SOME CHEMICAL REACTIOSS OF COLLOIDAL CLAY BY RICHARD BRADFIELD
At the first Colloid Syniposium seven years ago, the author had the privilege of presenting a paper on the same subject1 as the above. I n preparing this paper the first was reread with considerable interest and amusement. A friend, noted for his frankness, once remarked that the best part of the earlier paper was the discussion which followed! As this discussion was reread I wondered if our committee had acted wisely in abandoning the policy adopted by Professor Mathews of publishing a t least a r6sumB of the discussions in the Symposium Annual. An opinion the earlier paper shared by many was that it was rather futile to attempt to obtain an understanding of the reactions of a substance as complex as a clay. “Simple” substances like silicic acid, aluminum and iron hydroxides, pure carbon, and pulverized soil-forming minerals were suggested as being much more suitable objects of study. The author was not completely indifferent to this advice and a t times has used all of these “simpleJJsubstances in comparative studies with clays; nor was he convinced that it was wise to abandon clays completely for these simpler systems. The economic importance of clays to both agriculture and industry seems sufficient to justify some study of this material in spite of its complexity. Most comparisons of the behavior of natural clays with these simpler, supposedly similar substances, have served but to show how different the two really are. For every property found in common, several can be found in which they are very different. This has led to the feeling that the only way to find out about clays is to study clays. Synthesis is often the capstone of a chemical research problem but analysis logically precedes it. If we learn enough about the behavior and properties of clays we may eventually be able to make synthetic imitations. But first we must know a little more about what we are trying to imitate. The term clay means very different things to different people. TOthe soil scientist a clay is that inorganic fraction of the soil whose particles are less than z microns in diameter. Colloidal clay is that fraction of the clay whose particles have an effective diameter of less than I O O millimicrons. Since it represents a fractionation based solely on size of particle it may be applied to fractions which differ greatly in both chemical composition and in mineralogical make up. Since the classic work of OdBn and Svedberg many attempts have been made to determine the size-distribution of clays. The chief difficulty encountered is to secure suspensions of stable unit particles. Most of the clay exists naturally, fortunately for the farmer, in the form of rather stable aggregates. The various methods of deflocculation commonly used were not 1
Bradfield: Colloid Symposium Monograph, 1, 369 (1923).
SOhlE CHEMICAL REACTIONS OF COLLOIDAL CLAY
361
I O O per cent efficient and consequently the older mechanical analyses represented neither the size-distribution of the naturally occurring aggregates nor of the unit particle, but something in between these extremes. Recent studies of base-exchange phenomena and their effect upon the electrokinetic potential of clays, which will be discussed later, have pointed out the way to obtain more efficient deflocculation. It is now quite definitely proven that the amount of colloidal material in clays is much larger than earlier workers believed. Most of them, Schloesing, Ehrenberg, Gedroiz, et al, felt t h a t the amount of truly colloidal material in clays was a t most only a few per cent. Recent work has revealed clays with 80-90 per cent colloidal material, while from 20-50 per cent is quite common. I n fact there is a decided tendency t,o regard all clay particles under 2 microns in diameter, as defined by the International Society of Soil Scientists, as being colloidal. The Bureau of Soils’ workers were the first t o suggest that the limit should be placed a t I micron. Joseph, working on African clays, came to the same conclusion. Some experiments made by DeYoung in the author’s laboratory furnished strong evidence that pract’ically all the particles, in certain clays a t least, that are under z microns in diameter are in reality made up of aggregates which may be broken down to particles about 10-20 millimicrons in diameter by merely churning with distilled water, centrifuging out particles larger than z o millimicrons by means of the supercentrifuge, resuspending the coarser particles in distilled water, recentrifuging, etc. The original purpose of this experiment was to prepare a series of fractions of clays with different specific surfaces and to study the absorptive power as a function of external surface. It proved