American Mineralogist. Volume 81. pages 213-225. 1996

The origin and diagenesis of grain-coating serpentine-chlorite inTuscaloosa Formation sandstone, U.S. Gulf Coast

P.C. RvlN* AND R.C. RrvxoLDS JR.

Department of Earth Sciences, Dartmouth College, Hanover, New Hampshire 03755, U.S.A.

Ansrucr

Randomly interstratified serpentine-chlorite (Sp-Ch) is ubiquitous in sandstones of thesubsurface Tuscaloosa Formation. The Sp-Ch occurs as grain coatings, pore fillings, pel-oids, and infillings. The presence ofpeloids and infillings suggests that the Sp-Ch originatedas odinite, a 7 A mineral that forms at the sediment-seawater interface in shallow, tropicalmarine sediments. We conclude that during the very earliest stages of diagenesis, odinitepeloids and infillings transformed to mixedJayer Sp-Ch, forming grain coatings with pres-ervation of some of the original odinite textures. The formation of grain-coating Sp-Chduring early diagenesis apparently preserved high primary porosity in medium-grainedspecimens. Fine-grained and silty sandstones, however, contain poorly formed grain coat-ings and are tightly cemented by quartz, implying that grain size also has some effect onauthigenic quartz growth.

With increasing burial depth between 1702 and 6216 m, the proportion of Sp layers inSp-Ch decreases from 20.7 ! 1.1 to 1.3 + 0.10/0, indicating that Sp layers transform toCh layers with increasing diagenetic grade. The polytype of the Sp-Ch is essentially IDD atdepths <2000 m. At depths >2000 m, the polytype is randomly interstratified lbb-Iaa,and the proportion of IbD layers decreases from - 100 to 490lo over the interval studied.At depths <2000 m, orthohexagonal Sp transforms tolbb Ch, but at depths >2000 m,orthohexagonal Sp transformstolaa Ch. There is essentially no change in Sp-Ch chemicalcomposition with depth, implying that temperature is the dominant control on the struc-tural transformations. Mean crystallite thickness also remains relatively constant withincreasing depth, indicating that diagenetic transformations in Sp-Ch do not involve crys-tal-growth mechanisms typical of most clay minerals. Rather, the mineralogic transfor-mations appear to proceed on a layer-byJayer basis, with individual layers dissolving andreprecipitating within the confines of the crystallites.

INrnonucrroN

Authigenic chlorite (Ch) is a common authigenic min-eral in sandstone, often comprising the majority of theclay-size fraction. The occurrence of Ch as a coating ondetrital grains is particularly interesting because the coat-ing may inhibit quartz cementation, resulting in the pres-ervation of abnormally high primary porosity in deeplyburied rocks (e.g., Heald and Anderegg 1960; Pittmanand Lumsden 1968; Thomsen 1982). Grain-coating Chhas been the subject of several recent studies (e.g., Thom-sen 1982; Reynolds et al. 1992; Ehrenberg 1993; Hillier1994; Spdtl et al. 1994), but the origin ofthe Ch is un-certain, as are the effects ofburial diagenesis on polytyp-ism and mixed layering of Ch and interstratified 7 A lay-ers like those of serpentine (Sp) (e.g., Reynolds 1988;Walker 1993). Accordingly, the primary purposes of thisstudy are (l) to determine the origin of diagenetic ser-

t Present address: Environmental Sciences Department, Sa-lish Kootenai College, Pablo, Montana 59855, U.S.A.

0003-004x/9610 r 02-02 I 3$05.00

pentine-chlorite (Sp-Ch) in Tuscaloosa Formation sand-stone and (2) to examine the effects of burial diagenesison mixed layering and polytypism of Sp-Ch in the Tus-caloosa Formation. (In this paper, the 7 A layers that areinterstratified with Ch are termed Sp, rather than the morespecific mineral name berthierine, to avoid genetic im-plications associated with the name berthierine.)

In sandstone, authigenic Ch and related mixed-layerminerals commonly occur as grain coatings and pore fill-ings. They may form during the dissolution of kaolinite(Boles and Franks 1979; Burton et al. 1987) or volcanicrock fragments (Thomsen 1982), when Fe and Mg areliberated during illitization (Hower et al. I97 6; Boles andFranks 1979), or by the transformation of a 7 A phase to14 A Ch (Nelson and Roy 1958; Bailey 1988a; Hillier1994). Two 7 A clay minerals, berthierine and odinite,are structurally and chemically similar to Fe-rich Ch andoccur in similar rocks. Odinite is a Sp-group, Fe-richmineral that forms in tropical, shallow marine sedimentsat the seawater-sediment interface. It occurs as pelletsand infillings and has site occupancy intermediate to those

2t3

2t4 RYAN AND REYNOLDS: MIXED-LAYER SERPENTINE-CHLORITE

of trioctahedral and dioctahedral layers because of highamounts of Fe3+ (Odin 1985; Bailey 1988a). Berthierineis a trioctahedral Sp-group mineral that commonly oc-curs as ooids in unmetamorphosed iron formations andsome shallow marine sandstones (Velde et al. 197 4;Bhat-tacharyya 1983; Velde 1989), as well as in coal measures(Iijima and Matsumoto 1982) and altered igneous rocks(Ruotsala etal. 1964; Slack et al.1992; Jiang et al.1994).

Interstratification of Ch (14 A) and Sp (7 A) in sedi-mentary rocks has been documented by both X-ray dif-fraction (XRD) and transmission electron microscopy(TEM). Brindley and Gillery (1954) used XRD to docu-ment randomly interstratified kaolinite-chlorite, whichwas later interpreted as Sp-Ch by Brindley (1961). Morerecently, Dean (1983), Moore and Hughes (1990, l99l),and Reynolds et al. (1992) studied grain-coating Ch insandstone, and they attributed broadening of Ch odd-order 00/ peaks to randomly interstratified Sp layers.Walker and Thompson (1990) and Hillier (1994) alsomeasured peak broadening and showed that the propor-tion of Sp layers in mixed-layer Sp-Ch decreases as tem-perature increases. Ahn and Peacor (1985) used TEM todocument mixed-layer berthierine-chlorite in shales ofthe Gulf Coast Basin and indicated that the interlayeredberthierine may be a metastable precursor to Ch underburial diagenetic conditions. These occulrences of mixed-layer Sp-Ch may represent intermediate stages in theprograde transformation of a 7 A phase (e.g., odinite,berthierine) to Ch. The transformation requires the in-version of a tetrahedral sheet and may occur with nochange in composition (Brindley 1961; Jiang et al. 1994).Nelson and Roy (1958) showed that synthesized 7 A,phases transform to Ch with increasing temperature, andIijima and Matsumoto (1982) and Velde (1989) indicatedthat berthierine ooids transform to Ch during burial dia-genesis ofcoal measures and ironstones. Bailey (1988a)found that 7 A odinite transforms to Ch of the lba po-lytype during early diagenesis of tropical marine sedi-ments. These conclusions, when considered in conjunc-tion with the findings of Walker and Thompson (1990)and Hillier (1994), suggest that the transformation of 7A phases to Ch could be used as an indicator ofdiagenesisand low-grade metamorphism in sandstone, ironstone,and coarse-grained limestone, rocks in which illite-smec-tite is often a minor or absent phase and thus is of no usein determining thermal history.

The naturally occurring Ch polytypes are laa, Iba (Ib

0 : 97),lbb (Ib P : 90), and IIb (Brown and Bailey1962; Bailey 1988b; Walker 1993). Brown and Bailey(1962) first derived the structures ofthe Ch polytypes andshowed that the IIb polytype, which commonly occurs ingreenschist facies rocks, has the most stable atomic con-figuration. The type-I polytypes have less stable atomicconfigurations and should transform to the IIb polytypewith increasing grade. Brown and Bailey (1962) also in-dicated that \aa Ch is stable at slightly higher tempera-tures than the lbb polytype, citing the prograde recrystal-

Iization of the IbD polytype to the Ica polytype inironstone. Hayes (1970) studied Ch polytypes from a widevariety ofrock types and geologic environments and pro-posed that the IIb polytype is stable above 200 'C. Thelower temperature polytypes, in order of increasing sta-bility, are Ib (disordered),Iba, andlbb, which, in theory,should sequentially transform to the stable IIb polytypeif sufrcient energy is added to the system. However, fewfield studies have documented Ch-polytype transforma-tions (Karpova 1969; Hillier I 994; Spdtl et al. 1994), andIIb is often the only Ch polytype in diagenetic and low-grade metamorphic sequences (Walker 1989; Walker andThompson 1990; Ryan and Thompson l99l; Hillier1993). Walker (1989) noted that type-I chlorites are rare-ly found in fine-grained rocks and suggested that the typ€-Ipolytypes require large pores or vugs in which to crystal-Iize (e.g., sandstones, coarse-grained limestones, andveins). On the basis of studies of sedimentary and low-grade metamorphic rocks, it appears that Ch polytypismdepends on more than temperature and is probably alsocontrolled by factors such as P"ro, chemical composition,and time (Walker 1993). For a thorough discussion andreview of Ch-polytype geothermometry, the reader is re-ferred to Walker (1993) and de Caritat et al. (1993).

