294
J O H S I\-. J O R D h S
(9) K L E S O \ \ . , 0.: .lcta hem. dcaiiciiiiavia. 1, 328 (194i). (10) ~ I C B A I SJ ., W . : J . hem. Soc. 101, 2042 (1912) ; 106, 957 (1‘314). (11) ;\IcBars, J. W., FORD, T. F . , M I ) WILSOX,D . A , : Tcolloid-%.78, 1 (1937). (12) PEREPA: A . ,J.: s a ~ a i\rcd. i m i . 48, 40 (1948). (13) PRATT, R., DCPREYOY, J . , . i s i ) STRAIT, I,. -1.:J . I3nct. 55, 75 (1948). (14) REGKA, P . : Chemistry & Industry So. 18, p. 275; S o . 18, p . 295 (May,1948). (15) SHEDLORKI-, I,.: Ann. S . I-.A c a t l . Sci. 46, a r t . 6, 4 2 i (1948). (16) ST.ACI)INGER, H . : Oryanisc/ic Kolloidchcttcl‘c, F. Iyic\i-cg uiitl $ohii, ErauIischweig (1941) ; lithoprintetl by Ed\vards 13rorhcrs, I n c . , Ann . I r h r , XIichigan (1945). (17) STAT-DINGER. I%.:Jlaki.oi)iolcir/lnrc C’hc.itiic itnd Biologic. K c I p f and Co. Vcrlag, Basel (1947). (18) w O O D B C R Y , D. rr.,.\SI) I{(), - H I . V N , ( ‘ , : ,J. Uiol. (’hcm. 171, 447 (1917).
()
r ~ ~ ~ A ~ xLlC o P 13LI:s‘I’oSI‘~E;s. III t SI1 F.LLIYG IT
()RG \XI(
1,IQVIDs’
.JOIIS I\- JORDAISL l ~ c / / o Iii ~i c i i t i i i e , I ’ i i t ~ b ~ l i q 13. h Petiitsi/hailin
l < c ( c i i c d . l i c y t t s t 19 1.948
Smith (14,15) has prepared organic compounds ot bentonite, arid Hauser (6) has discovered that certain organic. compounds of bentonite have the property of sn-elling and dirpcrsing in organic liquids. The present paper is the report of an investigation of some of the factorb involved and of the extent of the conversion of the clay from an originally highly hydrophilic condition t o a hydrophobic and organophilic condition. Results of the research indicate a rather surprising change in the character of the clay upon conversion t o the organophilic condition with proper choice of organic base. &IITIJRIALS AND JlETHODb
For this investigation bentonite from the Satioiial Lead Company property a t Clay Spur, Wyoming, 11a5 used. Thermal and petrographic analysis along with x-ray diffraction indicated the clay fraction of this material t o consist principally of the clay mineral montmorillonite. The organic compounds T\ ere obtained for ihc most part from commercial sources. It was found that an organophilic product may be prepared by beveral methods. Investigation of these method? indicated that the following procedure, Presented at the TI\ ciit? -sccoiid S a t i o i i a l Colloid SJ mpobium. \vliich \\as held uiidei the auspices of t h r Division of Colloid Chemistry of t h e . h e r i c a n Chemical Society at Cairihridge. Xlassachuset ts, Junc 23-26, 1948 benior Frlloii , Fellonship on Lrad sustnincti by the Satioiial 1.t~stlConipanv a t \Irlloii Institutr Pittsburgh, Pcnns\ Ivmia
O K G l S O P H I L I C BESTOSITES.