a very easy matter to secure large quantities of clay with part,icles between 2 0 0 0 and I O millimicrons; but it did not prove possible t o make satisfactory fractionations of this material. After 8-10 fractionations all of the original kilogram of clay with particles under 2 0 0 0 millimicrons was reduced to a fraction with particles from 10-20 millimicrons in diameter. The amount of stable particles between 1000 and 2 0 millimicrons was negligibly small in this clay. The 2-5 microns fraction was very stable and easily washed free from all smaller particles. It possessed, however, none of the properties usually associated with clays but was decidedly silt-like in nature. Thomas, working in Utah, and Joseph, working in Africa, have observed that the smallest clay particles in the samples studied by them were of the same order of magnitude as those found above. The almost total absence of particles of the intermediate sizes may prove quite significant if found to be generally true. It is hoped that the investigation can be extended to other clays using an improved technique. Shape of Particles. A large proportion of the particles of most clays are plate-like in form. The stream lines observed when a clay sol is gently stirred is evidence of this. The particles can be oriented also by means of an electric field. The number of particles visible in the ultramicroscope can often be increased from 20-40 per cent by applying an electric field perpendicular to the direction of the illumination. These oriented sols are strongly doubly re-
362
RICHARD BRADFIELD
fracting, the amount and sign of the double refraction varying with the nature of the cation saturating the clay. The surface of clays is tremendously increased as a consequence of this plate-like shape. Variations in plasticity may frequently be associated with the ext,ent of this plate-structure development. X - r a y and Chemical Analysis. Results obtained by the X-ray analysis of clays were discussed before this symposium last, year. Kaolinite which many used to consider the predominant mineral in clays seems to be rather rare in agricultural soils. Minerals of the nontronite, beidellite, and montmorillonite types seem to be much more abundant. Chemical analyses of the colloidal fraction of clays have revealed the rather interesting fact that clays formed under similar climatic conditions tend to be very much alike in spite of great differences in the origin and nature of the parent material. The colloidal fraction of most of the soils of the corn belt regions of this country has a silica-sesquioxide ratio of about 3 to I . Other ratios are found, of course, but this one is most common. As we approach the tropics the proportion of sesquioxides increases and in extreme cases, the so-called lateritic clays, the silica disappears almost completely. All gradations between these extremes can probably be found. These considerations render extremely improbable the view held by some, that there is a single base-exchange complex or a single alumino-silicate responsible for soil acidity phenomena. Thus far an attempt has been made to portray the physical and chemical make-up of some of these colloidal clays as a preface to a discussion of some of their chemical react,ions. All the studies to be considered were made on the colloidal fraction of clays with particles all under I O O millimicrons which apparently were rather instable and capable of being broken down easily into particles from 10-20 miilimicrons in diameter. These particles are crystalline and are for the most part plate shaped. Mineralogically they seem to belong to either the montmorillonite or beidellite group. Cation Exchange Reactz'ons. The long-known ability of soils to exchange a certain definite amount of their cations for the cations of neutral salts is associated largely and in some soils almost exclusively, with this colloidal fraction. This exchange capacity varies commonly from 0.3 0-1.oo milliequivaImts per gram. The clays differ from the synthetic permutits, with which they are so often compared, in that a much smaller percentage of their total cations are easily exchangeable. In the permutits which have the general formula I base.hlzOa-3Si0?.xH~0almost I O O per cent of the cations present can be rather readily exchanged for the cations of a neutral salt. This corresponds to from 3 . 2 t o 1.j milliequivalents per gram depending upon the amount of hydration. The total amount of cations in the permutit and clay are almost identical. The exchangeable fraction constitutes then from but 10-30per cent in clays ordinarily in comparison with 9;-100 per cent in permutits. This difference is commonly explained by assuming a much more porous structure in the case of the permutit, the so-called permutoid structure of Freundlich in which the entire interior surface is readily accessible to the ions