Gnor,ocrc SrrrrxcThe Lower Tuscaloosa Formation is an ideal sedimen-

tary unit for the study ofdiagenetic Ch. Ch is abundantin Lower Tuscaloosa Formation sandstone, typically oc-curring as 5-10 pm thick coatings on detrital grains(Thomsen 1982). Petrographic evidence indicates that theCh formed during early diagenesis and that it may haveresulted in the preservation of abnormally high primaryporosity (22o/o at 6600 m) by coating detrital quartz grainsand inhibiting growth of quartz cement (Thomsen 1982;Wiygul and Young 1987). Porous, friable sandstone typ-ically contains well-formed Ch grain coatings, whereassamples with poorly formed Ch coatings are usually tight-ly cemented by secondary quartz (Thomsen 1982).

The Tuscaloosa Formation was deposited in deltaic,shallow marine, and alluvial environments in the GulfCoast Basin during a low sea-level stand in Late Creta-ceous (Cenomanian) time (e.g., Stancliffe and Adams1986; Hogg 1988). It is a tabular unit that strikes ap-proximately east-west, with outcrop extending from SouthCarolina to Mississippi. It dips about l" to the south be-neath Louisiana and Mississippi and attains a maximumburial depth ofapproximately 6800 m beneath southernLouisiana (Fig. l), thus spanning a wide temperature range(Alford 1983). Lower Tuscaloosa Formation sandstone isfine to medium grained, well sorted, and ranges from sub-litharenite (Stancliffe and Adams 1986) to quartz lithar-enite and quartz arenite (Hamlin and Cameron 1987).Detrital clasts are predominantly quartz, but volcanic andmetamorphic rock fragments compose up to 27o/o of lhebulk volume of the rock (Berg and Cook 1968; Stancliffeand Adams 1986).

RYAN AND REYNOLDS: MIXED-LAYER SERPENTINE-CHLORITE 215

Mnrrroos

All the samples analyzed in this study from depths<6000 m were core samples provided by David R. Pev-ear of Exxon Production Research Company, Houston,Texas. Paula Hansley of the U.S. Geological Survey,Denver, Colorado, donated core samples from 6214 to6216 m. The cores were sampled from l4 drill holes thatpenetrated Lower Tuscaloosa Formation sandstone overa wide range of burial depth (1702-6216 m) (Fig. l). Thetemperature interval represented by these samples is ap-proximately 80-175 .C (Alford 1983). Most of the sam-ples are porous, friable sandstone, but a few are well-cemented, fine-grained sandstone. The cores were washedin distilled water to remove drilling by-products andcrushed in an iron mortar and pestle without grinding(Bailey 1988b). The powdered sample was immersed indistilled water and agitated for 12 h in a sample shaker.The <0.5 and 0.5-2 prm (esd) fractions were then sepa-rated by repeated centrifugation and analyzed for mixedlayering, polytypism, and compositional variation inSp-Ch.

Oriented powder aggrcgates were made by concentrat-ing 50- 100 mg of clay in distilled water and dropping theclay slurry onto glass slides (2.7 x 4.6 cm). XRD patternsof air-dried and ethylene glycol-solvated oriented mountswere recorded between 2 and 35 20 with 0.05" steps and5 s counts to analyze the mineral content ofthe clay-sizefractions. XRD data were obtained using an automatedSiemens D-500 diffractometer with CuKcr radiation (40kV and 40 mA), a graphite monochromator, and a Da-tabox microcomputer.

The proportion of Sp layers in mixed-layer Sp-Ch wasdetermined by the method of Reynolds et al. (1992) andverified by comparison with calculated NEWMOD dif-fraction patterns (Reynolds 1985). A l" divergence slitwas used in conjunction with a 0.05" aperture slit andtwo 2" soller slits (Reynolds et al. 1992). The ch 004 and005 peaks were scanned, air-dried, with 0.05" steps andcount times ranging from 20 to 60 s. Analytical 004 and005 peaks were fitted to pseudo-Voigt functions by a least-squares method, and peak breadth (B) was measured atone-half height. Interstratification of 7 A (SpJike) layersand 14 A (Cn) layers can be difrcult to detect becausedoo,Ch t 2doo,Sp, and Sp peaks essentially overlap witheven-order Ch peaks. However, random interstratifica-tion ofSp and Ch causes broadening ofthe odd-order Chpeaks proportional to the amount of 7 A material, mak-ing Sp-Ch interstratification identifiable and quantifiable(Reynolds et al. 1992). Accordingly, residual line broad-ening (B.) caused by interstratified 7 A Sp layers and 14A Ch layers was determined by the equation

P,: (P[&; - Pr2s)a/t2s,

(Reynolds et al. 1992), where Booo and Boo, are the peakwidths at one-half height (in degrees 20) for the Ch 004and 005 peaks. Instrumental distortion of the Ch 004 and

TEXAS

Frcunr l. Location map ofwell sites that produced the coresamples studied (after Berg and Cook 1968; Hoee I 988). Dashedlines and numbers indicate depth isopachs to the base of theLower Tuscaloosa Formation (in meters).

005 peaks is assumed to be very similar and thus negli-gible in these calculations. The percentage ofSp was de-termined by the empirical equation (Reynolds et al. 1992)o/oSp : -0.51 + 24.270,. Duplicate analyses and com-parisons with NEWMOD calculated patterns indicate thatthis method produces results that are accurate to withinapproximately 50/o of the reported value.

To determine polytypism of Sp-Ch, random powdermounts were prepared by freeze-drying solutions of ap-proximately I g of powder in 50-100 mL of distilledwater. The freeze-dried powders were packed into side-loaded sample holders to maximize random orientationof the clay crystals for XRD analysis of polytypes (Reyn-olds 1992). The random powder mounts were scannedfrom 32 to 64' 20 with 0.05' steps and count times of l5to 20 s per step. The differential X-ray diffraction methodof Schulze (1982) was used to remove illite, anatase, andkaolinite interference. This method records an initial XRDscan of the Ch-, illite-, and kaolinite-bearing randompowder mount over the 20 range of interest. The powderis then heated in warm (80 .C) I N HCI for I h, whichdestroys Fe-rich Ch (Brindley 196l). An XRD trace ofthe HCl-treated residual containing illite, anatase, andkaolinite is then made over the same 20 range, after whichthe HCl-treated XRD pattem is digitally subtracted fromthe initial scan to give a Ch pattern (Fig. 2).

Ch and Sp polytypes were identified by comparisonwith published patterns (Brown and Bailey 1962; Balley1969, 1988b) and calculated patterns (Reynolds 1993).The proporti on of Ibb polytype layers in mixedJayer lbb-Iaa was determined by comparing analytical XRD pat-

2t6 RYAN AND REYNOLDS: MIXED-LAYER SERPENTINE-CHLORITE

o2e

Frcunr 2. An example of differential X-ray diffraction. Sp-Ch : serpentine-chlorite, ILL : illite, and An : anatase.

terns with calculated XRD patterns (Reynolds 1993).Atomic coordinates for the calculated Ibb structure arefrom Shirozu and Bailey (1965), andlaa coordinates werederived from the Ibb coordinates of Shirozu and Bailey(1965) by shifting octahedral layers +a/3 relative to ad-jacent tetrahedral layers.

The terminology for describing mixed-layer propor-tions is that used by Reynolds (1988). Ibb(0.65)-Iaarep-resents a mixed-layerlbb-Iaa polytype with 650/o lbblay-ers. Similarly, Sp(0.15)-Ch is mixedJayer Sp-Ch with l5oloSp layers.

Scanning electron microscopy (SEM) and optical mi-croscopy were used to determine textures, including thedistribution and morphology of diagenetic Sp-Ch and therelative proportions ofother diagenetic phases and detri-tal grain composition. Sp-Ch morphologies (e.g., graincoatings, peloids, infillings, pore fillings, and replace-ments) were verified by energy-dispersive X-ray (EDX)spot analysis. SEM and EDX instrumentation and oper-ating conditions are given below.

