combination bands in the infrared spectroscopy …clays.org/journal/archive/volume...

Clays and Clay Minerals, Vol. 46, No. 4, 466 477, 1998.

COMBINATION BANDS IN THE INFRARED SPECTROSCOPY OF K A O L I N S - - A DRIFT SPECTROSCOPIC STUDY

RAY L. FROST AND URSULA JOHANSSONt

Centre for Instrumental and Developmental Chemistry, Queensland University of Technology, 2 George Street, GPO Box 2434, Brisbane Queensland 4001, Australia

Abstraet--Kaolini tes with varying degrees of defect structures have been studied by both mid-infrared (IR) and near-IR diffuse reflectance spectroscopy (DRIFT). Difference bands were observed in the 2650- to 2750-cm I region. This region coincides with the kaolinite-deuterated hydroxyl stretching region. Sum- mation bands were observed in the near-IR spectra in the 4500- to 4650-cm -~ and in the 7050- to 7250- cm t region. Each of the spectral regions of the summation and difference bands is both kaolin polytype and sample dependent. It is proposed that each of these sets of bands arises from the combination of the hydroxyl stretching frequencies and the hydroxyl deformation frequencies and, to a lesser extent, the silicon-oxygen symmetric stretching vibration of the siloxane layer. Additional difference bands of very low intensity were also observed at 2930 and 2856 cm i. Combination bands were observed in all kaolinites at 2137 and 2227 cm-L Each of the 3 major combination spectral regions was composed of 5 component bands corresponding to the 4 IR active and the 1 Raman active kaolinite hydroxyl stretching frequencies. Combination bands were also observed at --1932 and 1821 cm-L

Key Words--Combinat ion Bands, Difference Bands, DRIFT, Kaolins, Kaolinites, Mid-Infrared, Near- Infrared, Raman.

I N T R O D U C T I O N

There are n u m e r o u s techniques ava i lab le to the clay scient is t for ob ta in ing molecu la r informat ion . These t echn iques inc lude bo th R a m a n and IR spectroscopy. A m o n g the IR spect roscopic techniques , there are 1) the absorp t ion (KBr) technique , 2) IR emis s ion spec- t roscopy (IES), 3) the ref lectance techniques , inc lud ing dif fuse and specular ref lectance, 4) a t tenuated total re- f lec tance (ATR), 5) IR mic roscopy and 6) photoacous- tic Four ie r t r ans fo rm IR (FTIR) spec t roscopy (PAS). In particular, the se lec t ion of the t echn ique depends on the par t icular p r o b l e m at hand. For example IES is sui table for s tudy ing kaol in i te dehydroxy la t ion and the rmal changes in kaol in i tes at e leva ted t empera tu res (Frost and Vassal lo 1996). The press ing of a K B r pel- let w i th c lays may alter the spec t rum th rough absorp- t ion or exchange o f the K into the clay structure. ATR is useful for powde r s but is l imi ted by the s ignal- to- noise rat io de t e rmined by the energy throughput . PAS is a nondes t ruc t ive t echn ique useful for hand l ing clays but is l imi ted by the n u m b e r of samples that m ay be hand l ed in a g iven per iod o f t ime. Specula r ref lectance is no t readi ly appl icable to clay minera ls , as the tech- n ique no rma l ly appl ies to flat, shiny po l i shed surfaces, a l t hough specular IR ref lec tance s tudies have been used to s tudy the rmal ly t reated Li and Cs mon tmor i l - loni tes (Karakass ides et al. 1997). D R I F T is more ap- p l icable to powder s bu t is l imi ted to some extent by the Res t rah len effects where the part ic le size and the

t Permanent address: Department of Chemical and Metallur- gical Engineering, Division of Inorganic Chemistry, Lule~ University of Technology, SE-971 87 LulegL Sweden.

inc ident IR wave leng ths create in te r fe rence effects. Such effects are m i n i m i z e d by mix ing the clay wi th K B r at the 5% level. The advan tage of us ing the D R I F T techn ique is that it p rov ides a rapid t echn ique for a n a l y z i n g s a m p l e s w i t h o u t any i n t e r f e r e n c e th rough sample preparat ion. The D R I F T t echn ique avoids the poss ibi l i ty of ion exchange of the K ion and n o n - r a n d o m l y or iented samples . The t echn ique is par t icular ly useful where the Res t rah len effects are min imal . These effects appear toward the low-frequency region, no rma l ly be low 1000 c m ~. Thus, the D R I F T techn ique lends i tsel f to the s tudy of the hydroxyl s t re tching reg ion of kaol in i tes whe re such effects are not observed . Also, the D R I F T techn ique is mos t sui ted to the near - IR region where the hydroxyl s u m m a t i o n bands occur.