SOME CHEMICAL REACTIONS O F COLLOIDAL C L S P
363
of the common salts. The clay particle is apparently denser in structure and the rapid exchange of ions is limited to those situated near the surface. Given sufficient time, suitable concentrations, and higher temperatures, the exchange can go to completion also in the case of certain closely related ions which are capable of fitting into the crystal lattice of the particle, as was shown over j o years ago by the studies of Lemberg. Kelley2 has found that the percentage of readily exchangeable cations in a bent'onitic clay could be greatly increased by prolonged grinding. The effect of the grinding was apparently due to the opening of fresh surfaces. The additional base obtained was largely Mg. There is some evidence that this may be rather generally true. Ca and H on the other hand constitute the largest proportion of t'he exchangeable cations of the soils of the humid regions. In the alkaline soils of arid regions Xa appears in the place of H and, in extreme cases, of Ca also. The proportion of total Ca which is readily exchangeable is usually quite high, in many cases almost I O O per cent, indicating that it may be the product of a secondary exchange reaction rather than a part of the surface of the original particle. The fact that Ca occurs in the drainage waters of our soils of the humid region in greater abundance than the other cations, makes such an hypothesis probable. Acid Clays. In humid climates any reserve of CaC03 which mag have been present in the soil material is eventually leached away. After this reserve is exhausted the bases on the surface of the colloidal particles are gradually replaced by hydrogen ions supplied by the carbonic and other acids which are formed largely as a result of bacteriological activities. The extent to which this replacement of the basic ions by hydrogen ions has proceeded is a measure of the degree of weathering of the clay. Natural clays are found in which exchangeable hydrogen constitutes over 60 per cent of the total exchangeable cations. The division between exchangeable and non-exchangeable cations is not an extremely sharp one but in most cases it is sufficiently well defined that comparable values can be obtained by very different replacement methods. Almost identical values can be obtained, for example, by extracting the clay with tenth-normal solutions of strong acids, or normal solutions of appropriate neutral salts. T i t h i n reasonable limits t,he amount of cations replaceable is independent of the concentration of the replacing solution provided the extraction is continued to completion. Electrodialysis. If most clays are subjected to electrodialysis a rather definite quantity of bases can be removed. The endpoint is usually rather sharp. The amount removed is found to be identical with that which can be removed by the acid or neutral salt extraction methods. The resulting clay is saturated with hydrogen ions. It is free or practically so from soluble salts and non-colloidal acids. Any ion which may have been present which was small enough to pass through a parchment membrane has been removed by the prolonged application of the electrical potential. This electrodialyzed
* In a
paper presented before Am. Soc. Agron., Chicago, Nov. (1929).
364
RICHARD BRADFIELD
colloidal clay represents an attempt to obtain a simpler system by methods which are not sufficiently drastic to cause any deep-seated change in the colloidal particle. As evidence of this, it has been found possible to put an electrodialyzed clay through a cycle of chemical reactions, for example, to neutralize it with Ca(OH)2, replace the Ca by prolonged leaching with a neutral NaCl solution, and then electrodialyze again, obtaining a product apparently identical with that with which we started. I t has the additional advantage that it can be brought easily to any desired degree of saturation with any of the important cations by the addition of the proper amount of the appropriate hydroxide. Characteristics o j Hydrogen Clays. I . Influence of the solid phase upon the hydrogen-ion concentration of clay sols. I t has been long known that the potential of the hydrogen electrode, when immersed in a suspension of a carefully washed acid clay, was markedly influenced by the concentration of the suspension. The colloid-free aqueous extract of such soil was frequently found to be almost neutral. In earlier studies on the acidity of a very acid colloidal clay it was found that the relationship between the pH value of the clay sol and its concentration was very similar to that observed with weak acids such as acetic3 On the basis of