The chemical composition of Sp-Ch from four samples(1937,3878,4054, and 5470 m) was determined by EDXspot analysis. Between eight and I I EDX analyses wereperformed on each sample. These four samples span the

range of composition with respect to polytypism and in-terstratified Sp. The sample from 4054 m is silty and wellcemented by secondary quartz, whereas the other threeare porous, friable sandstone. For SEM and EDX anal-yses, polished sections were coated with carbon and ex-amined in a Zeiss DSM 962 SEM instrument with anEDX analyzer at the Dartmouth College Electron Micro-scope Laboratory. Samples were analyzed for FeO, MgO,AlrO3, SiOr, MnO, KrO, NarO, CaO, TiOr, and PrOr,and analyses were normalized to a Ch unit cell withO,.(OH)8. The compositions of 2:l octahedral layers andhydroxide interlayer octahedral layers are assumed to beidentical. Standards were pure oxides for every elementexcept Na, for which albite was used. Biotite and a graniteglass were also analyzed as secondary standards to checkreliability. The accelerating voltage was 15 kV, the beamcurrent was 6 nA, spot size was l-2 prr], and the countingtime was 100 s per analysis. The spot size of l-2 pm issmaller than the thickness of Sp-Ch rims (5-10 pm) andsignificantly smaller than the size of peloids and infillings(25-50 pm), and so we are confident that the EDX anal-yses are relatively free of contamination from other min-erals. However, intergrowths of other authigenic miner-als, particularly illite, have been known to contaminateanalyses (e.g., Whittle 1986; Jiang et al. 1994). Thus, allanalyses with >0.50/o KrO were discarded to avoid illitecontamination (Hillier 1994). All Fe was assumed to beFe2*, and Ti was excluded from the Sp-Ch structural for-mulas because (l) it occurs in trace quantities in mostsamples, and (2) it probably occurs as anatase (TiOr).Analyses with significant amounts of TiO, are assumedto be intergrowths of Sp-Ch and anatase. Na, Ca, and Pare also not included in structural formulas because theyare present only in traces (typically less than one standarddeviation) and probably occur in halite, carbonate, andapatite.

The ratio of Sp-Ch to kaolinite was estimated by com-paring peak areas of the Sp-Ch 006 and kaolinite 003peaks obtained from analytical XRD scans with corre-sponding peak areas determined from calculated NEW-MOD patterns (Reynolds 1985). Quantitative estimateswere based on XRD pattems of oriented, <2 prn fractionpowders, and to minimize size segregation in the quan-titative estimates, powders were mounted onto glass slidesby the filter-peel method (Drever 1973;Reynolds 1992).

Rnsur,rsBulk mineral content

Sp-Ch, illite, kaolinite, and anatase are present in allsamples. The ratios of Sp-Ch to anatase and illite to an-atase are relatively consistent and show no systematicvariation, suggesting that the amounts of Sp-Ch, illite,and anatase remain relatively constant. The amount ofkaolinite, however, decreases with increasing burial depth.At 1702 m burial depth, kaolinite makes up about 800/oof the <2 rrm fraction, whereas between 1714 and 3600m. the <2 pm fraction contains 30-40o/o kaolinite, and

RYAN AND REYNOLDS: MIXED-LAYER SERPENTINE-CHLORITE

Sp-Ch 005

BURIAL DEITTH

2 t 7

Sp-Ch 004

20 79o Sp1937 m(63e3)

f = 0.30'

24 "Ze 26 30 "Ze

Frcunr 3. The 004 and 005 peaks of Sp-Ch. Dots are exper-imental points and solid lines are least-squares-fitted pseudo-Voigt profiles. Note the decrease in peak breadth (B) ofthe Sp-Ch 005 with increasing depth.

at depths greater than 3600 m, kaolinite makes up < 100/oof the <2 prm fraction. Weak calcite peaks were evidentin some samples, reflecting the presence of fossil frag-ments, calcite cement, or both. Thin-section analysisshowed that hematite and pyrite are present in all sam-ples; hematite is prevalent in shallow samples, and itscontent decreases with burial depth, whereas the propor-tion of pyrite increases with burial. In some shallow sam-ples (<2000 m), hematite occurs as grain coatings.

Texture

Thin-section studies show that in samples from <3600m, Sp-Ch occurs as subequal combinations of grain coat-ings, pore fillings, peloids, and infillings. With increasingburial depth, Sp-Ch grain coatings generally increase incontinuity and thickness, whereas the peloids and infill-ings appear to decrease in abundance. Sp-Ch in deepsamples (> 3600 m) most commonly occurs as prevalent,well-formed isopachous rims on detrital grains; pore fill-ings, peloids, and infillings are less common in the deepsamples. In the 4054 m sample we found two ribbonlikeoccurrences of Sp-Ch that may be pseudomorphous afterbiotite.

In all cases, Sp-Ch grain coatings appear to have pre-dated kaolinite and illite, and in samples that containcarbonate cement, Sp-Ch grain coatings apparently pre-

BURIALDEI'TH(103 ft.)

2 0 1 5 1 0 5 0

7o SP in SP-Ch(<0.5 pm)

Frcunr 4. Relationship between burial depth and Sp contentin mixedJayer Sp-Ch in the <0.5 pm fraction. Boxes indicatewell-cemented, fine-grained specimens.

ceded the carbonate cement. Samples from <2000 m

contain localized hematite grain coatings and Sp-Ch grain

coatings. Quartz cement is very minor or absent in most

samples, although one of the intermediate depth samples(3573 m) contains euhedral quartz overgrowths inter-

spersed with discontinuous Sp-Ch grain coatings. In ad-

dition, the five samples that exhibit poorly developed or

absent Sp-Ch rims contain abundant quartz or carbonate

cement (bur ia l depths : 2732,2782,3765,4054, and4056 m). In hand sample, these samples are fine grained

and well cemented, and XRD analyses indicate that they

are characterized by a Sp:Ch ratio or polytype composi-tion (or both) that is not consistent with the general trends.

Structure

All Ch analyzed in this study contains broad odd-order00/ peaks, indicating the occurrence of randomly inter-stratified Sp-Ch in all samples (Fig. 3). In the <0.5 pmfraction, peak width at one-half height of the Sp-Ch 005peak (Boor) decreases from 1.05 to 0.40' 20 between 1702and 6216 m burial depth, whereas Booo is relatively con-stant (0.29-0.38" 20) throughout the sequence (Fie. 3).The decrease in Boo, with increasing burial reflects a de-crease in the proportion of Sp layers in mixedJayer Sp-Ch from a maximum of 20.7 ! l.lo/o at 1937 m to aminimum of 1.3 + 0.lo/o at 6214 m (Fie. a). In all sam-ples, 0ooo values of Sp-Ch for the 0.5-2 y,mftaclion (0.22-0.33 20) are less than Booo values for the <0.5 pm fraction(0.29 to 0.38 20), indicating that mean crystallite size isgreater in the coarser fraction. The percentage of Sp inSp-Ch does not differ significantly between the two sizefractions. Variations in Boo. with depth are discussed be-low.

Nearly all Sp-Ch analyzed in this study contains ran-domly interstratified lbb andlaalayers (Fie. 5). At depths<2000 m, the polytype is essentially Ibb (Brown and Bai-ley 1962), although the diminished intensity ratio of the202:060 peaks may indicate traces of interstratified Iac(Fig. 5a). Composite peaks between the pure Ibb and pureIaa polytype peaks are found in deeper samples, and cal-

0= 1 0s '

1937 m (6393) BI'RIAL DEPTH

218 RYAN AND REYNOLDS: MIXED.LAYER SERPENTINE-CHLORITE

202(rbb)

| -rbbiqzo.luo spl'

tlbb(0.67)-laa

i 1o.rz, sp;

:rbb(O.49)-Iaa(4.67o Sp)

ilbb(O.53)-Iaa

(3.8% sp)

34

Frcunrpolytypesdepth.

Burial Deptlr7937 m(6393)

3573 m(1 1791)

4554 m(1s028)

"20

6. Randomly oriented XRD patterns of Ibb-Iaashowing decrease in IDb component with increasing

culated XRD patterns of randomly interstratified Ibb-Iaaclosely match the analytical patterns (Figs. 5b and 5c).Comparison of analytical and calculated patterns showsthat the IDD component decreases from - 100 to 49010 overthe interval studied (Figs. 6 and 7). There is essentiallyno difference in polytype composition between the <0.5and 0.5-2 pm fractions. The percentage of Sp in Sp-Chis proportional to \bb content in lbb-laa, whereas thepercentage oflbb and even-order peak breadth (0ooo) varyinversely (the percentage of Iaa in lbb -Iaa is proportionalto 0ooo) @ig. 8).

