C anadian. MineralogistVol. 18, pp. 529-533 (1980)

FIOLTANDITE AND LITHIOPHORITE IN BURROWS OF LATE CRETACEOUS AGE,NORTHEASTERN SOUTH GAROLINA

MILAN J. PAVICH, HBLAINE W. MARKEWICH eNo EDWARD J. DWORNIKU.S. Geological Survey, Reston, Va. 22092, U.S.A.

ABSTRAcT

Hollandite BaMn3O1" and lithiophorite (Al,Li)MnOz(OH)z con$titute tle cement of burrows ofSantonian Age in northeastern South Carolina. Theburrows were formed at the interface betweengravels of the Upper Cretaceous Middendorf For-mation and the underlying saprolite of felsic meta-volcanic rocks of tle Piedmont physioeraphicprovince. The hollandite and lithiophorite are con-sidered to be primarv minerals formed while theburrowing organisms inhabited these tubes. Theabsence of these minerals in ironstone in the sapro-lite and at the saprolite-Middendorf contact sup-ports this conclusion.

Keywords: hollandite, lithiophorite, burrows, UpperCretaceous,

Soutrenr

La hollandite BaMnsOm et la lithiophorite (Al,Li)MnOg(OH)2 constituent le ciment de tubulesd'8ge santonien dans le nord-est de la Caroline dusud. Ces tubules se pr6sentent au contact des gra-viers de la formation Middendorf (Crdtac6 sup6-rieur) avec la saprolite sous-jacente form6e auxd6pens des roches m6tavolcaniques felsiques de laprovince physiographique du Piedmont. Hollanditeet lithiophorite sont i consid6rer ici comme min6-raux primaires, datant de l'occupation de ces tubu-les par les organismes fopisseurs, conclusionqu'6taye leur absence des s6diments ferrugineux dela saprolite et de la zone du contact Middendorf-sa'prolite.

(Iraduit par la R6daction)

Mots-clas: hollandite, lithiophorite. tubules. Cr6-tac6 sup6rieur.

INrnonuctroN

During field reconnaissance of basal CoastalPlain sediments near the Fall Line in north-eastern South Carolina, five burrows were foundat the interface between the Coastal Plain sedi-ments and the underlying weathered metavol-canic rocks of the Piedmont province. Scanningelectron micrographs (SEM) and powder-diffrac-tion patterns of sediment from two burrowsreveal the presence of major amounts of hol-

landite and lesser amounts of lithiophorite asconstituents of the cemented burrow filling.This article will describe the geological settingin which the hollandite and lithiophorite occurand present a hypothesis regarding their origin.

Gnorocrcar SBrrrNc

The burrows discussed herein were found ina borrow pit on the north side of Route 348,9.6 km northwest of Cheraw, South Carolina(Fig. l). At this locality, the Middendorf For-mation of Santonian age (R. A. Christopher,verbal comm. 1979) rests unconformably onsaprolite derived from felsic metavolcanic rocksof the Carolina slate belt of the Piedmont(Fig. 2). There, the Middendorf is an inter-bedded sequence of cross-bedded feldspathicsand and gravel and laminated kaolinitic clay.Although no microflora or microfauna are pres-ent in the Middendorf beds at this locality, amarine or marginal-marine origin for the bur-rows is suggested ( 1) by the reported presence offoraminifera in Middendorf beds in an outcropless than lO km to the southeast (Siple er a/.1956), and (2) by the presence of the burrowsdiscussed here, at the contact between the gravelof the Middendorf Formation and the under-lying saprolite.

The saprolite is weathered from metavolcanicrocks of early Paleozoic age, and ranges inthickness from 2 to 5 m. Results of an X-ray-fluorescence analysis of the saprolite are givenin Table l.

Burrows observed at this locality extend fromthe Middendorf-saprolite contact into the sapro-lite and are filled wih sand of the Middendorfthat has been cemented by manganese oxideinto thin- to thick-walled tubes (Fig. 3). Theburrows are brown, coherent, cylindrical, noded,straight and nonbranching and do not have striaeor linings (Fig. ).

