J . Phys. Chem. 1986, 90, 641-644
641
The Interaction of Alkylammonium Salts with Synthetic Clays. A Fluorescence and Laser Excitation Study' T. Nakamura and J. K. Thomas* Department of Chemistry, University of Notre Dame, Notre Dame, Indiana 46556 (Received: July 12, 1985; In Final Form: September 25, 1985)
The interaction of surfactants with synthetic clays has been studied by a steady-state and a time-resolved fluorescencequenching method. The fluorescence quenching of 4 4 1-pyreny1)butyltrimethylammoniumion (PN') adsorbed on colloidal laponite clay by coadsorbed alkylpyridinium ions showed unusual behavior. Increasing the quencher concentration at first led to an efficient quenching of PN' fluorescence, but on increasing the quencher concentration further a reverse effect is observed, whereby the fluorescence started to recover, only to be followed by a smaller degree of quenching. The degree of recovery was strongly dependent on the chain length of the alkylpyridinium ions, the longer the chain length the larger the degree of recovery. PN' fluorescence was also quenched by laponite clay itself or cupric ion in copper-clay, where cupric ion was constructed in the clay lattice. However, coadsorbed hexadecyltrimethylammonium ions (CTAB cations) dramatically reduced the quenching. These results can be explained in terms of the change of geometrical arrangement of PN' adsorbed on clay. Quenching of PN' fluorescence by dimethylaniline, nitrobenzene, and nitromethane in the CTAB-laponite system obeyed Poisson type kinetics, indicating that the adsorbents exist in the form of clusters or zones on the clay surface.
Introduction A large body of literature indicates that clay systems have quite pronounced effects on many chemical reactions.z For the most part, these reactions involve an adsorption of the material of interest onto the clay followed by either a thermal or photochemical reaction to produce a colored product. A prime example is tetramethylbenzidine which becomes brightly colored when placed in contact with many clay system^.^ In this system, as in many others, electron-transfer reactions are involved whereby an electron is transferred from the material adsorbed onto the clay to a metal ion which is part of the lattice structure of the clay. The cations formed are strongly adsorbed at the anionic sites of the clay, and permanent chemical reactions is often achieved. Recently, there have been several fluorescence studies of molecules adsorbed on clay materials and, in particular, fluorescence quenching by the clay and by coadsorbed molecules via electron-transfer reaction^.^-^ The photochemical systems studied have been previously well established in micelles and similar colloids.68 For the most part fluorescent cationic materials such as ruthenium trisbipyridine and pyrenebutyltrimethylammonium bromide have been used to examine the anionic surfaces, use being made of the electrostatic attraction of the probe and surface. The cationic groups of the probe molecules are adsorbed strongly to the anionic exchange sites of the clay, thus locating the chromophores in close proximity to the clay surface. The spectroscopic properties of the probe molecules then report back on the nature of the clay surface or on events such as electron-transfer reactions which occur there. Neutral molecules may also be adsorbed to clay surfaces although these may not be adsorbed as specifically as the cations. In order to improve these latter systems it is possible (1) The authors thank the Army Research Office via Grant No. DAAG 29-82-K-0129 for support of this work. The authors thank Laporte Industries and Dr. R. Harrop for the laponite samples. (2) Theng, B. K.G. 'The Chemistry of Clay-Organic Reactions"; Adam Higler: London, 1978. (3) (a) Tennakoon, D.T. B.; Thomas, J. M.; Tricker, M. J.; Williams, J. 0.J. Chem. Soc., Dalron Trans. 1974, 2207. (b) Kovar, L.;DellaGuardia, R.; Thomas, J. K. J. Phys. Chem. 1984,88, 3595. (4) (a) Dellaguardia, R.; Thomas, J. K. J. Phys. Chem. 1983,87,990. (b) Ibid. 1983, 87, 3550. (c) Ibid. 1984, 88, 964. (5) (a) Schoonheydt, R.A.; Pauw, P. D.; Vliers, D.; De Schrijver, F. C. J. Phys. Chem. 1984.88, 5113. (b) Ghosh, P. K.;Bard, A. J. J. Phys. Chem. 1984, 88, 5519. (6) Turro, N.J.; Gratzel, M.; Braun, A. M. Angew Chem. 1980, 19,675. (7) Fendler, J. H."Membrane Mimetic Chemistry"; Wiley: New York, 1983. (8) Thomas, J. K. "Chemistry of Excitation at Interfaces"; American Chemical Society: Washington, DC, 1984;ACS Monograph No. 181.
