Characterization of an alkylammonium- montmorillonite-phenanthrene intercalation complex by carbon-13 nuclear magnetic resonance spectroscopy

Low molecular weight polycyclic aromatic hydrocarbons can intercalate from the solid phase into montmorillonite (Mt) saturated with quaternary alkylammonium ions. However, the interaction and relationship between guest and host organic molecules in the interlayer space of the clay are not well understood. We have intercalated phenanthrene into tetradecyltrimethylammonium (TDTMA)-montmorillonite by a solid-solid reaction. The basal spacing of the original TDTMA-Mt complex is close to 1.8 nm, indicating the presence in the interlayer space of a double layer of TDTMA ions with the alkyl (polymethylene) chains lying parallel to the silicate layers, and the carbon zig-zags adopting an all-trans conformation. After intercalation of phenanthrene the basal spacing increases to about 3.4 nm, indicating a change in orientation of the alkyl chains with respect to the silicate layers. 13C-NMR spectroscopy shows that adding phenanthrene to TDTMA-Mt leads to a displacement by -3 ppm of the -(CH2)nsignal for TDTMA. This signal and that for interlayer phenantbrene are also broadened relative to the respective pure compounds. These observations, together with measurements of nuclear spin relaxation time constants, strongly suggest that in the complex with phenanthrene the polymethylene chains of TDTMA extend away from the silicate layers, and no longer assume a rigid all-trans carbon zig-zag conformation. Rather, the TDTMA chains become relatively disordered and intimately mixed with phenanthrene. Polycyclic aromatic hydrocarbons (PAHs), of which phenanthrene is a member, are widespread in the environment. In particular, soils near disused gas works contain high levels of PAHs arising from the incomplete combustion of coal and oil (Byers et al., 1994; Chen et al., 1996). Concern about the persistence and genotoxicity of PAHs has stimulated much research into the bioremediation of contaminated sites (Durant et al., 1995; Grosser et al., 1995). In these sites PAHs are concentrated in coal tar and, therefore, can presumably be taken up, if not intercalated, into an organo-mineral matrix. However, the potential for the in situ containment of PAHs by entrapment or intercalation has not been explored. The adsorption of alkylammonium ions by 2:1 type layer silicates, notably montmorillonite and vermiculite, has been well documented (Theng, 1974; Lagaly, 1993). The process is essentially one of exchange between the inorganic cations (e.g. Na), initially present at the clay surface, and the organic cations in solution. Being hydrophobic, alkylammonium-exchanged clays have a larger propensity for sorbing non-ionic organic compounds (e.g. PAHs), than the parent (unmodified) materials (Jaynes & Boyd 1990; 1991). Nevertheless, the amount that can be taken up from aqueous solutions is limited because such compounds have a low solubility in water (Smith et al., 1990). This limitation may be overcome by mixing the non-ionic compound as a solid with dry alkylammonium-exchanged clays. In this way, Ogawa et 9 1998 The Mineralogical Society 222 B. K. G. Theng et al. al. (1992; 1995) were able to intercalate up to 50% by weight of naphthalene, anthracene, and pyrene into montmorillonite saturated with long-chain quaternary ammonium ions. They suggested that the process was driven by hydrophobic interactions between the guest PAH molecules and the host alkylammonium ions. If so, we might expect the conformation and motion of the interlayer guest and host organic species to be substantially altered. Nuclear magnetic resonance (NMR) spectroscopy has been useful in probing the dynamic behaviour of guest molecules in zeolites (Pfeifer, 1976; Lechert & Basler, 1989; Hong et al., 1993; Qiang Xu et al., 1996). To our knowledge, however, this technique has not been previously applied to characterize organic compounds in layer silicates. One obstacle has been the need to provide evidence that the NMR signals arise principally from the intercalated species rather than from the portion that is attached to external crystal surfaces. We have overcome this problem by intercalating phenanthrene from the solid phase into montmorillonite (Mt) saturated with tetradecyltrimethylammonium (TDTMA) ions using the method of Ogawa et al. (1992). Following Cheung & Gerstein (1981) the intimacy of mixing (association) between phenanthrene and TDTMA in the interlayer space of the clay has been assessed through proton spin diffusion using NMR spectroscopy. We should stress that it is the proton spin information that diffuses, and not the protons themselves. This information, which is exchanged between pairs of neighbouring protons through interactions between the two magnetic