VAPOR PRESSURE-WATER CONTENT RELATIONS FOR CERTAIN TYPICAL SOIL COLLOIDS' LYLE T. ALEXANDER
Bureau of Chemistry and Soils, U.S. Department of Agriculture, Washington, D . C . AND
M. M. HARING
Department of Chemistry, University of Maryland, College Park, Maryland Received June PO, 1936
Several attempts have been made to utilize the information obtained by allowing air-dry soil to absorb moisture at various relative humidities. The water absorbed by soil colloids when allowed to come to equilibrium over a sulfuric acid-water mixture containing 3.3 per cent sulfuric acid by weight has been made the basis for a method of estimating the quantity of colloid present in a soil (6). The water held under this condition (99 per cent relative humidity) falls below the hygroscopic coefficient. The British soil workers (8) have also used a determination of the moisture held at 50 per cent relative humidity as a criterion of soil properties. More recently, workers in the Bureau of Chemistry and Soils of the United States Department of Agriculture ( 5 ) have made determinations of the amounts of water held over sulfuric acid-water mixtures of various concentrations by different soil colloids. An attempt was made to correlate the ratios between some of the values so obtained with the chemical composition of the soil. The vapor pressure-water content curves of a number of soils have been studied by Thomas (13, 14)and by Puri, Crowther, and Keen (11). They covered the entire range from oven-dry a t 105°C. to saturation. It was found in both investigations that an inflection point in the curves occurred near 50 per cent of the vapor pressure of pure water. No breaks were found; this indicated no sudden change in the nature of the forces holding the water. Puri, Crowther, and Keen reached the conclusion that the curves were all of the same type, but that the general slopes of the curves were decreased with increases of clay and organic matter content. Thomas 1 Abstracted from a thesis submitted by Lyle T. Alexander to the Graduate School of the University of Maryland in partial fulfillment of the requirements for the degree of Doctor of Philosophy, May, 1935. Presented a t the Twelfth Colloid Symposium, held at Ithaca, New York, June 20-22, 1935. 195
196
LYLE T. ALEXANDER AND Y. M. HARING
also reached the conclusion that the slope of the curve is influenced by the quantity of fine material present, but concluded that the organic matter played a minor r6le in water vapor absorption. Brown and Byers (5) and also Anderson and Mattson (3) have called attention to the correlation between the avidity of a soil colloid for water and its chemical constitution. Since it has been shown that the coarser fractions of a soil only serve as a framework or as diluting material for the colloid, it seemed advisable to study the colloid extracted from the soil rather than the soil itself. This would eliminate the variable factor mentioned by Puri, Keen, and Crowther concerning the change in slope of the vapor pressure curve with clay content.
FIQ.1. Vapor pressure apparatus EXPERIMENTAL
The soils selected for this study represent a wide range of progressive weathering. The Barnes soil is a black dry land grass soil from North Dakota. It has been formed from calcareous glacial till. It has not been subjected to severe hydrolysis because of the low rainfall. The Carrington soil is a fertile prairie soil of Iowa. Like the Barnes soil, it has been developed from calcareous glacial till, but under conditions of more rainfall, and therefore its degree of hydrolysis is greater. The Miami, a gray-brown podsolic soil from Indiana, has been developed under somewhat higher rainfall than the Carrington. This is a timber soil and not a grassland one. The fourth soil selected is the Cecil, a red soil from North Carolina, that has been developed from decomposed granites and gneisses under condi-