The value of douo systematically decreases with increas-ing burial from 1.556 A in nearly purelbb to 1.549 A inIb\0.49)-Iaa, as does d,ar (2.698-2.665 A). Measured doo,progressively increases with increasing burial from l4.l 1to 14.16 A, and peak breadth at one-half height of the0010 peak simultaneously increases from 0.37" 2d in

(-Frctrnr 5. Comparison of calculated diffraction patterns of

polytypes (Reynolds 1993) and randomly oriented mounts: (a)lbb, (b) lbb(0.67)/laa, (c) IbbQ.aD/ha. Characteristic hkl peaksof the pure Ibb and Iaa polytypes are shown. A : anatase.

38'78 n (12'197'\ BURIAL DEPTH

204(leo)

CALCULATED lbr(O 6't)/Iaa R = 0

202(taa)

4554 m (15028') BURIAL DEPTH

RYAN AND REYNOLDS: MIXED-LAYER SERPENTINE-CHLORITE

2

BT]RIALDEPTH

(km)

2r9

5

BURIALDEPTH

1ro3 rt.; lo

7o lbb nlbb-Iaa

(<0.5 pm)

Frcunr 7. Relationship between burial depth and polytyp-

ism, <0.5 pm fraction. Boxes indicate well-cemented, fine-grained

specrmens.

nearly pure Ibb to 0.56' 20 in Ibb(0.49)-laa, suggestingthat the values ofd00r for the lbb and Iaa phases are notthe same. The value of doo, for the Ibb phase, measuredfrom XRD patterns, is 14.l t A. ttre value of doo, for theIaa component could not be directly measured from XRDpatterns (it makes up <510/o of the Ibb-Iaa), and so weused NEWMOD to model broadening and migration ofthe 0 0 I 0 peak caused by interstratification of two phaseswith different doo,. Using this method, we find lhat doo,of the Iaa phase is 14.35 A.

Decreasing 00/ peak breath is primarily an indicator ofincreasing crystallite thickness in clay minerals and is of-ten used to interpret crystal growth during diagenesis andlow-grade metamorphism (Kubler 1964; Frey 1970; Eberlet al. 1990; Hillier 1994). In Tuscaloosa Formation Sp-Ch, there is no increase in crystallite thickness (along c*)with increasing depth, and peak breadth (Booo) actuallyincreases slightly with increasing burial (Fig. 3). We at-tribute the increase in Booo to interstratification of 14.1 ILlbb and 14.35 A Iaa phases. The increase in Booo par-allels decreasing Ibb component (in lbb-Iaa), and crys-tallite thickness appears to be nearly constant throughoutthe sequence.

Chemical composition

SEM and EDX results are presented as oxides in TableI and as Sp-Ch structural formulas in Table 2. In allcases, oxide totals are less than the expected 85-880/o forCh with an O,'(OH), unit cell, a common problem forCh and mixed-layer Ch minerals (e.g., Sp-Ch and corren-site). Hillier (1994) attributed low oxide totals obtainedby microprobe analyses to microporosity between Chgrains, and to test whether or not low oxide totals ad-versely affect structural formulas, he compared a stan-dard metamorphic Ch (CCa-l, available from the ClayMinerals Society) prepared as a polished section to thesame specimen prepared as a polished 2-5 pm powderand epoxy resin mixture. The mean oxide total for thepolished section was 860/0, whereas the mean oxide totalfor the epoxy-powder section was 54o/o, with many totalsbetween 20 and 500/0. Although dispersion about the mean

<.

-O

o

Vo lbb nlbb-laa

Frcunr 8. Relationship between polytypism and even-orderSp-Ch peak breadth (0ooo).

was greater for the low-oxide data, the average structuralformulas obtained from the polished section and the ep-oxy section were nearly identical, suggesting that errorsresulting from low oxide totals are random and that meanvalues provide reliable results.

Standard deviations associated with SEM and EDXanalyses of Sp-Ch in this study were generally low: 0.8-l.l0lo for Si and Al, 1.0-1.50/o for Fe, and 1.9-4.3o/o forMg. Analytical errors were greatest for Sp-Ch from 1937m, in which weight percent oxide totals were much lessthan in deeper samples. The low oxide values in the shal-lowest sample may be due to a greater amount of micro-pore space between Sp-Ch crystals that is not so pro-nounced at greater depths.

The high amount of TiO, at 1937 m (1.050/d and theelevated TiO, value at 3878 m (0.360/o) are attributed tointergrowths of anatase and Sp-Ch in pore fillings. Indi-vidual analyses of pore fillings contained TiO, values of6.44 and 1.99o/o at 1937 m and 2.10 and 1.24o/o at 3878m, whereas all grain coatings, peloids, infillings, andpseudomorphous replacements contained only traces ofTior.

DrscussroN

Origin and occurrence of Sp-Ch

The decrease in the proportion of interstratifled ser-pentine-like layers with increasing grade raises the pos-sibility that Sp-Ch in the Tuscaloosa Formation origi-nated as a 7 A mineral that subsequently transformed tomixed-layer Sp-Ch during diagenesis. The presence of Sp-Ch peloids and infillings, combined with an inferred trop-ical marine depositional setting, strongly suggests that theoriginal material was odinite, an Fe-rich 7 A mineral thatforms at the sediment-seawater interface in Recent, trop-ical marine sediments (Odin et al. 1988; Odin 1990).Odinite is unstable outside its environment of formation,

220 RYAN AND REYNOLDS: MIXED-LAYER SERPENTINE-CHLORITE

Tlele 1. SEM and EDX analyses (weight percent) and atoms per formula unit of Sp-Ch

Burial depth (m)No of analyses

54708

4054o

1 937a

38781 1

sio,Alro3FeOMgoMnOTio,K"oNaroCaOPrO,

TotalSi

rrrAl16rAlFeMgSum (Vl)'

8.99 + 0.107.50 + 0.08

11.76 + 0 .201.49 + 0.060.05 + 0.071.05 + 0.070.05 + 0.070.13 + 0.070.11 + 0.050.05 + 0.08

3 1 . 1 92.931.071.783.140.725.64

16.71 + 0 .1313.33 + 0.1021.92 + 0.223.89 + 0.090.06 + 0.090.36 + 0.060.08 + 0.090.29 + 0.080.14 + 0.050.07 + 0.08

56.852.911.091.583.181.005 . / b

17.56 + 0.1314.85 r 0.1225.69 r 0.253.25 + 0.080.09 + 0.100.10 + 0.060.05 + 0.070.26 + 0.080.25 + 0.070.14 + 0 .10

62.232.801.201.603.43o.775.80

18.82 + 0 .1416.60 + 0.1323.92 + 0.264.84 + 0.090.05 + 0.100.07 + 0.060.04 + 0.060.31 + 0.080.14 + 0.050.08 + 0.09

64.862,801.201.722.981.055.75

. Number of octahedral cations (per six sites).

and it transforms to a mixed-layer 7-14 L phase duringthe first stages of diagenesis (Bailey 1988a; Odin et al.1988). In light of this, odinite probably formed withinthe upper decimeter of Tuscaloosa Formation sedimentsoon after deposition in a tropical marine environment(Odin et al. 1988). Ultrabasic volcanic rock fragments,which are ubiquitous in the Tuscaloosa Formation(Thomsen 1982), and possibly Fe-rich river water (Odinet al. 1988), provided cations that precipitated as poorlycrystallized odinite shortly after deposition. During veryearly diagenesis, perhaps within a few meters of the sur-face (Bailey 1988a), the odinite peloids and infillings dis-solved and reprecipitated as Sp-Ch. This transformation

TABLE 2. Structural formulas for different morphologies of SpCh

Sp-Ch structural formula Texture

resulted in the formation of Sp-Ch coatings, although someofthe original odinite textures were preserved. The trans-formation of odinite to Sp-Ch coatings during early dia-genesis is consistent with petrographic evidence indicat-ing that Sp-Ch grain coatings preceded all other diageneticphases with the possible exception of hematite. The earlydiagenetic transformation of odinite to grain-coating Chhas also been observed in Jurassic sandstones of the Nor-wegian continental shelf (Ehrenberg 1993).

The presence of mixed-layer Sp-Ch in all samples isprobably due to (l) the widespread occurrence ofvolcanicrock fragments in Tuscaloosa Formation sandstone(Thomsen 1982; Genuise and McBride 1990) and (2) atropical, shallow marine depositional environment thatfavored odinite formation (Odin 1990). Anatase, whichwas detected in all samples by XRD analysis, probablyformed when Ti was excluded from odinite during crys-tallization following the dissolution of mafic fragments(cf. Eggleton and Banfield 1985), or it formed during thetransformation of odinite to Sp-Ch. The coexistence ofhematite and Sp-Ch indicates that conditions during Sp-Ch crystallization were slightly oxidizing and that pH wasnearly neutral (Garrels and Christ 1965).