these and other experiments the hypothesis was advanced that the bulk of the acidity found in soils was due to acids whose anions were of colloidal dimensions on using the terminology of Michaelis, to acidoids. As the clay used in these earlier experiments were natural clays, only jo per cent of whose exchangeable cat,ions were hydrogen, it seemed that further work should be done on the clays saturated with hydrogen ions. The studies served naturally to magnify the differences between the clay and its aqueous extract. With an electrodialyxed bentonite, for example, it has been found that while the clay-paste collecting on the membrane of an ultrafilter and containing about I O per cent of oven dry clay had a p H value of 2 . 2 as measured with the quinhydrone electrode, the clear ultrafiltrate had a pH value of 5.2. In a second experiment a I . j per cent electrodialyzed bentonite sol was placed in a collodion bag and the bag set in a volume of distilled water equal to that of the clay. After standing 24 hours so that equilibrium might be established it was found that the bentonite sol on the inside of the bag had a pH value of 2.8 while the water 011 the outside gave a pH value of 5.4. The hydrogen, quinhydrone, and antimony electrodes all gave pH values in satisfactory agreement, with these sols. It is felt that the function of the membrane in these experiments was merely mechanical, preventing the diffusion of the colloidal clay anions. Any Donnan effect resulting from traces of any diffusible acids which might have been present would serve to decrease the differences in hydrogen ion concentration observed on the two sides of the membrane. Any other mechanism for separating colloidal clay from the medium in which it was 3Bradfield: J. Phys. Chem., 28, 170 (1924)
SOME CHEMICAL REACTIOXS O F COLLOIDAL CLAY
365
suspended would probably give identical results. Similar results have been obtained by using centrifugal force for the separation. The differences in pH value between the clay thrown down and the supernatant liquid were not as great as in the case of the ultrafiltration experiments due to the fact that all of the clay particles could not be thrown down with the centrifugal force available. The simplest explanation of these observations is that a part of the hydrogen on the surface of the acid-clay particle is ionized and is far enough removed from the particle to act as a n ordinary “free” hydrogen ion when brought in contact with the electrode, but these ions are restrained by electrostatic forces from moving farther than this distance.
FIQ.I
Titratable Acidity o j Electrodialyzed Clays. As has been shown earlier, even the natural clays give fairly distinct inflection points when titrated potentiometrically or conductometrically with solutions of standard hydroxides. The use of electrodialyzed clays tends of course to make these endpoints more distinct. A group of curves obtained by titrating I O O cc of I per cent electrodialyzed clay sols with 0.1N NaOH using the hydrogen electrode are shown in Fig. I . The curve for Putnam clay is of the type most commonly found in agricultural soils of the corn belt section. It resembles the monobasic type in having but one endpoint. This endpoint is not as sharp as those commonly obtained with simple monobasic acids. The reason for this, of course, is that these clay acids are very complex. For example, if we calculate the number of hydrogen ions t h a t must be supplied by a particle with a radius of I O mp in order to account for the amount of NaOH neutralizedup to pH 7.0,
366
RICHBRD BRADFIELD
we find 7 1 jo. To account for this observed titratable acidity the particles of a true monovalent acid must have a radius of only 0 . j p , a value much smaller than those observed. One clay has been found, and so far as the author has been able to ascertain this is the only one reported which gives a curve having two rather distinct inflection points. (Fig. I , curve 2 ) . In appearance this bentonite from Cheyenne, Vyoming, was, superficially a t least, very similar to the one from Rock River. The Si0, content of the two is almost identical-56 per cent. Minerals having two distinct types of crystal patterns have been found in bentonites. I t may be that the dibasic character of the Cheyenne bentonite curve can be correlated with its mineralogical composition. The marked similarity between curves z and 3 a t pH values above 6.j indicates that these clays may have some constituents in common. A comparison of the silicic acid curve 4 with any of the clays shows how little they have in common and consequently how futile it is to attempt to obtain an understanding of the acidity of clays by studying “pure substances” which are so different from clays. The colloidal