482
Langmuir 1990, 6, 482-486
Study of the Adsorption and Polymerization of Functionalized Organic Ammonium Derivatives on a Clay Surface B. Kunyima,? K. Viaene,? M. M. Hassan Khalil,? R. A. Schoonheydt,$ M. Crutzen,? and F. C. De Schryver*$+ Laboratory of Molecular Dynamics and Spectroscopy, Department of Chemistry, K. U. Leuven, Celestijnenlaan 200 F B, 3030 Leuven, Belgium, and Laboratorium voor Oppervlaktechemie, K. U. Leuven, Kardinaal Mercierlaan 92 B , 3030 Leuven, Belgium Received J u l y 13, 1989. I n Final Form: September 18, 1989 In this contribution, results are presented suggesting a reorganization of the adsorbed molecules at a given loading on clay surfaces which could be ascribed to hemimicelle or admicelle formation. The size of these detergent aggregates could be obtained by two methods. In the first method, the fluorescence decay of [3-(l-pyrenyl)propyl]trimethylammoniumbromide, P3N, was analyzed by using the analytical functions of the fluorescence quenching of molecules solubilized in micellar media. From these data, an aggregation number of 24 for CTAC is obtained. In the second method, clusters of ((methacryloy1oxy)ethy1)methyldidodecyla"onium chloride are polymerized. When the molecular weight of the obtained polymer is determined, an aggregation number of 7 is obtained. In addition, some measurements on the stability of diluted clay suspensions were performed. The experimental results indicate a face-toface aggregation of the clay particles.
Introduction The adsorption of organic molecules on a clay surface can be studied by means of specifically designed fluorescent probes.'-' Indeed, studying the photophysical properties of the adsorbed probes allows information on the adsorption and distribution at the clay surface of organic molecules to be obtained. The attractive features of clay minerals are their large surface area, their abundant occurrence in nature, and their catalytic proper tie^.^^" In a number of studies, cationic pyrene derivatives have been used to examine the nature of the adsorption process, since these molecules are strongly bound to the negatively charged clay particles. Also, the influence of coadsorbed cationic surfactants on the photophysical properties of the adsorbed probe molecules has been investigated. In all these studies, it was found that the adsorbed molecules form clusters on the surface in aqueous suspension^.^^^ The driving force of this aggregate formation was ascribed to hydrophobic interactions between the adsorbed detergent-like mole c u l e ~ . ~This . ~ cluster formation of adsorbed [ 3-( l-pyrenyl)propyl]trimethylammonium bromide, P3N, molecules results in ground-state interactions between the pyreLaboratory of Molecular D namics and Spectroscopy. Laboratorium voor Opperdaktechemie. (1) Della Guardia, R. A.; Thomas, J. K. J . Phys. Chem. 1983, 87,
+
990. (2) Nakamura, T.; Thomas, J. K. Langmuir 1985,1, 568. (3) Della Guardia, R. A.; Thomas, J. K. J. Phys. Chem. 1984, 88, 964. - - -.
(4) Schoonheydt, R. A.; De Pauw, P.; Vliers, D.; De Schryver, F. C.
J. Phys. Chem. 1984,88, 5113.
(5) Nakamura, T.; Thomas, J. K. J . Phys. Chem. 1986,90, 641. (6) Viaene, K.; Ji, C.; Schoonheydt, R. A.; De Schryver, F. C. Langmuir 1987, 3, 107. (7) Viaene, K.; Schoonheydt, R. A.; Crutzen, M.; Kunyima, B.; De Schryver, F. C. Langmuir 1988,4,749. (8) Viaene, K.; Schoonheydt, R. A.; Crutzen, M.; Kunyima, B.; De Schryver, F. C. Prog. Colloid Polym. Sei. 1988, 266, 242-246. (9) Theng, B. The Chemistry of Clay-OrganicReactions;Adam Hilger: London, 1978.
