748
W. II. SLABAUGH
that similar transitions are to be observed for reactions of higher order. Thus Marcz5found the velocity of crystallization a t low temperatures to be of the second order, .cvhilst at higher temperatures the process is first-order. Peshkov26 and (25) R. Marc. Z. phyaik. Chem., 73, 685 (1910). (26) €3. l'eshhov, Acta P h y w o c l n n . U. R. 8.S., 21, 100 (1946).
VOl. 56
Bnrkhiiy~en?~ are but two of the workers who have postulated a transport mechanism for the crystallization process; this, together with the work reported s t the recent Faraday Society Discussion 011 Crystal Growth, suggests that the process is one of intermediate type. (27) F. H. C. Barhtiiiysen, Cltern Il'eekbwd, 43, 234 (1947).
TIIF, SYNTHESIS OF ORGANO-BENTONITE ANHYDRIDES RY W. H. SLABAUGH Depart men1 of Chemistry, Kansas State College, Manhattan, Kansas Received September 17, 1861
The acylation of bentonite clay has been achieved by treating both the acid form and various metal forms of the clay with benzoyl chloride arid acetyl chloride in anhydrous media. Several types of analyses have been made in order to identify the products thus formed, arid methods for the estimation of the extent of acylation have been developed. The availahility of the base-exchange ion for this type of reaction has been studied and several postulates have been made to account for the variations observed.
Several successful methods have been reported by which certain organic groups have been attached to the laminar clays of high base-exchange capacity. Among these are the amine salt product1-6 which has received by far the most attention up to the present. Methylated c ~ m p l e x e shave ~~~ been made by introducing the methyl group with azomethane. An esterification products has been made by the action of an alkyl halide on bentonite clay. In all these studies, it is the conclusion that the attachment of the organic group to the laminar silicate structure is at the base-exchange site. The present work is concerned with the reaction between bentonite and acyl halides wherein it again appears that the reaction site is identical with the base-exchange site. From the standpoint of chemical and structural composition, bentonite may be considered as a giant molecule of laminar structure which consists of numerous cells, each cell bearing a negative charge as the result of the isomorphous substitution of a magnesium atom for an aluminum atom in the gibbsite layer.9 . The negative charge of this cell is balanced by a cation which is commonly termed the exchangeable ion. This cation is usually an alkali ion in natural bentonite, but it may be substituted with various other metal ions as well as hydrogen. The exact position of this cation as well as the exact location of the negative charge carried by the clay lamina are conjectural. The ionic system is undoubtedly governed by Pauling's on the electrostatic valence principle-neutrality smallest possible scale-even though the bond distance is modified by ionic size, charge and steric factors. (1) J. E. Gieseking, Soil Sci., 47, 1 (1939). (2) E. A. Hariser, Chem. Reus., 37, 307 (1915).
(3) E. A . Haiiser in Jerome Alexander, "Colloid Chemistry." Val. V I I , Reinhold Puhlishing Corp., New York. N. Y., 1950, pp. 431-432. (4) E. .4. Hauser. U. 6. Patents 2,531,427 and 2,531,812. (5) C. R . Smith, J. A m . Chem. SOC..66, 1561 (1934). (6) G. Berger, Chem. Il'eakblad, 38, 12 (1941). (7) G . Berger, Compl. rend. con/. pedol. medaterran. 119 (1947). ( 8 ) R. Bart and S. Gusman, "A Study of the Esterification of Bentonite," 5. €3. Thesis, Massachusetts Institute of Technology,
Camlwidge, Mass., 194F. (9) U. Hofinwn, I< Endell and D. Wilin, Z. Krist., 86, 340 (1933).