acids were prepared in a similar way, namely, the electrodialysis of their salts until all possible bases were removed. The curve for silicic acid shows a strong buffer action between pH values 9 and I I , indicating dissociation constants of the order of 10-10: a value in satisfactory agreement with those found in the literature. I t has, however, no buffer capacity in the acid region. Attempts have been made to prepare aluminosilicic acids by combining sodium aluminate and sodium silicate in the same ratios that they are found in clays, then converting the sodium salts into the acid by prolonged electrodialysis. The resulting product had no appreciable buffer action a t pH values less than 7 . Every other attempt made thus far to prepare synthetic substances analogous to clays has resulted in failure when subjected to this acidity test, even though the base-exchange reactions of clays and permutits are quite similar. Some have claimed that the differences in p H values of electrodialyzed permutits and electrodialyzed clays were due to the much finer particles of the latter. This might be true as far as pH measurements are concerned but the ability to neutralize KaOH in an acid medium should not be affected by particle size in substances as permeable to Na ions as are permutits, especially if sufficient time is allowed for the reaction. Since clays which are made up predominantly of silicates and aluminosilicates are so different from any pure artificial substances we have been able to prepare, the question naturally arises as to whether or not other acidforming substances might not be present in sufficient quantity to account for the results obtained. The most common substances in clays other than Si02 that might contribute to the acidity are C, S , S, and P. The analyses of a series of electrodialyzed colloidal clays are shown in Table I. The bentonites are very low in all of these elements. The nitrogen content is not given but in clays it is usually only one-tenth as abundant as carbon. If one considers that all of the carbon is present as carboxyl groups, all the S as H,SO,, and all the P as
SOME CHEMICAL REACTIONS O F COLLOIDAL CLAY
367
TABLE I Content of Acid-forming Elements in some Typical Electrodialyzed Clays' Colloidal Clay C S P I . Bentonite-Cheyenne. Finest 0 073 0 011 Trace 2 . Bentonite-Cheyenne. Regular 0,198 0 00; Trace 3 . Bentonite Rock River 0.ojj o 019 Trace 4. Putnam 0 014 0,040 0.977 0.184 0.010 0.129 j. Putnam (H202 treated) 0.667 0.014 0.065 6. Susquehanna 0 . j66 0.01; Trace 7. Boone 8. Sharkey 0.02j Trace 0.591
H3P04,which represents the maximum possible contributions of these substances to the acidity of the clays, we find that their combined acidity equals only about I O per cent of that found by the titration curves. This seems to force us to the conclusion t h a t the acidity of these clays is due to aluminosilicic acids. The differences in properties between the natural clays and the common synthetic alumino-silicates must be due to differences in structural arrangement. Characterization of the Clay Acids. Many studies have been made of the quantitative factor of soil acidity, or amouiit of exchangeable hydrogen, but the intensity factor has received scant attention. The most obvious way of getting a t this intensity factor is by the use of some expression which is analogous to the dissociation constants of ordinary weak acids, which the author has termed the apparent dissociation constant or, if expressed in the form of the negative logarithm, as the apparent pK value of the acid. It is obtained from the mass law equation: pK
=
pH
+ log salt/acid
At the point of half neutralization the last term becomes o and the apparent p f l value is numerically equal t'o the pH value a t that point. Such a treatment is of course not strictly rigid but it has the virtues of simplicity and usefulness and enables us to make comparisons with other acids. By inspection of the titration curves we see that the apparent pK value of the Putnam clay is about 5.6, that of the Rock River bentonite 3.8, the Cheyenne bentonite 3.6 and 5.9, while that of silicic acid is of the order of IO. It has been pointed out in a n earlier paper that the relationship between the hydrogen ion concentration of a clay and its concentration is similar to that observed with weak acids such as acetic and that by the use of the simplifying assumption that the concentration of the unionized acid is equal to the total titratable acidity, the relationship can be calculated with a fair degree of accuracy from the mass law. The pK values calculated from the titration curve and from the pH concentration relationship are in satisfactory agreement, Analyses obtained through the kindness of Mr. C. S. Schollenberger.