197
VAPOR PRESSURE AND WATER CONTENT OF SOIL COLLOIDS
TABLE 1 Relation between the vapor pressure and water content of the Barnes colloid at 86°C. Sample No. 10307. Dry weight of sample = 9.101 g. VAPOR PRESSURE
mm. H g
23.1 22.6 22.6 22.1 21.8 21.2 21 .o 20.6 19.9 19.5 19.0 18.5 18.1 17.3 16.6 15.7 15.4 14.1 13.2
WEIGHT OF WATER LOST B Y SAMPLE
PER CENT OF
oram8
per cenl
0.031 0.156 0.259 0,352 0,498 0.623 0.647 0.722 0.856 0.934 1.020 1.072 1.134 1.200 1.280 1.368 1.459 1.560 1.619
33.4 32.0 30.9 29.8 28.2 26.9 26.6 25.8 24.3 23.4 22.5 21.9 21.2 20.5 19.6 18.7 17.7 16.6 15.9
SAMPLE
1
WEIORT OF WATER LOST BY SAMPLE
PER CENT OF WATER IN SAMPLE
mm. Hg
grams
per cent
12.8 11.8 10.9 9.9 9.3 8.7 8.0 7.0 5.9 5.3 4.0 2.8 1.8 1.o 0 .o
1.641 1.702 1.760 1.825 1.869 1.902 1.961 2.027 2.112 2.143 2.283 2.363 2.478 2.583 3.069
15.7 15.0 14.4 13.6 13.2 12.8 12.1 11.4 10.5 10 1 8.6 7.7 6.5 5.3 0.0
P
l
S
E
1) Additional point by desiccator method I/ 23.2 I I 35.4 I
TABLE 2 Relation between the vapor pressure and water content of the Carrington colloid at 96°C. Sample No. 10084. Dry weight of sample = 8.977 g. WEIQHT OF WATER LOST BY BAMPLE
PER CENT OF WATER I N SAMPLE
VAPOR PREBBURQ
Hg
Oroms
per cent
25.7 23.1 22.4 22.1 20.7 19.0 16.4 15.8 12.2
0.019 0.137 0.293 0.499 0.731 0.974 1.233 1.343 1.492
26.9 25.6 23.8 21.6 19.0 16.6 13.4 12.2 10.5
VAPOR PRESSURE
mm.
WEIQHT OF WATER LOST BY SAMPLE
PER CENT OF WATER I N SAMPLE
mm. Hg
grams
per cent
10.1 6.5 5.3 1.4 1.2 0.0
1.621 1.819 1.898 2.130 2.158 2.432
9.0 6.8 6.0 3.4 3.1 0.0
11 Additional point by desiccator method
II
23.3
I
1
26.2
tions of high rainfall and temperature, where the hydrolysis has been severe. The colloid was so extracted that the upper limit of particle size was TEE JOURNAL OF PHYSICAL CHEMIMBT. VOL.
40. NO. 2
198
LYLE T. ALEXANDER AND M. hl. HARING
TABLE 3 Relation between the vapor pressure and water content of the Miami colloid at 25°C. Sample No. 10342. Dry weight of sample'= 9.437 g. WEIGHT OF WATER LOST BY SAMPLE
PER CENT OQ WATERIN SAMPLE
mm. Hg
grams
per cent
24.9 22.7 22.7 21.8 21.7 21.9 21.5 20.9 20.3 20.3 19.6 18.8 18.0 16.9 15.7 15.2 13.7
0.134 0.212 0.296 0.346 0.403 0,523 0.606 0.692 0.820 0.895 0.956 1.062 1.125 1,181 1.257 1.309 1,376
19 9 19 1 18 2 17 6 17 0 15 8 14 9 14 0 12 6 11 8 11 2 10 0 9 4 8 8 80 74 6 7
VAPOR PRESSURE
1
plig&E
1 ~
1
WEIGHT OF WATER LOST B Y SAMPLE
PER CENT OF WATER IN SAMPLE
mm. Hg
grams
per cent
11.4 10.4 9.6 7.8 6.3 4.7 4.4 3.6 3.0 2.5 1.9
1.443 1.456 1.515 1,561 1.597 1.659 1.672 1.703 1.737 1.744 1.781 1.845 2.013
6.0 5.8 5.2 4.8 4.4 3.7 3.6 3.3 2.9 2.8 2.4 1.8 0.0
1.1
-
0.0
A d r r a l p;int by desiccator method I 2 5 0
TABLE 4 Relation between the vapor pressure and water content 0.f the Cecil colloid at 25°C. Sample No. 9415. Dry weight of sample = 9.910 g. WEIGHT OF WATER LOST B Y S.4MPLE
PER CENT OF WATER IN SAMPLE
YAPOR PRESSURE
mm. Hg
grams
per cent
24.3 22.7 22.7 22.4 22.1 22.1 22.1 21.7 21.2 20.4 20 .o 19.8 18.2 17.1 14.3 12.3
0 096 0 274 0 394 0 529 0 669 0 794 0 947 1 071 1 223 1 360 1 469 1 587 1 708 1 803 1 909 1 966
22 3 20 5 19 3 17 9 16 5 15 3 13 7 12 5 11 0 96 8 5
VAPOR PRESSURE
~-
5.1
'
WEIGHT OF WATER LOST B Y S.4MPLE
PER CENT 06 WATER IK SAMPLE
mm Hg
grams
per cent
10 2 8 3 7 0 5 9 50 3 7 2 9 2 3 1 4
2.014 2,055 2.085 2,106 2.130 2.162 2.182 2.187 2.214 2.229 2.246 2.262 2.309
30 2 5 2 3 20 18 1 5 1 3 1.2 1 0 0 8 0 6 0 4 0 0
11
09 09 00
VAPOR PRESSURE AND WATER CONTENT OF
aom
COLLOIDS
199