(10) Spoeito, G. Surface Chemistry of Soils; Oxford University Press: New York, 1984.
nyl groups and very efficient static excimer formation. The adsorption of detergents on various surfaces, such as silica, alumina, and Fez03,has been e~amined.'~-~' In these studies, it was found that, at a certain concentration of the detergent, clusters of these surfactants are formed on the surface. The driving force of this aggregation of the adsorbed detergents can be compared to the driving force for micelle formation in aqueous media. Although the structure of these aggregates is not yet elucidated, it was initially thought that at this concentration a local monolayer is formed at the surface. These local monolayers are referred to as hemimicelles, and the concentration a t which these aggregates are formed is the critical hemimicelle concentration. Recently, however, calculations showed that the formation of local double layers (admicelles) would be energetically more favorable.'g~20In this paper, an attempt to determine cluster size at a given percentage of surface coverage and a study of reorganization at very low loadings of organic molecules on a clay surface are discussed.
Experimental Section The synthesis of [3-(1-pyrenyl)propyl]trimethylammonium bromide (P3N), cetyltrimethylammonium chloride (CTAC), didodecyldimethylammonium chloride (DDAC), and ((methacryloy1oxy)ethyl)methyldidodecylammonium chloride (DDMEMEC)has been described elsewhere!J1J2 (11) Viaene, K.; Verbeeck, A.; Gelade, E.; De Schryver, F. C. Langmuir 1986, 2, 456. (12) Voortmans, G.; Jackers, C.; De Schryver, F. C. Br. PO~YM. J. 1989,21, 161.
(13) Boens, N.; Van Den Zegel, M.; De Schryver, F. C. Chem. Phys. Lett. 1984, 111, 340. (14) Gao, Y.; Du, J.; Gu, T. J. Chem. Soc., Faraday Tram 1 1987,
83, 2671. (15) Gaudin, A.; Furstenau, D. Trans. Am. Znst.Min., Metall. Pet. Eng. 1955,202,958. (16) Gaudin, A.; Furstenau, D. Tram. Am. Zmt. Min., Metall. Pet. Eng. 1955,202, 959. (17) Scamehorn, J.; Schechter, R.; Wade, W. J. Colloid Interface Sci. 1982, 86, 463. (18) Yeskie, A.; Harwell, J. J. Phys. Chem. 1988, 92, 2346. (19) Somasundaran, P.; Furstenau, D. J. Phys. Chem. 1966, 70, 90. (20) Somasundaran, P.; Healy, T.; Furstenau, D. J. Phys. Chem. 1964, 68, 3562.
0 1990 American Chemical Society
Langmuir, Vol. 6, No. 2, 1990 483
Adsorption of Organic Molecules on Clay
Table I. Surface Properties of the Clays external Na+ CEC, clay surface, m2/g pequiv/g FezOnrwt hectorite 63 526 0.02 barasym 133 464 0.05 laponite
360
733
%
6t
+ + + + + + +
+
+
I
I
+
0.06
Three clays were studied: hectorite, barasym, and laponite. Barasym is a synthetic mica-type montmorillonite while laponite is a synthetic hectorite. These clays are chosen because of their low iron content and their difference in surface properties (Table I). Laponite was used as such, while hectorite and barasym were saturated with Na+ by repeated exchange with NaCl solutions (1M). The fraction containing particles smaller than 2 Km was collected by centrifugation, salted out with NaCl, and washed Cl- free. Stable suspensions, fit for spectroscopic studies, containing particles smaller than 0.3 pm were prepared by centrifugation. The dry weights of the suspensions were determined by drying a known volume at 383 K; 10 mL of these suspensions was exchanged with probe and detergent,the amount of which depended on the desired exchange level. In most cases, the amount of adsorbed probe (P3N) was 0.75% of the Na cation exchange capacity (CEC) of the clay. The detergent concentration was limited to 25% of the CEC to avoid flocculation at higher loadings. All samples were studied in a closed system in which the air was removed by bubbling argon through the solution for 15 min prior to the measurement. Fluorescence and excitation spectra were recorded on a Spex fluorolog. Fluorescence decay parameters were measured with a Spectra Physics cavity-dumped,mode-locked, synchronously pumped R6G dye laser with time-correlated single photon timing detecti~n.'~ Raman spectra were recorded on a Coderg T800 spectrophotometer with a Spectra Physics 164 Ar laser as the excitation source. Viscosity measurements were made by using a calibrated Ubbelohde viscosimeter. The polymerization was performed by irradiating a diluted clay suspension with 250nm light, after adsorption of DDMEMEC (27% of the CEC) and after removing the oxygen by bubbling argon through the suspension for 1h.