368
RICHARD BRADFIELD
Distributzon of a Base between Two -4cids. If the apparent pK value has the significance attached to it above, it should be possible to calculate the pH value that would result when an acid clay is treated with an equivalent amount of the salt of a second acid of known pK value. Under these conditionss the relationship, - =
1-x
KZ
holds, in which x represents t h e
amount of base combining with the acid, whose dissociation constant is K1, and I -x, is the amount combining with the second acid. I t is very easy to test this equation in the case of the colloidal clay acids because they can be separated from the second acid formed by merely centrifuging and titrating the clear supernatant liquid. The results of a series of such experiments are shown in Table 11. The agreement between the calculated and measured pH values are in most cases as good as one could expect. I t is possible then, to calculate ( I ) the relationship between the pH value of clay suspensions and the concentration of such
TABLE I1 Distribution of a Base between an Electrodialyzed Bentonite and Certain Organic Acids Concentration of Salt Millimols per liter IO
Monochloracetic pK = 2.81 PH Acid freed Found Calcd Found Calcd
Lactic pK = 3.85 PH Acid freed Found Calcd Found Calcd
3.50
3.42
2
13
2.01
3.93 4.29
3.88
4.14
4.11
5.12
3.96 4.95
4.03 3.97 4.57 4.36 4 73 4 65 j . 1 1 4.96
25
50 IO0
3.60 5.58 6.10 -
4.33 5 90 6.80 5.75
Acetic PK = 4.74 PH Acid freed Found Calcd Found Calcd
suspensions, ( 2 ) the pH values resulting when clay acids are treated with various increments of standard hydroxide solutions, and (3) the reaction resulting when clay acids of known combining weight are treated with the salts of acids of known pK value. This use of the pK value as an expression of the intensity factor of the acidity relations of clay acids is admittedly only an approximation but we are aware of no other method of treatment which enables us to predict as many of the reactions of clays. 5
W. C. McC. Lewis: “A System of Physical Chemistry,” 2nd Ed., 1, 237.
SOME CHEMICAL REACTIOSS O F COLLOIDAL CLAY
369
There is unfortunately one complicating condition, the value of the apparent dissociation constant obtained from titration curves is influenced noticeably by the nature of the base used. This is shown by the curves in Fig. 2 which are taken from a recent study made by Baver6 in the author's laboratory. The curves were obtained by adding increments of the hydroxide solutions t o fixed amounts of electrodialyzed Putnam clay. After standing for several days the pH value was measured with the quinhydrone electrode. The usual lyotropic series is quite evident. In the experiments cited above the
IC
8
6
4
2
-M l l L /fQU/VALfWS Ob BHS€ -FER 1
/OO GMS. !
!
CLRY t
The change in reaction of clays containingvarious amounts of different cations.(After Baver).
pK values obtained with KOH were used and K salts were used in the distribution studies. Natural clays contain more exchangeable Ca than any other ion. The calcium pK value would probably more nearly represent natural conditions. T h e Effect of the A m o u n t and Nature of the Exchangeable Catzons on the Physical Properties of Colloidal Clays. It has long been known that the physical properties of clays were greatly influenced by the nature of the exchangeable cations. Homoionic clays have usually been prepared in the past by prolonged leaching with a neutral salt of the desired cation. Two objections may be raised against this method: ( I ) It is difficult to obtain a Missouri Agr. Expt. Sta. Research Bull. 129 (1929).
3i o
RICHARD BRADFIELD
definite predetermined amount of replacement and ( 2 ) it is difficult to remove the last traces of the neutral salt. These objections may be overcome by preparing a stock sol of the hydrogen clay by electrodialysis and then adding the proper amount of the desired cation in the form of the hydroxide. Time does not permit a discussion of all the work that has been done in this field in the last few years but attention mill be called to the work of Baver whose results are in general quite typical. As changes in the physical properties of clays seem to be correlated usually with changes in electrophoretic potential let us first consider the effect of
FIG.3 The migration velocities of clay sols as affected by the amount and nature of exchangeable cations. (After Baver).
adding increments of different bases to an electrodialyzed clay (Putnam) on the migration velocity, expressed as v/sec/volt/cm. The measurements were made ultramicroscopically in a cell of the Tuorilla type. The most striking thing in the curves shown in Fig. 3 is the great difference in effects of the monovalent and divalent cations. With the monovalent ions there is a gradual increase to a maximum, then a sharp decline. The maximum comes a t the same concentration with each cation and it is identical with the saturation values obtained by conductometric and potentiometric titrations. There is some overlapping of the curves of the monovalent series but in most cases the relationship is what one would expect. The sharp decline in the curves is probably a common ion effect. With the divalent ions a gradual retardation of the velocity is obtained.