The v ibra t ional spec t roscopy o f kaol ins has at tract- ed cons iderab le interest due to its appl ica t ion to the kaol in structure. Hydroxy l bands are ve ry sens i t ive p robes for d i s t ingu ish ing be tween kaol in clay minera l s and de te rmin ing thei r s t ructure (Frost 1997). FT Ra- man spec t roscopy has already p r o v e n a ve ry power fu l t echn ique in the s tudy of kaol in minera l s (Frost et al. 1993; Frost 1995). The a s s ignmen t of the hydroxyl s t re tching bands r ema ins unde r cons tan t review, and di f fer ing v iewpoin t s conce rn ing the pos i t ion and ori- en ta t ion of the hydroxyl groups exist (Giese 1988; Hess and Saunders 1992). Kaol ini te , as wi th the o ther kaol in polytypes , conta ins 2 types of hydroxy l groups: 1) the outer hydroxy l groups, or so-cal led " i n n e r surface h y d r o x y l s " , des igna ted " O u O H " and 2) the inner hydroxy l groups, des igna ted " I n O H " . The O u O H groups are s i tuated in the outer upper, unsha red plane,

Copyright �9 1998, The Clay Minerals Society 466

Vol. 46, No. 4, 1998 Drift spectroscopy of kaolins 467

whereas the InOH groups are located in the plane shared with the apical oxygens of the tetrahedral sheet. In an ordered sample such as the Georgia kaolinite (KGa-1), the 4 distinct IR bands are observed as fol- lows: the 3 higher-frequency vibrations (v I, v2, v3) are assigned to the 3 outer (OuOH) hydroxyls and the v5 band at 3620 cm J to the inner hydroxyl (InOH) (Farmer 1974; Frost and van der Gaast 1997). The fourth band at 3685 cm -~ (v4) is Raman active and IR inactive. The commonly accepted view is that the v~ and v2 bands are the coupled antisymmetric and symmetric vibrations (Brindley et al. 1986; Michaelian 1986). The coupling of these bands means that over- tones and combination bands may also be observed. The assignment of the v3 band is open to question but the suggestion has been made that the band is due to symmetry reduction of the OuOH hydroxyl (Farmer and Russell 1964).

Kaolinites may be classified according to the ratio of the 2 0 u O H hydroxyls at 3685 and 3695 cm (Frost and van der Gaast 1997). The intensity of the 3650- and 3670-cm -~ bands was found to vary con- comitantly with the 3685- and 3695-cm -~ bands, re- spectively. It was hypothesized that the 3650-cm band is the out-of-phase counterpart of the in-phase vibration at 3685 cm-~ and the 3670-cm-~ band is the out-of-phase analogue of the in-phase vibration at 3695 cm -~. Linear relationships were found between the intensities of these in-phase and out-of-phase vibrations. Linear combinations of these vibrational modes mean that combination bands involving the Ra- man active modes will also be possible. A hypothesis based on theoretical studies suggested a model of hydrogen bonding to explain the reason for 2 0 u O H bands at 3695 (vl) and 3685 (v4) cm -I. This hypothesis was based on models for hydrogen bonding between the inner surface hydroxyls of one layer and the siloxane units of the next adjacent kaolinite layer. The 3685-cm -~ band is IR inactive and Raman active, al- though in some very highly ordered kaolinites a band component may be observed in this position. In the near-IR spectra, combination bands may also arise from this type of vibration. Other hydroxyl bands are observed at 915 and 935 cm -~. These bands are attrib- uted to the hydroxyl deformation modes of the inner hydroxyl and the inner surface hydroxyl modes, re- spectively (Frost 1998). Such bands may also couple with other bands to give overtone and combination bands in the near-IR region.

While IR spectroscopy has been extensively used for the study of clay minerals, many of the newer techniques have yet to be applied to the study of clay mineral structures (Lazarev 1972; Farmer 1974). For example, IES has been used to a limited extent to study dehydroxylation of kaolin polymorphs (Frost and Vas- sallo 1996). IR spectroscopy has been used extensively for the characterization of kaolinites, particularly with

respect to the determination of order, crystallinity and energy of cohesion between kaolinite sheets (Cases et al. 1992; Cruz-Cumplido et al. 1992; Yvon et al. 1992). Near-IR spectroscopy has also been used for a considerable length of time and, more recently, for the study of the kaolinite clay minerals (Matsunaga and Uwasawa 1993; Tanabe et al. 1994). Indeed, remote sensing has identified minerals using near-IR spectroscopy (Hunt and Ashley 1979; Hunt and Hall 1981). Near-FTIR has been used for the studies of structural iron in kaolinites (Delineau et al. 1994), the quantita- tive analysis of kaolin mixtures (Crowley and Vergo 1988), the presence of kaolinite in altered rocks (Hunt and Hall 1981) and the study of the deuteration of kaolinite (Michaelian et al. 1987; Johansson et al. 1998). Two IR absorption band regions were observed at 4465 and 7025 cm-l. Explanation of the origin of these bands was limited (Hunt et al. 1973) and direct proof of the origin of the bands was not given. Dif- ference bands have not been reported. The objective of this research is to report the FTIR hydroxyl summation and difference bands of kaolinite and to deter- mine the origin of these bands. Furthermore, this work reports the overlap of the kaolinite difference bands with the deuterol stretching bands of kaolinite.

EXPERIMENTAL

Clay Mineral Samples

Clay minerals were obtained from The Clay Min- erals Society Repository and from the Australian Commonwealth Scientific and Industrial Research Or- ganisation, Glen Osmond, South Australia. Minerals were collected from a number of Australian mineral deposits. Samples were analyzed for phase purity using X-ray diffraction (XRD) techniques before DRIFT spectroscopic analysis. Table 1 shows the origin and source of the kaolinites and other kaolin polytypes.