Results and Discussion Fluorescence Spectroscopy of P3N on Clays. In the following experiments, the spectroscopic properties of P3N are examined as a function of the concentration of P3N and/or detergent, in order to find out if, for the adsorption of detergents on a clay surface, evidence for hemi- or admicelle formation can be obtained. 1. Spectroscopic Properties of P3N as a Function of Loading. The ratios of the intensities of the excimer and monomer fluorescence of P3N, adsorbed on laponite, hectorite, and barasym, are given in Figure 1 as a function of the loading. For laponite and barasym, three coverage regions can be distinguished. In the first region, a t very low concentrations, for which at the most 2.3% of the CEC is covered, the ratio I J I , is almost constant with increasing loading. In a second, very narrow region (2.3-3.7%), a steep increase of this ratio is observed. After this steep increase, the ratio remains again almost constant. In the case of hectorite, however, the ratio I,/ I, initially increases and remains constant above 2.3%. 2. Adsorption of P3N Measured by Fluorescence Spectroscopy. The fluorescence intensity of the equilibrium solution (after removal of the clay) as a function of the total concentration of P3N added to a laponite suspension is given in Figure 2a and compared with the I J I , ratio of the suspension in Figure 2b. The observation of a fluorescence in the supernatant equilibrium solution indicates that a t the low loading investigated the adsorption is not 100%. The fluorescence of the super-
I
I
I
I
I
I
I
23 43 (7-CEC J
03
0
63
I 83
Figure 1. Ratio of the intensities of the excimer and monomer fluorescence of P3N adsorbed on laponite (o), barasym ( O ) ,and hectorite (+) as a function of the loading (% CEC). 5 O
I
0
0.3
L.3
2.3
ctot
O
O
6.3
107 M
Figure 2. (a) Fluorescence intensity of the equilibrium solution as a function of the total concentration of P3N added to a laponite suspension. (b) Ratio of the intensities of the excimer and monomer fluorescenceof P3N added to a Laponite suspension as a function of total concentration of P3N. natant first increases monotonically with P3N concentration to reach a maximum at the minimum of I J I , at a total concentration of 1.3 X lo-' M. However, even at the maximum in Figure 2a the nonadsorbed amount is less than 1%of the total amount of P3N. At higher concentrations, the fluorescence of the supernatant rapidly decreases to zero, indicating quantitative adsorption. The onset of quantitative adsorption is accompanied by the steep increase of IJI, (Figure 2b). For hectorite, the adsorption of P3N is quantitative at all loadings. 3. Influence of Detergent Concentration on & / I , of P3N Adsorbed on Laponite. Coadsorption of detergents, with a sufficient chain length, results in a decrease The detergent of the I J I , ratio of adsorbed P3N.acts as a diluent and separates the P3N molecules, resulting in a decrease of the relative excimer intensity. In Figure 3, In I J I , for P3N, adsorbed on laponite, is given as a function of DDAC and CTAC concentration. In the case of DDAC, this ratio is always smaller than that for CTAC, indicating that DDAC is a more efficient diluent than CTAC.' Below a loading of 2.5% of the CEC, In (&/I,) remains almost constant. Above that loading, In (IJI,) decreases monotonically with increasing loading of detergent.
Discussion The experimental results show a discontinuity in the dependence of the spectroscopic properties of P3N on the loading for laponite and barasym suspentions. First
484 Langmuir, Vol. 6, No. 2, 1990
*
Kunyima et al.