SOME CHEMICAL REACTIOSS OF COLLOIDAL CLAY
371
There is a marked similarity between the electrophoretic velocities and the viscosities of the sols. The rate of flow of the 2.35 per cent suspensions through the Washburn modification of the Ostwald viscometer is shown in Fig. 4. The relative viscosity of the hydrogen clay was 1.42. For some reason as yet unknown, the maximum of the viscosity curves occurs a t a lower concentration of the bases than the maximum of the electrophoresis curves. This may be due to the differences in concentration. It was necessary to use a high dilution, less than 0.01per cent, in the cataphoresis studies.
Viscosity
Baver).
The initial drop observed with the monovalent cations is due to a dispersion of the aggregates of the hydrogen clay. This effect is most marked in the
case of the K sol. The initial increase is undoubtedly associated with the hydration of the cation as pointed out earlier by Wiegner and others. The divalent cations cause a decrease in viscosity up to the saturation point. Further additions cause incipient flocculation. The relative. magnitudes of the changes in viscosity caused by the hydration of the particles and by the increase in volume of aggregates due to water entrapped in the micells, can be seen in Fig. 5 . An estimation of the particle-sizes in the clays in the N a and Ca series was made by the ultramicroscopic method. The particles were very large due
RICHARD BRADFIELD
372
FIG.5 Viscosity of clay sols containing different amounts of bases. (After Baver).
I 50 (I,
$22
I30
G?
23
& f
110
81 lurn %a 90 Gs k M / L L l€QfJ/VHL€Nrs0’5RsE
PER
loo
GMS. c L R r
FIG.6 The effect of exchangeable Ca and ?u’aupon the size of particlesof colloidal clay. (.4fter Baver).
S O U E C H E Y I C A L R E A C T I O S S O F COLLOIDAL CLAY
373
in part to the failure to remove all of the particles over I O O I.( in the preparation of the clays. Satisfactory count,s could bp made with the Ca-clay but with the Sa-clay it ivas quite evident that all t’he particles were not sufficiently visible to permit accurate counts. The shape of the Sa-clay curve in Fig. 6 is additional evidence of this. I t is not the smooth type to be expected in the light of the other studies. Another striking illustration of the difference in size between the particles of the Sa-clay and the Ca-clay is shown in Fig. ;. These curves represent the velocity of ultra filtration through a collodion membrane under a pressure of
In
e
I*
.7
*UTI$
FIG.7 The effect of evchangeable Ca rind Xa upon the filtration velocity of clays. (After Baver). 100 pounds per square inch. The Ca-clay is much more permeable. An attempt t o calculate the size of the pores in the clay membrane by the formula of Bjerrum and Manegold’ indicates that the cross section of the pores of the Ca-clay are about eight times as large as those of the Na-clay The greater permeability of Ca-clays has long been observed in the field. The unproductiveness of many of our irrigated soils in arid regions has been found to be due to a bad physical condition caused by the replacement of c1a by S a in the colloidal fraction. The formation of clay pans in the humid region, which reduces the productivity of the soils over them to lese than jo per cent of the normal expectation of the region is likewise to be attributed to a replacement of Ca by H. The study of the colloidal behavior of clays is admittedly bristling with difficulties but the results obtained in the last decade seem to justify a continuance of the work in spite of the dire predictions of many prominent colloid chemists, made at the first Colloid Symposium seven years ago.
Ohio State C n i w r s i t y ,
Columbus, Ohio. K o l l o i d Z , 43, 5
(1927).