0022-3654/86/2090-064 1$01.50/0
to coadsorb cationic surfactants onto the clay surface thus to make an organoclay system. The adsorbed surfactants provide a hydrophobic site for adsorption of hydrophobic molecules such as pyrene, which also can be used as a luminescent probe of the organoclay surfaces. The present study investigates the nature of the cationic adsorption process on clays and, in particular, the effect that the surfactant has on other material that is adsorbed onto the surface. The studies comment on the geometries of the clay-surfactant structures that are formed.
Experimental Section 4 4 1-Pyrenyl)butyltrimethylammoniumbromide (PN'Br-) was made from pyrenebutyric acid.9 A series of alkylpyridinium bromide, except hexadecylpyridinium chloride, were-made by refluxing an equivalent mixture of pyridine and the corresponding alkyl bromide under nitrogen for several hours. The oily or solid products obtained were washed or recrystallized from a mixture of ethanol and ethyl ether. Hexadecylpyridinium chloride and hexadecyltrimethykmmonium bromide (CTAB) were obtained from Sigma Chemical Co. and were used as received. Nitrobenzene (NB) from Eastman Kodak Co. was distilled under vacuum; N,N-dimethylaniline (DMA) from Eastman Kodak Co. was distilled over zinc powder; nitromethane (NM) from Eastman Kodak Co. was used as received. Laponite RD, which is a synthetic hectorite, was obtained from Laporte Industries and was used without further purification. A synthetic montmorillonite-like clay containing copper was made by a method similar to that of Graquist and Pollack.loJ1 In the synthesis copper to 25% of total equivalent of Al(OH)3 was added in the form of Cu(OH),. The reaction time and temperature was 12 h and 300 OC, respectively. A PRA pulsed Nz laser with pulse duration 500 ps, energy 2.5 mJ/pulse, and wavelength 337.1 nm was used to excite PN', and fast (C1 ns) spectrophotometry was used to observe the fluorescence decay. The details of the system and method of observing the fluorescence decay have been published previ~usly.~ The steady-state fluorescence data were measured on a PerkinElmer M P F 44B spectrofluorimeter, and absorption spectra on a Perkin-Elmer 55 1 spectrophotometer. Clay samples were prepared in the colloidal form with distilled deionized water by mixing a preweighed amount of the material (9) Wheeler, J.;Thomas, J. K. 'Inorganic Reactions in Organized Media", Holt, Smith L., Ed.; American Chemical Society: Washington, DC, 1982; ACS Symp. Ser No. 177,p 97. (IO) Granquist, W. T.; Pollack, S. S . Clays Clay Miner. 1967, 52, 212. (1 1) We thank Dr. J. Wheeler for making the synthetic clays.
0 1986 American Chemical Society
642 The Journal of Physical Chemistry, Vol. 90, No. 4, 1986
with the water followed by sonication for 0.5 min. Samples were deoxygenated by bubbling with prepurified nitrogen for several minutes immediately before use. All measurements were performed at 23 f 1 O C . The amount of DMA and NB adsorbed onto a CTAB-laponite system was calculated from the spectral absorbance of the supernatants obtained after centrifugation. The details of the method have been described elsewhere.I2
Results and Discussion Adsorption of PN+ on Laponite Clay. PN+ binds strongly to negatively charged clay surfaces and in the present system under the conditions used all the PN+ was bound completely to the laponite or synthetic clays. The effective concentration of PN' at the clay surface is greatly increased over that in the homogeneous solution; for example, at a clay concentration of 5 g/L the local concentration of the adsorbed PN' is some 200 times higher on the colloid surface, than if the cation was dispersed into the bulk of the aqueous system. This calculation assumes a thickness or a diffuse layer of 10 A on the clay surface where the adsorbed PN+ is located. With such large local concentrations of PN+ the excimer of PN+, (PN+)2*,which emits in 480-nm region is readily observed even though the overall concentration M is a concentration which gives only excited PNe of 4 X PN+, (PN+)*, in homogeneous solution. (PNf)*
+ PN'
Nakamura and Thomas
0
H \ H
0' 0. 0. 0. 0
0.1
0.2
CRP'I
0.3
/ mM
0
lRP*l_> 0.1 mM
s (PN+)2*