XRD of powder patterns of the kaolinites was obtained using a Philips PW 1050 X-ray diffractometer using CoKoL radiation operating at 35 kV and 40 mA. A 1 ~ divergence and scatter slit were combined with a normal focus Co tube at 6 ~ takeoff and a 0.2-mm re- ceiving slit. The samples were measured in stepscan mode from 2.0 ~ with steps of 0.05 ~ up to 40 ~ and a counting time of 2 s. XRD was used to deter- mine the Hinckley Index of the kaolinites, which was used as a measure of the defect structures of the kaolinite (Hinckley 1963). The Hinckley Index of kaolinite crystallinity is determined by measuring the ratio of the intensities of the 110 and 111 peaks above a line drawn from the trough between the 020 to 110 peaks to the background beyond the 110 peak. These Hinckley indices are shown in Table 1. Kaolinites were analyzed for chemical composition using induc- tively coupled plasma-atomic emission spectroscopy (ICP-AES) techniques. Deuteration of the Amazonian

468 Frost and Johansson Clays and Clay Minerals

Table 1. Table of kaolinites, their origin and the source of supply.

Hinck- Kaolinite ley

polymorph Index Locality Origin

Kaolinite 1.35 Mesa Alta, Mexico Wards Natural Sci- #9 ence establish-

ment Kaolinite 0.42 Amazon Brazil Kaolinite 1.38 Lal-lal, Victoria, M. Stielow Enter-

Australia prises Pty Ltd. Kaolinite 0.25 Williamstown, Commercial Miner-

South Australia als Pty Ltd. Kaolinite 1.4 Birdwood, South Commercial Miner-

Australia als Pty Ltd. Kaolinite 0.12 Weipa 1 Comalco Pty Ltd. Kaolinite 0.20 Weipa 2 Comalco Pty Ltd. Kaolinite 0.5 Pinong English China

Clays Pty Ltd. Kaolinite 0.65 Eyres Peninsula, Commercial Miner-

South Australia als Pty Ltd. Kaolinite 0.1 Mt. Hope (South Commercial Miner-

Australia) als Pty Ltd. Halloysite - - New Zealand New Zealand Chi-

na Clays Pty Ltd.

Halloysite - - Eureka, Utah Clay Mineral Soci- ety Standard

Dickite - - San Juanito, Mexico Wards Natural Sci- ence establish- ment

kaol in i te was unde r t aken by mix ing the dr ied kaol in i te in deu te r ium oxide in a sealed vesse l at 60 ~ for 5 weeks. The ex ten t of deutera t ion was de t e rmined by the decrease in in tensi ty of the hydroxy l bands .

D R I F T Spec t roscopy

Drif t spect ra were ob ta ined us ing a Pe rk in -E lmer F T I R spec t rome te r (2000) equ ipped wi th a t r iglycl ine sulfate (TGS) detector. Spec t ra were recorded by ac- cumula t i ng 128 scans at 8 c m ~ reso lu t ion us ing a mir- ror ve loc i ty of 0 .2 cm/s for the near - IR and 1024 scans at 4 c m -1 reso lu t ion and a mi r ro r veloci ty of 0.3 cm/ s in the mid-IR. A p p r o x i m a t e l y 5 % kaol in was dis- pe rsed in oven-d r i ed spec t roscopic grade K B r wi th a ref ract ive index of 1.559 and a par t ic le size of 5 - 2 0 Ixm. B a c k g r o u n d K B r spectra were ob ta ined and spectra ra t ioed to the background . A series of kaol ini tes , ha l loys i tes and dicki tes of d i f ferent c rys ta l l in i ty and par t ic le size were analyzed. Spec t ra were recorded in bo th the mid - IR and near - IR regions . The spectra were t r ans fo rmed us ing the K u b e l k a - M u n k a lgor i thm to p rov ide spect ra for compar i son wi th absorp t ion spectra. The spectra l man ipu la t ions of base l ine adjus tment , smoo th ing and no rma l i za t ion were pe r fo rmed us ing the Spect raca lc sof tware package G R A M S (Galact ic Indust r ies Corpora t ion , N e w Hampshi re ) . Band com- ponen t analys is was carr ied out us ing the peakfi t soft- ware by Jande l Scientific. Loren tz -Gauss cross-prod- uct func t ions were used th roughou t and peak fitting

carr ied out unt i l squared corre la t ion coeff ic ients wi th r 2 greater than 0.995 were obta ined.

R E S U L T S

The IR spectra of the kaol in i te p o l y m o r p h s are com- pl icated by the p resence of c o m b i n a t i o n bands , w h i c h are obv ious in the near- IR spectra bu t less obv ious in the mid-IR. Three ma jo r sets o f bands were o b s e r v e d in regions a round - 2 7 0 0 , 4600 and 7050 cm i. F igure 1 i l lustrates the 2700 -cm 1 reg ion (d i f ference reg ion 1), F igure 2 the 4500 -cm J reg ion ( s u m m a t i o n reg ion 1) and Figure 3 the 7050 -cm -~ reg ion ( s u m m a t i o n re- g ion 2) for the kaol in polytypes. The di f ferences in the spectra may be corre la ted to the d i f ferences in the hydroxyl s t re tching region of the kaol in poly types (Hunt et al. 1973; L indbe rg and Snyder 1972). Each kaol in has its own character is t ic near - IR spect rum. The near- IR spec t rum of wate r was o b s e r v e d in this reg ion at 5263 c m -~ (1.9 Ixm). The near- IR spec t rum of ad- sorbed wate r is absent in the spectra repor ted in this research. Hun t et al. (1973) repor ted that the kaol in po ly types had a s ingle p r o m i n e n t b a n d at 4545 c m 1 (2.2 ~m). This band was character is t ic of 2- layer di- oc tahedra l minerals . Bands were also o b s e r v e d at 6756 cm -~ (1.48 Ixm) and 6993 cm -~ (1.43 t~m). Prev ious work has not repor ted d i f fe rence bands in the mid - IR spectra of kaol in i tes (Hunt et al. 1973; Br ind ley et al. 1986). F igure 1 i l lustrates the D R I F T spectra of the hydroxyl d i f fe rence bands observed . It may be clearly o b s e r v e d that d i f ference bands are c o m m o n to all of the kaol in i tes s tudied in this work.