dipoles, can diffuse over dimensions of the order of tens of nanometres over milliseconds, or 100 nm over a period of about 1 s (Zumbulyadis, 1983). Distinct proton spin relaxation constants for guest and host molecules can therefore be taken as evidence for a lack of mixing or heterogeneity on a scale of tens of nm or more. The test for heterogeneity can be further enhanced by observing proton spin diffusion processes through the effects on 13C NMR signals (Zumbulyadis, 1983; Tekely et al., 1989). The improved chemical shift dispersion of signals is then combined with the sensitivity of proton spin relaxation to probe changes in molecular dynamics. M A T E R I A L S A N D M E T H O D S The montmorillonite used was a Na-rich specimen supplied by Kunimine Industries Co., Japan marketed as 'Kunipia-F', and commonly abbreviated to 'KpF'. Its structural formula (Iwasaki & Onodera, 1995) is: Nat.71Cao.o6Mgo.ol4(SiT.soAlo.2o) TM 3+ 2+ V! (A13.14Feo.2oFeo.o4Mgo.62) O2o(OH)4 Tetradecyltr imethylammonium bromide of -~99% purity was obtained from Sigma Chemical Co. (St. Louis, Missouri, USA), and reagent-grade phenanthrene (>98%) from Merck-Schuchardt (Hohenbrunn, Germany). Both compounds were used as received. The cation exchange capacity (CEC) of the KpF sample was determined by leaching with 1 M ammonium acetate at pH 7, washing with 90% ethanol, displacing the ammonium with 1 M NaC1, and measuring the amount displaced with an autoanalyzer (Blakemore et al., 1987). A CEC of 113 cmol(+) kg -1 was measured, in good agreement with the value of 115 cmol(+) kg -1 given by the manufacturer. Conversion to the TDTMAexchanged form was achieved by shaking the clay with an aqueous solution of TDTMA-bromide for which the concentration of TDTMA ions was equivalent to the CEC. After removing excess electrolyte by repeated washing with deionized water, the TDTMA-Mt complex was dried in an oven at 30~ The TDTMA-Mt-phenanthrene intercalate was prepared by mixing three parts of the TDTMA-Mt complex with one part of solid phenanthrene for 10 min in an agate mortar and pestle, as described by Ogawa et al. (1992). The amount of TDTMA and phenanthrene present in the respective complexes was checked by carbon and nitrogen analysis using a Leco FP-2000 instrument and EDTA as a calibration standard. X-ray diffractometry (XRD) was carried out by air-drying an aqueous suspension of the samples on a glass slide, and scanning at a rate of 2 ~ 20 min -~, using a Philips PW1010 diffractometer and Co-Kct radiation (Whitton & Churchman, 1987). Infrared (IR) spectra were obtained from KBr discs using a Digilab FTS-7R Fourier-transform spectrophotometer. The J3C NMR spectra were obtained by packing 0.25-0.32 g of air-dry samples in a 7 mm diameter silicon nitride rotor, sealed at both ends with KeI-F caps, and spinning at 5 kHz in a magic-angle spinning (MAS) probe made by Doty Scientific (Columbia, SC, USA). Spectra were run at a frequency of 50.3 MHz using a Varian Inova-200 spectrometer. Each 90 ~ proton preparation pulse of Alkylammonium-montmorillonite-phenanthrene intercalate 223 5/as duration was followed by a 1 ms crosspolarization contact time, and 30 ms of data acquisition. The recovery delay was 1 s for both the TDTMA-Mt and TDTMA-Mt-phenanthrene complex, 2 s for TDTMA-bromide, and 600 s for phenanthrene. Signals were averaged over periods between 20 and 90 min. Chemical shifts were expressed relative to tetramethylsilane with the methyl signal of solid hexamethylbenzene at 17.4 ppm serving as a secondary reference. Proton spin-lattice relaxation time constants, Tj(H), were measured by applying a 10 tas 180 ~ pulse, and waiting for a variable relaxation interval before applying the cross-polarization sequence. Proton rotating-frame relaxation time constants, TI0(H), were measured by inserting a variable proton spin-locking interval between the 90 ~ proton preparation pulse and the cross-polarization contact time. Values of the cross-polarization transfer time constant, TCH, were measured by varying the contact time, and fitting 13C NMR peak heights to a double-exponential function (Alla & Lippmaa, 1976): Height(t) = K{exp(-t/Tlp(H)) exp(--t/TcH)}/{1/TcH -1/Tip(H) } (1) where K is an arbitrary scaling factor. R E S U L T S A N D D I S C U S S I O N Chemical and X-ray diffraction analyses Table 1 gives the C and N contents of pure TDTMA-bromide and of TDTMA-Mt, both before and after intercalation of phenanthrene. The XRD patterns of TDTMA-Mt and its intercalate with phenanthrene are shown in Fig. 1. The C and N contents of TDTMA-Mt agree to within 2% of the values calculated on the basis that TABLE 1. Carbon and nitrogen contents of tetradecyltrimethylammonium bromide (TDTMA-bromide), TDTMA-exchanged KpF montmorillonite (TDTMA-Mt), and TDTMA-Mt-phenanthrene intercalation complex. (a)