Composition

Compositions determined by SEM and EDX are rela-tively consistent throughout the sequence (Table l), al-though infillings of Sp-Ch appear to contain more Fe andless t6rAl than the other textures (Table 2), and Sp-Chfrom 4054 m contains about 100/o more Fe than the othersamples. The total number of octahedral cations (per sixsites) ranges from 5.6 to 5.8 and shows no systematicvariation with depth. The apparent octahedral site va-cancies may be due to greater amounts of t6tAl than tatAl(Foster 1962; Whittle 1986), although a recent analyticalelectron microscope (AEM) study by Jiang et al. (199a)indicated that octahedral vacancies in Ch minerals areanalytical artifacts caused by inclusions of other minerals.

Structural formulas of Tuscaloosa Formation Sp-Ch

1937 m burial depth

(Fes 6Mgo75All s4)(Sir 11Al1 @)O1o(OH)8(Fes 01 Mgo 7r Alr sr)(Sis,3Alo e)O1o(OH)r(Fe3 {Mgo sAll 6r)(Si, ?sAll rr)O1o(OH)B(Fes loMgo 74All 7e)(Sir rrAll @)Oro(OH)6(Fe3,oMgo 7rAl1,s)(Si, $Al1 o?)O1o(OH)E

3878 m burial depth

(Fe3 o?Mgo eoAl, s)(Sir r6Ah s)O,o(OH)a(Fer rrMg, @Al, ?6)(Si, 8?Al1 1s)O1o(OH)8(Fes sMgo sAl, 3s)(Si, sAll 1?)O1o(OH)o(Fe3 1rMg1 osAl, 61)(Si, ssAl, 1r)Olo(OH)o(Fe3 18Mgl ooAl, ss)(Sir 11Al1 or)Olo(OH)6

4054 m budal depth

(Fe3 37Mgo 7rAl, 6sXSi, 6aAlr i2)O jo(OH)B(Feo €Mgo61Alr 56)(Si, @Al1 ,o)Oro(OH)a(Fe3 @Mgo z6A11 5o)(Sir,6A11 rr)Oro(OH)e(Feo eMgo 77Alr 67)(Si, sl All 1r)O1o(OH)e(Fe3 sMgo ?oAlr s)(Si, ?sAll 2r)Olo(OH)o(Fe3 €Mgo77All 60)(5i, ooAll ,o)O1o(OH)B

5470 m burial depth

(Fe, osMg, reAll 65)(512 s6Al1 1i)Oro(OH)s(Fe, soMg, oTAll so)(Si, osAll 1r)Olo(OH)s(Fe. *Mg' eAl, es)(Si, ??Alr a)O1o(OH)s(Fer.uMg, orAl,,r)(Si, olAll 1e)O1o(OH)6

Note: PF : pore fillings, pE1 : peloids, INF : infillings, 66 : graincoatings, and RPL: replacements; N: number of analyses.

PF (A/ : 2)PEL (N : 2 )rNF (N : 2 )GC (N: 3)avg.

PF (N : 3 )P E L ( N : 1 )rNF (N : 2 )GC (rV: 5)avg.

P F ( N : 1 )PEL (N : 2 )RPL (rV: 2)rNF (N : 1 )GC (N : 3 )avg.

P F ( N : 1 )rNF (N : 2 )GC (A/ : s)avg.

RYAN AND REYNOLDS: MIXED-LAYER SERPENTINE-CHLORITE 22r

determined by SEM and EDX in this study are very sim-ilar to both mixed-layer berthierine-chlorite sampled fromGulf Coast mudstones and analyzed by AEM (Jiang et al.1994) and mixed-layer 7-14 L minerals obtained from avariety o"f, sandstones and analyzed by microprobe (Hill-ier 1994).

Structure of Sp-Ch

The decrease in the proportion of Sp-like layers inmixed-layer Sp-Ch with increasing burial depth indicatesthat SpJike layers transform to Ch layers with increasinggrade, a finding consistent with similar reports by Nelsonand Roy (1958), Karpova (1969), Iijima and Matsumoto(1982), Walker and Thompson (1990), and Hil l ier (1994).The Sp-to-Ch transformation is essentially a polymor-phic transition involving compositionally similar or iden-tical 7 and 14 A phases (Xu and Veblin 1993; Jiang etal. 1994). Polytypes are defined as different stacking ar-rangements ofsuccessive layers that are identical in bothcomposition and structure (Moore and Reynolds 1989),and so strictly speaking, Z A Sp-tite layers cannot beconsidered polytypes of 14 A Ch layers. However, thetwo phases are crystallographically similar (Fig. 9), andthe 7 to 14 A transformation only requires inversion ofa tetrahedral sheet (Brindley 1961; Reynolds 1988).

Previous studies have shown that both Sp and Ch oc-cur as orthohexagonal poll'types in sandstone (e.g., Brownand Bailey 1962; Karpova 1969; Walker 1993). Karpova(1969) referred to both the orthohexagonal Sp polytypeand the orthohexagonal Ch polytype asIbb. Bailey (l 969)indexed the I ?" Sp polytype on a C-centered, orthohex-agonal cell. Other than the lack of 14 A 00/ peaks, itspowder diffraction pattern is virtually identical to that ofIbb Ch (Brown and Bailey 1962; Bailey 1969), and it isprobably the same polytype referred to aslbb Sp by lfur-pova (1969). In the Tuscaloosa Formation, XRD patternsindicating the occuffence of the orthohexagonal IbD struc-ture in mixed-layer Sp-Ch with ll.8-20.7o/o Sp layerssuggest that both Sp and Ch occur as orthohexagonalpolytypes. Sp most commonly occurs as an orthohexa-gonal polytype (Bailey 1969), and interstratification ofIbb Ch and orthohexagonal (Ibb or I D Sp with similaratomic geometries (Brindley 1961) would produce a goodfit between adjacent layers along the basal plane. Accord-ingly, we conclude that the mixed-layer Sp-Ch consists oforthohexagon al (Ibb) SpJike layers interstratifi ed with Chlayers ofboth the lbb andlaa polytypes and can be en-visioned as a three-component interstratification of IbbSp, Ibb Ch, and Iaa Ch. Shallow samples (<2000 m) aredominated by interstratified Ibb Sp and Ibb Ch, and thedeep samples are characterizedby interstratified laa Chandlbb Ch with minor IbD Sp. To our knowledge, this isthe only documented occurrence of interstratified Chpolytypes.

A model for the diagenetic Sp-Ch reactions is as fol-lows. After the early diagenetic transformation of odiniteto Ibb Sp-Ch, Ibb Sp layers transformtolbb Ch layers asgrade increases with progressive sedimentation and buri-

(Fe, Al, Mg, OH) "->

Frcunr 9. Schematic depiction of the possible lbb Sp-to-IaaCh transformation.

al. At depths greater than 2000 m, some of the Irb layerstransform to Iaa layerc, creating crystals of randomly in-terstratified Ibb and Iaa polytypes. The Ibb'to-Iaa trans-formation occurs simultaneously with the Sp-to-Ch re-action, implying thar lbb Sp layers transform to laa Chlayers at depths >2000 m. The consistent chemical com-position of Sp-Ch, as determined by EDX, indicates thatthe reactions of IbD Sp to Ibb Ch and lbb Sp to Iaa Chare isochemical and thus are polymorphic transitions(Jiang et al. 1992; Xu and Veblin 1993). This finding issimilar to those of Jiang et aI. (1992) and Xu and Veblin(1993), who observed isochemical 7-14 A transforma-tions, and Brown and Bailey (1962), who documentedisochemical Ibb -to-Iaa transformations.

An alternate interpretation of the gradual transforma-tion of Sp layers to Ch layers with increasing burial isthat all the Sp-Ch formed simultaneously during a ba-sinwide fluid-flow event, a phenomenon that has beenobserved in illite-smectite of the U.S. Gulf Coast Basin(Morton 1985). During punctuated or episodic diagene-sis, all minerals form simultaneously with mixed-layerproportions that reflect current burial temperatures, thusproducing more evolved minerals at depth, whereas dur-ing burial diagenesis, individual layers or crystallites con-tinually transform as temperatures increase with progres-sive burial, also producing more evolved minerals atdepth. We conclude that the Sp-Ch reactions occurredprogressively as burial depth increased on the basis ofpetrographic evidence indicating that (1) the Sp-Ch graincoatings formed during very early diagenesis at shallow

Ibbch

rbbSp

IbbSp

Ibbch

IbbSp

Ibb Sp-Ch Ibb -> laa lbb-laaSp-At

222

burial depth (Thomsen 1982), and (2) the precursor toSp-Ch appears to be odinite, which transforms to Ch orSp-Ch within a few meters of the surface (Bailey 1988a).