about 0.3 micron in diameter (5). The colloid was air-dried a t room temperature to avoid any irreversible dehydration at the elevated temperature. The apparatus used for determining the detailed vapor pressure-water composition curves is essentially that used by Wales and Nelson (15). A diagrammatic representation of it is shown in figure 1. After a 10-g. sample of air-dry colloid ground to pass a 100-mesh sieve had been placed in an evacuated desiccator over 3.3 per cent sulfuric acid for a period of five days, the sample was weighed and transferred to the bulb of the apparatus shown in figure 1. The apparatus was then evacuated through the phosphorus pentoxide tube until about 0.1 g. of water was collected. The stopcocke were closed and the whole apparatus allowed to
FIQ.2. Water content relation for soil colloids a t 25"C,
stand until equilibrium was reached. The phosphorus pentoxide tube was then weighed, and the difference in level of the two legs of the manometer read by means of a cathetometer. The apparatus was again evacuated and the process repeated. The experimental data obtained by this method are given in tables 1 to 4. Also given in these tables is the value for water held, at 25OC.,by the colloid in an evacuated desiccator containing aqueous sulfuric acid, with a water vapor pressure of 23.3 mm. Hg. Most of the values were taken after allowing twenty-four hours for equilibrium to be reached. Although this may not be a real equilibrium point, no further change in pressure could be noted by allowing three or four days time. The curves for the four soil colloids are shown in figure 2. These are all of the same general form. They are similar to the curves for gelatin given by Kats (7) and to those for wood found by Stamm and Loughborough (12). They are also similar to that for aqueous sulfuric acid. Data are available in the International Critical Tables for the sulfuric acid. The first few values on each curve are undoubtedly too high because of
200
LYLE T. ALEXANDER AND $1. M. HARING
removal of adsorbed air. The very low ones are a bit uncertain because the mercury manometer was not sensitive to small changes in pressure in the very low range. MATHEMATICAL EXPRESSIONS FOR THE CURVES
Katz (7) expressed the vapor pressure-water content relation for an elastic gel by the equation
where 01' and p are empirical constants, h is the ratio of the vapor pressure of the gel to that of pure water a t the same temperature, a is the water content in grams, per gram of dry colloid, T i 0 is the specific volume of water, and R and T have their usual significance. This equation was tested to see if it would fit the curves for the soil colloids studied. The results were not satisfactory. However, a satisfactory equation for two of the colloids may be obtained as follows. Consider the reaction Soil colloid . xH2O
+ water -+soil colloid . yHzO
where x is always less than y. The decrease in free energy for this reaction may be obtained from
RT M
AF = - I n -
P Po
where M is the molecular weight of water, P is the equilibrium pressure of the system, POis the vapor pressure of water a t the same temperature, and AF is the free energy decrease a t this temperature when 1 g. of water is added to an infinite amount of soil colloid (xH20). These values have been calculated for vapor pressures taken from the curves of figure 2, and are presented graphically in figure 3. As a rule such curves (figure 3) will be found to fit an equation of the type y =
(2)
where a and b are constants and y and 5 represent the ordinates and abscissae, respectively, and e = 2.7183. Letting x = percentage of water in sample and y = AF, we have AF = ae-bz
Equating equation 1and equation 3
(3)
VAPOR PRESSURE AND WATER CONTENT O F SOIL COLLOIDS
201
Since R, T, M , and a are constants, we may combine them with the conversion factor to Briggsian logarithms and obtain
P log -
PO
ate+*
(5)
a’ and b may be evaluated by throwing the equation into the logarithmic form
P
log log - = log a’
PO
P Po
log log -
log a’
- bx log e
(6)
- 0.4343 bx
(7)
FIG.3. Change in free energy of soil colloids a8 a function of water content at ’25°C.