0 0 0
I 1650
c
T-'
Figure 4. Raman spectra of DDMEMEC on laponite surface before (upper curve) and after irradiation (lower curve). Table 11. Percent Conversion of the Double Bond as a Function of Irradiation Time
Since at very small concentrations the fluorescence intensity is proportional to the concentration, the following equations can be written:
-
If
If
If
-
[MI,,,
(4)
[MIaddK'
(5)
(1/(1+ K'))C,,
(6)
with
+ [Mlso~n
(7) Together with the data of Figures 1and 2a, it shows that the discontinuity in the spectroscopic properties occurs a t a characteristic loading of 2.3% of the CEC or 10.7 and 16.9 pequiv of P3N/g for barasym and laponite, respectively. Below that loading, adsorption is not quantitative; above that loading, it is. Such a sudden transition from nonquantitative to quantitative adsorption loading has been observed for several detergents on various surface^'^-^' and has been ascribed to hemi- or admicelle formation. It seems appropriate to assume that in the case of the adsorption of P3N molecules on clay surfaces the discontinuity can be ascribed to cluster formation, eventually, of a hemi- or admicelle nature on the clay surface. The increase of the ratio I,/ I , can be explained by assuming that the reorganization results in a higher local concentration of the adsorbed P3N molecules. I t was also found that at a certain concentration of coadsorbed detergent molecules (2.8% CEC) the decrease of the Ze/Imratio of P3N becomes more pronounced. This also can be ascribed to a reorganization of the adsorbed detergent molecules, resulting in a higher local concentration. The clusters that are formed solubilize P3N efficiently and decrease the concentration of excimer. An interesting result is the observation that this reorganization of the adsorbed molecules is only observed in laponite and barasym suspensions. In a hectorite suspension, the spectroscopic properties do not change abruptly Ctat =
Mads
time, min
absorbance
% conversion
0 1 2 5 10 15 20 30 40
0.76 0.68 0.60 0.43 0.26 0.21 0.18 0.16 0.10
0 10 19 38 58 65 68 70 78
with the loading. The difference between laponite and barasym on the one hand and hectorite on the other is the particle size. The external surface of hectorite in suspension is significantly less than that of laponite. As a consequence, the critical concentration for cluster formation is reached at such low loadings that they fall beyond detection of our experiments. The relation between the loading, necessary for the onset of clustering, and the surface areas of the clays in Table I suggests that, at the low loading of the present investigation, the adsorption phenomena occur at the external surfaces of aggregates of clay particles in aqueous suspension. Photopolymerization of DDMEMEC on a Laponite Surface. To prove that polymerization took place, Raman spectra of the suspensions were recorded before and after irradiation for several hours (Figure 4). Before irradiation, a band at 1650 cm-' is observed, which can be ascribed to the double bond of the adsorbed monomer. After irradiation, this band disappears, indicating that polymerization on the clay surface took place. Additional proof for the polymerization has been obtained by UV spectroscopy. The adsorbed monomer shows an absorption in the region 209-230 nm, which decreases as a function of irradiation time (Table 11). In table 11, for example, the percent conversion of the double bond has been calculated as a function of irradiation time minutes) for a DDMEMEC concentration of 14% CEC. From the slopes of the initial straight portions of the percent conversion versus irradiation time curves, the rates of polymerization at low coverages were calculated. In Figure 5, these values are plotted as a func~~~
~~~~
(21) Gelad6, E. Doctoraataprwfschrift K. U. Leuven, 1983; p 242. (22) Almgren, M.; Grieser, F.; Thomas, J. K. J. Chem. Soc., Faraday Trans. 1 1979, 75, 1674. Almgren, M.; Grieser, F.; Thomas, J. K. J. Am. Chem. SOC.1979,101, 279. (23) De Mayo, P.; Natarajan, L.; Ware, W. Chem. Phys. Lett. 1984, 107, 187. (24) Van Olphen, H. An Introduction to Clay Colloid Chemistry; Wiley: New York, 1963.
Langmuir, Vol. 6, No. 2, 1990 485
Adsorption of Organic Molecules on Clay
for the intermolecular excimer formation of pyrene adsorbed on a silica surface. In absence of coadsorbed molecules, he observed ground-state complexing leading to excimer emission, while after addition of alcohols dynamic excimer formation was reported. Assuming no exchange of excited P3N molecules between the detergent clusters, the physical meaning of the parameters A, to A , is
2 30
2 11 8L 2
1/
A, = I , A, = k ,
IO -10 2
-100
-980 -950 I n LDDMEMEC]
-9LO
-920
Figure 5. Rate of polymerization (R,) as a function of
DDMEMEC concentration (M).