The initial excitation of PN' gives rise to (PN+)* which emits in the 400-nm spectral region and very rapidly forms the excimer in the clay system due to the large local concentration of PN+ at the clay surface. The time-resolved fluorescence decay of PN+ adsorbed on laponite is not first order even when the overall concentration of PN+ is only 5 X M, a condition when the local concentration on the clay surface is sufficiently small so that no excimers are observed. The fact that the decay of excited PN+ is not exponential is due to quenching by the clay itself. In earlier systems we found extensive quenching of excited adsorbed molecules due to the iron which is in the clay lattice.'* Laponite contains only traces of quenching materials, but apparently the small amount of quenching material is located at regions of the clay where PN' is adsorbed. Also, the heterogeneity of adsorption sites on the clay surface may be responsible for the nonexponential decay. Increasing the overall PN+ concentration beyond 4 X la-6 M leads to excimer formation but the decay profile of (PN+)* was not changed. This indicates that under the conditions used excimer formation is rapid and indicates a high local PN+ concentration. The (PN+)2*formation is instantaneous which requires that two PN+ molecules are adjacent on the clay surface. The excitation spectra of PN+ observed at 470 nm were virtually the same as those observed at 400 nm under the conditions of [PN'] C 0.4 mM, indicating that the excimer is formed dynamically although the excimer emission growth was within the laser pulse. However, increasing the PN' concentration to more than 0.4 m M the excitation spectra observed at 470 nm were different from those observed at 400 nm, indicating that ground-state aggregates of PN+ are a source of the excimer. Quenching by Various AlkyIpyridinium Ions. Alkylpyridinium ions which strongly quench pyrene and PN' fluorescence are also strongly adsorbed onto clay surfaces. Typical results of the quenching of the fluorescence of PN+ with various alkylpyridinium ions are shown in Figure 1A. The quenching behavior is unusual and nothing Like that observed in homogeneous solution. Increasing the quencher concentration first leads to an efficient (PN+)* quenching, but on increasing the concentration of alkylpyridinium salt a reverse effect is observed and the fluorescence tends to recover and increase indicating a smaller degree of quenching. A maximum quenching effect is reached at about an overall concentration of 0.02 m M alkyl pyridinium salt. The effects observed are strongly dependent on the chain length of the (12) Nakamura, T.; Thomas, J. K. Langmuir, in press.
[RP'I 20.02mM
_
+-p
_
_
Rf-I
-
-
-
Figure 1. (A) Influence of the alkyl chain length of various alkylpyridinium ions on the PN+ fluorescence quenching behaviors in laponite colloids (5 g/L): (0)ethyl-, (A) pentyl-, (0)octyl- (*) dodecyl-, (+) hexadecyl-; [PN+] = 5 X lo-' M. (B) Schematic model for PN+ fluorescence quenching by alkylpyridinium ions: Py+.-, alkylpyridinium ions; +-P, PN'.
alkylpyridinium ions; the longer the chain length the larger the degree of the unusual quenching effect. With short chain alkyl groups such as the ethylpyridinium ion the quenching behavior is conventional, i.e. a simple continuous decrease in the fluorescence yield is observed with increasing concentration of ethylpyridinium ion. All data may be explained as follows. In the quenching concentration range below 0.02 mM, PN' is oriented so that the molecule lies along the clay surface, the ammonium group being attached to the negative site on the clay surface with the alkyl chain and pyrene moiety stretched along the surface. Addition of the pyridinium ion leads to a similar situation, where at low concentrations the pyridinium ion or quenching group is on the clay surface with its alkyl chain also lying on the surface. At low concentrations the pyrene moiety and the pyridinium ion are in the same plane and efficient quenching is observed. However, as the pyridinium concentration increases further and beyond 0.02 mM, the interaction of PN+ and the alkyl chain part of the quencher becomes much more significant. Accordingly, the orientation of PN+ may be changed from that lying parallel and close to the clay to one where the butyl group and pyrene moiety extend away from the clay surface. Under these conditions the pyridinium quenching group still lies at the clay surface with its alkyl chain away from the surface. However, the pyrene moiety is attached to the surface by its ammonium group and the pyrene moiety and the butyl chain lie away from the surface, hence the pyridinium quenching group, and the pyrene moiety are no longer in the same plane but are separated in two layers along the clay surface which are approximately 4 A apart. Quenching is relatively inefficient and as more alkyl surfactant is added to the system the separation of quencher and probe becomes more pronounced and quenching disappears. A schematic explanation
Surfactant-Synthetic Clay Interaction
The Journal of Physical Chemistry, Vol. 90, No. 4, 1986 643
LsgO LA
0.2 0.1 0
-
0
0 9 r b 0 0 1
, 2
0
\
I
0.21
[NBleq /mM
[CTABl/mM
Figure 2. Plots of relative fluorescence intensities of PN+ as a function of CTAB concentrations in clay colloids: (0) laponite (5 g/L), ( 0 ) M. copper-clay (1 g/L); [PN'] = 5 X