MrNrnntocv oF BuRRows

SEM photomicrographs indicate the presence

529

530

Frc. l. Index map showing location of outcropwith burrows at the contact between sands ofthe Middendorf Formation of Late Cretaceousaee (Km) and saprolite of the felsic metavol-canic rocks of the Carolina slate belt of thePiedmont province. Triassic rocks are denoted byTr, the Yorktown Formation (Pliocene) by Ty.

Frc. 2. A schematic drawing of the outcrop show-ing the irregular contact between the saprolite(b) and the highly jointed felsic metavolcanicrocks (a). Irnses of irons:rne (c) follow thejoints in the saprolite and are discontinuous alongthe contact between the saprolite (b) and theoverlying sediments of the Upper CretaceousMiddendorf Formation. Burrows (d) are withinthe saprolite and have linings of Fe and Mnoxides consisting of the minerals hollandite andlithiophorite. The Middendorf Formation (Km),here composed of cross-bedded gravels (e),sands (g) and interbeds of laminated clay (f),is considered to be of marine origin.

THE CANADIAN MINERALOGIST

TABLE'I. X-MY.FLUORESCENCE ANALYSIS*

Contact lronstone 2mSaprollte Burreu imnstone below contactLB-t90 79UCF27 7E$CF25 79WCF26

Nelght percent

oxldes:

si02 69.741203 13.3Fe203 1.4Cao .28

90 1.2 ' t't102 .631,4n0 .05Pzos .0B

TOTAL 86.7

Traceelmnts:

Ba 450

Cu

Zr

AsPb

Rb

Sr

ZrCo

30,000

500

>200*100

100100500

tr

*Analyst, R.,lohnson, U,5, Gsloglcal Survey*Thls concentratlon ms hlgher than any of the standards available

for cdnDarison.

of hollandite and lithiophorite as the only crys-talline components of the manganese oxidecement (Figs. 5, 6). Bystriim & Bystriim (1950)suggested the formula (Ba,K,Pb) r-"(Mn,Fe) e-,(O,OH)tu(HrO)" for hollandile. The chemicalanalysis of the burrow material (sample no.79WCF27, Table 1) shows that the major ele-ments reported by Bystriim & Bystriim (1950)are present in this sample. The energy-dis-persive X-ray (EDAX) graph (Fig. 5B) showsthe predominance of manganese and barium,and their relative percentage$, in the sample.The X-ray-diffraction patterns of the same sam-ple correspond to X-ray powder data for hol-landite compiled by Fleischer (1964). Theprincipal spacings in A and their intensities,obtained using Ni-filtered Cu radiation, are3.48 (80) , 3.12 ( l0O), 2.4O (9O) ard 2.14( 80 ) .

Wadsley (1952) gave a formula of (Al,Li)MnO, (OH), for lithiophorite; Fleischer & Faust(1963) noted that later analyses showed lithiumto be only a minor constituent. The EDAX scan(Fig. 68) of the crystals shown in Figure 6A in-dicates the presence of Mn, Al and Co in thecrystal strucfure. X-ray powder data cor.respond

47.0

'17 .0

28.4

4 . 0

47 .3

460100'100

fr

46

3482

22

t501 7

182

l 0

l-" cottwtum

At - . f

b

)"N2irs- r ' 1

HOLLANDITE AND LITHIOPHORITE IN BURROWS 531

9.50 (50), 4.71 (lOO),2.36 (7O) and 1.88 (70).Hollandite was identified as the major phase

in two burrows, and lithiophorite as a minorphase in one burrow. Neither hollandite norlithiophorite was indicated from powder-diffrac-tion patterns of the saprolite or sediment of theMiddendorf Formation. Ironstone samples fromthe Middendorf-saprolite contact (sample no.79WCF25, Table 1) and from 2 m below thecontact in the saprolite (sample no. 79WCF26)1re predominantly iron oxyhydroxide (e.g.,hematite and goethite) and contain low per-centages . of manganese. No manganese oxideswere identified in the powder-diffraction pat-terns of the two ironstones. Barium and oihertrase elements are present in the ironstone sam-ples in generally lower concentrations than inthe hollandite cement of the burrows. Bariumconcentrations in the saprolite (LB-190, Tablel) and in the ironstones are one to two ordersof magnitude lower than in the hollandite.