tion of monomer concentration by using27
R, = k[M]" (8) At low loadings, [DDMEMEC] < 15% CEC the expected linear dependence of In R, on the monomer concentration is observed with 0.4 as the slope. If the polymerization is performed in the presence of P3N, an increase of the I e / I m ratio was observed after polymerization. The same result is obtained when the same concentration of P3N is added to the clay suspension before and after the polymerization. Again the I,/ I , ratio of P3N is larger when P3N is added to the suspension after irradiation. This suggests that a kind of segregation takes place during polymerization resulting in a higher local concentration of P3N. This can be compared to the separate clusters formed by detergents and P3N molecules on a clay surface, when the chain length of the detergent is not long enough. Determination of the Size of the Detergent Clusters. As the detergents form micelle-like aggregates on the clay surface, an attempt was made to describe the fluorescence decay of adsorbed P3N molecules using decay laws utilized in the analysis of fluorescence quenching in micellar media.20y21 These measurements were performed on a laponite suspension in the presence of CTAC molecules. The fluorescence decay of a laponite suspension with a total coverage of 12% CEC (using a [CTAC]/ [P3N] ratio of 200) was analyzed. Even at this ratio, an excimer emission at 480 nm is observed. The fluorescence decay of P3N, monitored at 378 nm, could be fitted to I ( t ) = A , exp(-A,t) - A3[l - exp(-A,t)]
(9) This equation describes the fluorescence quenching in micellar and allows, within certain experimental conditions, the determination of the aggregation number of the detergent clusters. The intermolecular excimer formation is, in the presence of CTAC molecules, a dynamic quenching process. Indeed, the fluorescence decay of the excimer emission shows a growing-in. It should be noted that the excimer emission of P3N adsorbed on a clay surface in the absence of detergent molecules is not due to a dynamic quenching, indicating cluster formation of the P3N molecules.6 This phenomenon was also observed by de Mayo et al.23 (25) Turro, N.; Kumar, C.; Grauer, Z.; Barton,J. Langmuir 1987,3, 1056. (26) Bandrup, J.; Immergut, E. H. Polymer handbook, 2nd ed.; 1966; p IV-21. (27) Lougnot, D. J.; Fouassier, J. P. J. Polym. Sci., Polym. Chem. Ed. 1988,26, 1021-1033. (28) Solomon, D. H. J. Maccromol. Sci. Reo. Macromol. Chem. 1967, 1, 179.
(10) (11)
A3 = [P3Nl/[Ml (12) A, = k, (13) Where [MI is the micelle concentration (in the first approximation proportional to the detergent concentration), I, the fluorescence intensity at time zero after excitation, k , the reciprocal of the fluorescence lifetime of P3N molecules which are not quenched, and k, a first-order rate constant related to the excimer f~rmation.'~From these parameters, the aggregationnumber of the detergent clusters can be calculated:
Nagg= A3 [CTAC]/[P3N] (14) These experiments were performed at two P3N concentrations of respectively 2 X and 7.8 X M mainM). taining the CTAC concentration constant (4 X In Table 111, the results of global analysis at two different wavelengths for the two concentrations of P3N on laponite are given. An average aggregation number of 24 has been obtained. The aggregation number obtained suggests rather small detergent clusters on the surface. Gao et al. determined also the aggregation numbers of clusters of alkylammonium acetates with a chain length of 10-18 carbons on silica and found an indication that small detergent clusters are formed." Further work is in progress to evaluate the limits of the use of the analysis of the fluorescence decay of P3N adsorbed on a clay surface. An alternative method to obtain the aggregation number of the clusters is the determination of the molecular weight of the polymer formed by polymerization of detergent clusters. In order to obtain the molecular weight of the polymer derived from DDMEMEC, formed on the clay surface, the polymers had to be detached from the clay surface. It is known that desorption of polymers is very difficult since one polymer molecule is attached to the clay surface by several positive charges. Another difficulty for the desorption is the insolubility of the polymer in an aqueous medium. It was, however, possible to desorb the polymer by chemical modification. The polymer-clay mixture was added to a concentrated NaOH (3 M) solution and refluxed at 100 "C for 24 h. This basic hydrolysis results in the formation of poly(methacry1ic acid), which is soluble in basic media. The clay particles, together with the ammonium residues, are precipitated by centrifugation (4000 rpm). After separation, the polymer is precipitated from the supernatant solution by adding 3 M HC1. The IR spectrum of the obtained polymer is identical with the IR spectrum of poly(methacrylic acid)," and the molecular weight of the obtained polymer was determined by viscosity measurements in a NaOH solution (2 M) at 25 "C by using eq 8: 101 = K(MW") (15) Where [ q ] is the intrinsic viscosity, K = 42.2 X lo-' dm3/ g, and a is 0.64.'* The obtained molecular weight (MW) corresponds to seven monomer units. This result also
486 Langmuir, Vol. 6, No. 2, 1990
[CTAC],1 0 b M
[P3N], IO'M
4
Table 111. Global Analysis. A,x lo-' A3x 10
X,nm
A,x 10
A,x lo2
A,
2,'
380
6.5320
2.0322
1.2584
4.2161
15.6
1.278
385 378
6.3967 8.3046
2.0322 1.9707
1.2584 4.2439
4.2161 9.9578
15.6 15.6
1.6886 1.257
385
8.2954
1.9707
4.2439
9.9578
15.6
2.639
2
4 a
Kunyima et al.