With these ideas in mind, the reaction between bentonite and an acyl halide may he represented by the equations
+ RCOCl + RCO-Bentonite + HCI RS-Bentonite + RCOCl + RCO-Bentonite + MCI
H-Bentonite
(1) (2)
In equation (2), M is any cation, assuming that proper allowance is made for its valence. Polyvalent ions will satisfy more than one base-exchange site. In view of the analogy to ordinary anhydride formation, the reaction product is termed a mixed anhydride. Experimental Materials.-A typical Wyoming bentonite which showed base-exchange capacities of 68 meq. per 100 g. bentonite by titrat,ion of the hydrogen-bentonite form and of 88 meq. per 100 g. bentonite by the ammonium acet,ate absorption niet8hod10was used. The raw clay WRS converted to the hydrogen forin by electrodialysis. Similar results were obtained with hydrogen-bentonite prepared by leaching the clay with 0.05 N HCI. However, it was somewhat easier to obtain uniform samples of the prepared clay by the former niethod. The metal forms of the clay were obtained by titrating the hydrogen form with the appropriate base to a pH of 8.5. The acetyl chloride and benzoyl chloride used in this work were of Eastman Kodak Company practical grade. Procedure.-Two grams of the prepared bentonite, dried at 105" for 2 hours, was dispersed in 15 ml. of dioxane and refluxed with 1 ml. of the acyl halide for 1 hour. The ratio of acyl halide to bentonite was not critical, except that when larger amounts of the acyl halide were present it was necessary to reduce the t)imeof refluxing in order to prevent carbonization. Acetyl chloride reacted more rapidly than benzoyl chloride, the products otherwise being essentially identical. This study was based for the most part upon the benzoyl chloride reaction. The product was washed with 80% et,hyl alcohol on the filter and dried overnight at 105" in order to remove all traces of alcohol. The product was very similar in appearance to the original clay. Proof of Formation of an Organo-bentonite Anhydride (l).-The alcohol washings from t,hc separation of the product from the reac,tion mixture were concentrated, ignited, taken up in water and then titrated with AgNOa in oJder to determine thc amount of metal chloride formed during thc reaction. Barring side react.ion, this result Rhould be an indication of the amount of acyl groups which have reacted with and become attached to the clay. These data are col(10) Robert P. Grahrtin and ,J, D. Sullivan. J. Am. Ceram. SOC.,21, 176 (1938).
SYNTHESIS OF OI~GANO-BENPONITE ANHYDRIDES
June, 1952
lected in Table I. When bentonite was refluxed with pure dioxane, entirely negative results were obtained for all three tests listed in Table I. (2).-The mixed anhydrides, when treated with NaOH a t room temperature slowly underwent hydrolysis. At 65", this hydrolysis reaction occurred more rapidly with about 5 minutes being sufficient for the system to equilibrate. Figure 1 shows a curve obtained by tit8ratingbenzoyl bentonite anhydride with NaOH a t 65". The plateau region of the curve at A corresponds to the hydrolysis reaction with the end-point at B. Up to 30% of the base-exchange capacity can be associated with anhydride formation.
440
II
10
pn. 9
8
12
7 10
6
E
V
VH.
[---A 4
i
5
10 15 ml. 0 110 N NoOH.
20
3 4 ml. 0 . 1 0 8N nci.
5
6
7
Pig. 2.-Detcrniination of hydrolysis value by backtitration of hydrolysis mixture of 10 ~ n l of . 0.099 iV NaOII and 1 g. of benzoyl bentonite anhydride prepared from: A, Na-bentonite; B, Ca-bentonite; C, K-bentonite. D, using untreated metal bentonite, serves as control.
6
0
2
I
25
Fig. 1.-Potentiometric titration of 3 g. of benzoyl bentonite anhydride with 0.11 N NaOH at 65", with 5-minute intervals between additions of the alkali; end-point of titration a t B.
ently reacts with a significant amount of NaOH, hence a comparison to a blank sample was necessary. Table I contains these data. (5).-Combustion analysis of the mixed anhydrides gave re roducible results which indicated a definite composition. Tfis method of analysis was based primarily on the carbon content inasmuch as the hydrogen determination were seriously influenced by the cation in the base-exchange posit,ion and the condensibility of tJhe hydroxyl groups on the clay structure, which in turn affects the amount of water formed in the combustion analysis. These data are included in Table I.