In the D R I F T spectra repor ted in this work, more clear ly reso lved bands were observed . For the first s u m m a t i o n reg ion cen te red on 4545 c m -~ (2.2 Ixm) (Figure 2), the 2 hal loysi tes have sl ightly d i f ferent spectra wi th d i f ferences in the 4625-cm-~ region. This d i f ference may be a t t r ibuted to the amoun t of wate r in the hal loysi te structure. P rev ious studies repor ted b road un reso lved spectra for hal loys i tes tha t were de- penden t on the wate r con ten t of the ha l loys i te (Hunt et al. 1973). The kaolini te , in this example a h ighly ordered kaol in i te f rom South Austra l ia , shows a p rom- inent band at 4527 c m -~ toge ther wi th bands at 4558, 4610 and 4632 c m -t. The dicki te spec t rum shows a ve ry di f ferent spec t rum wi th a p r o n o u n c e d band at 4590 c m ~. For the second s u m m a t i o n reg ion cen te red at 6995 c m ~ (1.43 Ixm), the 2 hal loys i tes show weak but s imi lar spectra (Figure 3). The spectral pa t tern is qui te complex wi th the mos t p r o m i n e n t b a n d o b s e r v e d at 7067 c m ~ (1.415 Ixm). Kaol in i tes show bet ter-reso lved bands in this reg ion wi th 2 features at 7067 and 7168 c m -1. Dicki tes show bet te r reso lu t ion wi th 2 p r o m i n e n t bands at 7067 and 7225 cm -1. Prev ious work showed 2 wel l - reso lved bands for d icki te at 1.43 (6993 cm -~) and 1.48 p~m (6756 c m -1) wh ich are at f requencies s o m e w h a t less than o b s e r v e d in this work (Hunt and Ash ley 1979; Hun t et al. 1973). The spec-


r

1400 1600 1800 2000 2200 2400 2600 2800 3000

W a v e n u m b e r / c m "1

Figure 1. DRIFT spectra of the hydroxyl difference bands in the 1400- to 3000-cm -t region of kaolinites from (a) Birdwood (South Australia), (b) Mesa Alta (Mexico), (c) Lal-lal (Victoria, Australia), (d) Eyre Peninsula (South Australia) and (e) Amazon (Brazil).

(e)

4400 4450 4500 4550 4600 4650 4700

W a v e n u m b e r / c m "l

Figure 2. DRIFT spectra of the hydroxyl summation bands in the 4400- to 4700-cm -~ region of (a) halloysite (Eureka, Utah), (b) halloysite (NZ), (c) kaolinite (Birdwood, Australia) and (d) dickite (San Juanito, Mexico).


(d)

(b)

6950 7000 7050 7100 7150 7200 7250 7300

Wavenumber /cm "l

Figure 3. DRIFT spectra of the hydroxyl summation bands in the 6950- to 7900-cm ~ region of (a) halloysite (USA), (b) halloysite (NZ), (c) kaolinite (Birdwood, Australia) and (d) dickite (San Juanito, Mexico).

t rum of s u m m a t i o n reg ion 1 for the kaol in i te shows a s t rong r e s e m b l a n c e to that o f its hydroxy l s t re tch ing region. The dicki te spec t rum is ve ry different , showing cons iderab le complexi ty . Two wel l - reso lved bands are obse rved .

F igures 4, 5 and 6 show the spectra for a series of kaol in i tes wi th d i f ferent defect structures. In t e rms of re la t ive intensi t ies , the bands in the 4500- to 4650- c m -1 reg ion are approx imate ly 1 ten th of the in tensi ty o f the hydroxy l s t re tch ing bands be tween 3620 and 3695 c m -1 (Figure 7 and Table 2). T he in tensi t ies of the set o f bands be tw een 7050 and 7200 cm-~ are ap- p rox imate ly 0.75 of the in tens i ty of the 4500- to 4650- cm-~ bands . The spectra for kaol in i tes show some res- o lu t ion of c o m p o n e n t bands wi th the kaol in i tes char- ac te r ized by the band at 4525 c m - L Each set of bands shows a cons tan t pa t te rn for each of the spectral reg ions and each spectra l r eg ion m ay be d e c o m p o s e d in to 4 m a j o r c o m p o n e n t bands (Figures 8 and 9). Table 2 repor ts the f requenc ies of these c o m p o n e n t bands . In the 4500- to 4 7 0 0 - c m -~ region, 5 bands were ob- s e rved wi th approx imate ly equal intensi t ies , as deter- m i n e d by b a n d areas at 4527, 4558, 4610 and 4632 c m -~ wi th a less in tense b a n d at 4649 c m -~. The 4527- c m -~ b a n d is the mos t p r o m i n e n t in this spectral region. In the 7050- to 7 2 0 0 - c m ~ region, 4 in tense bands were o b s e r v e d at 7067, 7118, 7168 and 7182 c m -~ wi th a smal le r c o m p o n e n t at 7156 c m -~.