7.8
ZxZglob
N,,
1.995
25
2.652
22
A, is the lifetime of the reference (isopropylcarbazol) in ns. Definition of A,-A, in text.
2
I
5
L
3 time
days]
Figure 6. Ratio of the intensities of the excimer and monomer fluorescence of P 3 N adsorbed on laponite as a function of time (days). 09LO I
I Laponite
Laponi t e + Probe1 P3NI
," 0935 1
e 0 920 1
0
I
1
I
I
A
I
2
3
4
5
6
7
8
Time [ d a y s )
Figure 7. Viscosity of a laponite suspension with and without P3N as a function of time (days). suggests rather small detergent clusters on the clay surface. Stability of Dilute Clay Suspensions. In Figure 6, the ratio I e / I m of P3N, adsorbed on laponite, is plotted as a function of time (days). In this figure, an increase of this ratio is observed with time. In Figure 7, the viscosity of a laponite suspension with and without adsorbed P3N molecules is given, again as a function of time. It is clear that the viscosity of a laponite suspension decreases with time until it remains almost constant after several days at 25 "C. When P3N molecules are adsorbed on the clay particles, the decrease of the viscosity is more pronounced. The observed changes in the viscosity of a dilute laponite suspension are indicative for aggregation of the clay particles, either face-to-face, face-to-edge, or edge-to edge. Because of the small particle size of laponite (average plate dimension is 30 nm), no clear distinc-
tion can be made between these three types of aggregation. For large particles, it is known24that only a faceto-face aggregation results in a decrease of the viscosity. The P3N molecules are initially adsorbed on the external surface of the aggregates. Aggregation diminishes the external surface area, and the excimer concentration must increase because of the increase of local concentration of P3N. Also, the electrical double-layer repulsion is smaller in the presence of P3N. As a consequence, the rate of aggregation is higher in the presence of P3N. Turro et al.25also found a change of the spectroscopic properties of Ru(bpy),2+ as a function of time. In this study, an increase of the fluorescence lifetime after several days was observed. These authors explained this phenomenon also by an aggregation of the clay particles. They assumed a more hydrophobic environment between the clay platelets, after aggregation took place, resulting in an increase of the fluorescence lifetime of adsorbed RU(bPY)aP+. When detergents are coadsorbed on the clay surface, the ratio &/Imdecreases with time. This can also be ascribed to the aggregation of the Laponite particles. Indeed, the local concentration of detergents will increase upon aggregation, resulting in a better separation of the P3N molecules. These experimental results indicate that even very dilute laponite suspensions undergo aggregation phenomena, which also influence the spectroscopic properties of the adsorbed molecules. Therefore it is necessary to adhere to a strict protocol in the preparation of the sample, but when this procedure is followed, the experimental data are perfectly reproducible.
Conclusions In this contribution, some results are presented which indicate a reorganization of the adsorbed molecules at a given loading, which can possibly be ascribed to hemior admicelle formation. The size of these detergent aggregates could be obtained by two methods. In the first method, the fluorescence decay of P3N was analyzed, using the analytic function derived for the fluorescence quenching of molecules solubilized in micellar media. From these calculations, the aggregation number of a CTAC cluster at 12% coverage is obtained. In the second method, clusters of a polymerizable detergent are polymerized. When the molecular weight of these polymers is determined, the aggregation number of the clusters is obtained. In both cases, the size of the cluster is rather small. In addition, some measurements on the stability of diluted clay suspensions were performed. The experimental results indicate a face-to-face aggregation of the clay particles. Acknowledgment. We are very grateful to ERO, the FKFO, and the ministry of Scientificprogrammation (GOA and UAPS) for continuous financial support to the laboratory. Registry No. C T A C , 112-02-7; D D A C , 3401-74-9; DDMEMEC, 114199-09-6; P3N, 105763-23-3; hectorite, 1217347-6; laponite, 53320-86-8; barasym, 37220-90-9.