.
. .
. .
AND THE
AND
'75 179
185 '95 209 212 214 227
230
NATURE
OF TfHE
ASH
The great difficulty of obtaining preparations of protein bodies free from ash-constituents is well-ktnown to all observers who have had occasioni to make the attempt to do so. It is, indeed, a matter of doubt whether a true protein absolutely free from iniorganic electrolytes has hitherto beeni in the hands of investigators.' It is easy by dialysis, or other means, to remove a portioni of these constituents, but beyond a certain limit it appears to be a matter of extremely proloniged labour to get rid of the remaininig small amount. i. See Die
Gr-sse des Eiweissmolekuls, by Fr.
N. Schulz, 1903, p.
9.
BIO-CHEMICAL JOURNAL
176
One is therefore justified in some degree in distinguishing with Schulz' between the 'essential' and 'accidenital' ash-constituents of proteins. The discussion as to the state of combination of these constituents in the proteini molecule concerns in fact only the former, that part namely which it appears difficult or impossible to remove without destroyinig the character of the protein itself. Now of recenit years a considerable amount of research has been unidertaken with regard to the niature of the pheniomena kniown by the niame of 'absorption' or 'adsorption '. Too much space would be taken up in givinlg a complete list of this work, and I must conitent myself with menitioninlg the niames of Van Bemmelen, Biltz, Schmidt, Walker and Appleyard, and Craw. The maini result of these investigationis, so far as conicerns us for the presenit purpose, is the discovery of what we may call the 'Law of Adsorption.' Suppose we have a series of solutionis of such a dye as Conigo Red, for example, in progressively diminishing concenltratioll, anid in each of these we place the same amount of filter-paper, we filnd that a part of the dye is taken up by the paper and in relatively larger proportion the more dilute the solution. To take some examples from the paper by G. C. Schmidt.3 Picric acid and cellulose: Concentration of.picric acid
Amount adsorbed
o889
...
o0340
...
0111o O°°95
Eosin and silk Concentration of eosin
o666 O0376
Amount adsorbed
... o-84 ...0-75
Iodine and carbon: Amount a(dsorbed
Concentration of iodine
O-i6i1
o-o87
...
...
3 275
2-958
I. Loc. cit., p. I I. 2. Notwithstanding the suggestion of.Zsigmondy (Zur Erkenntniss der Kolloide, Jena, 1905, p. 6i, footnote) to make uise of the name ' absorption' in honor of Van Benmmelen, the thief worker on the question who uses this form, I think that, in English at all events, ' adsorption' is less likely to lead to confusion, and ' absorption ' can be still used for phenomena like the solution of gases in water. 3. Zeitsch. f. Physik. Chtmie., XV, p. 6o. I894.
ASPECTS OF ADSORPTION PHENOMItNA
If t
These numbers show that there is some kind of 'affinity' between the bodies adsorbed and those which take them up. Put in other words the law states that there is not a proportionality between the concenitrationi of the solution or partial pressure of the solute anid the amount adsorbed. Ostwald' inideed points out that there is no hard and fast distinctioni to be drawn hetweetn chemical affinity and adsorption which lattter he sometimes speaks of as 'mechaniical affinity.' He calls attenltion to the complete series of transitionis between the two pheniomenia, and, in referring to Van Bemmelen's researches, he remarks that they frequenitly leave it a matter of unicertainty as to whether the facts brought forward should be regarded as of a chemical or physical niature. In any case all the evidence shows that electrolytes when adsorbed are noni-ionized and no loInger take part in the electrical conductivity of the solutionl. This circumstance renders the determiniation of coniductivities a rapid and cotnveniienit method of inivestigatinig certain aspects of the problem as will be seenl later. In the course of a series of experiments oni the chaniges of conductivity in protein solutions brought about by the action of elnzymes I have had occasioni to take into conisideration the possibility of the splitting-off of inorganiic electrolytes unider these conditions. The results of some preliminiary experiments suggested to me that the most satisfactory explanationi of the niature of the 'combinationi' between proteini anid ash was that the constituenits of the latter are present in adsorbed form rather thani in true chemical union. Before passing oni to the description of the experimenits made to throw light oni this problem, 1 may, in order to make clear the poinit of view, give the results of a preliminary experimenit on ani unidoubted case of adsorption, viz., congo-red and filter-paper. Experiment: Seven flasks were taken, each containinig 50 c.c. of a solutioni of conigo-red in IO per cent. alcohol. These solutions were of a regularly diminishinig concentrationi from 0o014 to 0002 per celnt. Congo-red forming a colloidal solution is very readily precipitated1 from watery solution by traces of electrolytes, so that ani alcoholi t. Lehrbuch d. Allgem. Chemie.,
.te. Aufl. Bd. 1, p. Io84,
et seqq.
1310-CHEMICAL JOURNAL
i79
solution is more easily worked with. For the same reasons, the filterpaper used was Schleicher anid Schill's analytical paper, and of this a disc I2S5 cm. diiameter was added to each of the above flasks. After 24 hours stanlding the depth of colour in each solutioni was determinied by Gallenkamp-Heele's colorimeter, the results beinig as follows solution
Proportion of dye in solution
Proportion of dye in paper
014
40 4%
6o %
0'0I2
20 %
800
0010
9 3 %
90°7 %
oooo8
4% I3 %
96 % 98 7 % practically all
Concentration of
o-oo6 o C04
trace
0 002
The figures show at once that the amount of dye takeni out of solution by the paper is not directly proportionial to the concentration of the solution ; nor, on the other hanid, is it inidepenident of this conicenitration as would be the case if a true chemical compound were formed. In fact, from a 0014 per cent. solution 40 X =2l4- 00028 gram is taken Up by a certaini amount of paper, whereas from a ooo8 per cenit. solutioni only 4 X '0°'4 - 0I00208 gram, or onie-tenth of the former, is taken up by the same amount of paper, although there was still a quantity of free conigo-red in the solutionI. The peculiarity of these phenomenia, which I have spokeni of above as ' the law of adsorption,' is well showtn by the series of numbers in the third column. Now, if we put these data inito the form of a curve as is done in Fig. i, where the ordinates represent the percenitage of dye left in solution and the abscissae the original conicentrationi of the solution, we see that the curve forms part of a hyperbola, anid onily approaches the axis (i.e., zero concenitration) asymptotically.' In other words, however diluted the original solutioni may be, there will practically be always a certain amount of the dye left unadsorbed by the paper. It follows I. This hyperbolic form of the curve is pointed out by Ostwald, Al/gemein. Chemic., Bd. 1, p. I,o96. Walker and Appleyard (Jour. Chem. Soc., 69, i896), p. 1334, also find a logarithmic formula for the adsorption of picric acid by silk.
ASPECTS OF ADSORPTION -PHENOMENA
1-79
-----------------
3o
251 2 o0 /5 /0
0 0,1A4
IO/
0/°
*006
'°° 8
FIG.
*004
°°2
I
also that if we attempt to wash out the dye by meanis of io per cent. alcohol it will require, theoretically, all infinite number of changes of the solvent to effect complete removal of the dye. I found, in fact, that if the piece of paper which was stained deep-red in the 0'002 per cenit. solution was allowed to soak in i o per cent. alcohol, a trace of dye onlly was removed each time a fresh supply of alcohol was added ; the alcohol became very fainitly pink, the colour only to be detected in a deep layer, while the paper did not perceptibly lose in depth of colour after several extractiolls. II. WASHING-OUT OF ELECTROLYTES
it will be seen, on a little considerationi, that the great difficulty of removing the last small part of the ash-constituents of proteinsmay be readily explained on the hypothesis that the case is one of adsorption. The hyperbolic form of the characteristic curve shows
180
1310-CHEMICAL JOURNAL
that, as the asymptote is approached, a less and less percenitage ot these constituents will be taken away by repeated changes of the extracting solvent. In order to test this hypothesis, I have taken commercial gelatin and extracted it with repeated changes of distilled water at room temperature. The comparative amounts of electrolyte removed by each change were estimated by taking the electrical conductivity. Gelatin at room temperature not being dissolved by water this latter can be poured off and replaced by fresh with ease. Of course, the water imbibed by the gelatin in swelling cannot be changed ; this does not affect the result -to any serious degree, merely making the process of extraction a more prolonged one. The experiment was performed as follows :-IO grams of Coignct's gold-label gelatin were allowed to soak in distilled water unitil swollen anid sufficient water added to enIable ioo c.c. to be poured off. About i8o c.c. were nieeded in all. A little toluol was added, and the mixture left to stand for twetntyfour hours. The watcr was then poured off and its conductivity determined. A fresh IOO c.c. of water were then added, and the process repeated every twenty-four hours. After the first three or four extractions the conductivity of the extract was so low that I found, with the apparatus used, that more accurate readings were obtained by concentrating the fluid in a platinum capsule over the water-bath to a definite volume (I2 c.c.) before measurinig its conductivity. Even when their conccntrationi is thus inicreased, the electrolytes may be looked uponI as practically completely dissociated, so that no appreciable error can be put down to this circumstance. It might be objected that impurities in the distilled water would also be concenltrated, but I found on testinig this possibility that the conductivity of the distilled water used, which had originally a conductivity of 5 gemmhos (= reciprocal megohms) did not increase by concentratinig go c.c. down to I2 c.c. Presumably, this is to bc accounted for by the fact that the slight conductivity was almost entirely caused by carbon dioxide, which would be driven off on heating. The numbers obtained were as follows:-
ASPECTS OF ADSORPTION PHENOMENA No. of extract
Conductivity in gemmhos
No. of extract
18I
Conductivity in gemmhos 23'25 I8z25
1
1094
...
7
2
4755
...
8
3
2323
...
9
4
1204
...
10
15 94 I 545
5 6
62v8 29.41
...
II
I5.40
We sce at once that after a certain time each further chatnge of water conitracts only an infiniitesimal amount of electrolytes. Since the conductivity of the distilled water was 5 gemmhos, this must be subtracted from the values giveni in the table, leaving only io gemmhos for the electrolytes washed out from the gelatin. These electrolytes conlsist inl all probability chiefly of calcium sulphate, and a calcium sulphate solutioni of 0o25 per cent. has a conductivity of 2039 gemmhos, so that, nieglectinig differences of dissociation, a solution with a conductivity of Io gemmhos will conitaini only 0o25 X -L3-= O-OOI3 per cent. It may be said that at this stage the ash was nearly all washed out. In order to test this possibility, I determined the ash of the gelatin. The incinieration was performed in a platinum capsule, care being taken inot to let the temperature get above dull redness, and niot to prolong the heating after all charred portions had lost their blackness. The original gelatin contained o 55 per cent. of ash, atnd after extracting with water eleven times it still contained oo36 per cenit. Therefore, when the rate of extraction had already become almost negligible, there was still left in the gelatin a considerable amount of ash. On the hypothesis that the electrolytes are merely mixed with the gelatitn anid washed out by diffusion, the rate of diminution of conductivity of the successive extracts would have been much greater. This can be seen by calculating on the basis of the amount washed out by the first and second extractions, what would have been, e.g., the conductivity at the seventh extraction. The proportion removed at each chanige would, obviously, have been the same. If the above calculation be made it will be founid that the conductivity of extract No. 7 would have been 7-4 gemmhos inistead of 23'25 gemmhos as found,
182182I
BIO-.CHEM[CAL JOURNAL
Fig. 2 gives the curve of this experiment showing the same hyperbolic form as the adsorption curve of conigo-red, Fig. i.
00
J5>.f
6
T FIG. 2
8
9,!.