for the unusual quenching behavior is also shown in Figure 1B. With shorter alkyl chains such as ethylpyridinium, the alkyl chain length is not sufficient to cause the above ordering effect on the clay surface and the quenching is normal, i.e. increased quencher leads to increased quenching and the data follow Stern-Volmer kinetics. Quenching by Lattice Copper Ions of a Synthetic Clay. Synthetic clays containing copper in the lattice rather than adsorption sites strongly quench the fluorescence of PN+ which is adsorbed at the clay surface. The quenching occurs via electron transfer from the excited pyrene chromophore to the lattice copper producing cuprous ions in the lattice and pyrene cations at the surface. However, addition of CTAB to the system leads to a marked increase in the fluorescence yields or a decrease in quenching efficiency as shown in Figure 2. This figure also shows that the quenching of PN' on laponite is also decreased on addition of CTAB. The fluorescence decay of PN+ is nonexponential in the clay system but becomes a single-exponential decay in the presence of CTAB indicating that the surfactant decreases or eliminates quenching by the clay material. These results again may be satisfactory explained in terms of an orientation of PN+ away from the clay surface on addition of the added surfactant CTAB. In the absence of CTAB the pyrene chromophores are in contact with the clay surface, a condition which leads to efficient quenching by metal ions such as copper af lattice sites in the clays. Addition of CTAB produces a layered structure on the clay surface and the pyrene chromophore of PN+ is displaced from a configuration which is close to the clay surface to one where the chromophore is separated by several angstrom from the clay surface. Quenching of the pyrene chromophore is inefficient in this geometric arrangement. The situation is similar to that observed in the pyridinium systems. Quenching by DMA in CTAB-Laponite Systems. Coadsorption of CTAB with PN' onto colloidal laponite provides a simple system where the pyrene chromophore is attached to the surface of the clay under conditions where the PN' fluorescence exhibits a single-exponential decay and no quenching by laponite takes place. Addition of DMA to this system gives rise to efficient quenching of the PN' fluorescence. A small but significant fraction of the added DMA is also adsorbed onto the CTAB cosurfactant which surrounds the PN', and the adsorption isotherm of DMA on the CTAB-laponite system is shown in Figure 3A. The adsorption of DMA rises almost linearly with the DMA in solution at low concentration and then shows a tendency to saturate at higher concentration. Figure 4 shows a steady-state Stern-Volmer plot of PN" fluorescecne quenching by adsorbed DMA; the amounts of DMA adsorbed on the clay surface were obtained from Figure 3A. The Stern-Volmer plot shows curvature but a semilogarithmic plot is linear indicating that the quenching kinetics are of a static type. The model suggested is a Poisson distribution of quencher molecules among reactive sites, i.e. a
Figure 3. Adsorption isotherms of DMA on NB on laponite: (A) DMA; (B) NB, (X) calculated values based on eq 3.
I
\0
I
[ D MA ladsd. /m M Figure 4. Stern-Volmer plot of PN+ fluorescence quenching by DMA in the IaponiteCTAB system and plot based on eq 1: [PN+] = 5 X IO" M, [laponite] = 5 g/L, [CTAB] = 0.2 mM.
random distribution of quenching species about the location of the fluorescing molecule. The situation is similar to that used in micelles:" In Io/Z = [Q]/[site]
(1)
where I,, I, [Q], and [site] refer to the initial fluorescence intensity, the fluorescence intensity in the presence of quencher at concentration [Q], and the site concentration, respectively. From the slope of this straight line the site concentration is calculated M. This value is 2% of the negative charge conas 8.1 X centration of laponite a t 5 g/L. These results lead to a model in which at least some coadsorbed CTAB molecules are associated in the form of clusters on the clay surface. This is not unexpected as a single CTAB molecule does not solubilize DMA in solution, while micellar aggregates of CTAB are effective host systems. Each cluster is an independent system where the PN+ is located and where quenching by DMA takes place. A quantitative analysis of the time-resolved fluorescence, which would help to confirm this model, is difficult due to the DMA luminescence which overlaps the pyrene chromophore emission in these systems. However, Figure 2 shows that the effect of CTAB addition on the PN+ fluorescence yield saturates at about 0.1 mM CTAB which is close to that of the site concentration. It should be added also that in the quenching study with long-chain alkylpyridinium ions, the point of maximum recovery of the PN' fluorescence yield (13) Turro, N. J.; Yekta, A. J . Am. Chem. SOC.1978, 100, 5951.