DrscussroN

Hollandite and lithiophorite are commonminerals in supergene manganese oxide depositsand have been found in many locations in thesouthern Appalachians (Fleischer 1964, Flei-scher & Faust 1963, Pierce 1944). These min-erals have not been reported, however, fromsediments of the Atlantic Coastal plain or fromburrows of marine origin. The apparent absenceof these manganese oxides from marine sedi-ments of the Coastal Plain is curious in tharmanganese-nodule formation has been shown tobe an active process at the sediment-water in-terface in a variety of marine settings (Glasby& Read 1976). Reduction below the interfacemobilizes manganese, which diffuses upwardand reprecipitates when oxidized at the sedi-ment surface (Lynn & Bonatti 1965).

The stratigraphic relationships between bur-rows and substrate and the chemical data sug-gest that the manganese oxides formed as prim-ary cementing minerals in the bunow linings.Several observations support this hypothesis.( 1) The manganese oxides are present onlywithin the burrows in this outcrop. The absenceof manganese oxides from the enclosing sedi-ment and the saprolite provides evidence thatthese minerals were not formed here by second-ary supergene solutions as they were elsewheroin the southeastern United States (Pierce lg44).In addition, the supergene ironstone that formedat and below the sediment-saprolite contactshows no evidence of manganese oxide forma-

Frc. 3. Closeup of bunows at the contact betweenthe Middendorf Formation (above), and thesaprolitized felsic metavolcanic rocks. The bur-rows, of assumed marine origin, extend fromthe Middendorf-saprolite contact into the sapro-lite and are filled with sand of the Middendlorf.Same location as Figure 2.

Frc. 4. Cylindrical, straight burrow with an en-larged spherical termination. Surface containsclosely spaced vemrcae. Burrow is filled withsand that has been cemented by manganese oxides.Included in the manganese minerals are hollanditeand lithiophorite. Burrow from locality describedin Figure 2. Bar - I cm.

I Jhgse of lithiophorire crystals compiled byFleischer

.& Faust (1963). The principal spac-rngs (in A) and intensities for this sampte are

532 THE CANADIAN MINBRALOGIST

Frc. 5. (a) SEM photograph of hollandite crystals (2000x). (b) EDAX scan of hollandite crystalshowing Mn and Ba peaks.

Frc. 6. (a) SEM photograph of lithiophorite crystals (2000x). (b) EDAX scan of lithiophorite crystalshowing Mn, Al and Co peaks.

tion. The manganese oxides and iron oxyhy-droxides appear to have formed at distinctlydifferent times and in different hydrologicalenvironments. (2) Aller & Yingst (1978, p. 247)demonstrated that the innermost linings of poly-chaete burrows in Barnstable Harbor, Massa-chusetts, become ". . .enriched to a small extentin Fe and Mn oxides which accumulate bydiffusion from the more rcduced portions ofthe burrow wall . . . ." In the burrow lining,these oxides act as scavengers of barium, cobalt,zinc, copper and other metals, as has beenshown in various aqueous envitonments, e.9.,

in bogs in Sweden (Ljunggren 1955)' in aridenvironments like desert surfaces (Engel & Sharpl95S) and in dendrites (Potter & Rossmanl97S). Aller & Yingst (1978) postulated, more'over, that these metals tend to concentratethrough time in the burrow lining as it thickens.The burrows they studied contain oxide innerlayers surrounded by older Fe and Mn sulfideouter layers. The burrows found in this studyprobably were formed either in an environmentsimilar to that Barnstable Harbor or in a moreopen marine envfuonment, and the burrows werealso probably loci for accumulation of manga-

nese oxide and associated metals. (3) A post_depositional weathering environment would notbe favorable for barium adsorption by manga-nese oxide. Solutions involved in near-surfaceweathering generally are acidic. At pH lessthan 7, barium adsorption by manganese oxideis weak (Hem 1964). A marine environment ata pH of about 7.5 would provide more favor-able conditions for the uptake of barium andother metals by manganese oxides.