The coincidence ofincreasirglaa content and decreas-ing kaolinite content raises the question ofwhether or notkaolinite dissolution was related to formation of the Iaapolytype. Boles and Franks (1979) found a similar rela-tionship between kaolinite content and bulk Ch contentin sandstone of the Wilcox Formation and suggested thatkaolinite dissolves and combines with Fe and Mg (pro-vided by the illitization reaction) to form Ch. In the Tus-caloosa Formation, the consistent nature of Sp-Ch crys-tallite thickness and the lack of systematic variation inthe Sp-Ch to anatase ratio suggests that no new Sp-Chforms upon kaolinite dissolution. However, the dissolu-tion of kaolinite may reflect increases in either pore-waterpH (Garrels and Christ 1965) or temperature that causedthe transformation of Ibb layers to Iaalayers.

Alford (1983) and Beskin (1984) identified the IL Chpolytype in deeply buried Tuscaloosa Formation sand-stones, as did Ryan and Reynolds (1993). However, afterusing differential diffraction (Schulze 1982) to remove il-lite and anatase interference (Fig. 2), we now believe thatthe presence of illite and anatase peaks with d valuessimilar to those of IID Ch polytype peaks (e.g., 201,202,and 202) led to the IIb misinterpretation. We also mag-netically separated some of the deeply buried samples(Tellier et al. 1988; Walker and Thompson 1990) andfound no evidence ofthe IIb Ch polytype, indicating thatthe differential diffraction technique did not preferential-ly discriminate against IIb polytype identification.

Implications of the mineralogic reactions

The trend of increasing Iaa component in mixed-layerIbb-Iaa may represent a prograde progression from pureIbb to pure Iaa through an intermediate mixed-layerlbb-Ica series. This trend is especially noteworthy when con-sidered in conjunction with the occurrence ofthe laa Chat <250'C in Appalachian shales (Weaver et al. 1984)and the prograde transformation of lbb to Ica in ironformations (Brown and Bailey 1962). lt is interesting tonote that there appears to be almost as many documentedexamples of \bb-to-Iaa transformations as the expectedIb-to-llb transformation (Hillier 1994; Sp6tl et al. 1994).Perhaps the transition from type-I to type-Il Ch is inhib-ited by compositional differences between Fe-rich Ibb andthe more Mg-rich IIb, or by the necessary rotation ofhydroxide layers, or both (Bailey 1988b; Walker 1993).

The lack of variation in mean crystallite thickness im-plies that the reaction ofSp layers to Ch layers does notinvolve crystal-growth mechanisms typical of other com-mon clay minerals. In illites, kaolins, and some chlorites,crystallite size commonly increases with increasing erade,reflecting simultaneous dissolution of fine-grained crys-tals and crystallization on coarser existing crystals (Ku-bler 1964; Eberl et al. 1990; Jahren l99l; Warr and Rice1994; Hillier 1994; Spdtl et al. 1994).If Sp-Ch crysralsdissolved and reprecipitated during the Sp-to-Ch reac-

RYAN AND REYNOLDS: MIXED-LAYER SERPENTINE.CHLORITE

tion, crystallite size should increase with increasing grade(Eberl et al. I 990), but we have not observed such a trend.Sp-Ch crystals are probably not destroyed during the re-action, and the simultaneous Sp-to-Ch and Ibb-to-Iaatransformations probably involve disaggregation and re-organization of individual layers on a layer-by-layer ba-sis, a process that has been observed in transformationsofbiotite to Ch (Eggleton and Banfield 1985; Li and Pea-cor 1993) and berthierine to Ch (Ahn and Peacor 1985;Xu and Veblin 1993) and inferred for illite-smectite (Al-taner and Ylagin 1993) and Sp-Ch (Ryan et al. 1994). Byreacting in this manner, individual layers transform andthe integrity of the existing crystal is preserved (Fig. 9).

There are distinct differences between transformationsin Sp-Ch and illite. The Ibb-to-laa polytype transfor-mation occurs simultaneously with the reaction of Sp toCh, but in mica, nearly all smectite reacts to illite beforethe lMopolytype transformsto 2M, (Maxwell and Hower1967; Moe et al. I 996). The reaction of lMo to 2M, llliteis a dissolution-reprecipitation reaction (Reynolds I 96 5)that produces physical mixtures of lMo and 2M,, but nointerstratificationof lMoand2M,layerc (Moe et al. 1996),whereas the lbb-to-laa reaction produces randomly in-terstratified polytype layers (Ibb-Iaa) and does not resultin crystal growth. Also, the illitization reaction requireschanges in tetrahedral and interlayer composition (Howeret al. 197 6), but reactions in Sp-Ch appear to be isochem-ical (Ahn and Peacor 1985; Jiang et al. 19921, Xu andVeblin 1993).

Relationship between Sp-Ch and porosity

Our findings support the work of many others (e.g.,Heald and Anderegg 1960; Pittman and Lumsden 1968;Thomsen 1982; Ehrenberg 1993), who have indicated thatthe formation of chloritic grain coatings during early dia-genesis is responsible for the preservation of abnormallyhigh primary porosity. By forming during early diagene-sis, the grain coatings covered potential nucleation sitesand inhibited the growth of quartz cement. However, thegrowth of quartz cement may be influenced by grain size.At depths >3600 m, Tuscaloosa Formation sandstonesthat contain fine-grained quartz and detrital clay are wellcemented by secondary quartz, whereas sandstones withlittle or no quartz cement are medium grained and con-tain no evidence ofdetrital clay. The fine-grained detritalquafiz, which is more reactive than medium-grainedquartz, may have dissolved and reprecipitated as cementin situ, or the high concentration ofpotential quartz nu-cleation sites in the fine-grained rocks may have fosteredgrowth of quartz cement from solution. There is littleevidence of dissolution of carbonate or framework grains,implying that most of the porosity is primary, althoughkaolinite dissolution may have created some secondaryporosity.

Potential for Sp-Ch geothermometry

The reaction of Sp layers to Ch layers and IDb layersto laa layers raises the possibility of Sp-Ch and lbb-Iaa

geothermometry in sandstone. Grain-coating Ch is com-mon in sandstones, especially those deposited in near-shore and shallow marine environments. Other occur-rences of Sp-Ch in nearshore marine sandstones (Ahnand Peacor 1985; Walker and Thompson 1990; Hillierand Velde 1992; Hillier 1994) may be intermediates insimilar mixed-layer series, indicating that the use of sucha geothermometer could be widely applied. Sp-Ch is morecommon than illile-smectite in some sandstones, and aSp-Ch geothermometer could be more useful at higherdiagenetic grades than illite-smectite. The reactions of Sp-to-Ch and Ibb-to-Iaa polytype appear to be isochemical,suggesting that temperature is the dominant control onthe reactions.

Scatter in the Sp-to-Ch trend may indicate that tem-perature is not the only controlling factor on Sp-Ch struc-ture, but some ofthe scatter appears to be related to rocktype. The three samples that are most anomalous withrespect to Sp content and polytype composition are finegrained, silty, and well cemented, whereas all other sam-ples analyzed in this study are medium grained and fri-able. The presence of detrital clays in the anomalous sam-ples may have created a different environment (e.g.,chemistry, pressure, permeability) in the fine-grainedsandstones and siltstones than that present in the coarsersandstones. However, even discounting the three fine-grained samples, there is still scatter in the Sp-to-Ch andIbb-to-Iaa trends, suggesting that factors other than tem-perature are involved in the Sp-to-Ch and IDb-to-Iaa rc-actions. Sp layers may form metastably in the field of Ch(Jiang et al. 1992; Bailey et al. 1995), and both Ibb andIaa may form metastably in the field of IIb Ch (Bailey1988b), indicating that the reactions observed here maybe partially controlled by kinetic factors.

Thus, although it appears that temperature is the dom-inant control on the Sp-Ch and \bb-laa reactions, strictinterpretation of temperature on the basis of Sp:Ch ratio,lbb:Iaa ratio, or both must be approached with caution.The proportions ofSp layers in Sp-Ch andlbb layers inIbb-Iaa would probably be best used as estimates of rel-ative diagenetic grade rather than absolute temperature.The reader is referred to articles by Walker (1993), deCaritat et al. (1993), and Jiang et al. (1994) for discus-sions on some of the problems inherent in clay mineralgeothermometry.