21°
7r-7
Since the last values are the unost initeresting, and if the whole had been plotted out, the final course of the curve would have been oni a much smaller scale than in the figure, the first two values are omitted, the curve commencinig at the third. The ordinates represent specific conductivities of the extract in gemmhos, the abscissae are at equal initervals. The value No. 6 is obviously an experimental error. Iu dlrawing the curve the conductivity of the distilled water used is takeni as the axis of abscissae. The positioln of the other asymptote is not showni since onily part of the hyperbola has been drawn. Before leavinig this experiment I may mentioni that an attempt was made to re-dissolve the ash and determine its conductivity. A quantity of the gelatin after the last extraction was dried in a toluol oven and weighed o075 grams. This was incinerated and the ash redissolved in the quantity of water correspondinig to that in whijh th$
ASPECTS OF ADSORPTION PHENOMENA
183
gelatin had soaked, viz., I 35 c.c. It dia niot, however, enitirely go inito solution ; yet notwithstanditng this, the cond(uctivity was 29 gemmhos as compared with the IO gemmhos of the water in which the gelatin had beetn soaked. The signification of this fact is that it confirms the deduction drawn from the previous experimenit that electrolytes are held by the gelatin in such a maniner that they can only be separated by an enormous number of changes of water. If gelatin which has been washed free from the greater part of its electrolytes be placed in a dilute potassium chloride solution, it takes up a certain quantity of this electrolyte, which call be washed out by repeated changes of water, and the curve of the coniductivities of the series of extracts is of exactly the same form as that of Fig. 2. The fact mentionied in the last paragraph makes it of initerest to see whether direct evidenice can be obtained of the adsorption of electrolytes by gelatin. This caIn be done in the followinig way Experiment. 30 c.c. of NO KCI were foutnd to possess a conductivity of i 6,6oo gemmhos at 39 70 C. At room temperature 2'3 grams of 'washed' gelatini were placed therein and allowed to remain for 24 hours. The coniductivity of the solutioni was niow found to be diminished to 14,590 gemmhos at 39.7°. This means that at least 12 per cent. of the potassium chloride had been takeni out of solution. The fact can also be shown in other ways though in not so obvious a mannier. If we take distilled water and add successively equal amounits of Kci, each increment produces a slightly less increase of conductivity than the previous one. The reason is that as the solutionl becomes more cotncentrated the KC1 becomes less dissociated ; to a less degree also increased iotnic frictioni will conitribute to the result. If, oll the other hatnd, we add to the water first some 5 per cenit. or so of gelatin and afterwards the successive doses of KCI the result is differenit. In a particular experiment IO grams of gelatin were placed in 200 c.c. of distilled water, the gelatin having beeni previously washed fairly free from electrolytes. The conductivity of the water was raised N to 26-9 gemmhos. When 2 c.c. IO KCI were added and allowed to stanid for 24 hours the conductivity was increased by 200 gemmhos,
I84
BIO-CHEMICAL JOURNAL
Another 2 c.c. were added, atid after 24 hours the coniductivity was foutnd increased by 208 gemmhios. That is, as I interpret it, a part of the first dose of KCI was taken up by the gelatin. It is impossible to compare directly these figures with those obtainied on a similar experiment with distilled water without gelatin, sinice the proportion of water takenl up in the swellinig of the gelatin is unknowtn. Again, if gelatin has the power of taking up electrolytes, it is to be expected that commercial gelatin and ' washed ' gelatin would show a difference in their power of takilng up more, since commercial gelatini has already a larger percenitage. The following experiment was made to find out if this is so: Experiment: I2 grams (air-dry) of commercial and 'washed' gelatin were soaked for 20 hours in distilled water. The conductivity of the water was then found to be Commercial gelatin: 553 gemmhos. ,, Washed : 7-7 gemmhos.
Two solutionis (very dilute) of calcium sulphate and phosphate were made equal in conductivity to each of these extracts by adding, drop by drop,.a saturated solution of these two substances in distille(d water. Of these solutionis a volume was taken equal to that of the water in two flasks in contact with the two kinds of gelatitn, this was found to be (by weighing) 176 grams. Theni to each of- the four flasks i c.c. of the above-mentiotned saturated solution of calciumn sulphate anid phosphate were added, anid the increase of conductivity so produced was determined. If a part of the added electrolytes were takein up by the gelatin there should be a greater rise of conductivity in the cases of the watery solutions thani in those where gelatini was present. The results actually obtainied are rather difficult to interpret. To take first the. ' washed ' gelatini and the watery solution equal to it in conductivity. The rise of conductivity in the latter by the additioni of I5 c.c. of the calcium sulphate solutioni was I63 gemmhos at I00. Where the gelatin was present thie first effect was a rise of 238 gemmhos, owilng, no doubt, to the electrolytes not at onice diffusinlg into the water contained in the swolleni gelatin, this steadily diminished, and in about forty-five minutes had become 128 gemmhos. But then
ASPECTS OF ADSORPTION PHENOMENA
I85
the coniductivity commeniced to go up again, finally becomiing i 6o gemmhos. If, therefore, we take this last as the true value, the amounit of electrolyte takeni up by gelatin was very small, a difference of 3 gemmhos only betweenl the two cases. If, oni the other hand, we take the lowest value to which the coniductivity fell, the difference is conisiderable, viz., 75 gemmhos. IIn the case of the commercial gelatin anid its eqtuivalent calcium sulphate solutioin, the rise produced by the additioni of IS c.c. of saturated calcium sulphate anid phosphate was 68 gemmhos in *the watery solutioin anld 56 gemmhos in that where the gelatin was present, a differenice of I2 gemmhos. The additioni of the electrolyte at first caused a rise of io6 gemmlhos where the gelatini was present, it fell to 72 in one hour, but the final value was niot reached for several hours, as opposed to the first case, where it was reached in two-anid-a-half hours. This experiment affords a certaini amounit of evidence that electrolytes are takeni up by gelatin, but does Inot distinguish betweeni adsorption and chemical combinationi. As we shall see later, the effect of heat shows it to be, in all probability, a case of the former. III.
THE KINETICS
OF
ADSORPTION
Velocity of Reaction. Very little informationi is to be obtainied as to the rate at which bodies are takeni up by adsorbing substanices. A few experimenits that I have made on this question may therefore be of interest. i.
Experiment: 50 c.c. of 20 per cenit. alcohol in test-tube lined with filter-paper. In thermostat at 39O. Wheni warmed, i c.c. warmed o 5 per cent. congo-red added, stirred, conductivity takeni at intervals by pipette electrodes. The curve (Fig. 3) shows the timne course of the diminution of conductivity. It will be noted that the curve is hyperbolic anid that it arrives niearly at the asymptote in io minutes. At this temperature the velocity is considerable. The first part of the curve in all probability escaped observationi, ow;ing to the great initial velocity of the reaction as shown by the shape of the curve.
t86
6BIO-CHEMXICAL
JOUTRNAL
FIG. 3
Iln aniothier experimenit, in which the proportion of paper to dye was greater thani inl the previous onie, equilibrium was reached in six minutes. So that the rate of chanige is proportional in the first place to the conicenitrationi of the adsorbinig body. I have niot enough data, however, to warratnt aniy further discussioni of this poinit. The reactioni is onie takinig place in a heterogeneous system anid its mathematical expressioni a complex one. It is probable that diffusion plays a considerable part as indicated by the results of the niext experiment, which was made to obtaini some idea of the temperature coefficient of the reactioni-velocity. It was founid that, for some reason or otherperhaps some action of the platiniized electrodes oni the dye-at low temperatures the electrical conductivity method did not give satisfactory results, so that the colorimetric method, though more laborious and less senisitive, was made use of. Experiment: 50 c.c. of o oo5 per cenit. conigo-red in' 5 per cent. alcohol ini each of two flasks, a circle of filter-paper added to each, and
ASPECTS OF ADSORPTION PHENOMENA
I87
onie kept at 50° C, the other at io° C. After a certain time the solution was poured off and a new piece of paper and fresh dye solution, previously warmed, or cooled respectively, added and a differetnt time of actioni allowed. The flasks were thoroughly shakeni at frequent intervals to allow access of the solutioni to the whole surface of the paper. This shaking caused the separationi of bits of paper which renidered it niecessary to allow the solutionlI to deposit before makinig colorimeter observatiols ; neglect of the precaution caused the loss of several of the earlier observations. Filtrationi is naturally iniadmissible. The results were as follows, the colorimeter readings expressing the relative percenitage of dye left in solutioni, the original solution IOO0.
Colorimeter reading. Duration of action 5 8 20 33
minutes
200
zOo
95
circ. 98
,,
90
...
,,
85
...
,,
4°0 I00
500
,, ,,
5 hours 7 ,, 14
,,
24
,,
...
90
8o
...
77
70
...
6o
75 * 75 ...
FIG. 4
5 48 45
I88
BIO-CHEMICAL JOURNAL
The curves (Fig. 4) reproduce these two series up to seven hours. What we notice is that at the higher temperature the velocity is greater until the equilibrium position at this temperature is approached. The question of equilibrium will be discussed in the next section. In estimatinig the value of the temperature coefficient of the velocity, some uncertainty exists, since it will differ according to what stage of the reactioni we take. If we take as criterion the time takeni to reduce the concenitrationi of the solution to a given percentage, say go per cent. at io°, this time is about 2-5 times that at 500, if we take 8o per cent. it is about 1-3 times. In either case, however, it is extraordinarily low ; taking the higher value at Ioo it takes twenty-five minutes to become go per cent., and at 50° it takes teni minutes, a difference of fifteen minutes for 400 ; so that, assuming uniformity of inicrease for the total interval, the time taken at 40° would be only IS -3-6 minutes more than at 5Q0, or ten miniutes anid I3-6 miniutes respectively. This gives a temperature coefficient of the extremely low value of 1366- I-36. In the table giveni by Van t' Hoff' there are otnly two values below this, viz., the dissociationi of PH3 and AsH3 probably depending on the high temperatures at which the observations were taken, since, as Van t' Hoff poinits out, the velocity-ratios for ioo usually diminish as the temperature rises. The value found in the presenit case inidicates that the theory of Nernst2 as to the part played by diffusion-processes inl heterogeneous reactions applies herc ; the correspondinig value found by Brunner3 for the dissolutioni of benzoic acid in water is, in fact, I A few observations on the effect of alcohol and certain electrolytes oni the velocity of reaction may be shortly referred to. Conigored is the sodium salt of a substituted aromatic sulphoniic acid, anld is less soluble in alcohol thani in water, so that it seems probable that more dye would be taken up from a solution containiing alcohol than from a watery solution of the same concentration. This was found T. Vorles. fiber Theor. und Physik. Chemie., 2te Aufl. p. 2. Zeitsch.f. Physik. Chemie., XLVII, I904, p. 52.
3. Zeitsch.f. Physik. Chemie., XLVII, I904, p. 62.
zz5.
ASPECTS OF ADSORPTION PHENOMENA
.I 89'
to be, the case. Two solutions containing o oos per cent. of conigored were taken, one in 50 per cent. alcohol, the-other in water. From the former a circle of filter paper adsorbed at 140 in seventy minutes 45 per cent. of its conitenits, and from the watery solution only 30 per cent. The actioni of a neutral electrolyte, sodium chloride, was similar, but more pronounced. From a solution containiing o00042 per cent. dye and 2 per cent. NaCI, 75 per cent. was taken up in fifty minutes, and from a solution of the same concentrationi in dye, but without NaCL, Ollly some 25 per cent. was taken up at the same time. The action of strong acids is, of course, to set free the sulphoniic acid from its compounid, the free acid while more soluble in alcohol is less so in water, so that again we have a favouring effect oIn adsorption as follows :-Two flasks of the usual 0o005 per cent. solution of dye with each one circle of paper to 50 c.c. solution were prepared, one remained neutral, the other was acidified by the addition of one drop of 5 per cetnt. HCI. Though the solution became blue Ino precipitation of the colour-acid oc curred in this weak concentrationi. After two hours at I3° to 140, 35 per cent. was taken up from the acid solutioni and I2 per cent. from the neutral one. Before taking the colorimeter reading the acid solution was neutralized by a drop of ammonia. It ap pears, then, that bodies which tend to diminish the solubility of conigo-red favour its adsorption by the substances immersed in its
solutioln. The experiment with gelatin and calcium sulphate, previously rela ted, shows that the rate of adsorption in that case is' considerably greater than that of congo-red by paper, as would be expected from the greater mobility of the ions concerned. At ii° equilibrium was attained in less than two-ancd-a-half hours, whereas at the same temperature at least twenty-four hours was required for congo-red anid paper. 2. Equilibrium The most interesting fact under this head is the influence of temperatnre. As the curve (Fig. 4) shows, at the higher temperature,
BIO-CHEMICAL JOURNAL
Igo
less of the dye is contained in the paper and more in the solution. The curve (Fig. 5) shows the results of a few observations at various temperatures, ordinates being percentages in paper and abscissae temperatures. It will be noted that a straight line is formed. The values at the lower 'temperature are not quite so accurate as at the higher,
-s
S-
0
b
-
2o
Lr
3o
3f
40
4j-
Jo
FIG. 5
sinice the former were obtained by allowing the flasks to stand in the laboratory, the temperature of which was not constant within 2° or 30 during the time necessary for equilibrium to establish itself. The upper values are accurate, beinig taken in a thermostat. Attention has already beeni directed to the letngth of time necessary at low temperatures for the establishmetnt of equilibrium in these dilute solutions. Ostwald also notes this fact.' The dissociation of adsorption compounds by raising the temperature can also be shown in the case of gelatin and inorganic electrolytes, though complicated in this instance, by the passage from the state of hydrogel to that of hydrosol. There is, however, as I finid, no sudden change in passing from the one state to the other as regards conductivity. If one warms a gelatin hydrogel from I5° to 40o determining I, Loc. cit., p. 1093.