These lines of evidence all support the hypo-thesis of a marine origin for the manganeseoxide minerals in the burrow cement. The pro-cess of crystallization of hollandite and lithio-phorite may have postdated burial of the bur-rows.

CoNcr,usroNs

This and other occurrences of hollandite andlithiophorite indicate that these minerals canform in a variety of surficial or sedimentaryenvironments. This occurrence also indicatesthat burrows have been sites of manganese oxideand trace-metal accumulation in past sedi-mentary environments. As shown by Aller &Yingst (1978), burrowing organisms can havea profound effect on the distribution of ele-ments at the sediment-water interface.

ACKNOWLEDGEMENTS

We thank several of our colleagues at theU.S. Geological Survey for their help in thisstudy. Michael Fleischer and J. P. Owensmade several suggestions concerning manu-script preparation and reviewed the final article.R. A. Christopher helped in preparing the firstdraft of the manuscript. L. M. Force found theburrows. Robert Johnson and R. B. Finkelmansupplied chemical and powder-diffraction data.Deborah Dwornik photographed the cementedburrow for Figure 4.

RenenrNcss

ALLER, R.C. & YrNcsr, J.Y. (1978): Biogeochemis-try of tube dwellings: a study of the sedentarypolychaete Amphitrite ornata (I*idy). l. Mar.Resour. 36, 2Ol-254,

533

Bvsrndu, A. & Bvsrntiu, A.M. (1950): Thecrystal structure of hollandite, tle related man-ganese oxide minerals, and 6y-MnOi, Acta Cryst.3, t46-154.

ENcnr., C.G. & SHARp, R.P. (1958): Chemical dataon desert varnish. Geol. Soc. Amer. Bull. 69,487-518.

FLETScHER, M. (1964): Manganese oxide minerals.Vm. Hollandite. /a Advancing Frontiers inGeology and Geophysics, a Volume in Honourof M.S. Krishnan (A.P. Subramaniam & S.Balakrishna, eds.). Indian Geophys. Union,Hyderabad. India.

& Feusr, G.T. (1963): Studies on man-ganese oxide minerals. VII. Lithiophorite.Schweiz. Mineral. Peftog. Mitt. 4r, 197-216.

Glessy, G.P. & Rreo, AJ. (1976): Deep-sea man-ganese nodules. /n Handbook of Strata-Boundand Stratiform Ore Deposits. T. Att,U, Fe, Mn,Hg, Sb, W and P Deposits (K.H. Wolf, ed.).Elsevier, Amsterdam.

Hru, J.D. (1964): Deposition and solution of man-ganese oxides. U,S. Geol. Surv. Water Sup. Pap,1667-8.

L.tuNccnrN, P. (1955): Geochemistry and radio-activity of some Mn and Fe bog areas, Geol.Fiiren. Stockholm Fdrh. 77,33-44.

LvNN, D.C. & BoNarn, E. (1965): Mobility ofnranganese in diagenesis of deep-sea sediments.Mqr. Geol. 3, 457-474.

Ptrncr, W.G. (1944): Cobalt-bearing manganesedeposils of Alabama, Georgia, and Tennessee,U.S. Geol. Surv. Bull. 940-1.

Potrrn, R.M. & Rossvrer, G.R. (1978): Deserlvarnish: mineralogy of the ferromanganeseoxides. Anter. Geophys. Utrion Tratrs, 58, 1242(abstr.).

Srelr, G.E., BRowN, P.M. & LsGneNp, H.E.(1956): Stratigraphic significance of foramini-fera from an outcrop of the Tuscaloosa Forma-tion at Cheraw, South Carolina. Geol. Soc.Amer. Bull. 67, 1757-1758 (abstr.).

Weosr-rv, A.D. (1952): The structure of lithio-phorite, (Al,Li)MnOr(OH)r. Acta Cryst. 6,676-680.

Received lune 1980, revised manuscript acceptedSeptember 1980.

HOLLANDITE AND LITHIOPI{ORITE IN BURROWS