I
295
hich \ m s used cwlu4vely for obtaining the data herein reported, \\-as most efli'ectivcl for general purpohes, 11hile modificatiorii might be advantageous for specific rases The montmorillonite laminae TI ere separated by dispersion of 118 g. in 3 lit or^ of 11 ater; after standing for 1 111.. the dispersion was decanted, leaving behind approiim:itely 8.5 g. of coarse non-clay sediment conii,ting of leld+par\ and cjunitz. Computing to a u-ater-free hasis (8 per cent loss at 103°C. for clay as rcceived) the clay material remaining for reaction approsirnnted 100 g 'Illis quantity was chosen for convenience becaube the b:tsecwhangc ~ i i p ~ t c l tofy the clay, so treated, \ \ > I \ foulid t o he 100 milliequiv. ' 100 g. by the ammonium acetate method (4). The organic bases ]\ere converted t o salts, if not received a b such, by the addition of acetic or hydrochloric acid, dissolved in miter, and added to the clay dispersions in the ratio of 100 milliequiv. '100 g. of clay except \ h e r e otherwise specified. Flocculation u-ab in most cases very thorough, and the precipitates 11 ere washed by filtering and repulping for removal of salts, then filtered, dried in an air-circulating oven at G"C., and pulverized in a laboratory hammer mill. The teqt method used t o determine the effectiveness of the organic cation in converting the bentonite t o the organophilic condition consisted in hlowly adding 2.00 g. of the pulverized product t o an organic liquid contained in n 100-ml. or 2.50-ml. graduated cylinder, and allon-ing each portion to solvate and settle before adding the nest. The apparent volunie of the settled gel \\as obzeived after itanding 24 hr This method n a b bused upon the finding that in liquids such a\ nitrobenzene the rate of wetting \I a3 very slon- and the gel volume large. and that the organic ammonium bentonite samples appeared t o disperse more or less readily upon shaking, exhibiting slight tendency t o settle even after long standing. I n cases \\-here the material welled without forming permanent dispersions the samples 11ere added rapidly and the cylinders and contents shaken vigorouqly and allon-ed to stand for the usual 24 hr., this method being considerably more rapid than the other from thc standpoint of the operator'il time
11
I?i:>cL'rb
~ Y D DIXUSSIOS
Solr m l p r i m a y aliphatic amnioniiim bentonite complew\ wei e pi epired mcl gel volumes (letermined in nitrobenzene, benzene, and isoamyl alcohol. Figure 1, in \\hich gel volume iz plotted :I.; a function of thc number of carbon Itoms in the amine rliain up t o and including eiqhtepn, 41on-.; that organophilic woperties are negligible until an amine chain of ten carbon :Itoms is reached and hat, twelve carhons are required for mavimum siwlling Further, it is obvious hat tlic liquid of high dielectric constant, nitrobenzene. i5 rrmnrliably highci in dvatiiig :Ibility than benzene or isoamyl alcohol, Traces \T ere obtained foi theye same :ill;ylapmoniuni bentonite\ \ n t h a ;rigti.-counter x-r:ly spectrometer. I'sing 9.6 A. ai the 001 o r ha7:iI plane ,pacing of the montmorillonite laminae, btepn-ise separation of thc flakes vas obieivcd as the length of the amine chain attaclied t o the cln>- \vas increabed. rhese steps were of the order of 4 -1 01 approximately the v m del, Waals di-
296
J O H S W. JORDAT
ameter of a methyl group. Gieseliing (3) and Hendriclis (11), using a similar technique, have pointed out that organic molecules attached t o montmorillonite through base-exchange reaction tend t o be attracted more or less in their entirety onto the surface of the mineral plates, the non-cationic portions being held by adsorptive forces. Ctilizing Rendricks' (12) demonstration that about 80 per cent of the exhangc positions of montmorillonite are on the basal plane surfaces n-ith the remainder on the edges of the flakes, it was computed that the average area per b:se-eschange position on the basal plane surfaces should be on the order of 165 h.?for a clay haying a bLize-eschangc capacity of 100 milliequiv./100 g. I-Iauser and Reed (10)haye pointcd out that the exchange capxcity of montmorillonite is practically independent of particle size. For the relatively simple case of a normal primary aliphatic aTine reacted with bentonite the stepwise separation of the plates by units of 4 -1. has been taken to indi-
I a
o Nitrobenzene 0 Benzene P
Isoamyl Alcohol
v)
i
I
NUMBER O F CARBON ATOMS I N AMINE C H A I N
FIG. 1. Effect of aininc rhain length on si\-elling or organic ammonium bentonites organic liquids.
