Quantitative Analysis of Silica in Silica-Kaolin Mixtures by Photoacoustic and Diffuse Reflectance Spectroscopies RAGHOOTTAMA S. PANDURANGI* and MOHINDAR S. SEEHRAt Department of Physics, West Virginia University, Morgantown, West Virginia 26506-6315 The techniques of photoacoustic spectroscopy (PAS) and diffuse reflec- tance spectroscopy (DRS) are investigated for the quantification of silica using the IR bands at 1080, 797, 693, and 1875 cm -1. It is observed that, in both techniques, signal intensity increases linearly with mass initially. However, signal saturation sets in at much lower masses (>2 mg) in PAS than in DRS (>30 mg), showing that saturation is strongly tech- nique dependent, in addition to having dependence on the absorptivity of the sample. These findings are applied to the quantification of silica in laboratory-generated silica-kaolin mixtures. It is shown that, with the use of the weaker band at 1875 cm-' in DRS, satisfactory results for the quantification of silica are obtained. Index Headings: Analysis for silica; Infrared; Photoacoustic spectros- copy; Reflectance spectroscopy; Spectroscopic techniques. INTRODUCTION Silica and kaolin (a layer aluminosilicate) are among the major constituents of most coal mine dusts. 1Whereas kaolin is cytotoxic in vitro but nonfibrogenic in vivo in humans and experimental animals, 2~4 the cytotoxic and fibrogenic potential of silica is rather well established) Wallace and co-workers 6-s have studied the cytotoxie po- tentials of both native and surface-modified silica and kaolin and inferred that retoxification is slow in the case of occluded silica vis-a-vis native silica. Recent epide- miological studies indicate that silica concentration alone in a mine dust, as determined by transmission IR spec- troscopy (TIRS), does not correlate well with the pro- gression of coal workers' pneumoconiosis.9 This obser- vation has led to the suggestions6~ that silica particles occluded by clays such as kaolin may not be fibrogenic. This proposal requires accurate estimations of kaolin, silica, and occluded silica. The current method of silica analysis in a mine dust sample employs TIRS of the 797- cm -~ band of silica, although its accuracy, as well as that of x-ray diffraction studies, is no better than 25 %, even in a laboratory-generated dust. 9 In this paper we have employed the techniques of diffuse reflectance spectros- copy (DRS) and photoacoustic spectroscopy (PAS), in conjunction with an FT-IR spectrometer, for the quan- titative determination of silica in laboratory-generated silica-kaolin mixtures. One of the main advantages of DRS and PAS is that, in these techniques, problems arising from sample prep- aration of the KBr-based disks used in TIRS are avoided. There are, however, several other considerations (e.g., particle size effects and saturation phenomenon) which complicate the use of DRS and PAS for quantitative Received 30 September 1991; revision received 15 May 1992. * Present address: Center for Radiological Research, University of Mis- souri, Columbia, MO 65203. t Author to whom correspondence should be sent. work. 1°-25In this paper, we have carried out a systematic comparison of the quantitative abilities of the two tech- niques, as applied to the silica-kaolin mixtures, with par- ticular attention to the saturation behavior of several bands with different absorption coefficients. In partic- ular, we show that a weaker band for silica at 1875 cm -1 is suitable for quantitative purposes, and that quanti- tative accuracy is higher in DRS than in PAS. Details of this work are given below. EXPERIMENTAL Studies were carried out on two forms of silica: (1) Mini-U-Sil, crystalline silica with particles of an average size ~5 #m, and (2) Cab-O-Sil, consisting of amorphous silica particles ~0.05 #m in size. In a recent paper, 26 we have reported on the electron microscopy, particle-size distribution, and cytotoxicity studies of these particles. The kaolin used here was obtained from Georgia Kaolin Mills, Augusta, GA, and has particles of about 5 urn, with a surface area of ~ 13 m2/g.26 The silica-kaolin mixtures were prepared by dispersing the appropriate amounts in isopropyl alcohol, followed by drying. The Kbr used in these experiments was dried at 150°C. For PAS, the car- bon black used as reference was dried at 150°C. Porosity values and vibrational assignments for all the bands of silica were taken from earlier reports. 27,2s Spectra reported here were taken with a Mattson In- struments Cygnus 100 FT-IR spectrometer. For PAS, the Model 100 MTEC photoacoustic cell and, for DRS, the Model FXA-530A by Analect Instruments were used. The spectrometer is continuously flushed with dry air with the use of a Balston air filter system. For PAS, all spectra were taken at a mirror velocity of 0.08 cm/s. In all cases, data were collected for the range 400 to 4000 cm -1 with the use of 64 scans and a resolution of 4 cm -1. The software supplied with the Cygnus 100 allows us to determine relative intensity in PAS (carbon black as reference) and to convert relative reflectance spectra of DRS into the KM (Kubelka-Munk) units. RESULTS AND DISCUSSION To compare the spectra obtained with the three tech- niques (TIRS, DRS, and PAS), spectra obtained with 1 mg of sample are shown in Fig. 1A for Cab-O-Sil and in Fig. 1B for Min-U-Sil. For TIRS, 1 mg of the samples was used to prepare a KBr disk by the standard methods and, for DRS, 1 mg of the samples was mixed with KBr (about 300 mg) to fill the cup to a 3-mm depth. For PAS, 1 mg of the samples was used alone. Comparing the signals observed in the three tech- niques in Fig. 1, it is apparent that the broad and intense Volume 46, Number 11, 1992 0003-7028/92/4611-171952.00/0 APPLIED SPECTROSCOPY 1719 © 1992Society for Applied Spectroscopy