ASPECTS OF ADSORPTION PHENOMENA
I 9I
conidLuctivity at frequent initervals, the curve, which is in fact a straight linae, shows no kink of aniy kinid at the poilnt where the hydrogel liquefies to a hydrosol. The experimenit previously described on the relationis of gelatini to calcium sulphate was continiued as follows :-The two flasks containinlg gelatini aind calcium sulphate were warmed to 530, the rise of conductivity in the case of the washed gelatin was 397 2 geinmhos, ill the commercial gelatin 1317 gemmhos. Now the point of initerest is, whether this rise is greater than would be the case if nio separation of additionlal electrolyte had takeni place under the inifluenice of heat. To decide this, solutioni of calcium sulphate was prepared, having at 100 the same coniductivity as the solutions in contact with the two kinids of gelatin. 'Fhese were then warmed to the same temperature as the gelatini. The rise of conductivity in the more dilute amounited to 376 gemmhos, anid in that of the stroniger to 1446 genimhos. There is therefore anl inicrease in electrolytes in the case of the washed gelatini corresponidinig to a differenice in conductivity of 2I gemmhos. It is to be remembered, -moreover, that gelatini in the state of hydrosol has a slight effect similar to that of other nion-electrolytes in diminishing the coniductivity of solutionis in whlich it is presenit. This is, no doubt, due to an increase of itnternial friction in some way or other, and amounts, according to determinationi I have made, to 3-4 per cenit. diminution of coniductivity for each i per cetnt. of the gelatin presenit. So that each of the values for the gelatitn solutions should be inicreased by, approximately, 3-4 x 4 9 per cenit., the gelatin solution being 4+9 per cenit. in concenltration. This increases the two values to 463 and I536 gemmhos. The differences in favour of the gelatin come out niow to be 87 gemmhos for the washed, anid go gemmhos for the commercial s-ample. The closeniess of these two numbers serves to
strenigthen the view that the elecrolytes causing the increase were in some way separated off from the gelatin under the influence of heat. It has beeni shown above that the conlductivity of a solution of the ash from a givenl sample of gelatin is greater than that of the solution with which the gelatine is in equilibrium. I thought it interestinig, therefore, to compare the aslh of the particular ' washcdc'
BIO-CHEMICAL JOURNAL
192
gelatine of the above experiment, as regards its conductivity, when re-dissolved to original volume, with the conductivity of the gelatin solution at I0° and at 530. The ash from I00 c.c. of the gelatin solutioni weighed 0o0262 gram, it was re-dissolved in 96-6 c.c. of distilled water (to allow for the volume of the gelatin), the whole of it not going into solutioni. Now the conductivity of the gelatin solution was found at 6.20 C. to be I4I gemmhos, that of the redissolved ash was 270 gemmhos. It is obvious that this is more than sufficient to account for the additional electrolytes split off on heatilng. It also confirms the view already expressed that inorganic electrolytes are held in some kind of combination by gelatini. Since they are also separated to a certain degree by warming, the combination in question is in all probability of the nature of adsorption. The various substances whieh increase the velocity of adsorption also affect the final equilibrium position but in the opposite direction to that in which it is affected by heat. The action of neutral salts is the most initerestinig and of the greatest practical and theoretical importance. Sodium chloride has a very striking effect, even in very low concentration, as the following experiment shows Solution of congo-red 0o0o4 per cent. containing various percentages of sodium chloride. After a circle of paper had lain in each for 2 hours the following colorimeter readinigs were obtained Percentage of NaCl
Percentage of dye left in solution
Percentage of
Percentage of dye left in solution
NaCi
0
85
00005
77
0102 0.1
33 25
olOO I
7I
0'2
I6
0.005
50
The effect of sodium chloride, then, appears to be proportional to its-concentration but not in direct linear proportion. The fact that in so -weak a concentrationi as o ooo5 per cenit. an obvious effect is produced indicates that the action of inorganic electrolytes is rather to be brought into rclation with the well-known effect of these bodies on typical colloids, rather than with the precipitating action of alcohol, which requires a much higher relative concenitratiotn in order to produce a comparable effect, In inivestigating this action of electrolytes
ASPECTS OF ADSORPTION PHENOMENA
I93
the phenomena met with, although of considerable interest, showed themselves to be of a very complex nature, so that it will be better to devote a special section to the distussion of them later oni. At first sight the action of heat in dissociating the adsorption compound, as above described, appears at variance with the common practice of histologists in staining their preparations. I find that the beneficial effect of heat in this case is probably due to the fact that a piece of paper, for example, dyed with congo-red and exposed to a temperature of ioo0 C. loses the power to a large extent of giving up the dye to water, in other words it is fixed more firmly thani if merely stained at room temperature. According to the well-known laws of Van't Hoff, since the compound of cellulose and congo-red is dissociated more and more as the temperature rises, the formation of the compound should be accompanied by evolutioni of heat. I have attempted in various ways to detect such a production of heat but have been unable to do so. The two methods used were: (i) Immersioni of a Beckmanin thermometer in the cenitre of a series of co-axial cylinders of filter paper in a large beaker filled with NaCi solution to facilitate adsorption. A test-tube of I per cent. solutioni of congo-red was also immersed in the solutioni, and wheni the whole was at the same temperature, the dye was mixed with the saline solution. No temperature chanige was to be observed, although by aid of a lens, the thermometer could be read to IjI th of a degree; (2) A disc of paper moist with NaCl solution was allowed to rest oni the face of an ordinary bismuth-antimony thermopile conitained in a vacuum-jacketed vessel and coinnected to a low-resistance d'Arsonval galvanometer. The vessel also contained a small tube of concentrated congo-red solution, closed at the bottom by a glass rod with inidiarubber tube around its lower end. When the galvaniometer was steady the glass rod was raised and the dye No deflection indicatinig the allowed to flow over the paper. production of heat was observed. It is to be nioted, however, that at the temperature of these experiments, the process would niot be very rapid, and it is possible that a slight heat productioni would b. conducted away as fast as it was formed.
I
194
BIO-CHEMICAL JOURNAL
3. Reversibility The question as to how far the process is a reversible otne is of some theoretical interest. It has been already metntionied that if the temperature is raised up to I0oo C., a part of the dye is fixed in the paper, so that under certaini coniditions, the reversibility is not complete. At ordinary temperature it appears complete. If a circle of paper. be dyed with conigo-red, rinised with water and theni placed in water along with another similar circle of paper, it will be founid that after some tinme the two pieces of paper will be of the same depth of colour. At the same time, as already poinited out at the betgiinniing of this paper, to completely extract all the dye from a piece of paper nieeds an enormouis nLliber of repeated changes of water. As regards temperature reversibility is apparenitly complete. Experiment: Two flasks cotntainin1g conigo-red anid filter-paper were placed, one at Soo the other at i I1 ; inext day both flasks were placed in a thermostat at 25°, when equilibrium was attained after about 30 hours, the colorimeter readings were From So' - 5 3 % ,,
11
-
57
%
Againi, if sufficienit time be allowed for equilibrium to be reached, a preparationi may be alterniately warmed and cooled, anid nearly the same readitng,s obtainied at the same temperature. It appears, however, that after several days conitact the dye becomes partly fixed. For example, in onie case the readinig at 38' was 66 per cenit., it was then cooled to I30 with a readinig of 48 per cent. ; onl again warming to 380, 50 per cent. only was reached, and oni coolinig again to I3°, after two days the value became 25 per cenit. Gelatin, also, shows fairly complete heat reversibility as regards its adsorbed electrolytes. At i I1 C. a particular hvdrogel had a specific conductivity of I62 gemmhos, after being hieated to 530 anid againi cooled to i io C. its coInductivity was I57-8 gemmhos. Aniother preparationi cointainin1g more electrolytes- had anl iniitial colnductivity of 609 gemmhos, after heatinig to 55 5° anid coolingr again to I 10, it became 542 gemmhos, or less than its original conlductivity. Perhaps inl the first measuremenit equilibrium was niot completely attainled.
AsPE:cTs OF ADSORPTION PHENOMENA
t95 When adsorptioni has taken place under the actioni of electrolytes a very colnsiderable degree of fixation occurs, so that the dye is nlot given up again to water, or onily to a very slight degree. Experiment: Two similar pieces of paper dyed to same depth of colour, onle in solutioni of NaCl 0.02 per cenit., the other inl distilled water. Of course the latter solutioni was more conicenitrated as regards the dye. These pieces were rinised in water, pressed betweeni blottingpaper, and theni placed in equal volumnes of distilled water alonig with aniother piece of paper each. After 24 hiours the onie which had been dyed in distilled water hiad giveni up a considerable amounit of pigment to the water, and the two pieces of paper were very nearly equal in depth of tinit. The paper dyed in the presence of NaCl gave up nio colour to be detected by the eye, although the second piece of paper was fainitly pink, showing a slight extractioni from the dyed piece. IV. THE ACTION
OF ELECTROLYTES
Wheni we conisider that a solutioIn conitaining onily o0ooo5 per cenit. of NaCi has a distitnct effect in augmentinig adsorptioni of conigored by paper, viz., 23 ptr cent. takeni up as againist I6 per cent. from distilled water, it is plain that the phenomenioni is niot of the nature of a ' salting-out,' comparable to the precipitation say of egg-albumetn by ammonium sulphate. The precipitation of such solutions as those of the colloidal metals and hydroxides is rather suggested. Now, there is considerable evidence that a large number of the aniilin dyes exist in watery solutionis in a colloidal form. Congo-red, having a molecular weight of nearly 700, would be expected to be one of these. Its solutions, in fact, do not diffuse through Schleicher and Schilll's parchment-paper thimbles, although they do so, very slowly, through ordiniary parchment-paper. Accordinig to Michaelis' under the ultramicroscope they arc heterogeneous, being resolvable into submicroscopic particles. In the electric field the dye migrates to the anode, so that the particles are negatively charged. It is, however, somewhat difficult to make oneself certain of this fact oni account of I. Deutsche Med. Wochenschr!ft, 1904, No. 42.
I96
13IO-CHEMICAL JOURNAL
the electrolytic decomposition. Colloids do behave as electrolytes, as is well-known', and I have made one rough determinationi, by Whetham's boundary method, of the velocity of the coloured ioIn in congo-red solutioni. At I30 the boundary, which was niot very sharp, moved i i mm. in one hour unider a potential fall of 3 volts per cm. Hardy1 finds for globulin about 7 mm. in the same time, and for methylenie-blue considerably more, about 40 mm. The electrical conductivity of conigo-red solutionis is comparable with that of inorgatic electrolytes, a 0M solutioii at 400 having a specific conductivity of 56oo gemmhos. I do not lay aniy stress on the absolute value of this measuremenit sinice the preparation was not specially purified, and possibly containied a small amount of inorganic salt. On the whole, theni, we may regard congo-red as being a negatively charged colloid. As such it would be specially sensitive to di- and tri-valent kations, and this is in fact the case. All the experiments to be described in this section, except when otherwise stated, were made in the same way, viz., to So c.c. of the A dye solutioni and one piece I2'5 solution of electrolyte IO c.c. of 1000 cm. diam. of Schleicher anid Schull's extracted- filter paper added. After about 24 hours the amount of dye taken up by the paper was estimated by taking the colorimetric value of the dye left in solution. It is easily seeni at onice that calcium salts are much more active in promotinig adsorptioni thani those of the monovalent alkali metals, e.g., the amount taken up from a solution conitaining -5caso, 500 was 85 per cent., from a M- KCI 67 per cenit. Considering this fact the effect of tap-water is not surprising, since New River water contains an equivalent in Ca+ + ions of about caso4. From tap-water equal amounit to the contenit therein of -M 300 85 per cent. was adsorbed in onie experiment, anid from distilled water only 27 per cent.3 Moreover, these facts show that in order to obtain any reliable data, special care must be taken as to the purity of the paper used. As an illustration, I give the following experiment: I. Hardy, Journal of Physiology, XXXIII, 1905, p. 292. 2. Loc. cit., p. 29I and p. 289. 3. The absolute amounts in different experiments can only be compared by taking theirra'io to the amount taken up from distilled water in each case, since the temperature was not the same in all.