(J).-When t'he benzoyl bentonite anhydride was hydroTABLE I1 lyzed with NaOH, the sodium bensoat,e which was formed BASE-EXCHANGE CAPACITY OF A BEIDELLITECLAYWITH was convert,ed to benzoic acid and conclusively identified. ALKALIIONSI N THE I k C I i A N G E POSITION This involvement apparently produced no structural changes in the benzoyl group. Cation Baee-exchange capacity (4).-A hydrolysis value for the mixed anhydride was deI,i 89.7 t,ermined by heat)ing l g. of the product a t 80" for l hour wit>h 10 ml. of 0.1 N NaOH, t,hen t>itrat,ingt,he excess alkali with Na 89.4 HC1. Figure 2 shows a typical set of potentiometric tit,raI< 76.9 tion curves obtained in this analysis. A comparison to the Rl) 54.4 curve for the untreated bentonite a t a p H of 7.0 gives the cs 43.8 amount of NaOH which evidently was required for the hydrolysis reaction. The unt,reated metal bentonite appar-
EXTENT OF
TABLE I FORMATION OF BENZOYL BENTONITEANHYDRIDE
Evchangeahle ion on bentonite lariiina
Li Na K Itl) (IS 130
.\lg ( :n
8r
Ba
TI (ic) 4"
CsHbNH, HOCzHjNH3
H
Meq. benzoyl per 100 g. bentonite hIetal Hydrolysis Carbon cliloride value coin bus tion
34 40 28 27 18 11 26 29 30
12.0 18.4 12.0 12.5 12.5 10.5 15.4 15.4 17.0 17.0 11.6 27.2 21.2 22.5 26.3
4.0
11.0 5 .6 6.2 F.5 4.0 4.8 15.2 15.8
15.8 19.4 19.2 46.0" 16.9 21.1
Considerable carbonizat,ion is indicated by this high rcsult. a
Discussion of .Results It is readily noted that there are considerable differences among the three methods of analysis given in Table I. However, if the data for the alkali and alkaline earth ions are plotted, as in Figs. 3 and 4, there appears to be a good degree of correlation. The data based upon combustion analysis are probably most dependable. Between these data arid the hydrolysis values are considerable differences, although these differences are almost constant a t 7.0 1.0 meq., per 100 g. of bentonite for the alkali ions. The results for the othci. metals, particularly Be and Mg, show considerable anomalies. Evidently the organo-bentonite structure reacts with a certain amount of the NaOH above that required for the hydrolysis of the acyl groups. The surfaces of the laminar particles may react with NaOH either through surface adsorption of OH groups, neutralization of weakly acidic groups, or decomposition of the clay to give soluble silicates. 11
*
(11) R. P. Mitra. Bull. Indian SOC.Soil Sei.. 4, 41 (1942).
W. H. SLABAUGH
750
40
35
30 W
t
0
5
25
m
0.50
1.00
1.50
RADII OF EXCHANGEABLE IONS IN ANGSTROMS.
Pig. 4.--Estent of formation of benzoyl bentonite anhydride from alkaline earth bentonites. Analyses are bused upon;. D, amount of metal chloride found in reaction misture; E, hydrolysis values; F, carbon combustion analysis.
It has been suggestedi2that five factors determine the extent of ion exchange, namely, diffusion, ionic size, charge, steric availability and exchange position. Since the formation of organo-bentonite anhydrides involves the base-exchange sites, these same factors would determine to a considerable extent, the amount of anhydride formed with an acyl halide. With the exception of sodium, in Fig. 3 there is a definite trend which shows that as (12) 10. C (lY44).