III
cate that the hydrocarbon chains lie flat along the surface with the planes of the zig-zag chains parallel with the plane of the mineral. The areas covercd by such molecules have been calculated according to the listing in table 1 and caressed graphically in figure 2. Figures in the last column of table 1 for the area of clay covered by organic matter appear t o correlate well n-ith basal plane spacings when it is postulated that, in the case where the organic chain occupies no more than half of the available area, the organic molecules on the top surface of one lamina may fit into the gaps between those on the bottom surface of t h r lamina directly above i t , so that the resulting sep5ration of the tivo laminae is only the thickness of one hydrocarbon chain, or 4 X. Where the chains are longer nnd occupy morc than 50 per cent of the surface area, it is obviotir that adjacent flakes will bc unahle to appol'oacli any more closely than the thicline-s of t\vo hydrocarbon rhains, or 8 This fact may be illustrated tn-o-dimenqionally according t o thc
sketches in figure 3. I t \\-auld appear from these data per cent coverage is required for the development of character. Since the dodecyl complex gave optimum results in diverse types (figure I ) , its behavior \vas observed in a
that approximately 50 definitely organophilic three liquids of rather wide variety of liquids
TABLE 1 Spatial y e l a t i o m j o i . homologous alk~lammoniici~i bentotiites I
CARBOP ATOYS Ih’ CEAIN
001 SPACISGS
s
~
~
LAYLRS ~ OF~AXISE ~
CALCULATED , I \ ~~I N EAREA
C4LCUL4TED AREA
~OF CLAY ~ COATED E BY AXIXE
0
c,
3.9 3,s 3.;
10
9.6 13.5 13.4 13.3 13.6
I 1 1 1
12 1.1 16 1s
17 1 17.4 17.5 17.6
7.5 7.S 7.0
0 3 4
s
0
vi196
& O
I9
265
I
I
0
34
57 61 72 79
69
2 2 2
415
94 I06 119 131
44
‘1
so
SI
0 23 27 42 19
0 38
49
565
64
P
u
7’1 5
79 IO
z
3
V
176W
a
J -1
156-
a m
/
2
w 136-
L z
/
_J
i l l 6 -
/
0
I
P
u
0
d
/
’
8 96d
gz
s
/
_1
-6
rr0
I + z
/ -4
B LL
0
/
0
2
-Os
/
z
CL
*
z
/
-2
2
l-
a n
/ I
I
,
I
B
F I G .2 . Effect of amiiir chain l e ~ i g t h011 ~ i ~ o n t ~ n o r . i l l o hi ~siat el p l a n t spaciiiys ib libted in table 2. This table suggeits that solvation may be 1 0 1 ~in liquids of [ion-polar nature such as the aliphatic and aromatic hydrocarbons. Geiierally ‘he gel volume appears t o increaw with the dielectric constant of the liquid, dtliough the correlation 13 not perfect. --Iqualitative observation vould indi,ate that the most effective liquids are tho.? nhich combine highly polar n-ith iighly organophilic rharncteriqtics, good c3~1111pl~~ being nitrobenzene and
~
298
JOHK W. JORDSP;
beneonitrile. hcetonitrile, relatively ineffective, satisfies the polarity requirement but is deficient in organic character; on the other hand toluene, also ineffective, is highly organic in nature but not sufficiently polar. I t is hypothcsized that, in the case of an incompletely clad amine bentonite complex, :idLESS T H A N 50% COVERAGE BY AMINE 4.0
A.
9.6
Octylomrnoniurn ion Montmorillonite Octylornmonium ion
9.6
Montmoril Ion ite Octylommonium ion
50 -100% COVERAGE 4.0 %. Hexadecylammonium f9.6
Montrnoril l o n i te
t4.0
Hexadecylammonium
44.0
h.6
I,
Montmorillonite
MORE T H A N 100%COVERAGE
9.6%.Montmorillonite Dimethyl-
$::
ammonium Montmorillonite
FIG. 3. Diagrammatic edge view of montniorilloiiite platelets having organic ammoniuiii compounds attached by hnse-e\-rh;inge mactioii.
sorption of a highly polai. organic licpid onto the uncoated portion- of the clay flakes greatly enhances solvation of the micelle in the remainder of the liquid. This picture 1 3 supported by the fact that octadecylammonium bentonite, in which the clay is approximately 80 per cent coated, ssn-ells only slightly in toluene
OHG.\SOPHILIC U E S T O S I T E S .
299
I
or ligroin and only moderately in ethyl acetate, but is highly solvated in mixtures of toluene or ligroin with small proportions of the ester, as illustrated in
GcL r ~ ~ l i i m cu,Ts 2-y. ...........
Isoiiiplcq
TABLE 2 of dodec!/laiiiinorziic,n bentonite in carious liquids
. . . . . . .