ASPECTS OF ADSORPTION PHENOMENA
197
Equal weights of the following samples of filter-paper were placed in 50 c.c. distilled water with IO C.C. m dye as usual. The amounts taken up by each are given ... ... i. Dreverhof's ordinary, No. 333 washed in distilled water ,, ,, 2. ... ... 3. Dreverhof's extracted paper ... ... washed ... ,, ,, 4. 5. Schleicher and Schull's extracted paper
95 % 59 % 46 % 46 % 46 %
I may mentioni here that Dreverhof's extracted paper turnis blue when a drop of dilute congo-red is placed otn it, this does not happetn with Schleicher anid Scholl's, nor with Swedish paper. As the above experiment shows this slight acidity does niot seem to affect its adsorbinig power, anid as will be seen later H+ ions have comparatively little favouring effect. The result of this experimenit seems to suggest the possibility, that if one could obtain a complete absenice, of electrolytes no adsorption would take place. As regards trivalent kationis I have onily tested aluminiium 1M precipitation took sulphate, but as even in a concentrationi of 1000 place the result was valueless. It is scarcely necessary, perhaps, to remark that if the dye is precipitated by the electrolyte nio adsorption by the paper takes place, the large particles merely float about in the liquid. What is needed is the local concentration of electrolyte oni the surfaces of the adsorbing solids.' Although the greater favouring power of divalent as compared with monovalent kations is easy to demonstrate the order in each of these classes is difficult to decide since the action of each member is so nearly the same as that of the others. From a niumber of experiments the following seems to be the order: H < Li < K < Na < NH4 < Mg < Ca
I. Notwithstanding the fact that actual precipitation must not take place in these experiments the addition of electrolytes to congo-red, for example, causes an increase in the size of the colloidal particles even when no actual precipitation occturs; so that the solution is on its way to precipitation even when this does not actually occur. The specimen of congo-red used in all my experiments showed the Tyndall phenomena very faintly in solution in distilled water; but when NaCl was added the beam of light became much more distinct and the light reflected from it at right angles was polarized. When CaC2 was added in sufficient amount to slowly precipitate the dye (about M ) the polarization ceased as the particles became larger than the mean mQve-length of light,
BIO-CHEMICAL JOURNAL
I98
This is very niearly the samne as that foutnd by Pauli' inl the case of precipitation of proteinis. It is a matter of some interest to see whether in this case of adsorption the anionis have a retarding action like that showni by Pauli in the above-mentioned experimenits. That there is some action of this kinid can I think be shown by considerationi of the comparative actioni of chloride anid sulphate of the same metal, say potassium. The difficulty lies in the fact of the great preponlderanice of the opposite actioni of the kationi, probably on accouint of the niegative charge of the conigo-red. If we take equimolecular solutions of KCI and K,S0o we find their effect almost exactly the same within limits of experimetntal error. Now there are, in the dilute solutions employed, very niearly twice the number of kations in the K,2SO than inl that of the KCI, niecessarily, therefore, the so, iotn has a greater retarding power thani the ci ion. In determininig the order of the anionis in this respect the same difficulty presenits itself, as in the case of the opposite effect of the kations, but in a more marked degree. The approximate order as regards power of inhibition is the following OH < CNS < acetate < Cl < F
The results of these experiments will be best shown in a table anialogous to those giveni by Pauli. The niumbers give the percentage amounit of dye adsorbed at room temperature under the inifluence of the body formed by the combination of kation and anioni showln vertically above and horizonitally at the left-hand side of each number, anid in 200 m- concenitration. CONGO RED z6
Li 58
CNS
...
*.
Acetate
...
H
OH
..
CF
Oxalate
SO,, P04
K
Na
...
68
(m)
Mg *'
Ca 89 ...
...
*-.
64
86
58
70
8I
...
...
70
75
72
...
91
...
65 78
...
...
91
> IOO
..
I2 4-
71
80 74-
...
...
...
..
NH4 79
78
74
...
X, HIofmcister's Beitrage, III, p. 225; V, P. 27; Vi, p. 233; and VII, p 531.
ASPECTS OF ADSORPTION PHENOMENA
199
There are some irregularities especially as regards NH4 salts. The valuLe put against CaSO4 means that the conicentrationi had to be taken at 500 to avoid precipitationi, and at this strength the value was 84 per cent. The actioin of O- anld H ionls needs a little further ekplanatioii since it seems to vary according to conicentrationi in both cases. It is somewhat remarkable that in coIncenitrations of Im NaOH aid Naci have almost exactly the same amount of favourinig action ; in I on the conitrary, while Naci has a distinict favourirg actioni, viz., 58 per cent. adsorbed from Naci as agaillst 30 per cenit. from water, NaOH has a very slight inhibiting actioni, viz., 25 per cent. against 26 per cent. from water. In one experimenit I took three solutions of equal conitent in OH '
ionS, viz. 1000
m
NaOH
...
NH40H
...
Na
Cos'
3 I%
79 %
89%
The amounit adsorbed in presence of these is shown opposite each one. I could not detect any inifluenice of anilini-water oni the process. Sulphuric acid 2 has a favouring actioti, 33 per cenit. agailnst 2 I per cent. from water, while in m it has the opposite effect as thie table shows. I am unable to suggest anly explanation of these facts. In conniectioni with the well-knowni mutual precipitation of oppositely charged colloids1 the behaviour of congo-red as a niegative colloid is of interest. Colloidal platinium prepared by Bredig's method does not precipitate, but has a chemical action, turninig the dye brownl. This platilnum sol is niegatively charged so that precipitation would niot be expected. Oni the other hand, according to Paulil' anid other investigators, the precipitating action of salts of the heavy metals on egg-white is due to the presence in their solutiotns of metallic hydroxide in the colloidal form and presumably positively charged. I. Sce Picton and Linder, Journ. of Chem. Soc., I892, page I48, etc., and W. Biltz, IBer dl. Deutsch. CAem. Ges., 1904, Bd. XXXVII, page I II i. 2. Hofmeister's Beitrage, VI, page 257, 1905.
13IO-CHEMICAL JOURNAL
200
This beinig so, the powerful actioni of suLch salts oni adsorption is Inot -id5000 Zn7O4 to hlave ani equal effect to be wonidered at. In fact, I find to that of 20 KCI. This actioni of zinic sulphate gives the opportunity for testilng by expetrimenit whether in this case, as in Pictoni anid Linider's' cases of precipitationi of colloids by electrolytes, the precipitatinig iOln is carried dowIl with the colloid to become attached to thep aper. Sinice the'precipitating' agenit in the adsorptioni of a niegative colloid is the positively-charged kation, anid the most coniveniienit method of detectilng a disappearance of ani ion, is the use of a conicenitration battery in the mainner of Nernist, it is plaini that we catnniot coniveniienitly determine the questioni where salts of alkali metals are conicet-nied. A conicenitration battery ill Zn + iolns is, oIn the conitrary, easily ar-raniged. In the presenit case I took two vessels, eacih conitainilng ani amalgamated zitnc electrode ma- ZnSO4 solution and(i conniected together by ani iniverted inmmersed in 500 U-tube filled with the ZnSO, solutioni. Such a battery beinig symnmetrical hias 1no E.M.F. A piece of filter paper was nlow placed in one of the vessels, the E.M.F. begani to rise anid attainied a value niearly equial to what it would be by the Nernlst formula, if all the Zn+ ionis had disappeared fronl this vessel, ViZ., o0o257 volt. This effect was, in all probability, dIuie to adsorption l)y the paper. Oni niow addinig 5 c.c. of i per cenit. conigo-red iMl 50 ZnSO, the E.M.F. went downi againi, due to the additionl of more Zn+ iOlnS. Presenitly, as the dye become adsorbed, the E.M.F. commenced to rise againi anid whenl plotted oni squLare(d paper showed thle usuLal form of the adsorptioni-curve. As far as it goes, theni, the experiment shows a dinliniutiotn in concetntratiOll of Zn + + ionls. But there are two circumstanices which deprive it of much value. In the first place I found that inext day thc E.M.F. of the battery had riseni to o0o94 volt, a value above that possible by rem-oval of all the Zn+ + ions; there nmust have beeni, therefore, some seconidary process set up by the congo-red. In the second place the possibility must be takeni into accounit, that wheni congo-red and zinicsuLlphate al-e mixed, a zinic-salt of the congo-red may be formed by double decompositioni wlhich might niot be conmpletely dissociated, this 4
i. Loc cit.
%
ASPECTS OF ADSORPTION PHENOMENA
201
salt might be the actual body adsorbed by the paper. Considerinig the great dilution of the solution, however, I do l1ot think the latter objectioni is very serious. Aniother property of colloids is that of being protected from precipitationi of electrolytes wheni a small amounit of a stable colloid, such as gelatin, is presenit. As poinited out by Zsigmonidy,' the fact was kniowni to Faraday, although its meaniitng was, of course, n1ot uniderstood at the time. The following experimenit shows that gelatini also protects conigo-red from the action of ani electrolyte A. 50 C.C. H1,O + I c.c. 0-5 Y. dye B. 50 c.c. oo5 % gelatin + I c.c. o % dye , C. 50 c.c. 0-038 % NaCl + D. 50 c.c. o0o38 ° NaC1 containing o o5 %. gelatin + I c.c. o 5 % dye Adsorbed in three hours A. 30 °/. C.74/ 1. 31 /c D. 34 %
It will be seeni that NaCI is practically without effect in presetnce of
gelatini. Egg-albutnini has a similar effect, but less marked, probably oni accounit of its conitent in electrolytes. I did niot subject it to dialysis. Gelatin is stated by Victor Heinri 2to be a niegative colloid; serumalbumini has recenitly beeni showni by Pauli to be positive in acid solutioln, anid niegative in alkaline solutioni. It was of interest therefore, to see whether any difference in the protective action of cgg-albumin was to be noted in the two itnstances. The result showed that electropositive albumin increased the actioni of CaS04, while electro-negative albumin diminished it A.
m Albumin o0o4 % + CaSo, 5c0 mc + H,S0O 1000 M
1B. CaS04 500
+
H
SO 1000
+ dye + paper as usual
Adsorbed: A. > So% (partially ptd-) B. 365 % i. Loc. cit., p. 65, footnote. 2. Rev. Gcn. des Sciences, 1905, p. 641. 3. Hofmeister's Beiirage, VII, p. 5 and p. 536.
BIO-CHEMICAL JOURNAL
202
m + NaOH m C. Albumin o004. % + CaSO,25~00 1000
D. CaSO4
m 500
+ NaOH
m 1000
+ dye + paper.
Adsorbed:
C. 75 D. 98
% %
It is difficult to compare with these the action of the neutral albumin sinice acid and alkali have themselves an inifluence. But it appears that the signi of the charge on the stable colloid determines the nature of its actioi, although the particular albumin in this experiment had also a protective action in neutral reaction. I am unable to state whether it was niaturally electroniegative or otherwise. Other electro-negative dyes, such asaniilini-blue anid nigrosin, under the action of electrolytes, givc similar results to congo-red. The colloidal conidition of the solution is, however, apparently a sine qud non. I could not detect any influenice of electrolytes oni picric acid, although it is difficult to observe a small differenice, because of the comparatively feeble colourinig power of this dye. I compared the depth of tint of the two solutions, one with electrolyte, the other without, in both of which the same amount of paper had been immersed, by placitng them in two cells of equal thickness in fronit of a piece of plantinotype paper alnd exposed to light. No difference could be detected in the depth of tint on developmenit of the paper, although, as is well kniown, this paper shows very small d;fferences of tone. Scharlach R. is initerestinig, because it is a so-called ' inidifferent dye, that is it does Inot form salts. It is insoluble in water ; but as Michaelis' poinits out, when an alcoholic solution is mixed with five to six times its volume of water, it is not precipitated, but becomes a colloidal solutioni. In this state it behaves towards electrolytes like a negative colloid, and, in fact, in the electric field, as I find, it walnders to the anode. Rosolic acid shows similar properties. I. Deutsche,
Mfed. Wochensch., 1904,
No. 42,
ASPECTS OF ADSORPTION PHENOMENA
203
In order to save space, I have put together in a table at the end of this paper a number of facts relatinig to various dyes in common use which have a bearing on the quLestioni under discussion. The otnly other poinlt I will refer to as regards the negative dyes is that eosini, although its colloidal properties are doubtful (see table), shows a much greater senisitiveniess to the anioni than the dyes hitherto menitioned, alkali in this case causes the taking up of less colour by paper than is taken up from water. We found in the previous casts that the action of the OH ion was overpowered by the opposite action of the kationi ( N). We turni niow to the electro-positive dyes like toluidin-blue. It would be expected that in this instanice the anions would have a favouring action, the kationls the reverse. This is, indeed, the case, but like the niegative dyes the effect of the kation is predominianit, so that the result of addinig any electrolytes, except alkalies, is to lessen adsorptioni. In the absetnce of electrolytes, as a rule, more dye is taken up by the paper thani from the solutioni of an electro-negative dye. This would appear to be at variance with the genieral use of conigo-red to dye cottoni, but it it is to be remembered that in practice there is always sufficienit electrolyte present to reverse the relative behaviour of inegative atnd positive dyes as shown when dissolved in pure distilled water. For details as to various positive dyes see the table.. Methylenie blue precipitates the electro-niegative colloidal platinium. I have niot beeni able to detect aniy marked. protective action of stable colloids in the case of electro-positive dyes. I n the circumstatnce that both the + and - dlyes are more senisitive to kations thani to aniionis they behave like albumin solutionis to salts of the heavy metals as shown by Pauli.' On referenice to the table at the end of this paper it will be nioticed that in all the cases inivestigated where the dye is in the form of a salt with an inorganiic base or acid, as the case may be, the sign of the charge of the colloid is determined by the crganic constituent of 1. Hofmeister's Beitrdge, VI,
pp.
233 to 249.