-
M a a‘ (and therefore a) and b may then be evaluated 2.303RT’ P P by plotting log log - against x or by plotting log - on semi-log paper Po P O against x. Log a‘ will be the intercept on the y axis and -0.43433 the slope. Curves for the four soil colloids studied, plotted by the first method, are given in figure 4. Straight portions are found for all the types, especially at lower water percentages. Equations for the straight portions of a’of course =
202
LYLE T. ALEXANDER AND M. M. HARINQ
each are given below. They are valid, naturally, only for colloids of water content lying on the straight portion.
P log - = 2.32e-0.2O5"
(Car rington)
Pa P log - = 2.2ie-0.311~
(Miami)
PO P
log - = 1.15e--0.49**
(Cecil)
PO
Possible explanations for the failure to obtain straight lines a t higher water contents will be offered later.
FIG.4. Log log of vapor pressure ratio as a function of water content a t 25°C.
Bsrnes. ..................................... Carrington. ................................. Miami. ..................................... Cecil. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
AF IN CALORIES PER
AH IN CALORIES PER
GRAM OF COLLOID
GRAY OF COLLOID
14
22 17 14 6
9 6
3
VAPOR PRESSURE AND WATER CONTENT OF SOIL COLLOIDS
203
The values for AH were obtained by measuring the heat of wetting of the soil colloids according to the method of Anderson (l),except that the samples were dried over concentrated sulfuric acid at 25°C. instead of in the oven at 105°C. RELATION TO COLLOID COMPOSITION
In recent years chemists have discovered a n important relation between the chemical composition of the colloid of a soil and its field characteristics. This relation is sometimes obscured by the presence of unweathered minerals in the extracted colloid. Such minerals are abundant in most soils (9). Since unweathered ground minerals such as orthoclase do not hold water appreciably in the hygroscopic range, they should, if present as an impurity, serve merely as a diluent, and hence affect the vapor pressure curves only slightly. An examination of the curves in figure 2 shows that the Miami falls nicely into its proper relation between the Carrington and Cecil and not between the Barnes and Carrington as its silica sesquioxide and silica alumina ratios would predict. The differentiation shown by this family of curves between the four groups of soils covered in this investigation is sharp and of even distribution. It appears that the hygroscopic moisture of the colloid is a characteristic function of the factors which determine the field characteristics of the soil, Further investigation may reveal some curves out of place, just as some of the ratios are out of place, but this will not invalidate the use of these curves as criteria of soil properties. THE NATURE O F THE HYGROSCOPIC WATER
The data shown in this paper indicate that the hygroscopic water of the
soil colloids is held in a similar manner to water of swelling in gelatin and in wood. Probably the most illuminating treatment of hygroscopic water in elastic gels is that of Peirce (10). He proposes a “two-phase” theory of absorption of water by cotton cellulose, which has water content-vapor pressure isothermrtls very similar to those of soil colloids. According to the Peirce theory the hygroscopic water can be considered as occurring in two different phases, or states, on the cellulose. The first of these is the a phase which is held to be chemically bound by the hydroxyl groups of the hexose units of the cellulose. The second or b phase is made up of the water molecules attracted by the water molecules of the a phase and by the colloid surface which is not reactive toward water. From theoretical considerations he arrives a t the following formula for the vapor pressure-water content isotherms for cotton and starch
204
LYLE T. ALEXANDER AND M. M. HARING