LIQTID
GEL V O L T X L
LIQUID
GEL VOLUJSE
Water (untreated bentonite). . . . . . . . . . . . . Water (dorlec~~lamrnonium bentonite). . , . , . Petroleum oil, Gulfpride S..i.E. 10. . . . . . . . . Petroleum oil, Gulfpride 8..4.E. -10... . . . . . DON-Corning Fluid # Z O O . . . . . . . . . . . . . . . . . . Petroleum e t h e r . . . . . . . . . . . . . . . . . . . Piperidine.. . . . . . . . . . . . . . . . . . . . . . . Naphtha .. . . . . . . . . . . . . . . Carbon disulfide.., . . ............... Carbon tetrachloride ...... Dibutylaminc , . . . . . Glycerol . . . . . . . . . Trihutylamine . . . . . . . . . .... .i\myl n i t r a t e . . . . . . . . . . . . . . a-Butylene txomidc . . . . . Eucalyptol. . . . . . . . . . . . . . . . . . . . . . . . Styrene. . . . . . . . . . . . . . . . . . Toluene. . . . . . . . . . . . . . . . . Bromohcnzcric . . . . . . . . . . . . Linoleic ncirl . . . . . . . . . . . . . . . . . CynicII c . . . . . . Aniline . . . . . . . . . . . . Cyclohcsanol . . . . . . Ethylene dichloritlc Benzenc. . . . . . . . . . . . . .
10 0 10 0 10 0
Dodecil .rlcohol 11cthvl e t h l l ketonc Diethxl ketone
20 0 20 0 21 0
gurc 4. It i h p(Jbtulatd herc that tlw highly polar c-ter, after :idsoiption on he uncoated areas of the clay flake,., lenders the individual flakes entirely rganophilic and compatible 71 ith thr h d r o c a i h n portion of the +oh-ating
liquid. This characteristic of binary liquid systems may be brought out by a wide variety of polar liquids, such as alcohols, esters, acids, ketones, and aldehydes, in combination with less polar liquids. For example, figure 5 indicates for mixtures of 10 per cent by volume of normal primary alcohols nnd acids in 50
I
I
IO
I
1 20
I
I
j
I
30 40 50 60 PER CENT E T H Y L A C E T A T E I N
1
I
1 70
ao
I
I
I 90
MIXTURE
F I G ,-i,Sivelling of ~ ~ c t a ~ e c y l s ~ n ~ n obentonite riium in binary riiixtures
NUMBER OF CARBON ATOMS IN ALCOHOL O R ACID C H A I N
F I G .5 , Swelling of r)ctadec~-l:iniiiioiiiumtjcntoiiitc in binary inixtures
toluene that the alcohols ~ i n dacids ai e highly effective 111 promoting The rolvation of octadecylammonium bentonite :~ndthat the effect diminishes markedly uith diniini>hing pularitg. I n figure ci thc gel volume of the same bentonite complex i i plotted again-t pcxrcentagc of nlcohol in mixture. of toluene with
ORGANOPHILIC BESTOXTES.
30 1
I
methanol, ethanol, butanol, and octanol. Here it is seen that only a. small percentage of alcohol is required in each rase to provide a remarkable increase in gel volume. 90
-
80
w
i 70
a
z
a fl
60
-j50 r W-40
I 13 A
9 30 J
20 IO
0
IO
20 30 40 50 60 70 80 PER CENT ALCOHOL I N A L C O H O L - T O L U E N E M I X T U R E
IO0
90
FIG. 6. Swelling of octadecylammonium bentonite in binary mixtures 100
I
I
I
I
I
I
I
Y
W
2 80w
0
$60I j 4 0 I
-
3
9J 2 0 W
W
I
I
0
2
4 6 8 IO 12 14 NUMBER OF CARBON ATOMS I N A M I N E CHAIN
FIG.7 . S;n.cliing of ory:iiiic
~ i i i ~ n o i i i u ihcntonit ii FS
16
18
i n tolucnr i90)-met htinol (10) m i s t u r c
Finding the mixture of toluene u-ith 10 volume per cent of methanol t o be highly effective in solvating octadecylammonium bentonite, this liquid mixture \vas used t o check the results plotted in figure 1 regarding the requirements in nolecular size of the cation for converting bentonite t o the organophilic conlition. The striking similarity of this c u r r e (figurc 7, relating gel volume of
302
JOHN W. JOIZD.\S
amine bentonites t o amine chain length) t o the curvc obtained with nitrobenzene confirms the previous finding that solvation is negligible until 2% chain length of ten carbons i s reached or until appro;iinately hLdf oi tiic clay siiiface is co.ited. I n toluene-ester mixtures the data suggest, as brought out in figurql 8, that in the solvation of octadecylammonium bentonite the acid portion of the ester is more important than the alcohol portion; e g., ethyl formate I S more effective than methyl acetate; methyl, ethyl and butyl acetate5 are nppro\;imately ~ q u d in effect ; and butyl stenrate is practically Xvithout influence A11 the systems discussed so far have involved the h i t o n i t c d t s of singlechain primary amines of less than sufficient size t o coat the mineral platelets completely with a single layer of hydrocarbon chain5 Bentonite x a s reacted n-ith quaternary ammonium salt.; having 11vo long alipliatic ('hnim, and gel 1-olumes were found to be generally good in :i .ccount for the solvntion phenomenon. The degree of solvntion depends upon at least three factors: ( 1 ) the extent of the surf'acae coating of the cia>- particles by organic matter; ( 2 ) the degree of sutimition of tlie base-exchange capacity of thc clay 13-j- organic cations; and ( 3 ) thc nature of the solvating liquid. The TI riter acknowledge> ITith thank- the a i s i s t a n c ~of Sylvia 11.David, E. I.. , ASD S~I,LIT..IS, J.D . : J , Ah. C'eram. S O C .21, 176-53 (1938).