BIO-CHEMICAL JOURNAL
20+
the salt. JIn this respect they differ from the globulini of Hardy and the serum-albumini of Pauli, which in acid or alkaline solutionis assume the signi of the charge of the H + or 0 H - ionis respectively. They are in fact more like certaini complex ions, such as the ferrocyan- iOl, Fe (CN),a which, although conitainiing an iron atom is, nevertheless, electro-negati ve. 1 have beetn uniable, for wanit of time, to test the behaviour of more than a few typical dyes of the two classes, but it appears that there is conisiderable differenice of degree in the sensitiveness to electrolytes. This property inideed seems to depetnd oni the degree of colloidality of their solution and the amoutnt of their electric charge. Each inidividlual dye requires separate itnvestigation. As to the explaniationi of the actioni of electrolytes, I thitnk the clue is giveni by the followinig considerationls. V. Henri anid Larguier des Banicels,' in their work oni colloids, observed that gelatin, as hydrogel, when immnersed in a solutionl of anilin-blue or conlgo-red in distilled watter took up little or nionie of the pigmenit, anid give as the reasoni for this that both the bodies are electro-negative colloids and(l therefore mutuially repel onie another. If, oni the conitrary, a solutioIn of a bivalent metal, such as barium niitrate, is added the gelatin. becomes deeply stained. The tnegative clharge of the aniiliniblue is nieutralized by the positive barium iolls so that it can niow freely attach itself to the gelatini. This case theni resolves itself into one of the muLtual precipitationi ofcolloids, andcl the onily doubtful poitnt about the explaniation is whether the niegatively-charged particles of the gelatini hydrosol may be regarded as retaininlg their charge when in the form of gelatine. There seems nio a' priori reasoni why this should niot be so, anid there is also experitnental evidence iln favour of it. Pictoni and( Lind(ler2 in refei-rilng to the adsorption affinity of the hydrogels of ferric hydroxide and arseniious sulphide for anilin dyes state that these hydrogels 'retain the same selective affinity for- the dyes which will coagulate them' as they possessed as hydrosols, viz., 'the hydrogel of ferric hydroxide for anilin-blue, that of arsenious C.R. Soc. de Biologie, LIX, p. 132, 1905, z, Journ, Chem. Soc., 88, 1905, p. 193+4
I.
ASPECTS OF ADSORPTION PHENOMENA
205
sulphide for nmethyl-violet. We regard this fact as evidence that the granular aggregates iln these hydrogels still retain some charge. In other words, the differenice of potenitial existing betweeni the molecular aggregates and the field in the hydrosol state is lnot destroyed by coagulanits, but onily reduced to a point at which the forces of cohesiotn are just able to overpower the forces of repulsion brought illto play by such difference of potential.' It will readily be- seen that if paper assumes a negative charge wheni immersed in water the pheniomenia I hiave described above fall inito linle with the behaviour of electrically charged colloids. We niotice that of a niegative colloid like conigo-ied the less is takeln up by paper the more prfect thle freedom of the water from electrolytes, there is nmutual repulsioni until the dye is discharged by a kationl. On the other hanid, the opposite holds as regards the positive dyes, there is attractioni of the niegative paper for the positive dye. The katiotn in this case probably acts by dischargiAg the paper, beinig attracted thereto by its opposite charge. Prof. Donniiani, to whom I propounided the questioni, refcrs nme to the work of Coeihni', who found that whenl various nion-coniductinig bodies are immersed in fluids of differenlt dielectric-constaints they assume a positive or niegative charge accordinig as their ownI dielectricconstanits are highier or lower thani the fluid with which they are in conitact. For iinstanice, glass (5-6) is inegative in water (80) or alcohol (26), whereas in turpentine (2.2) it is positive. Now, paper accordinig to Thwinzg2 has a dielectric-constant of 2-82, anid would therefore be electrically negative to water. Moreover, Quincke3, in- the course of his investigatioins oni electrical enidosmosis founid that, as a rule, all inon-coniducting bodies in water took oIn a niegative charge. Among the substanices tested by hinm were cottoni-wool an.d silk ; particles of these in water exposed to ani electric field wanidered to the aniode and were therefore negatively charged. T. 2.
Wied. Ann., 64, page 217, I898.
Zeitsch.f. Physik. Chemie., XIV, page 292, I894. 3. Pogg. Ann. 113, page 583, i86i.
2o6
BI2-CHEMICAL JOURNAL
Trhe cotnclusioni is, I thinik, justified that paper takes up.very little electro-negative dye because it is itself negatively charged, and that when the dye is discharged by the additioni of ani electro positive colloid or a kation, there is no loniger the same mutual repulsion between the dye anid paper. It would be of initerest to test the behaviour of paper in turpenitine, in which it would be electro-positive, if one could obtaini ani electronegative colloidal dye in solution in turpeniitne. This interpretationi is confirmed by the results of experiments otn silk Experiment. From a watery solution of congo-red a piece of paper took up 26 per cent., a piece of silk (well washed), of the same weight took up onily 5 per cenit. From a m MgS04 solution, paper took up 9I per cenit. and silk 98 per cent. From toluidini-blue in water, paper took up 85 per cent. Silk rapidly took up the whole, so that I added another double amouAt of dye solutioni, the colorimeter reading thus showed that 67 per cent. of the whole, that is 67 x 3 = 20I as compared with 85 in the case of paper was taken up. The actioni of neutral-salts on the adsorptioni of positive dyes by silk was of the opposite signi to their effect oni paper. For example : Paper: From water ... m ,, M9SO4
...
85 % adsorbed
...
510
Silk: From water (concentration of dye three times that in case of paper) 67 % adsorbed m MgSO .. 9I %
The results may be explained by the conisiderationi that silk, owilng to its lower dielectric conistant thani that of paper, would no doubt have a higher nlegative charge, which would make it less accessible to negatively-charged dyes, but more so to positively-charged dyes. I am unable to state why the action of electrolytes on the positive dye is opposite in the two cases of silk and paper, unless that for some reason silk is relatively more sensitive to the adjuvant action of the anion. If this were so, however, one would expect a more marked
ASPECTS OF ADSORPTION PHENOMENA
207
inihibitory effect of the aniion in the case of electro-niegativc dyes anid silk, whereas we see that the action of MgsO4 is rather greater in the case of silk thani in that of paper. The troublesome staini which appears oni the surface of glass containinlg electro-positive dyes is no doubt due to the niegative charge of the glass. This film is so adherent that water will not remove it, and I lhave founid it niecessary, when working with these dyes, to ritnse all flasks used with concentrated nitric acid, in order to obtain colncordant results. Aniother fact receives its explanationi from the negative charge of paper, viz., that from alcohol more negative dye is taken up thani from water, in onie experiment: From 50 % alcohol - 65 % taken up ivater - 53 % w, 9"
Sinice alcohol has a lower dielectric-constanit thani water it would nlaturally be expected, sinice the niegative charge of paper is due to the differenice betweeni its dielectric-constanit anid that of the fluid in which it is immersed, that the niegative charge would be greatcr the greater this differenice is, anid therefore in alcohol the charge would be less thani in water anid the attachmenit of a niegative dye less difficult. On the other hand, less + dye is takell up ini the presence of alcohol, thus: From SO % alcohol - s % taken up ,, - 8o % ,, water
Finally, I may refer to the action of acid in preventing the favouring action of electrolytes on adsorption of a niegative dye, for example: CONGO-RED From water m
NlaCl
200
0
200
NaCl +
... ...
_ 2000
... .
..
H
2SO
...
z7
% adsorbed
*
75
%
.
3I %
,,
This result is probably to be explained by the fact that in the prcsence of sulphuric acid the colour-acid is set free from congored. This colour-acid is, nlo doubt, rmore stronigly electro-negative
208
BIO-CHEMICAL JOURNAL
thani the salt, so that the amount of kation which would suffice to discharge the salt would be insufficienit to discharge' the acid, it would therefore be comparatively inieffective in promoting adsorption. In view of the theory set out in the preceding paragraphs it will be of initerest to give two curves showitng the form of the relation between the concentration of the kation atid its effect otn congo-red. The ordinates in Fig. 6 represenlt the percentages of congo-red taken up from a 6000 solution containinig electrolyte of the concenitration' given by the abscissate. The upper curve is that of caso4, the lower curve that of KCi.
FIG. 6
The commencemenit of the CaSO4 curve is irregular, owinlg to partial precipitation of the dye having taken place. The inclination to an S-shape at the beginning of the KCI curve is, probably, also an indicationi of incipient precipitation. The extreme senlsitiveness of
ASPECTS OF ADSORPTION PHENOMENA
coingo-red to electrolytes is also showtn by the curves; even at the low concenitration Of 6000 m of the electrolyte they do not fall to the level of the amounit takeni up from the particular sample of water used, which amount is represetnted by the horizontal line with an ordinate value of 20 per cenit. It is noticeable, otl the contrary, that the curves tenid to become asymptotes. I amn tnot prepared -to give anly interpretationi of the form of these curves. V.
CONDITIONS
OF
DISSocIATION
OF
ADSORPTION COMPOUNDS
Recent research has brought to light the importanit part played in vital phenomenia by the mutual relations of colloids to electrolytes and to other colloids. Since these relationis are apparenitly those of adsorptioni, it is of some importance to kniow what are the coniditionis unider which this associatiotn is broken downi. i. Temperature We have already seeni that dissociation is favoured by rise of temperature, and that, onl the other hand, if suddenily. exposed to the temper-ature of boilinig water, a dyed piece of paper becomes to a certaini degree fixed in. its conditionl, so that the dye can onily with difficulty be extracted by water. 2. Precipiiaiion It was noticed by Cramer anid Swale Vinicenit' that the inorganic conistituents of various organiic bodies are set fi-ee wlhetn these bodies are acted upoIn by precipitanits of bases. I have observed the same fact in the case of the precipitationi of gelatin by tanniiin. Experiment: A i5 per cenit. solutioni of gelatin had a specific coniductivity of 1570 gemmhos at 38-5. 20 c.c. of this solution were mixed with go c.c. water anld io c.c. of IO per cent. tannini solutioni, filtered, anid ioo c.c. filtrate concenitrated to 20 c.c. The coniductivity of this was nIow 2750 gemmhos. This 20 c.c. would represenit '0 of the first, sb that the conductivity should be nmultiplied by 3300 gemmhos, that is, the conductivity was doubled by the action of i. Journ. of Physiology XXXX, p. 150, 1903.
0MO-CHEMICAL JOURNAL
2IO
tannin. The amounit of tannini added was intenitionially less thani that required for complete precipitationi, so that the filtrate containied a little unaltered gelatini, but nio tannliin. 3. Action of Chloroform Moore anid Roaf' have showni that electrolytes are set free in blood by the actioni of chloroform. 4. Excitation, Injury, and Death Macdonald 2 found that, as a result of inijury, potassium salts are set free in the axis-cyliniders of nierve fibres. Unider certaini coniditions these salts may be takeni up again by the cblloids, anid Macdonald suggests a theory of excitationl anid inhibition oni this basis. Hoeber3 finids that the excitability, the capacity of stainintg, and(i the colloidal conisistency of nerve vary concurretntly. Wakeli n Barratt4 finids that chlorinie ionis are set free when uniicellular organiisms are inijured or killed. Brailsford-Robertson5, as the result of his work oni the pheniomenia of the heart-beat, comes to the coniclusion that these pheniomena are such as would be expected if there were a periodic driving out l)y aniotns of kationis from combinationi with proteid. Howell,6 havinig founid that the vagus canniot exert its action on the heart-nmuscle in the absetnce of potassium ioIns, brings forward the hypothesis that the inhibitioni produced by the vagus is due to splittinlg off potassium from combiniation in the muscle. In order to see whether there is aniy evidetnce of a suddeni separationi of electrolytes at the moment of death, I have made the
following experiment: The skinis of two frogs were immersed in isotonic cane-sugar solutionl. The electrical resistanice of the solutioni was measured, after soakinig for some time, anid found to be at i i° C.) 6407 ohms. The solutioni conitaininig the skinis was theni slowly warmed anid the resistanice determitned at frequenit intervals up to 56' C. The t. Proc. Roy. SOc., 73, p. 382, 1904, and 77 B, p. 98, I906. Thompson-races Lab. Reports, Vol. 4, p. 213, I90±. 3. Centralb.f. Physiologie., XIX, page 390, 1905. 4. Zeitsch. f/ Aligem. Physiologic, V, page 33, 1905. 5. Pfluger's Archiv. I 10, page 623, 1905. 6. 8merican Jcurn. of Physiology, XV, page 29I, I905. 2.