where P is the water vapor pressure of the cellulose (xHzO),Pois the vapor pressure of water at the same temperature, IC is a constant characteristic for each sample, C. is the concentration of moisture in the a phase, B is another constant characteristic of each sample, and Cb is the concentration of water in the 6 phase. This equation is similar in many respects to the equation arrived a t on page 201 for the soil colloids. Peirce pointed out that the equation would be applicable only to pure substances. It is probable that the failure of the vapor pressure-water content curves to rectify, as shown in figure 4, is due to some contaminating colloid which does not exert a noticeable lowering of the vapor pressure in the lower moisture content ranges, but which does make itself noticeable in the higher moisture content ranges, Iron oxide is a contaminant to which we might attribute this behavior. If then we apply the two-phase theory of Peirce to soil colloids, we may picture the a phase water as being combined with the alumino silicic acid complex, with the tendency toward further hydration as the attracting force. The b phase water is held on the surface of the colloid not occupied by a phase water and as outer layers on the a phase water. Concerning the a phase water, one must conclude that in the little hydrolyzed soils of the chernozem group the tendency to hydrate is very great, and that in the lateritic Cecil series the tendency is very small. The b phase water does not appear to be so much a function of the chemical composition of the colloid as of the surface. Further investigation is necessary to determine the quantitative relations between these two kinds of hygroscopic water. This idea of the two kinds of hygroscopic water is easily harmonized with the work of Anderson (2) and of Baver and Horner (4) on the effect of exchangeable ions on the hygroscopicity of colloids. It is interesting to note that the colloids having high base exchange capacities are the ones having much attraction for the a phase water. It is possible that the base exchange phenomena and absorption of a phase water are due to the same chemical affinity, that is, secondary valence, or it may be that the base exchange bases are held by primary valence bonds. Further investigations along this line would be necessary to decide which of these is responsible for the phenomena. SUMMARY
1. A study has been made of the vapor pressure-water content curves of four typical soil colloids. The curves are shown to be characteristically different for the different soil groups. 2. The change of free energy as a function of water content has been calculated, and the total free energy change on wetting has been determined approximately.
VAPOR PRESSURE AND WATER CONTENT OF SOIL COLLOIDS
205
3. An attempt has been made to correlate vapor pressure curves with soil classification. 4. Peirce’s two-phase theory of water absorption by cellulose is used to .picture the nature of the hygroscopic water of soil colloids. REFERENCES (1) ANDERSON,M. S.: J. Agr. Research 28, 927-35 (1924). (2) ANDERSON,M. S.: J. Agr. Research 38, 565-84 (1929). (3) ANDERSON, M. S., AND MATTSON, SANTE:U. S. Dept. Agr. Bull. 1452 (1926). (4) BAVER,L. D., AND HORNER,GLENM.: Soil Sei. 36, 329-53 (1933). (5) BROTVN, I. C., AND BYERS,H. G . : U. S. Dept. Agr. Tech. Bull. 319 (1932). (6) GILE,P.L., MIDDLETON, H. E., ROBINSON, W. O., FRY,W. H., AND ANDERSON, hf. S.: U. S. Dept. Agr. Bull. 1193 (1924). (7) KATZ, J. R.: Die Gesetze der Quellung. Eine biochemische und kolloidchemische Studie. 1 Teil: Die Quellung in Wasser ohne Komplikation. Kolloidchem. Beihefte 9, 1-182 (1917). (8) KEEX,BERNARD A,: The Physical Properties of the Soil. Longmans, Green and Co.,New York (1931). (9) MCCAUGHEY, W. J., AND FRY,W. H. : Bureau of Soils Bull. 91 (1913). (10) PEIRCE, F.T . : J. Textile Inst. 20, T133-T150 (1929). (11) PURI,A. N., CROWTHER, E. M., AND KEEN,B. A. : J. Agr. Sci. 16,68-88 (1925). (12) STAMM, ALFRED J., AND LOUGHBOROUGH, W. KARL:J. Phys. Chem. 39, 121-32 (1935). (13) THOMAS, MOYERD . : Soil Sci. 9, 409-34 (1921). (14) THOMAS, MOYERD.: Soil Sci. 17, 1-18 (1924). (15) WALES,H., AND NELSON, 0. A,: J. Am. Chem. SOC.46, 1657-66 (1923).