Nacliod and W. Wood, J . A m . Cheni. Soc., 66, 1380
\p.ol. 56
the cation becomes larger it is more readily replaced by the acyl group. It is noted in Fig. 4 that the divalent ions react more readily than the univalent cations. Inasmuch as the base-exchange sites are univalent, a divalent, cation will be associated with two of these sites,l3 with the result that the latter type of cation will exhibit a greater degree of steric availability. Because of their greater availability trivalent cations which occur in thallic bentonite are more available and reactive with an acyl halide than are the divalent cations. The differences of behavior of the alkali ions i n the exchange position are influenced by the laminar silicate structure. It has been fairly well establishedI4 that the structure of these clays consists of oxygen-populated surfaces composed of hexugonal arrangements of groups in the Si20r layer. The holes in this layer which are bounded by 6 oxygen atoms are of approximately the same radius as a potassium ion, namely, 1.33 8. Since the clay is dried at 105' before it is treated with the acyl halide, the exchangeable ions become more or less firmly attached in this hole, depending upon their sizes. It is assumed on the basis of Grim's that the negative charge of the silicate layer is situated in the vicinity of the bottom of this hole. If the lithium ion were dehydrated, it could easily drop into this hole and then form a more covalent bond than the other aJkali ions. The sodium ion is small enough (0.98 A.) to slip in and out of the hole easily. The potassium ion becomes more or less sterically attached to these holes, as has been shown by previous studies.'G Rubidium and cesium ions, only slightly larger than the potassium ion, become less hindered by the fact that they cannot penetrate into the hole as far as the potassium ion. With respect t,o the performance of Na- and K-clays i n this reaction, these results are in agreement with Page and Baver'c but disagree with the results of Bart and Gusman.* There are undoubtedly undetermined factors which influence the availability of cations for exchange, but in view of the generally recognized structure of these laminar, clays, this hypothesis accounts at least partially for the results of this study. In a recent electrochemical study17 of the bond energies in K- and Ka-bentonit>et'he activity of the potassium ion has been shown to he considerably less than that of the sodium ion. This observation substantiates the result,s o1)tained in the present, acylation study. The availability of the ammonium ion (1.43 A. radius) should be similar to that of t8hepotassium ion. However, the acyl halide can readily react with the hydrogen atoms of the ammonium ion with the formation of a peptide linkage. The same would be true for the clays containing aniline and ethanolamine, the latter providing an addi(13) W. 11. Slabarigli a i d .I. I,, Culberholi, Tllrs .JOITRNAL, 66, 744 (1951).
(14) C. E. Marahall, "Colloid Chemistry of Silicate Minerals," Academic Press Inc., New Tork, N. Y., 1949, p. 52. (15) Ralph E. G r i m , J . Oeol., 60, 225,(1942). (16) J. I% Page a i d L. D. Bal-er, Soil Sci. SOC.Am. Pmc., 4, I50 (1439). (17) S. K, h l i k h e ~ j e eand C. E. RIarsliall, THISJ O U I ~ N A66, L , 61 (lY.51).
OEGANICDERIVATIVES OF MONTMORILLONITE
Julie, 1952
tional reactive point at the hydroxyl group. These complexes contain groups that are both sterically and chemically reactive toward the acylating agent. The results in Table I bear out these arguments. The typical erratic and unpredicted behavior of hydrogen ions on clay is shown in the results obtained when acyl halide reacts with this type of clay. One concept offered in this respect isbased upon the idea that in H-bentonite, some of the hydrogen ions penetrate the silicate lattice and release an equivalent amount of aluminum ions. Since this system is highly reactive toward the acyl halide, there are apparently fairly reactive cations 011 the clay which may or may not be identical with the original base-exchange sites. In spite of efforts to maintain anhydrous reaction mixtures the presence of a small amount of moisture will produce HC1 from the acyl halide to exchange with the metal ions on the clay. These released metal ions would then contribute to the (18) S. B. Hendricks and L. 95 (1940)
T.Alexander, Proc. Soil S e i . A m . ,
5,
751
content of metal chloride found in the washings of the reaction products and thus give high results as noted in Figs. 3 and 4. The base-exchange capacity of the alkali clays themselves (Table 11) is reflected in the metal chloride analyses, which show an opposite trend to the hydrolysis values and combustion analysis data. Although the larger cations are individually sterically available for reaction with an acyl halide, there is statistically a smaller tendency for these cations to be available for base-exchange reactions. Thus, the data obtained from the metal chloride analysis is largely influenced by the base-exchange capacity of the metal-bentonite. The color of the product formed from the reaction between benzoyl chloride and the aniline bentonite complex was dark gray with a strong violet undertone. Since this color developed early in the reaction and the hydrolysis value of the product is relatively high, the color is probably not due entirely to carbonization. An investigation of the synthesis of anhydrides from organic-clay complexes is being planned.