ASPECTS OF ADSORPTION PHENOMENA
2III
values when plotted on squaied paper formed a regular curve, somewhat concave towards the axis of abscissae. There was no indication of ally kinik in the curve except for one determinationi at I6°, which was obviously ani experimental error. That electrolytes were given off is shown by the fact that the diminutioni ot resistance was conIsiderably greater than would have been the case if the diminutioni depended only on the effect of the rise of temperature on electrolytes present at the beginninlg. The temperature-coefficienit was, indeed, 4 i4 per cent. per degree, instead of, at the most, 2-75 per cent., according to the determinations of Arrneniius. The separation of electrolytes also showed itself to be irreversible, since cooling down to II C., the resistance only increased to 4I28 w. The shape of the curve shows that the rate of separation was greater at the lower than at the higher temperatures, but, as already remarked, there was tno poinit oln the curve to indicate that death occurred. It is obvious that the results may also be accounted for by gradual destruction of the impermeability of the livinig cell for most inorganiic salts, but, if so, this process must also bc a gradual one. To compare with this experimenit I made a smilar one with eggwhite, raising the temperature to 950 C. In this case the result was opposite to the above, viz., there was evidence of taking up of electrolytes. The physical state of the coagulated egg-white is, however, so differenit from that of the fresh substance that it is difficult to initerpret the meaninlg of the observation. S. Coagulation of blood. It was found by E. G. MartinI' that ca++ ioIns have a marked stimulatinig influenice in causinig freshly isolated strips of the tortoise venitricle to commence beatinig sponitaneously. This effect is shown by serum, but not by uncoagulated plasma. It appears, thus, that Ca+ ions are set free during the process of clottinig, so that there should be ani increase of electrical conductivity in the process. Robert T. Frank2 was uniable to detect any chanige of this kinid. I have, on thc contrary, succeeded in showing that there is such a change, not in i. Amer. Journ. of Physiology, XI, p. 1 17, 1904. 2. Amer. Journ. or Physiology, XIV, p. 466, I905.
13IO-CH4EMICAL JOURNAL the sense of an increase but a decrease, so that there is a disappearance of ions. This result is, I think, what would be expected from our ktnowledge of the relationi of calcium to clottinig, atnd the observations of E. G. Martini must be explainied in aniother way. My experimelnt was made oni fowl's blood. As showni by Delezeniie, if kept from conitact with the tissues this blood remaitns uniclotted for a long time. It was collected from the externial jugular vein and placed in a U tube furtished with ani electrode in each limb and immersed in a waterbathl at IO2° C. Wheni the temperature had becomne conistanit the specific coniductivity of the blood was 5382 gemmhos. A small piece of muscle from the fowl was theni dropped itnto one side of the tube. The coniductivity steadily fell to 4360 gemmhos, which value was attained in 4o mitnutes; no doubt the process was completed earlier thani this, but it was niot wished to distarb the apparatus unitil it was to be expected that clotting had ocurred. The diminution of conductivity therefore amounited to i8-5 per cenit. of the original value. 6. Action of enzymes I do niot purpose here to enter inito this subject in anly detail, since it will be dealt with fully in a forthcoming paper. Thcre is a certain amounit of evidence that the splittinig-off of inorganic conistituents contributes to the rise of electrical colnductivity observed in maniy cases, but the direct proof was found to be of
eonsiderable difficulty. In the actioni of reninet onI milk, I finid a slight increase of coInductivity. This is merely an additional fact showing the nature of this process to be quite differenit from that of clotting of blood. In all probability reinnet-actiotn is onily ani expressioni of pepsin-actioni in neutral or fainitly alkaline medium.' SPECIFIc ADSORPTION of adsorptioni show in this respect an approxi-
VI.
The pheniomena mation to true chemical combinationl. It was shown by Schonbeinti that wheni strips of filter-paper were immersed in solutions of various salts that the height to which the salts I. Pawlow and Parastschuk. Hoppe-Seyler's Zeitsch., 42, p. 415, 1904. z. Poggendorff, Ann., 114, p. 275, i86i.
ASPECTS OF ADSORPT'ION PHENOMENA
213
rose was less in the case of calcium and barium thatn in that of potassium or stronltium, the calcium, etc., being more completely held by the paper. In the case of a solution of iodinie in KI, while the latter rose nearly as far as water, the iodine was kept back in the lower thircd. These facts were explainied by SchOnbeink as capillary attractiotn, but, as Ostwald remarks, there is nlo doubt that it is really an adsorption process that causes the separation of the bodies in question, capillarity onily occasions the transport of the separated
fractiolns. These results of Schdnbeini have been extended anid employed as a meanis of anialysis by Goppelsroeder.' It is well know n Lo histologists how certain tissues take up particular dyes in preferenice to others. In certain cases, as for example, the differenices betweeni paper anid silk, as to their respective behaviour towards conigo-red anid methylene blue, this is probably to be accounited for by the fact of the electrical charge in the two dyes beinig opposite in sign combinied with the negative charge of the substanice to be staiied, as already pointed out. Acid fuchsini is used in Vani Giesoni's staini to show coninectivetissue. In view of this, I thought it of initerest to try the behaviour of gelatini in sheet towards this dye and toluidin blue respectively. The same amount of gelatin was founid to take up 72 per cent. of the acid-fushin anid only 32 per cent. of the blue. It is possible here that the electrolytes of the gelatini were responisible for the differenice, sinice they would increase the adsorption of the acid dye anid diminiish that of the basic dye. This possibility is excluded when we compare the amounits of conigo-red anid acid fuchsin, both electronegative dyes, taken up by gelatini or filter-paper. The followinig are the results of aIn experiment: Eqiiimolecular concentrations of dye. Equal wveights of gelatin and filter-paper. Filter-paper from Congo-red ,, acid-fuchsin ,, Congo-red Gelatin ,, ,, acid-fuchsin i. Mitt. d. techn. Gewerbe Museums, Wien,
analyse, Basel, 1904.
i899,
...
3' %
... ... ...
32 % 49 %
and Studien uber die
95 %
Ani-vtndung
d. Capillar
BI20-CHEMICAL JOURNAL
2I4
Gelatin, then, has a a specific adsorptioni affinity for acid fuchsin. The preferenice exhibited by silk for methylene-blue over colngored, previously described, may be explained in aniother way than by specific adsorptioni, viz., by the probable greater negative charge of silk.
That gelatini has a greater adsorptioni-affinity for calcium sulphate than it has for potassium chloride seems to be shown by the experiments on washing out electrolytes, given in the earlier part of this paper. It was found that KCI is more easily washed. away than calcium sulphate. VII. THE APPLICATION OF rHE FACTS OF ADSORPTION-PROCESSES TO VARIOUS PHENOMENA The theoretical interest of adsorptioni-processes has been pointed out by Ostwald,' who suggests the possibility that a mechaniical theory of chemical affinity may be developed oni the basis of a complete study of these phenomenia. From a more practical stanidpoint, it will be founid that they play a very important part in niumerous processes. I may refer briefly to a few of these. I. The Soil The power of the soil in holding back soluble salts anid other bodies is, no doubt, due to adsorptioni. By this means these bodies are prevenited from beinig readily washed away by the rain. 2. Purificazion of sewage In the filter-process of niitrificationi, it appears that the complex organic substances, which are iniimical to the nitrifying organisms, are kept back by adsorptioni in the upper layers of the filter.2 3. The ash-constituents of organic compounds I have already giveni evidence that these are present in an adsorbed form, and showni that the difficulty of removing the last fractions of iniorganic electrolytes is adequately explained on this basis. I. Op. cit., p. 1098. ;. See Dr. Harriette Chick, Proc. Roy. Soc., 77B, p. 247, I906
ASPECTS OF ADSORPTION PHENOMENA
2fS
It is scarcely niecessary to give inistances of this difficulty, anid I will merely refer to three instances. Macallum' in searching for the cause of the silver reaction of tissues met with cotnsiderable difficulty in completely removinig chlorides. Mellanby1 states that there is a molecular combiniationi between tneutral salts anid globulini. Lange3 founid that it was impossible to remove all ash-constituents from filterpaper by extraction with hydrochloric and hydrofluoric acids and washing.
4.
Dyeing
Oni the wlhole, I thitnk it must be takeni as the true explaniation of this process that it is, in the main, an adsorption. The very slight diffusibility of most dyes through parchment-paper seems to me a conisiderable objectioni to Witt's theory of solid-solutioni. If the dye is dissolved in the paper it should be readily giveni off againi to water oni the opposite side in the same way as hydrogeni passes through palladium in the experimnenits of Ramsay.4 The theory to which my experiments lead is very much the same as that to which Pictoni anid Linder have inidependenitly arrived.' According to these observers, there are two stages to be distiniguished. Stage I. The ' coagulationi stage ' in which single ionic initerchange takes place between the ' fibre substanice' (colloid) anid the dye, resultinig in the separation of insoluble dye derivatives retainiing a feeble charge. Stage 11. The 'colour absorption' stage, in which coagula produced in stage I attract and retain the oppositely-charged particles of the dye substance. Mv experimienits show that no actual precipitation of the dye must take place except in the substanice to be stained, anid I slhould be iniclinied to modify the above theory by omittinig stage 1, anid reading inistead of 'coagula produced in stage 1 ' simply 'colloids of fibrcsubstance.' These bodies usually having a negative charge, it is I. Proc. Roy. Soc. 76 B, p. 225, 1905. 2. Journ. o0 Physiology, 33, p. 359, 1905. 3. Ber. Deutsche. Chem. Ges., 1878, p. 823. 4. Phil. Mag., 38, pp. zo6, 2i8, 1894. 5. Journ. Chem. Soc., 88, p. I,935' 0905.
BIO-CHEMICAL JOURNAL
2I6
nlecessary, in order for them to take up an electro-negative dye, that the latter shall be discharged by the presenice of a kation. When the dye is electro-positive, onl the other hand, nieutral salts are unnlecessary and usually deleteriots, since both classes of dyes are most sensitive to the katioii which would be inihibiting in the case of positive dyes. In certaiin cases alkali inicreases the adsorption of these latter, presumably owilng to the powerful actioni of the OH iOt more thani counteracting the opposite effect of the kationi. The commoni practice of using basic dyes in stronig solution of borax anid acid dyes in acid solution is in agreemenit with this view. It appears, that in the case of certaitn acid dyes, such as conigo-red, the concenitrationi of the electrolyte used in practice is un niecessarily high. 5. The Staining of Histological Preparations Here again the balance of evidence is, as it seems to me, in favour of-the adsorptioin theory.' It may be that there are exceptional cases where true chemical combinationi takes place, but they appear to be rare. This being so, it is obvious that the part played by electrolytes must be taken inito account. If electrolytes are split off from livilng cells when they die or are injured it is clear why these cells readily take up acid dyes unider such conditions ; moreover, since electrolytes are unniiecessary wheni the substanice to be stained is electro-lnegative, it is also clear why livinig cells cani be stained with basic dyes, if, as seems probable, the structures in question have a negative charge. The reactioni of the tissues is oni the alkaline side of neutrality anid, as Hardy has showtn, globulin is electro-negative in alkalinie solutioln, anid Pauli has showni the same for albumini. I am niot forgettinig the work of Overtoni oni the solubility of basic dyes in the 'Plasmalhaut,' and the insolubility of acid dyes thereini. The action of electrolytes must also, nevertheless, make itself felt. An interestinig case, which at the same time presenits conisiderable difficulty as to its explanationi, is the stainiing of granules in the axiscylinider of nerve-fibres. These graniules are stained by nieutral-red as I. See
Alfred Fischer, 'Fixierung, Farbung, etc. d., Protoplasmas.' Jena, ;899.
ASPECTS OF ADSORPTION PHENOMENA
2
showni by Macdoniald', but only in the positioni anid neighbourhood of an injured spot. They corresponid to the distributioni of potassium ionis, and are interpreted by Macdoniald as precipitates of the dye caused by potassium. He finds, in fact, that fairly strong solutionis of potassium salts are capable of Throwing down the dye. Since nieutral-red is a basic dye these results seem at variance with the actionl of neutral salts on adsorptioni of these dyes as described in the precedinig pages. I thought it of initerest therefore to see-whether nieutral-red behaves differenitly to the basic dyes on which my experiments were chiefly performed. I founid it, however, to behave quite in the same way. I did not, indeed, niotice aniy tenidenicy to precipitationi eveni by 2-6 per cenit. Kc1. There are evidenitly, then, some other factors at work in the case of the staininig by neutral-red of the graniules produced by inijury in the axis-.cylinder- of -nerve--fibre. Since con-go-red slhowed itself to be narticularly sensitive- -to electrolytes it would seem a suitable reagenit for the detectioni of electrolytes, if split off from injured tissue. I have tested the behaviour of nierve-fibres to this dye. It is easily seen, utnder the microscope, that only the cut ends of the nierve are stained and not the uninijured parts of the fibre. I could niot detect any appearance of granules, but did not look for them unider a lhigh power. It was to be noted also that conniiective-tissue fibres took the staini, as' I imaginie because they- were nlon-livinlg. The fixatiotn of adsorbed dye by heat has already been poinited out. This fact is of initerest in coinnectioni with Altmann 's method of usinig acid-fuchsini. The fixationi of dye by electrolytes has also been meintioned. Recenitly Emil Mayer2 has showni that the affinity of the Nissl bodies of nerve-cells for basic dyes is abolished by previous treatmelnt with neutral salts. This is in complete concordanlce with the results I hiave described. It occurred to me that, conisiderinig the opposite actioii of nieutral salts oni electro-negative and electro-positive dyes it should be possible I . Prcc. Roy. Soc., 76, page 325, I905. i. Hofmecister's Beitrage. 7, p. 560, 1906.