D I S C U S S I O N

The ques t ion arises as to the r eason for the o b s e r v e d c o m b i n a t i o n bands . For kaol ini te , there are several possibi l i t ies . It was first cons idered that the bands in the 4500- to 4 7 0 0 - c m -1 reg ion arise f rom addi t ion of the hydroxy l s t re tching f requencies and the s i l icon oxygen s t re tching f requency. One poss ibi l i ty is that the 4 5 2 7 - c m * band is the addi t ion of the 3621 + 1021 c m -1 bands . Such a c o m b i n a t i o n ar ises f rom the ad- d i t ion of the inner surface hydroxy l s t re tching frequency and the Si-O s t re tching f requency of the si loxane layer. The o b s e r v e d b a n d at 4527 cm -* is some 120 c m ~ less than the expec ted f requency of 4642 c m -t. This is then an unl ike ly c o m b i n a t i o n of 2 v ibra- t ional f requencies even though the hydroxy l symmet - ric s t re tch and the s i loxane Si-O s t re tching f requency are coupled th rough hyd rogen bonding . It has been sugges ted that the 4 5 2 7 - c m i reg ion can be a t t r ibuted to the c o m b i n a t i o n of the hydroxy l s t re tch ing f requency and the low-f requency v ibra t iona l modes (Hunt et al. 1973). A second possibi l i ty is the combina t i on o f the hydroxy l s t re tching f requencies and the hydroxy l de fo rmat ion modes . The sharp b a n d at 4527 c m -~ arises f rom the addi t ion o f the hydroxy l s t re tch ing 3621- c m -* b a n d and the hydroxy l de fo rma t ion band at 915 c m -* (Frost 1998). This wou ld give an addi t ion of 4536 c m -~, wh ich is c loser to the o b s e r v e d f requency of 4527 c m ~. Thus, it wou ld appear that the hydroxy l


(e)

( b )

4400 4450 4500 4550 4600 4650 4700 4750 4800

W a v e n u m b e r / e m "1

Figure 4. DRIFT spectra of the hydroxyl summation bands in the 4400- to 4800-cm a region of kaolinites from (a) Mt. Hope (South Australia), (b) Lal-lal (Victoria, Australia), (c) Pittong (Victoria, South Australia), (d) Weipa (Queensland, Australia) and (e) Weipa (Queensland, Australia).

s t re tch ing f requency is c o m b i n i n g wi th the hydroxy l de fo rma t ion f r equency to give r ise to the com bi na t i on b a n d at 4527 c m -]. Similar ly , it is p roposed that the c o m p o n e n t bands at 4558 cm -~ arise f rom the com- b ina t ion o f the hydroxy l s t re tch ing f requency at 3652 c m -~ a t t r ibuted to the inner surface hydroxy l bands and the hydroxy l de fo rma t ion v ibra t ion at 915 c m - L The 2 bands at 4610 and 4632 c m -~ arise f rom the c o m b i n a t i o n o f the 3670- and 3 6 9 7 - c m -~ wi th the hydroxyl de fo rma t ion v ib ra t ion of the inne r surface hydroxyl groups. The com bi na t i on bands therefore orig- inate f rom the comb i na t i on of bands f rom vibra t iona l modes of the s ame group; that is, the com bi na t i on o f the hydroxy l s t re tching f r equency and the hydroxy l de fo rma t ion f requency ar is ing f rom the same hydroxy l groups. Thus, specific v ibra t iona l modes ar is ing f rom the mix ing of modes f rom the same hydroxy l g roup c o m b i n e to g ive the spectral profi le in the 4500- to 4 6 5 0 - c m -j region. A ve ry w e a k band c o m m o n to all kaol in i tes s tudied was o b s e r v e d at - -4195 c m -~ (2.38 v~m).

The bands in the 7050- to 7 2 0 0 - c m -~ reg ion in all p robabi l i ty arise f rom the doub l ing of the hydroxy l s t re tch ing f requencies . I f we take the inne r hydroxy l s t re tch ing f r equency at 3621 c m -] and double this number , the va lue o f 7242 c m -~ is obta ined, which is of the order of 175 c m -~ grea te r than the o b s e r v e d

f requency of 7067 c m -x. The IR symmet r i c s t re tch ing v ib ra t ion of the inne r hydroxy l at 3621 c m -~ has a b a n d w i d t h of 18 cm -~. The b a n d w i d t h o f the 7067- c m -~ band is 21.6 c m -~. These b a n d w i d t h s assis t in the conf i rmat ion that the 7 0 6 7 - c m -~ b a n d is the du- p l ica t ion in f r equency o f the 3 6 2 1 - c m -1 band. I f the f r equency o f 3697 c m -~ is doubled , a va lue of 7394 c m -1 is obta ined, w h i c h is 212 c m -~ grea ter than the o b s e r v e d f requency o f 7182 c m -t. I f the va lue o f 3670 c m -~ is doubled , then a va lue of 7340 c m -~ is obtained, wh ich is 172 c m ] greater than the o b s e r v e d f requency of 7168 c m -~. It is cons ide red that the d o u ' b l ing o f the 3 6 5 2 - c m -~ b a n d does not occur. The 3652- cm -~ band has been desc r ibed as an ou t -o f -phase vi- b ra t ion and it is unl ike ly that ou t -of -phase v ibra t iona l modes would c o m b i n e to g ive over tones (Fros t and van der Gaas t 1997). I f i t does, then the in tens i ty of the b a n d is too weak to observe . One way o f descr ib- ing the IR bands at 3652, 3670 and 3697 c m -~ is to a t t r ibute the 3697 -cm -~ b a n d as an in -phase v ib ra t ion and the bands at 3652 and 3670 cm -~ as ou t -o f -phase v ib ra t ions (Frost and Vassal lo 1996; Fros t and Van der Gaas t 1997). I f this is the correct descr ip t ion o f these lat ter 2 bands , then it is unl ike ly tha t ou t -of -phase vi- b ra t ions would couple to g ive bands in the 7050- to 7200-cm-~ region. A n o t h e r poss ib i l i ty is tha t the near- IR b a n d at 7050 c m -~ is due to the c o m b i n a t i o n of a


e~

(c)