644
The Journal of Physical Chemistry, Vol. 90, No. 4 , 1986
TABLE I: Kinetic Parameters for Fluorescence Quenching by NB and NM' quenchers [Q]/mM ti k0/1O6 s-' k,/106 s-I NB 0 0 0.2 0.46 6.1 & 0.3 19 f 6 1 .o 0.79 2.0 1.1
I n 5.0 -0 L =l \
2
3.0
NM
v LL
0
.o
1
2.5 5.0
-C 1.01 0
400
600
6.0 f 0.2
20 f 6
1 .o
800
number of quenchers contained in a cluster, and k , denotes the rate constant for quenching of a excited PN+ in a cluster containing one quencher. In the above model it is assumed that each cluster, which may contain a different number of CTAB molecules, offers the same environment regarding the quenching reactions. The time-resolved fluorescence decay data for NB and N M can be satisfactorily explained by eq 2 as shown in Figure 5. The kinetic parameter values used for the simulations are listed in Table I. The average number of quenchers contained in a cluster, ii, is a function of the amount of quenchers adsorbed in the CTABlaponite system, [Qladsd,as follows:
5.0 -0 L
a
\
3.0
U
LL
-C 1 .01 0
0 0.36 0.69
'The parameter values are calculated on the basis of eq 2.
200
T i me/nS
2
Nakamura and Thomas
200
400 T i me/nS
600
800
Figure 5. Time-resolved fluorescence decay of 5 X lo-' M PN* in 0.2 mM CTAB-laponite (5 g/L) system with various quencher concentrations. (A) NB: (1) 0, (2) 0.2, (3) 1.0, (4) 2.0 mM. (B) NM: (1) 0, (2) 1.0, (3) 2.5, (4) 5.0 mM. Broken lines are simulations based on eq 2.
also appears at 0.1 m M pyridinium ion. Quenching by Nitrobenzene (NB) and Nitromethane ( N M ) . The CTAB clay systems also adsorb small amounts of N B and NM. The amounts of NM adsorbed are very small and cannot be observed by simple adsorption measurements, but the NB data are shown in Figure 3B. The time-resolved fluorescence decay data can be satisfactorily explained by a Poisson type quenching kinetics as shown in Figure 5.12,'4 The model introduced is mathematically identical with that of micellar systems. Therefore, the time-resolved fluorescence decay of PN+ in the CTAB-laponite system should be expressed by the following equation: In I ( t ) / Z ( O ) = -kot - ii(1 - exp(-k,t)] (2) where I ( 0 ) and Z ( t ) denotes the fluorescence intensity of PN+ at time zero and t , respectively; ko denotes the fluorescence decay rate constant in the absence of quencher, ri denotes the average (14) (a) Infelta, P. P.; Gratzel, M.; Thomas, J. K. J . Phys. Chem. 1974, 78, 190. (b) Tachiya, M. Chem. Phys. Lerr. 1975, 33, 289. (c) Atik, S. S.; Singer, L. A. Chem. Phys. Lett. 1978, 59, 519.
In eq 3 it is assumed that all the quencher molecules adsorbed exist in clusters and directly participate in quenching reactions. Using eq 3 one can calculate [ Q l a a , since the cluster concentration has already been determined in the previous section as 8.1 X M. The results thus obtained for N B are shown in Figure 3B together with those obtained from the independent centrifuge method. Both results agree within the experimental errors. In the case of N M such a comparison could not be done because of the low adsorptivity of N M into the CTAB-laponite system. A quantitative treatment of the steady-state quenching data for N B was not attempted because of strong absorption by N B at the excitation wavelength. The quenching for N B and N M are identical in nature and occur via electron transfer from the excited pyrene chromophore to the NOz group of the quencher. Similar kinetic effects are observed in both systems, and it is expected that the mechanisms of the CTAB-laponite systems are similar.
Conclusion The data for the quenching of an excited adsorbed chromophore by an adsorbed quencher, on colloidal clays, show a strong dependency on the geometry or arrangement of the reactants at the clay surface. The geometry of the reactants depends on the nature of a coadsorbent, i.e. cationic surfactant. Reaction occurs at selected sites on the clay surface which indicates a zoning or clustering of the adsorption sites into selected sites. Fluorescence quenching studies provide a reasonable picture of the interplay of struction and adsorption and these effects on kinetics of reactions at clay surfaces.