218
1310-CHEMICAL JOURNAL
to vary the colour of paper stainied il a mixture of two dyes of opposite character by the additioni of electrolytes. This proved nIot to be so, the reasoni beinig the formationi of dye-comiipouaids, which appeared to be unaffected by electrolytes. One of these dye-compounlds, that of eosii and methylenie-bluc, is well-kniowni to histologists anid is generally regarded as a salt of the methylenie-blue base with the eosini acid. Similar cotnpounids are formned by most pairs of acid anid basic dyes. They are very insoluble in water though soluble in excess of either conmponienit. This latter fact suggests the possibility that they are lnot true compounids but adsorptioni-compounds of oppositelycharged colloids. As showni by Biltz mutual precipitationi occurs in such cases aid soluLtioll in excess of either colloid. Oni this accounIt I have devoted some attentioni to these bodies. In the first place, if true chemical compounids are formed by a process of double decompositiotn the inorganiic componienits should be founid combined together in the solutiot. I tested this in the case of anilini-blue and methyleneblue by taking abouLt 3 grams of the latter anid the molecular equivalent of the aniilin-blue. The precipitate was filtered off and the filtrate evaporated to dryness in order to estimate the chloride present. Unfortunately, by an omission to make a note of the amounit of the filtrate from the silver chloride, the exact value was niot arrived at. From the quantity of silver niitrate required to *completely precipitate the chloride, -it was obvious, notwithstaniding, that practically the whole of the chlorinie of the methylene-blue was conitainied in the filtrate. The precipitate was, therefore, a compounld of the methylene-blue base with the aniilin-blue acid, the sodium of the aniilini-bluc havinig at the same time combinied with the chlorine of the methylene-blue. I should hesitate to state definitely whether this kind of combiniation was inconisistent with a process of adsorptionl, especially in the light of vani Bemmelen's results ; oni the face of it, at the same time, it appears more like a true chemical compound. Dr. J. H. Scott iniforms me that he has nioticed that a nlew absorptionspectrum makes its appearance when solutionis of eosini and methyleneblue are mixed. I have repeated this observation in the following way A solutioni of methylenie-blue was takeni of such a strength as
ASPECTS OP ADSORPTION PHENOMENA
219
to show distinctly the two dark banids oni the red side of the D-linle in a certaini cell. Ani equivalenlt molecular solution of eosini was takeni, so that wheni mixed in equal volume there should be two molecules of mnethylenie-blue to onle of eosini. This eosini solutioni was place(d in a cell of equal depth to that conitaininig the methylenieblue anid the two cells theni placed before the slit of the spectroscope. The combined spectrum of the two dyes was seetn. Equal volumes of the two solutionis were theni mixed, a portioni of the mixture at onice placedl in both cells anid these theni observed. By taki.ng a double layer it would seem that there would be no chanige in the spectrum uniless chemical chaniges had takeni place. What onie sees is that the banid of methylenle-blue nlext the D-line has almost completely disappeared, leavinig a faint shade. No chanige is to be seen in the banid in the red, nior in the eosini banid at F. The same disappearanice of the one banid of methylenie blue may be observed wheni any other acid-dye such as aniilini-blue or conigo-red is takeni inistead of eosill. Oni the other haid, if thionlini or toluidini-blue is takeni instead of methylenie-blue, nio obvious chanige is produced in the spectrum until precipitationi occurs, when, of course the banids become much fainiter. I believe the explanationi of the disappearance of the methylenie-blue band is to be fotunid in the fact that it is the first to disappear on mere dilutioni. Now wheni ani oppositely charged colloid is added to a solutioni of the blue the precipitate does not fall at once and it might be supposed there was nlo importanit change ; but, if the experimenit is made in a vessel through which a beam of bright light is passing, the track of the beam is barely visible in the methylenie blue, but almost immediately oni addinig the other dye the beam beginis to appear, gradually becoming brighter as the particles inlcrease in size. In poilnt of fact then, methylenie blue is really taken out of solution, and the first banid to disappear would be the onie niext the D-linie. The other thiazini-dyes, so far as inivestigated, do not show the same behaviour oni dilutioni, the various banids all disappear or fade in nearly the same
proportioni. These dye-compounds, although almost insoluble in water, are soluble in alcohol, so that it might be supposed that by taking a strong
220
BIO-CHEMICAL JOURNAL
alcoholic solution of a methylene-blue compounid, onie Should see both the methylene blue bands. This is niot so, and is due to the fact that in alcoholic solutiotn of methylene-blue the band niext the D-line is not to be seen at all distinictly. If the solutionl showinig onily the other band is gradually inicreased iln depth or conicentrationi, it is seen that this band broadens out unitil it reaches the D-line, but there is no indication of a gap betweeni the banids. A fact which confirms the interpretationl I have giveni of the change in the methylenie-blue spectrum produced by combitnation with ani acid-dye, is that scharlach R. in weak alcoholic solutioni produces the same change, anid also a precipitate of a dyecompound. Now scharlach R. is ani inidifferenit dye, not forminig salts, but forms ani electo-negative colloidal solution ; so that it would seem that here, at all evenits, the precipitate must be ani adsorptionl compounid. A similar state of affairs is to be observed in the case of rosolic acid. Here the possibility of double decomposition is also absent, except, of course, that by mass-action a small proportion of methylene blue rosolate miglht be formed. The onily instanice I have found where there is a real chanige in the spectrum is in the mixture of methylelneblue anid picric acid ; picric acid is, however, a. powerful reagenit, anid much weight canniiot be laid oni this fact as far as conicernis the questioil before us. In .mixing methyl-violet atnd an acid-dye no change is to observed ini the absorption spectrum. The absence of chanige in the absorption-spectrum of these dyecompounds does niot, of course, exclude the possibility of their beinig true chemical compounds. With regard to the properties of these bodies, they are indiffusible, as such, through parchmenit-paper. In dilute watery solution they appear to be dissociated ; for, oni dialyzinig the eosin-methylenie-blue compoulnd, for example, it will be seeni that after a time eosin begins to pass through, an-d is ultimately followed by the methylenie-blue. Since this observation was made oni a solution of the pure body, which had been thoroughly washed with cold watcr, the eosini anid methyleneblue must have passed out as the free acid and base respectively. This dissociation is also shown by the fact that, at Iooo, the solution is a fairly good conductor of the clectrical currenit, and is electrolyzed by a cotnstant current.
ASPECTS OF ADSORPTION PHENOMENA
221
Although these compounds are, no doubt, colloids, they do not appear to carry a charge-or only a very small otne. Electrolytes have a very slight effect in the senise of retardinig adsorptio ti. In the boundary apparatus the behaviour ofanilin-blue-methylenle-blue uniderelectric stress is difficult to initerpret, but sinice the phenomenia are the same in neutral acid, or alkaline solution, it does nlot seem to be a question of electric charge. What happenis is this: after the current has passed for sotne time it will be seeni that electrolysis has occurred, so that on the anode side there is a layer of anilini-blue solution, and on the kathode side one of methylenie-blue. The upper boundaries of these two are at the same level in both limbs, but the methylene-blue layer being deeper than the anilini-blue it nmakes the level of the unialtered compounid on the kathode side lower than oni the aniode side, so that it appears as if nlegatively charged anid movinig to the aniode. The behaviour to electrolytes, if atny at all, is, on the conitrary, in the senise of a positive charge ; so that I am iniclinied to thinik that the behaviour in the electric field is, in somne way, due to differ-enit velocities of the two iolls. In the case of the eosin-methyleine-blue there was no similar difference in level of the unaltered dye, so that it may be that here the two ionis are more nearly equal in velocity. This absenice of proof of aniy definite electric charge makes the compounds less initerestinig from the theoretical poinit of view in some respects. But it is, I suppose, what might be expected if these bodies are formed by the mutual nieutralizationi of electro-positive and electronegative colloids. Oni the whole it seems impossible to give, in the presenit state of kinowledge, a decided aniswer to the questioni as to the nature of these dye-compounds. 6.
Antitoxins It has beeni shown by Craw' that the combiniationi between toxin and anititoxin follows more closely in its nature that of adsorptioni than that of chemical combination. The puzzlitng fact knowni as the 'Danysx-voni-Dungerni pheniomenoni' is, for exanmple, satisfactorily Journz, of Hygiene, 5, p. I15, I905, and Proc. Roy. Soc., 76 B, p. 179, 905.
BIO-CHEMICAL JOURNAL
222
explainied by what I have called in the earlier part of this paper the 'law of adsorptioz:.' In illustration I may cite onie experiment: Two filter-papers of equal size (I2 cm. diam.) were each cut into eight pieces. Two flasks, each contaillilng 50 c.c. of a dilute congored solutioni were taken, anid to one of these the whole of onie paper was added at onice, to the other piece by piece at initervals of about twelve hours. After all the paper had beeni added the amounits taken up from the solution were: ... ... Added altogether ,, piece by piece at intervals
... ...
37 % 48 %
It is advisable in this experimenit to wait a conisiderable time betweeni the additioni of each piece of paper oni account of the slow attainimenit of equilibrium in the case of conlgo-red alnd paper. 7. The part played in enyzme-action The interest of adsorptioni pheniomenia in this regioni is in coinnectioni with the combiniationi between enizyme anid substrate, which so much recetnt work inidicates as anl inidispenisable coniditioni of attack. The fact that enzymes are carried downi wheni a colloidal precipitate is produced in a solutioni conitainiing them, has beetn lonig kniown anid used as a means of preparation. But whether the uniotn between a more or less specific enlzyme anid the body hydrolyzed unider its influence is to be looked uponi rather as a true chemical combination is a matter of dispute. Certain experiments lhave been made by Dauwe' as to the takinig up of enzymes by colloids. These were donie chiefly oni pepsitn and various proteinis, a few oIn the taking up of pepsin by agar anld of emulsin by coagulated egg-white. Dauwe comes to the conclusioni that the process is not otne of adsorptioni, since the same weight of boiled eggwhite takes up the same amount of pepsin whether the egg-white is in large pieces or fine powder. He considers that the hypothesis of solid solution applies to the case better. It seems to me that this coniclusioni is scarcely justified. Coagulated egg-white is niot a homogenieous solid body, but porous, like charcoal, anid therefore its active surface must not i,
Hofmneister's Beitrage, Vi.
p. 426, 1905.
ASPECTS OF ADSORPTION PHENOMENA
223
be restricted to the externial surface of the pieces of various size. Porous solids like charcoal, eveni in lumps, adsorb gases, and that niot only on their surfaces but throughout their substanice. Dauwe, indeed, appears to have thought of this possibility, but regards adsorption as the same thing as solid solutioni in such cases.' The distinction is, surely, analogous to that betweeni true solution and colloidal solution. Although recenlt investigationis teach that we must not draw a hard anid fast linle between these two kiinds of solution, there is no doubt that the molecules or particles in suspensioni change their properties as their dimensionis increase beyond what are ordinarily called molecular, anid that they begini to have the properties, due to surface, of matter in mass. In the same way, as I think, we ought to keep the name 'solid solutioni ' for such cases as alloys of metals, and give the name adsorptioni not onily to the taking up of substanices from their solutiotns by surfaces, in the usual sense of the word, but also by suchi surfaces as those forming the walls of pores. The criterion by which any given case is to be decided, is, of course, the way in which the relative amounts taken up vary as the concenitrationi of the body taken up changes. Dauwe rejects the hypothesis of chemical combination oni the ground that the enizyme takeni up can be extracted again by an appropriate solvenit, for instanice, pepsin from coagulated egg-white by a solution of egg-white. He regards the process as one of solid solutioni, or relative solubilities of the enzyme in the substrate anid water, and brings forward itn support of this view the experiments of Reichel anid Spirot on the a?parent loss of renniet in the process of clottinig of milk. These observers showed that the disappearance of the enzyme was to be accounted for by the taking up of it by the clot. Now, as I uniderstand their results, they are rather in favour of adsorption. If it were a case of sDlid solutioni the percentage loss would be the same whatever the rennet concentration, but if onie looks at the table oni page 48i of the seconld paper referred to, it is seen that the percentage loss steadily increases as the concenitratiotn of the I. Loc cit., p. 443. z.
Hofmneistcr's Beitrage, VI, p. 68, 1905, and VII. p. 479, 1905.
2.24
BIO-CHEMICAL JOURNAL
enzyme falls ; in other words the more dilute the solution the more adsorption takes place. This is particularly well shown by the last experiment on the table. The mathematical expression given is also in accordance with this interpretation. If the uniioIn of etnzyme anid substrate follows the law of adsorption it would be expected that there would be found some effect of this law on the relation of activity of enizyme to its concentration. In my paper on The Kinetics of Tryptic iction,' I showed that the form of the function in question varied according to the stage of the reaction, but if we take the early stages in which the action is most rapid there seems to be some effect suggestive of an adsorption process. The initial linear stage is very short, so that we may omit it, and, if we compare the relative times taken by concentrations of trypsin varying as 8, 4, 2, and i to produce an increase of cotnductivity of i8oo gemmhos in caseinogeni solutioI, we find the values are Relative trypsin content 8
...4
4.
...
79
2
..
126
I
...