Acx-uowr,nocMENTs

Constructive reviews by Donald Peacor and Jeffrey Walker greatly im-proved this paper, as did an earlier review by Jeffrey Moe. We appreciatethe support ofDavid R. Pevear ofExxon Production Research Company,Houston, Texas, who provided core samples, thin sections, and financialsupport, and also Paula Hansley, U.S. Geological Survey, Denver, Colo-rado, for providing core samples from 6214 to 6216 m depth. Joel Blumand Melody Brown of the Dartmouth College Eanh Sciences Departmentassisted with SEM analyses, as did Charles Daghlian and Louisa Howardof the Dartmouth College Electron Microscope lab. Libby Stern is thankedfor helpful discussions and for sending files to P C.R in Montana. Pol-ished sections used in SEM and EDX analyses were prepared by MineralOptics Laboratory of Wilder, Vermont.

223

RrrnnnNcrs crrED

Ahn, J.H., and Peacor, D.R. (1985) Transmission electron microscopicstudy ofdiagenetic chlorite in GulfCoast argillaceous sediments. Clays

and Clay Minerals, 33,228-237.Alford, E.V. (1983) Compositional variations of authigenic chlorites in

the Tuscaloosa Formation, Upper Cretaceous ofthe GulfCoast Basin.

M.S. thesis, University of New Orleans, I-ouisiana.Altaner, S.P., and Ylagin, R.F. (l 993) Interlayer-by-interlayer dissolution:

A new mechanism for smectite illitization. Proglams and Abstracts,30th Annual Meeting of the Clay Minerals Society, San Diego, Cali-fornia, 88.

Bailey, S.W. (1969) Polytfpism of trioctahedral l:l layer silicates. Claysand Clay Minerals, 17, 355-37l.

-(1988a) Odinite, a new dioctahedral-trioctahedral Fe3*-rich l:1

clay mineral. Clay Minerals, 23,237-247.-(1988b) Chlorites: Structures and crystal chemistry. In Mineral-

ogical Society of America Reviews in Mineralogy, 19,347-403.Bailey, S.W., Banfield, J.F., Barker, w.W., and Katchan, G. (1995) Do-

zyite, a l:1 regular interstratification of serpentine and chlorite. Amer-ican Mineralogisr, 80, 65-77.

Berg, R.R., and Cook, B.C. (1968) Petrology and origin ofl-ower Tus-

caloosa sandstones, Mallalieu Field, Lincoln County, Mississippi.Transactions-Gulf Coast Association of Geological Societies, 10, 207 -

2r3 .Beskin, E.A. (1984) Compositional variations ofauthigenic chlorites in

the Tuscaloosa Formation, Upper Cretaceous, ofthe GulfCoast Basin.Programs and Abstracts, 21st Annual Meeting of the Clay Minerals

Society, Baton Rouge, Louisiana, 25.Bhattacharyya, D.P. (1983) Origin ofberthierine in ironstones. Clays and

Clay Minerals, 31. 173-182.Boles, J.R., and Franks, S.G. (1979) Clay diagenesis in Wilcox sandstones

of southwest Texas: Implications of smectite diagenesis on sandstonecementation. Journal of Sedimentary Petrology, 49, 55-70.

Brindley, G.W. (1961) Chlorite minerals. In G. Brown, Ed., The X-rayidentification and crystal structures of clay minerals,p.242-296.Ilfin'eralogical Society, Irndon, U.K.

Brindley, G.W., and Gillery, F.H. (1954) Mixed-layer kaolinite-chlorite.

Clays and Clay Minerals, 2nd National Conference, 349-353.Brown, B.E., and Bailey, S.W. (1962) Chlorite polytypism: I' Regular and

semi-random oneJayer structures. American Mineralogist, 47, 819-850.

Burton, J.H., Krinsley, D.H., and Pye, K. (1987) Authigenesis of kaolinite

and chlorite in Texas Gulf Coast sediments. Clays and Clay Minerals,35,291-296.

Dean, R.S. (1983) Authigenic trioctahedral clay minerals coating Clear-water Formation sand grains at Cold Iake, Alberta, Canada. Programsand Abstracts, 20th Annual Meeting of the Clay Minerals Society, Buf-falo. New York. 79.

de Caritat, P., Hutcheon, I., and Walshe, J.L. (1993) Chlorite geother-

mometry: A review. Clays and Clay Minerals, 41,219-239.Drever, J.I. (1973) The preparation oforiented clay mineral specimens

for X-ray diffraction analysis by a filter-membrane peel technique.American Mineralogist, 58, 553-554.

Eberl, D.D., Srodon, J., kalik, M., Taylor, B.E., and Peterman, Z.E.(1990) Ostwald ripening of clays and metamorphic minerals. Science,248, 474-417.

Eggleton, R.A., and Banfield, J.F. (1985) The alteration ofgranitic biotiteto chlorite. American Mineralogist, 7 0, 902-9 10.

Ehrenberg" S.N. (1993) Preservation ofanomalously high porosity in deeplyburied sandstones by grain-coating chlorite: Examples from the Nor-wegian Continental Shelf. American Association of Petroleum Geolo-gists Bulletin, 77, 1260-1286.

Foster, M.D. (1962) Interpretation of the composition and classificationofthe chlorites. U.S. Geological Survey ProfessionalPaper 414-4, l-33'

Frey, M. (1970) The step from diagenesis to metamorphism in pelitic

rocks during alpine orogenesis. Sedimentology, 15, 261-279.Garrels, R.M., and Christ, C.L. (1965) Solutions, minerals, and equilibria,

450 p. Harper and Row, New York.Genuise, J.J., and McBride, E.F. (1990) SEM petrography and XRD anal-

ysis of authigenic chlorite in Oligocene and Cretaceous sandstones of

RYAN AND REYNOLDS: MIXED-LAYER SERPENTINE-CHLORITE

1 1 A

the Texas/Louisiana Gulfcoast. Programs and Abstracts, 27th AnnualMeeting of the Clay Minerals Society, Columbia, Missouri, 5l.

Hamlin, K.H., and Cameron, C.P. (1987) Sandstone petrology and dia-genesis of Lower Tuscaloosa Formation reservoirs in the Mccomb andLittle Creek field areas, southwest Mississippi. Transactions-Gulf CoastAssociaticrn of Geological Societies, 37 , 95-104.

Hayes, J.B. (1970) Polytypism ofchlorite in sedimentary rocks. Clays andClay Minerals, 18, 285-306.

Heald, M.T., and Anderegg, R C. (1960) Diferential cementation in theTuscarora sandstooe. Journal of Sedimentary Petrology, 30, 568-577 .

Hillier, S. (1993) Origin, diagenesis, and mineralogy ofchlorite mineralsin Devonian lacustrine mudrocks, Orcadian Basin, Scotland. Clays andClay Minerals, 41, 240-259.

-(1994) PoreJining chlorites in siliciclastic reservoir sandstones:Electron microscope, SEM, and XRD data, and implications for theirorigin. Clay Minerals, 29, 665-67 9.

Hillier, S., and Velde, B. ( 1992) Chlorite interstratified with a 7 A mineral:An example from offshore Norway and possible implications for theinterpretation of the composition of diagenetic chlorites. Clay Minerals,27,475-486.

Hogg, M.D. (1988) Newtonia Field: A model for mid-dip lnwer Tusca-loosa retrograde deltaic sedimentation. Transactions-Gulf Coast As-sociation of Geological Societies, 38, 461-47 l.

Hower, J., Eslinger, 8.V., Hower, M.8., and Perry, E.A. (1976) Mecha-nism of burial metamorphism of aryillaceous sediments. GeologicalSociety of America Bulletin, 87,'125-737.

Iijima, A., and Matsumoto, R. (1982) Berthierine and chamosite in coalmeasures ofJapan. Clays and Clay Minerals, 30,264-274

Jahren, J S ( I 99 1) Evidence of Ostwald ripening related recrystallizationofdiagenetic chlorites from reservoir rocks offshore Norway. Clay Min-erals. 26. 169-178

Jiang, W.-T, Peacor, D.R., and Slack, J.F. (1992) Microstructures, mixed-layering, and polymorphism of chlorite and retrograde berthierine inthe Kidd Creek massive sulfide deposit, Ontario. Clays and Clay Min-erals. 40. 501-5 14.

Jiang, W.-T., Peacor, D.R., and Buseck, P.R. (1994) Chlorite geother-mometry? Contamination and apparent octahedral vacancies. Claysand Clay Minerals, 42,593-605.

Karpova, G.Y. (1969) Clay mineral post-sedimentary ranks in terrigenousrocks. Sedimentology, 13, 5-20.

Kubler, B. (1964) ks aryiles, indicateurs de metamorphisme. Revue, In-stitut Francais du Petrole, 19, 1093-l I 12.