(b)

(a)

I . . . . . . . . . . . . - A

6950 7000 7050 7100 7150 7200 7250 7300 7350 7400

W a v e n u m b e r / c m 1

Figure 5. DRIFT spectra of the hydroxyl summation bands in the 6950- to 7400-cm t region of (a) Mt. Hope (South Australia), (b) Lal-lal (Victoria, Australia), (c) Pittong (Australia), (d) Weipa (Queensland, Australia) and (e) Weipa (Queens- land, Australia).

R a m a n act ive m o d e and an IR act ive mode. For example , the bands in the 7050-cm-~ region have differ- ences of 7156 - 3652, 7168 - 3670 and 7182 - 3697 c m -1, w h i c h are 3504, 3498 and 3485 cm -I, respec- tively. The cons t ancy o f these d i f ferences cou ld then resul t f rom the c o m b i n a t i o n of a s ingle R a m a n m o d e and the var ious IR modes .

The bands o b s e r v e d in the 2650- to 2730 - cm 1 re- g ion (Table 2) are a t t r ibuted to d i f fe rence bands based on the subt rac t ion o f specific f requenc ies f rom the hydroxyl s t re tch ing f requencies . These bands are cen- tered a round 2690 c m -~ wi th 2 ma jo r peaks at 2699 and 2709 cm-~ and 2 peaks of less in tensi ty at 2671 and 2655 c m -~. This spectra l reg ion is complex and b a n d fitting does not fit this reg ion wel l (Figure 10). Never the less , 2 ma jo r bands were o b s e r v e d at 2698.5 and 2709.5 cm -~ wi th about equal bandwid ths of 15.5 c m -~. It is poss ib le tha t the 2709 .5 -cm -1 b a n d arises f rom the coup l ing o f the 3620- and 915-cm -1 f requencies. The d i f fe rence in these 2 f requencies is 2705 c m -~, wh ich co r responds wel l wi th the obse rved frequency of 2709.5 cm t. The 3620 - cm -~ R a m a n b a n d resolves into 2 bands at 3617 and 3628 c m 1 at l iquid n i t rogen tempera tures , so consequen t ly the 3620 - cm b a n d cons is t s of 2 ove r l app ing co inc iden t bands tha t only b e c o m e reso lved at low temperatures . The cou- p l ing wi th the 915-cm -1 band m ay also cause the res-

o lu t ion o f these 2 bands in the m i d - l R d i f fe rence spectra. The b a n d at 2671 c m -~ is of m e d i u m intens i ty for this set of bands . One poss ibi l i ty is that the b a n d arises f rom the d i f fe rence be tween the 3697- and 1020-cm -~ bands . The complex i ty of this spectral r eg ion may be a t t r ibuted to the subt rac t ion o f bo th the hydroxy l de- fo rmat ion f requencies and the s i loxane Si -O f requencies f rom the hydroxy l s t re tching frequencies . In a s imple analysis , there are 5 hydroxy l s t re tching fre- quenc ies (4 in the F F I R and 5 in the R a m a n ) that could combine wi th 2 hydroxyl de fo rmat ion f requencies and several s i loxane Si-O s t re tching f requencies . This wou ld resul t in some 20 poss ib le combina t ions , wh ich makes the d i f fe rence spectra complex .

W h e n kaol in i te is deuterated, the deuterol s t re tch ing f requencies occur in the 2660- to 2 7 4 0 - c m -~ reg ion (Table 2). For the kaol in i te f rom A m a z o n , 5 deuterol bands were obse rved at 2672, 2680, 2698, 2721 and 2732 c m t for the part ial ly deutera ted kaol in i te (Figure 11). It is obv ious that the deuterol spectral reg ion of the deutera ted kaol in i te over laps wi th the d i f ference bands of kaolinite. The 2705 -cm -t b a n d makes up 50% of the total band intensi ty of the d i f ference b a n d profi le of the un t rea ted kaolini te . This b a n d profi le coincides exact ly wi th the deuterol b a n d profile. This makes the analysis of the spectral reg ion o f the deuterated kaol in i te difficult, as bo th sets of bands occur


(b)

i r 4 i i

2650 2660 2670 2680 2690 2700 2710 2720 2730 2740 2750


Figure 6. DRIFT spectra of the difference bands of kaolinite in the 2650- to 2750-cm -~ region of kaolinites from (a) Birdwood (South Australia), (b) Mesa Alta (Mexico), (c) Lal-lal (Victoria, Australia), (d) Mount Hope (South Australia), (e) Eyre Peninsula (South Australia) and (f) Amazon (Brazil).

1.2

0.8

0.6

0.4

0.2

(d)

(c) . . == .= .