306
S Y D S E Y ROSS h S D C.
H. SECOY
( 5 ) GRIM,R. E., . ~ L L A W A T , W. H. A N D CUTHBERT, F. L . : J. Am. Cerain. S O C 30, . 137-42 (1917). (6) HAUSER, E. A . : Private communication. (7) HAUSER,E. A , , A X D LE BEAU,D . S.: J . Phys. Chem. 42, 961-9 (1935). (8) I ~ ~ s E E. R A , , A X D LE BEAI-,D. 8.:J. Phys. Chem. 43, 1037-48 (1939). (9) HAUSER, E. h., . ~ X DLEGGETT, 31.13.: J . Am. Chem. SOC.62, 1811-14 (1949). (10) HATXER, E. A . , BSD REED,C . E.: J . Phys. Chem. 41,011-31 (1937). (11) HESDRICKS, S. €3.: J . Phys. Cheni. 45, 65-81 (1941). (12) HENDRICKS, S.B . , SELSOS, It. A . , ASD -%L D E R , L. T . : J. Am. Chcm. SOC.62, 1457-64 (1940). (13) IIACETVBN, D . AI, C,.: S a t u r e 154, 577-8 (1914). (14) SMITH,C. R . : J. Am. Chem. SOC.66, 1561-3 (1934). (15) SMITH,C. R.: U. S.patent 2,033,856 (Xarch 10, 1936).
O S PHYSICAL ADSORPTIOS. 111 A SIMPLIFIED EQUATION OF STATEAT HIGHPRESSURES AND ITS APPLIC.i.rIos SURFACE FILMSON LIQUIDSAKD SOLIDS~~?
TO
SYDSEY ROSS AKD C. H. SECOY O a k Ridge N a t i o n a l Laboratory, O a k Ridge, Tennessee Received August 19, 1348
The results of E. H. Amagat ( 5 ) on the behavior of gases and liquids a t high compressions have been expressed by the equation
pV,,
=
b'p
+ iRT
(1)
The nature of this equation4 is readily appreciated when the product p l i , is plotted against p . A straight line is obtained -\Those slope is b' and whose intercept on the pV,-axis is expressed by i R T . It n7as observed by Amagat himself that the application of equation 1 t o his data was not perfect, as the 1 Presented rtt the Twenty-second Sational Colloid Symposium, which was held under the auspices of the Division of Colloid Chemistry of the American Chemical Society a t Canibridge, Massachusetts, June 23-25, 1948. 2 This document is based on lyorlc performed under Contract S o . W-7405, eng 26 for the Atomic Energy Commission a t Oak Ridge Sational Laboratory. It was presented in part a t the Twenty-second Sational Colloid Symposium and 111 part before theDivision of Colloid Chemistry a t the 114th lleeting of the American Chemical Society, Portland, Oregon, September 13-17,1948. Present address : Department of Chemical Engineering and Chemistry, \V:tlker Laboratory, Rensselaer Polytechnic Institute, Troy, S e w T o r k . 4 Equation 1 is often called the . h a g a t equation, although we can find 110 record t h a t Amagat ever virote i t . Amagat did, hon-evcr, suggest tvio equations of s t a t e , both of them very much more complex than equation 1. See 0 . D. Chwolson: Lehrbuch der P h y s z k , Vol. 3, pp. 811-12, BraunschJreig (1905). It is probable t h a t Amagat would have looked on equation 1 as a rather crude or even frivolous attempt t o express an equation of state. I t is a n empirical equation t h a t is true for only a portion nf each isotherm
