Mineralogical interference on kaolinite crystallinity index measurements

This study examines the influence of minerals and amorphous phases associat ed with kaolin and kaolinitic rocks on kaolinite crystallinity indices (KCI ) derived from X-ray diffraction (XRD) data in order to select the best ind ex for systematic studies of commercial kaolins or geological sequences. Fo r this purpose, 8 kaolins of differing structural order were chosen and use d to prepare mixtures containing different weight fractions of quartz, feld spar, illite, smectite, chlorite, halloysite and iron hydroxide and silica gels. An additional 17 samples of kaolin were also studied to test the resu lts and evaluate the restrictions. KCIs used included Hinckley (HI), Range and Weiss (QF), Littard (R2), Stoch (IK), Hughes and Brown (H&B) and Amigo et al. (full width at half maximum, FWHM), and the "expert system" of Planc on and Zacharie. Based on more than 15,000 KCI determinations. the HI and QF are influenced by quartz, feldspar, iron hydroxide gels, illite, smectite and halloysite. IK can be used in the presence of quartz, feldspar and iron hydroxide and s ilica gels. Also, R2 is the only KCI that could be measured in the presence of halloysite; FWHM indices should not be used in the presence of chlorite and/or halloysite; and H&B should only be used with pure kaolinite samples . The "expert system" of Plancon and Zacharie is strongly affected by the p resence of other mineral phases, particularly with more than 25% of well-or dered kaolinite. Their system is less sensitive to other mineral phases whe n only disordered kaolinite is present, and it should not be used with kaol inite of medium order-disorder because the well-ordered phase is present in an inappreciable proportion (<10%). KCI is only measurable in kaolinitic r ocks if kaolinite is >20 wt% and the precision increases with an increase i n the quantity of kaolinite. In all cases, the reliability will depend on t he other minerals present. When a KCI can be measured accurately, the other s can be obtained by using the empirical relationships reported in this pap er.