233
Time taken in minutes
so that the lower concentrationis of trypsin are relatively more active. It is somewhat remarkable that the value of the exponcent founid in aniother experimenit was II42 to i672 and this is the same as that giveni by Reichel and Spiro in their paper for the ' Teilungsfactor ' of reniniet, ViZ., I'5 to I-67. Whether this is merely an accidenital coinicidenice I am not prepared as yet to state, the subject being still under investigation.i I have myself made one or two experiments similar to those or Dauwe, which are perhaps of sufficient interest to give here. It was shown by W. A. Osborne that calcium caseiniogenate does not pass through a porous clay filter, trypsin, on the contrary, does do so. If then, we mix solutions of trypsin and calcium caseinogenate, the i. Archives des Sciences Biologiques. Tome XI, Supplt. p. z6I. St. Page 26 of the reprint,
z.
Petersburg, 1904.
ASPECTS OF ADSORPTION PHENOMENA
llg
' compound ' formed should prevent the appearance of trypsin in the filtrate. This I find to be the case. It may be said that what we have here is an instance of mutual actiotn of colloids, but, as I think I have been able to show in the previous pages, there is no esselntial differetnce between this actioni and adsorption by solids from watery solution. Oni the other haid, it might be held by some, that the case is onie of true chemical com)biniation. To test this view I performed a similar experimenit, taking caseiniogeni and malt amylase, arld found that this enzyme is also held back by caseinogen, so that it appears to be merely adsorbed. This initerpretationi of the uniion of enzyme and substrate does nlot, of course, exclude a certain degree of specific relation betweeni a particular enzyme and the substrate hydrolyzed by it, as we have seen there is conisiderable evidence of specific adsorption. It is probable, moreover, that it does not hold, in the same degree, for such enzymes as invertase, maltase, or lactase, wheni the substrate hydrolyzed is not colloidal, and where there seems to be a very close relationship between the chemical structure of the enzymes and the bodies split by them. In another way the results described in the presenit paper are of interest in conniiectioni with enzymes. I refer to the action of electrolytes. From the work of Cole,' McGuigan,z anid others, it follows that there is a certaini oppmSition between the actioni of katiolns anid anionis. This indicates that eniyzmes are possibly electrically charged colloids. Victor Henri,3 in fact, speaks of trypsini as a negative colloid. I have tested the behaviour of a solutiotn of GrQbler's trypsin in the bounidary apparatus, and founid that it does inideed move to the anode, but how far this negative charge is due to the enzyme itself it is niaturally impossible to decide until we have in our hands a pure preparation of the enzyme. It is possible, however, to attack the problem in aniother way. CaseiinogeIn in solutioIn in alkali is electro-niegative, if trypsini is also negative neutral salts, anid especially those of bivalenit kations, such i. Journal of Physiology, XXX, pp. 202 anid 28i, I904. American Journal of Physiology, X, p. 444, 1904. 3. C. R. Soc. di Biologie, LIX, p. I32, 1905.
2.
26
BIO-CHEMICAL JoURNAL
as calcium, should increase- the adsorption. In some preliminiary experiments I have made tllere seems to be some action of this kind, but it is difficult to obtaini the correct relative conicentrationis so as to avoid precipitation. Anl experimenit was also made on adsorption of trypsini by paper as follows Four crystallizing dishes were taketn, each containinlg a circle of filter paper, in two of them there was a watery solution of trypsiln, in the other two a solutioni of trypsini of the same strenigth in 350m0 CaSo4. After stanidinig in a cool place fol 24 hours, a paper froin the watery solutioni was takeni out, drainied for i miniute, placed in a flask, anid heated to ico° in a steam sterilizer ; the same was donie with a paper from the CaSo4 SOlutioI. i 6o c.c. of 5 per cenit. caseinogeni in ammonia were added to each, and they were theni placed in the thermostat at 390; when warmed to the first was added a drained paper from the CaSO, trypsin, and to the second a similar onie from the watery solution. The electrical cotnductivities were determinied at iintervals. There was not a very great differenice between the two, but what there was showed that rather more trypsini had been takeln up by paper under the actioni of caso4. The times takeni to reach a change of 8oo gemmhos were i65 minutes for the trypsini from caso4 solutioni, anid 192 miniutes for that from watery solutioni. A similar experiment taking a m-- solution of toluidini-blue instead of caso4 was unsuccessful onl accounit of mutual precipitation of the dye anid trypsini. The precipitate, wheni filtered off and washed, showed itself to have conisiderable tryptic power. This subject requires further investigation. The importanice of the study of adsorptioni anid surface-actiotn in getneral in connlectiotn with the enzyme-actioni is emphasized by Bredig' in his article on ' Chemical Kinetics ' in the Ergebnisse of Asher anid
Spiro. 8. Oligo-dynamic action It has been poinited out by Pauli2 that the puzzlinig phenomenia called 'oligo-dynamic actioni ' by v. Naegeli, anid conisistinig in toxic T. Ergebnisse d. Physiologie, I, p. 2115 I902. 2. Hofmeister's Beitrdge, VI, p. 257 (footnote), 1905.
ASPECTS OF ADSORPTION PHENOMENA
227
actionis of distilled water which has been in contact with polished metal, are, in all probability, due to the presence of colloidal metallic hydroxides. Various solids, such as paper, are knownl to remove this toxic property. It occurred to me, therefore, that perhaps the adsorptionl of congo-red which occurs to some extenit, even from distilled water, might be due to these hydroxides. I found that the ordinary laboratory distilled water caused considerably more adsorption of conigo-red thani that which I had prepared myself, usinig a tini conidenser. From this latter, nevertheless, 22 per cent. was taken up. By allowing filterpaper to soak in this water previously the value was reduced to i 8 per cent. It aid niot seem possible to get below this-no doubt the electrolytes dissolved from the glass anid the traces remaininig in the paper were sufficient to account for the effect. VIII. SUMMARY OF RESULTS i. The hyperbolic form of the curve of adsorption is conifirmed. 2. The curve of electrical conductivity of successive distilled water extracts of gelatini has the same form. 3. It is impossible to wash out all the electrolytes from gelatin except by a practically infinite number of changes of water, each change removing a less percetntage thani the previous one. 4. The electrolytes are, therefore, neither chemically combined nior merely admixed, but in the intermediate form knowni as
adsorption. S. When gelatini has been washed ilearly free from electrolytes it is capable of diminiishing the conductivity of solutionis of electrolytes in which it is placed. This it does by adsorbing them in a tionionized conditioll. 6. The rate at which conigo-red is taken up by paper is greatly accelerated by rise of temperature, but the total amount taken up when equilibrium is attainied is less the higher the temperature. 7. The temperature-coefficient of the reaction-velocity is extremely low, so that the thleory of Nernst as to the part taken by diffusion in heterogeneous reactions seems to apply to the case of
adsorptiotn.
228
22IO-CHEMICAL JOURNAL
8. At low temperatures equilibrium is attained very slowly, at room temperature at least 24 hours being required. 9. The adsorption-compounid of gelatin anid inorganic electrolytes is also dissociated as the temperature rises. IO. Raisinig the temperature rapidly to I000 tenids to fix conigored in paper so that it is afterwards extracted by water with considerable slowness. i I. No evidence was obtained of any production of heat in
adsorption. 12. The reaction between conigo-red and cellulose is reversible, as is also that between gelatin and electrolytes. I3. Dyes forming colloidal solutions are extremely sensitive to electrolytes in regard to their adsorption. The effect seems to be proportioinal to the degree of their colloidal nature or size of colloid particles. I4. The action of electrolytes may be expressed as follows In the case of electro-negative dyes, like congo-red, kations facilitate adsorptioni, anionis depress it. In the case of electro-positive dyes like toluidin-blue kations depress, anions facilitate. But in both cases the effect of anionis is very small compared to that of kationis. iS. The effect of bivalent kationis is considerably greater than twice that of univalent kations. i6. Salts of the heavy metals which form positively charged colloidal hydroxides in solution have a very powerful effect in promoting adsorptioni of electro-niegative dyes. I7. There is evidence of the carrying-downi with the adsorbed dye of the facilitating ion. i 8. The presence of a stable colloid, like gelatin, protects conlgored from the action of electrolytes. A negative charge seems necessary for this actioni, since egg-albumin, although exercisinig a similar action in alkalinie solution, had the opposite effect in acid solution. I9. The explaniation of the action of electrolytes, as well as that of other electrically-charged colloids is to be found in the niegative charge of nion-conductors, like paper, when immersed in water.
(Quincke).
ASPECTS OF ADSORPTION PHENOMENA
229
20. The different behaviour of silk and paper towards electronegative anid electro-positive dyes, as also the influence of alcohol on the process, is to be explained by the results of Coehn on the ilnfluence of the respective dielectric constants on the charge. 2I. Wheni gelatinl is precipitated by tanniin its adsorbed electrolytes are split off. 22. There is no evidence of a sudden separation of electrolytes at the moment of death. There is a gradual one when a living tissue is warmed from i i° C. to 56° C. 23. In the process of clotting of blood there is a diminution of electrical conductivity, so that ions (probably ca + +) disappear from
solutioln. 24. The adsorptioni affinity of gelatitn for acid-fuchsin is greater than that for conigo-red, while that of paper is the same for both. 25. A theory of dyeinig is suggested on the basis of adsorption in relation to electrically-charged colloids. ,26. Ani explaniation, apart from different permeability of the cell, is suggested for the staining of living cells by basic dyes and the nonstaininlg by acid dyes. 27. The niature of the compounds between acid and basic dyes is investigated. They appear to be uncharged colloids, but the evidence as to whether they are colloidal adsorption compounds or true chemical compounds is not decisive. 28. Evidence is brought forward to show that the uniion between enzyme and (colloidal) substrate is of the niature of adsorption. The adsorption of trypsin by paper is facilitated by calcium sulphate.
BIO-CHEMICAL JOURNAL
230
4i
I
u
j4)
4) *0
.0 0 co -
4)00
z
~~~~~~~~~1.
.~~~~~~~~~;r
i. . .
P!o1!O3
*.
AII!God P!0TI03
.
Ic
~
~
f i~O
0.
.4Cfi
.
JAV}U3N 0 0 0
Z
* 2 4.
*
1-
0
0.
:
.++
+I
0 co
+
+
:+
u
co 0 .0
±
+
+
+
+
+
0i
+
+
+
+
+
Al
toi
+
+
0
93
.4
4)
~~~4)
~ C-
0.
0 Coz
o
-
c
z
zQz
z
cn ..-
0
*
E
144
lo
0
0
CY%
*
1010
N
tn
to
U ._
H
0
4)
0i 0
o c4.1"
0
0
0
E 0
41)
" en
D
g0.
co Q C: Z u)
co
Cn
t .Q
o
0.
Q
O
C
eC
sNg
cop_ coQ
°0
X
0
°0
.0
EW
0
0
L0
00
00
°
0i
4.0
0~~~~~~~~~~~~~~~~~~~~~~1
00
0
4'. U
0
X
"
o
z
.4-
~
~~U
Z0 0
Z 4
0~~~~z
i0
vL 0
s
0 0 04
ASPECTS OF ADSORPTION PHENOMENA
231
qJ la .00 = VZ
P!O11°O
JAi}!9Od
*
P!O1°ot
2AEJ232N 0
0
.0
00
*
+ *
+
:
0 .
._0
u
u , 1-0
*
U
0
0
0
0
0
0
*
E0
0 0
e-
b
<0 oE
- D
o
0
0
o
0
°o
0*
00
3
0
0
.0
0
D
,
0
0
0
0 ,,2
a
0~~~~~~~~~~~~~~~~
M00
.
E
_
-
.
U
0°~
O
--.0
0
4
oU c
00
<_
4
_
o0
Uv
0 c
"
U .*_
'"
"
0
2.32
23
BIO-tCHEMICAL JOURNAL
NOTES TO TABLE OF DYES The molecular weight ot methyl-violet is given as that of the penta-methyl derivative. Di.fuuion. The words 'very slow' mean that a certain amount had passed through parchment paper by the end of 24. hours. The statements as to the ultra-microscopic behaviour of dyes in the column headed 'Size of particles' are taken from the work of Michaelis referred to in the text. 'Sub-microscopic' means that the solutions are resolvable into particles. 'A-microscopic' means that they are optically homogeneous. Electric migr-ation. It was difficult to- be certain in the case of methylene blue since bleaching occurred by the electrolytic' chlorine. The measurementsof eltctricalconductivity were taken at 39 C unlessotherwisementioned. They are not to be looked upon as absolutely correct, since the dyes were not specially purified, except the eosin methylenie-blue compound. State of solution. It is impossible to make definite statements as to certain dyes, eg., eosin, since they have some properties of colloids, but not all. The columns headed 'Kations? and v Anions' refer to the action of these bodies on adsorption by paper. It is to be understood that in all cases the action of Anions is very much less than that of Kations. The columns 'Negative' and: ' Positive colloid ' refer to actual precipitation, not
adsorption. The column ' Stable colloid ' refers to the effect of this on the facilitation or inhibition of adsorptiQn by kations.