Li, G., and Peacor, D.R. (1993) Sulfides precipitated in chlorite alteredfrom detrital biotite: A mechanism for local sulfate reduction? Ab-stracts with Progams, 1051h Annual Meeting of the Geological Societyof America, Boston, Massachusetts, 25, 146-147.

Maxwell, D.T., and Hower, J. (1967) High-grade diagenesis and low-grademetamorphism of illite in the Precambrian Belt Series. American Min-erafogist, 52, 843-856.

Moe, J.A., Ryan, P.C., Elliott, W.C., and Reynolds, R.C., Jr. (1996) Themineralogy, chemistry, and K-Ar dating of a K-bentonite in the Pro-terozoic Belt Supergroup, Montana. Journal of Sedimentary Research,rn press.

Moore, D.M, and Reynolds, R.C., Jr. (1989) X-ray diffraction and theidentification and analysis ofclay minerals, 332 p. Oxford UniversityPress, New York.

Moore, D M., and Hughes, R.E. (1990) The clay mineralogy of two Mis-sissippian sandstone reservoirs in the Illinois Basin. Programs and Ab-stracts, 27th Annual Meeting of the Clay Minerals Society, Columbia,Missouri, 90.

-(1991) Characteristics of chlorite interlayered with a 7A mineralas found in sandstone reservoirs (abs.). Programs and Abstracts, ClayMinerals Society, 28, 115.

Morton, J.P. (1 985) Rb-Sr evidence for punctuated illite/smectite diagen-esis in the Oligocene Frio Formation, Texas Gulf Coast. GeologicalSociety ofAmerica Bulletin, 96, ll4-122.

Nelson, B.W., and Roy, R. (1958) Synthesis of the chlorites and theirstructural and chemical constitution. American Mineralogist, 43, 707 -'125.

Odin, G S. (1985) La "verdine," facies granulaire vert, marin et cotierdistinct de la glauconie: Distribution actuelle et composition. CompteRendus de I'Acad6mie des Sciences de Paris, 301, I(2), 105-118.

RYAN AND REYNOLDS: MIXED-LAYER SERPENTINE-CHLORITE

-(1990) Clay mineral formation at the continent-ocean boundary:The verdine facies. Clay Minerals, 25,477-483.

Odin, G.S., Bailey, S.W., Amouric, M., Frolich F., and Waychunas, G.S.(1988) Mineralogy of the verdine facies In Developments in Sedimen-tology,45, 159-206.

Pittman, E.D., and Lumsden, D N. (1968) Relationship between chlontecoatings on quartz grains and porosity, Spiro Sand, Oklahoma. Journalof Sedimentary Petrology, 38, 668-670.

Reynolds, R.C., Jr. (1965) Geochemical behaviour of boron during themetamorphism of carbonate rocks. Geochimica et Cosmochimica Acta,2 9 , l l 0 l - l l l 3 .

-(1985) NEWMOD: A computer program for the calculation ofone-dimensional diffraction profiles of clays. Hanover, New Hamp-shire.

- (1988) Mixed layer chlorite minerals. In S.W. Bailey, Ed., HydrousPhyllosilicates (Exclusive of Micas), Reviews in Mineralogy, 19, 601-629. Mineralogical Society of America, Washington, D.C.

-(1992) X-ray difraction studies of illite/smectite from rocks, <1pm randomly oriented powders, and <l pm oriented powder aggre-gates: The absence oflaboratory-induced artifacts. Clays and Clay Min-erals. 40. 387-398.

-(1993) WILDFIRE-A computer prog.ram for the calculation ofthree-dimensional diffraction profiles of clays. Hanover, New Hamp-shire.

Reynolds, R.C., Jr., DiStefano, M.P., and Lahann, R.W. (1992) Random-ly interstratified serpentine/chlorite: Its detection and quantification bypowder X-ray diffraction methods. Clays and Clay Minerals, 40,262-267

Ruotsala, A.P., Pfluger, C.E., and Dow, G.M. (1964) Iron-rich serpentineand chamosite from Ely, Minnesota. American Mineralogist, 49, 993-1 0 0 1 .

Ryan, P.C., and Thompson, G.R. (1991) Structural and compositionalvariations in low-grade metamorphic illite and chlorite from the BeltSupergroup, Montana and Idaho. Programs and Abstracts, 28th An-nual Meeting of the Clay Minerals Society, Houston, Texas, 134.

Ryan, P.C., and Reynolds, R.C., Jr. (1993) Diagenesis of chlorite in sand-stone of the Gulf Coast Tuscaloosa Formation. Programs and Ab-stracts, 30th Annual Meeting of the Clay Minerals Society, San Diego,Califomia, 87.

Ryan, P.C., Chamberlain, C.P., Conrad, M.E., and Reynolds, R.C., Jr'(1994) Stable isotopic implications for diagenetic mineral reactions inTuscaloosa Formation sandstone. Programs and Abstracts, 3lst AnnualMeeting of the Clay Minerals Society, Saskatoon, Saskatchewan, 78.

Schulze, D.G. (1982) The identification of soil iron oxide minerals bydifferential X-ray diffraction. Soil Science Society of America Journal,45,437-440

Shirozu, H., and Bailey, S.W. (1965) Chlorite polytypism: III. Crystalstructure of an orthohexagonal iron chlorite. American Mineralogist,s0, 868-885.

Slack, J.F., Jiang, W.-T., Peacor, D.R., and Okita, P.M. (1992) Hydro-thermal and metamorphic berthierine from the Kidd Creek volcano-genic massive sulfide deposit, Timmins, Ontario. Canadian Mineralo-glst ,30, l l27-1142.

Sptttl, C., Houseknecht, D.W., and Longstaffe, F.J. (1994) Authigenicchlorites in sandstones as indicators of high-temperature diagenesis,Arkoma Foreland Basin, USA. Journal of Sedimentary Research, 464,553-566.

Stancliffe, R.J., and Adams, E.R. (1986) lower Tuscaloosa fluvial channelstyles at Liberty Field, Amite County, Mississippi. Transactions-GulfCoast Association of Geological Societies, 36, 305-3 I 3.

Tellier, K.E., Hluchy, M.M., Walker, J.R., and Reynolds, R.C., Jr. (1988)

Application of high-gradient magnetic separation (HGMS) to structualand compositional studies of clay minerals. Journal of SedimentaryPetrology, 58, 7 61-7 63.

Thomsen, A. (1982) Preservation of porosity in the Deep Woodbine/Tuscaloosa Trend, Louisiana. Joumal of Petroleum Technology, 34,1156-1 162

Velde, B. (1989) Phyllosilicate formation in berthierine peloids and ironoolites. In Geological Society Special Publication, 46,3-8.

Velde, B., Raoult, J.-F., and kikine, M. (1974) Metamorphosed berthier-ine pellets in mid-Cretaceous rocks from north-eastern Algeria. Journalof Sedimentary Petrology, 44, 127 5-1280.

Walker, J.R. (1989) Polytlpism of chlorite in very low gmde metamorphicrocks American Mineralogist, 7 4, 7 38-7 43.

-(1993) Chlorite polytlpe geothermometry, Clays and Clay Min-erals. 41. 260-267.

Walker, J.R., and Thompson, G.R. (1990) Structural variations in illiteand chlorite in a diagenetic sequence from the Imperial Valley, Cali-fornia. Clays and Clay Minerals, 38, 315-321.

Wan, L.N., and Rice, A.H.N. (1994) Interlaboratory standardization andcalibration of clay mineral crystallinity and crystallite size data. Journalof Metamorphic Geology, 12, l4l-152.

Weaver, C.E., Highsmith, P.B., and Wampler, J.M., Eds. (1984) Chlorite.Developments in Petrology, 10, 99-139.

Whittle, C.K. (1986) Comparison of sedimentary chlorite compositionsby X-ray diffraction and analytical TEM. Clay Minerals, 21,937-947.

225

Wiygul, G.J., and Young, L.M. (1987) A subsurface study of the LowerTuscaloosa Formation at Olive Field, Pike and Amity Counties, Mis-sissippi. Transactions-Gulf Coast Association of Geological Societies,37,295-302.

Xu, H., and Veblen, D.R. (1993) Periodic and disordered interstratifica-tion and other microstructures in the chlorite-berthierine series. Ab-stracts with Programs, l05th Annual Meeting of the Geological Societyof America. Boston. Massachusetts, 25, 146.

MlNuscnrrr REcETVED Jurv 8, 1994

Meruscnrm AccEprED Ocronen 2. 1995

RYAN AND REYNOLDS: MIXED.LAYER SERPENTINE-CHLORITE