(b)

0 I p

3300 3400 3500 3600 3700 3800


Figure 7. DRIFT spectra of the hydroxyl stretching of kaolinites in the 3300- to 3700-cm J region from (a) Birdwood (South Australia), (b) Mesa Alta (Mexico), (c) Lal-lal (Victoria, Australia), (d) Eyre Peninsula (South Australia) and (e) Amazon (Brazil).


Table 2. Band component analyses of the DRIFT spectra of Amazonian kaolinite.

Deuterol stretching

region Hydroxyl 2700-cm ~ stretching 4500-cm ~ 7050-cm ~ 2700-cm

region region region region region

2672 3621.5 4527 7067 2655 - - 3637 4558 7118 2671

2698 3652 4610 7156 2698.5 2721 3670 4632 7168 2703 2732 3697 4649 7182 2709.5

at s imi la r f requencies , par t icular ly for par t ia l ly deuterated kaol ini tes . There is, however , a cons iderab le dif- fe rence in intensi t ies . The deuterol bands are about 3 t imes as in tense as the d i f fe rence bands . Deutero l dif- fe rence bands were not obse rved . It was cons idered that these bands were o f ex t remely low intensity. Ad- d i t ional kaol in i te d i f fe rence bands o f ve ry low intensity were also o b s e r v e d at 2930 and 2856 c m ~ (Figure 1). B a n d s that invo lve the hydroxy l group have been ident i f ied in R a m a n spec t roscopy at 141 and 230 cm -~ (Frost et al. 1993; Frost 1995). Fur ther bands have been o b s e r v e d at 670, 707, 754 and 789 cm t and were a t t r ibuted to hydroxy l t rans la t ion v ibra t ions invo lv ing the f lexing o f the kaol in i te lat t ice (Frost et al. 1993; Frost 1995). The d i f fe rence be t w een the 3620- and 6 7 0 - c m -~ bands is 2950 c m -~, wh ich is c lose to

the obse rved b a n d at 2930 c m 1. A n a l ternat ive is the d i f fe rence be tween the 3695 -cm -1 band and the 754- cm 1 band of 2944 c m ~. Thus, it is p robab le tha t the bands obse rved at 2930 and 2856 cm -1 are the difference be tween the hydroxy l s t re tching v ibra t iona l frequency and the hydroxy l t rans la t ion f requency. Com- b ina t ion bands were also o b s e r v e d in all kaol in i tes at 2137 and 2227 c m -t (Figure 1). It is not k n o w n wh ich combina t i on of bands gives r ise to these f requencies , but it is p robab le that the f requencies are over tones of low-f requency v ibra t ions such as the doub l ing of the Si-O s t re tching f requency at 1020 c m -~. Deu te ra t ion of the kaol in i te resul ted in the r emova l o f these com- b ina t ion bands . Bands were also o b s e r v e d at 1932 and 1821 c m -~. These bands are a t t r ibuted to the combi - na t ion of the Si -O s t re tching (1020 c m - t ) and hydroxy l de fo rmat ion f requencies (915 c m 1) and the doub l ing of the hydroxy l de fo rma t ion f requency at 915 c m -t , respect ively.

C O N C L U S I O N S

Kaol ini tes have been s tudied us ing near- IR and mid- IR spectroscopy. Bo th s u m m a t i o n and d i f fe rence bands were obse rved , and a t tempts at ra t iona l iz ing the or ig in of these combina t i on bands were made. Sum- mat ion bands were obse rved in the near - IR spectra in the reg ion f rom 4500 to 4650 cm-~ and in the 7050- to 7 2 5 0 - c m 1 region. Di f fe rence bands were o b s e r v e d

r

4450 4500 4550 4600 4650 4700

W a v e n u m b e r / c m -1

Figure 8. Band component analysis of the hydroxyl bands of the DRIFT spectra of Amazonian kaolinite in the 4450- to 4700-cm -~ region.


~D

7000 7050 7100 7150 7200 7250


Figure 9. Band component analysis of the hydroxyl bands of the DRIPT spectra of Amazonian kaolinite in the 7000- to 7250-cm 1 region.

2650 2660 2670 2680 2690 2700 2710 2720 2730 2740

W a v e n u m b e r / e m "!

Figure 10. Band component analysis of the hydroxyl bands of the DRIFT spectra of Amazonian kaolinite in the 2650- to 2740-cm -1 region.


to

r

r

2600 2620 2640 2660 2680 2700 2720 2740 2760 2780


Figure 11. Band component analysis of the deuterol bands of the DRIFT spectra of Amazonian kaolinite in the 2600- to 2780-cm- ~ region.

in the 2650- to 2750-cm -~ region. This latter region coincides with the kaolinite deuterat ion-stretching region. Each of the spectral regions of the summat ion and di f ference bands is both po lymorph and sample dependent . It is p roposed that each of these bands arises f rom the combina t ion of the hydroxyl stretching f requencies and the other hydroxyl vibrations such as the deformat ion modes and the s i l icon-oxygen symmetric s tretching vibration o f the si loxane layer. Ad- ditional d i f ference bands of very low intensity were also obse rved at 2930 and 2856 c m - L Combina t ion bands were also obse rved in all kaolinites at 2137 and 2227 cm -~ and also at 1932 and 1821 c m - L

Importantly, care must be exerc ised in the study of kaolin spectra when low-intensi ty bands are found as occurs wi th deuterat ion and intercalation. Furthermore, whi le near-IR is used for the identification of the presence o f kaolins, care should be taken i f further inter- pretat ion is required. Certainly DRIFT spect roscopy in both the mid - lR and near-IR lends i tself to the study of kaolins.

