Interpretation of the infrared spectrum of the NH]-clays: application to the evaluation of the layer charge

Previous studies show that NH~ bearing 2:1 clays have two types of ammonium vibrational spectra. The first one, mainly observed for KBr pellets of NH4-smectites, gives absorption bands at 3150, 3020, 2840 and 1400 cm -~ (Petit et al., 1998). These NH~ bands, observed in the 1R spectra of many NH4 salts, correspond to vibrations of NH~ ions in the tetrahedral symmetry group (Nakamoto, 1963). According to Nakamoto (1963), Sherman & Smulovitch (1970) and Ryskin (1974), the intense absorption bands at 3150 cm 1 and 1400 cm -1 are due to u3 (stretching) and u4 (deformation) vibrations of the NH~ ion, respectively. The 2840 cm -a and the 3020 cm 1 bands are assigned to the overtone 2 ~2 and the (u4 + u2) combination vibration, respectively. The second type of ammonium vibrational spectrum shows bands in the 3300-3250 cm t region and near 1440 cm -1. Chourabi & Fripiat (t981) assigned these bands to the adsorbed NH~ ion in tetrahedral symmetry as well, but due to H bonding of NH~ with the clay structure, the bands are in slightly different positions compared to those observed for NH4 salts (i.e. at 3150 and 1400 cm-1). This second type of ammonium vibrational spectrum is observed in the case of: (1) self-supporting films of clays (Mortland et al., 1962; Mortland & Raman, 1968; Chourabi & Fripiat, 1981; Srasra et al., 1994) and (2) nonexpandable NH4-clay minerals such as illites or micas (Vedder, 1965; Sucha et al., 1998). Moreover, this second type of ammonium vibrational spectrum is also present as a superimposed component on the first type of spectrum and was observed in the spectra of undried or dried KBr pellets of some smectites (Petit et al., 1998). No detailed discussion of IR spectra of NH4-clay minerals is available in the literature, therefore the purpose of this paper is to explain the occurrence of these two types of IR ammonium vibrational 9 1999 The Mineralogical Society 544 S. Peti t et al. spectra for NH4-saturated smectites and to interpret this occurrence. The effect of layer charge on exchangeability of NH~ ions in KBr-clay mixtures is also discussed. M A T E R I A L S A N D M E T H O D S Three synthetic saponites and one smectite obtained from the Source Clay Repository of the Clay Minerals Society (SAz-1) were used (Table 1). The sample SAz-1 is a Mg-rich montmorillonite with the majority of layer charge in the octahedral sheet. The samples SAP0.35, SAP0.5, and SAP0.8 are synthetic saponites with layer charge arising from tetrahedral substitutions only. The range of their tetrahedral substitution varies from 0.35 to 0.8 atoms of A1 substituted for Si per half unit-cell. These were the samples used by Petit et al, (1998). Suspensions of the samples in 1 M ammonium acetate solution were obtained by agitation and were left to stand for 2 h. After centrifugation, this process was repeated once to achieve complete NH~ exchange. The samples were then washed until free of ammonium salts (until negative reaction with the Nessler reagent). The NH~-saponites were further re-saturated with Ca 2+ using solutions of 1 M CaCI2. This exchange permits identification of the irreversibly collapsed layers of high-charged smectites by XRD (Righi et al., 1998). In the present work it was used to distinguish between fixed and exchangeable NH~ ions by Fourier transform infrared (FTIR) spectroscopy. A Nicolet 510 FTIR spectrometer was used to record FTIR spectra at 4 cm 1 resolution in the 4000 400 cm -1 range. The spectrometer was continuously purged with dry air during scanning of the transmission spectra of KBr pellets. The pellets of 2 cm diameter were prepared by mixing 3 mg of sample with 300 mg of KBr. Diffuse reflectance infrared spectra (DRIFTS) with and without dispersion in a KBr matrix were recorded using a DRIFTS accessory 'Collector' from Spectra-Tech. The samples, analysed at room temperature, were placed loosely into a sample cup of -1 mm depth and 3 mm diameter. Diluting the sample in a non-absorbing matrix, the proportion of the IR beam that is diffusely reflected by the sample increases. However, the DRIFTS spectra of saponites and SAz-1 obtained without dilution were of good quality in the regions of NH4 absorption, and dilution was used to study whether KBr affects position of NH4 bands in the spectra. To compare samples quantitatively, the spectra were normalized using an internal reference band, either the main Si -O band near 1000 cm -1 or the OH-stretching band depending on which was more appropriate for the transmission or reflectance measurements. R E S U L T S A N D D I S C U S S I O N The study was focused on the ua NH~ band near 1400 cm l only, because the intensity of this NH~ band is linked quantitatively to the N content of the analysed samples (Ferriso & Hornig, 1959; Vedder, 1965; Petit et al., 1998). Unlike the NH~stretching region, where the bands are overlapped with OH-s t re tch ing v ibra t ions of water (near 3400 cm i), the NH~ deformation region is free of other absorptions, and no decomposition is necessary to extract the u 4 NH~ signal (Petit et al., 1998). The important point is that the u4 NH~ band characterizes unambiguously the type of NH~ vibrational spectrum. If the u4 NH4 band is at TABLE 1. Mineralogy, source, and cation exchange capacities (CEC) of the samples used. Samples SAz-I SAP0.35 SAP0.5 SAP0.8 Mineralogy Montmorillonite, minor silica phase Source Apache County, Arizona, USA CEC 125 (cmol(+).kg i) References Jaynes & Bigham (1987) Synthetic saponites 9 + S14_xAlxMg3010(OH)eNax CRSCM, Orl6ans, France O.-L. Robert)