ORGANIC DERIVATIVES OF MONTMOIlILLONITE’ BY W. F. SPENCER^
AND
J. E. GIESEKJNG
Department of Agronomy, University of Illinois, Urbana, Illinois Receiued November 16, 1961
The nature of the OH groups on montmorillonite was studied by attempting to replace these groups by organic groups. The phenyl derivative of montmorillonite was prepared from dry rnontn~orillonitechloride by treating it with phenylmagnesium bromide, The montmorillonite chloride was prepared, by treating dry H-montmorillonite with thionyl chloride, Acetyl derivatives of montmorillonite were prepared by refluxing dry H-montmorillonite with either acetyl chloride or acetic anhydride. The montmorillonite was converted from a highly hydrophilic condition to a hydrophobic and somewhat organophilic condition by the substitution of OH roups by organic groups. When the acetyl groups were removed from acetylated montmorillonite by hydrolysis with 0.2 Ba(OH)r the resulting material again exhibited hydrophilic properties. It was not possible to remove all of the organic radicals from phenyl montmorillonite by boiling with 35% KOH for 30 minutes. The results indicate that the phenyl radicals formed a silicone type of bond with the montmorillonite.
%
Introduction The montmorillonites possess large amounts of highly reactive surfaces. Berger3 first detected the presence of OH groups in the montmorillonite surfaces by esterification with diazomethane. Two general types of organic montmorillonite derivatives have been studied in the past. These types are: (1) clay salts with organic cations on the exchange positions and (2) complexes in which the organic group is attached to the clay through a covalent bond. The organic cation clay complex has been studied more extensively and systematically than the latter type of complex. Gieseking4 studied the mechanism of exchange reactions in the montmorillonite-type clays and demonstrated that the organic cations tend to be attracted onto the surfaces of the mineral plates. Jordan6 has studied extensively the swelling properties of organic amine (1) Contribution from the Department o i Agronomy, University of Illinois, Urbana, Illinois. Published with the approval of the Direclor. Presented before the Colloid Division of the Jubilee hleeting of the American Chemical Society in New York, September 3 - 7, 1951 ( 2 ) Atonlic Energy Commission Fellow. (3) G . Berger, Chem. K’eekblad, 38, 42 (1941). (-1) J. E. Gieseking, Soil Scienre, 47, 1 (1939). ( 5 ) J. W. .lordan, Ttirs JOIJRXAL,63, 20-1 (19451).
clay complexes. DeuelG succeeded in preparing some organic montmorillonite derivatives with a covalent ether- or ester-type linkage (C-0-Si). He found that approximately 60 meq. of OH groups per 100 g. of clay were esterified. Gieseking’ observed that montmorillonite, when treated with acetyl chloride, lost its hydrophilic properties. The purpose of this investigation was to prepare some covalent bonded organic derivatives of montmorillonite and to study their stability and properties in an attempt to add to the knowledge of the surface characteristics and crystal structure of montmorillonites. Preparations H-Montmorillonite.-Wyoming bentonitme,finer than onc micron, was the source of the H-montmorillonite used throughout this investigation. Sodium saturated Wyoming bentonite was washed by decantation with 0.1 M HCI to replace the sodium ions and excess IICl was iwnoved by washing three tinies with distilled water. A Sharples supercentrifuge was used to separate the clay. The hydrogen saturated montmorillonite was dispersed in 95% ethanol in a Waring blendor and washed three times with al,solute ~~
(6) H. Deuel, G. Huber and R . Iberg, Helu. C h i m . Acto, 33, 1229 (1950). ( 7 ) J. E. Gieseking, Advances i ?A~g ~ ~ n o n t 1, y , 1 3 (1949).