A C K N O W L E D G M E N T S

The financial support of the Queensland University of Technology Centre for Instrumental and Developmental Chemistry is gratefully acknowledged. L. Barnes, chief ge- ologist of Commercial Minerals, is also thanked for supply of several Australian kaolinites. The authors wish to thank J.

M. Cases and T. Kloprogge for useful and constructive com- ments.

R E F E R E N C E S

Brindley GW, Chih-Chun K, Harrison JL, Lipsiscas M, Ray- thatha R. 1986. Relation between the structural disorder and other characteristics of kaolinites and dickites. Clays Clay Miner 34:233-249.

Cases JM, Lietard O, Yvon J, Delon JE 1992. I~,tude des propriEtEs cristallochimiques, morphologiques, superficiel- les de kaolinites d6sordonn6es. Bull Mineral 105:439-455.

Crowley JK, Vergo N. 1988. Near-infrared reflectance spectra of mixtures of kaolin-group minerals: Use in clay mineral studies. Clays Clay Miner 36:310-16.

Cruz-Cumplido M, Sow C, Fripiat JJ. 1992. Spectre infra- rouge des hydroxyles, cristallinit6 et 6nergie de coh6sion des kaolins. Bull Mineral 105:493-498.

Delineau T, Allard T, Muller J, Barres V. 1994. FTIR reflectance vs. EPR studies of structural iron in kaolinites. Clays Clay Miner 42:308-320.

Farmer VC. 1974. The layer silicates. In: Farmer VC, editor. The Infrared spectra of minerals. London: Mineral. Soc p 331-363.

Farmer VC, Russell JD. 1964. The infrared spectra of layered silicates. Spectrochim Acta 20:1149-1173.

Frost RL, Vassallo AM. 1996. The dehydroxylation of the kaolinite clay minerals using infrared emission spectroscopy. Clays Clay Miner 44:635-651.

Frost RL, Fredericks PM, Bartlett JR. 1993. Fourier transform Raman spectroscopy of kandite clays. Spectrochim Acta 20:667-674.


Frost RL. 1995. Fourier transform Rarnan spectroscopy of kaolinite, dickite and halloysite. Clays Clay Miner 43:191- 195.

Frost RL. 1997. The Structure of the kaolin clay minerals- - An FT Raman study. Clay Miner 32:73-85.

Frost RL, van der Gaast SJ. 1997. Kaolinite hydroxyls--A Raman microscopy study. Clay Miner 32:293-306.

Frost RL. 1998. Hydroxyl deformation in kaolinites. Clays Clay Miner 46:280-289.

Giese RE 1988. Kaolin minerals: Structures and stabilities. Rev Mineral 19. In: Bailey SW, editor. Hydrous phyllosil- icates. Mineral Soc Am. Chelsea, MI: BookCrafters. p 29 - 66.

Hess CA, Saunders VR. 1992. Periodic ab initio Hartree-Fock calculations of the low symmetry mineral kaolinite. J Phys Chem 96:4367-4374.

Hinckley DN. 1963. Variability in "crystallinity" values among the kaolin deposits of the coastal plain of Georgia and South Carolina. Clays Clay Miner 11:229-235.

Hunt GR, Hall RB. 1981. Identification of kaolins and asso- ciated minerals in altered volcanic rocks by infrared spectroscopy. Clays Clay Miner 29:76-78.

Hunt GR, Ashley RR 1979. Spectra of altered rocks in the visible and near infrared. Econ Geol 74:1613-1629.

Hunt GR, Salisbury JW, Lenhoff CJ. 1973. Visible and near- IR spectra of minerals and rocks. VI. Additional silicates. Modern Geol 4:85-106.

Johansson U, Holmgren A, Forsling W, Frost RL. 1998. Iso- topic exchange of kaolinite hydroxyl protons--A diffuse reflectance infrared Fourier transform spectroscopic study. Analyst 123:641-645.

Karakassides K, Petridis D, Gournis D. 1997. Infrared reflectance studies of thermally treated Li- and Cs- montmoril- lonites. Clays Clay Miner 45:649-658.

Lazarev AN. 1972. Vibrational spectra and structure of silicates. New York: Plenum Pr. p 178-182.

Lindberg JD, Snyder DG. 1972. Diffuse reflectance spectra of several clay minerals. Am Mineral 57:485-493.

Matsunaga T, Uwasawa M. 1993. Near infrared diffuse reflectance spectra of clay minerals. Nippon Dojo Hiryogaku Zasshi 64:329-331.

Michaelian KH, Bukka K, Permann DNS. 1987. Photoacous- tic infrared spectra (250-10,000 cm -~) of partially deuterated kaolinite #9. Can J Chem 65:1420-1423.

Michaelian KH. 1986. The Raman spectrum of kaolinite #9 at 21 ~ Can J Chem 64:285-289.

Tanabe K, Uesaka H, Inoue T, Takahashi H, Tanaka S. 1994. Identification of mineral components from near infrared spectra by a neural network. Bunseki Kagaku 43:765-769.

Yvon J, Lietard O, Cases JM, Delon JE 1992. Mintralogie des argiles kaoliniques des Charentes. Bull Mineral 105: 417-581.

(Received 18 August 1997; accepted 5 January 1998; Ms. 97-076)

combination bands in the infrared spectroscopy …clays.org/journal/archive/volume...

Documents