American Mineralogist, Volume 70, pages 202JM, 1985

A review of the todorokitebuserite problem: implications to the mineralogy of marine manganesenodules: discussion

Ruoorr GrovnNort Laboratory of Electron Microscopy Institute of Inorganic Chemistry, University of Berne Freiestrasse 3, P.O.B. 140, CH-3000 Berne 9, Switzerland

Abstract

The author agrees that is a valid species. However, previous comparisons of todorokite and psilomelane (romanechite)are criticized. In addition, the lower-valence manganeseions in psilomelane cannot be Mn2+ but must be Mn3* ions, which explains the extraordinary coordination of the position M2 in the psilomelane structure as Jahn-Teller distortion. Lattice imaging in and psilomelane on synthetic material is discussed and confirmed as showing the tunnels with Ba2* ions as earlier published. In the case of todorokite, however, the published and accessibledata are not convincing and contradict X-ray difraction and other data. Experimental evidence from the synthesis of hollandite and psilomelane also makes it unlikely that the proposed large tunnels in the previous speculationson the todorokite structure can be stable. The content of these tunnels is also in doubt. It is emphasizedthat the reduction of MnzOrr ' 5H2O to yMnOOH has been checked by X-ray diffraction methods totally independentfrom electron microscopy.

Introduction and following Levinson's paper this sameconstituent was Todorokite is a manganeseoxide phasenamed after the more generally acceptedas being todorokite. These rela- Todoroki mine (Hokkaido, Japan) that was described by tions are listed here in order to clarify the history of given Yoshimura (1934). It was later found in Cuba (e.g., todorokite and l0A-manganate as by Burns et al. Straczeket al., 1960).More recently,Perseil and Giovan- (1983) in a slightly but distinctly diferent manner. In oli (1982)have described a samplefrom Ambollas (East- addition. there are several incorrect statementsin their ern Pyr6n6es,France). article which I correct here, while fully agreeing with Independently of Yoshimura, Wadsley (1950) synthe- much of the remainder of their text. sized a phase fresh Mn(OH)2 by rapid oxidation of a Todorokite is a valid mineral species suspensionusing molecular 02. This phase showed re- markable ion exchangeproperties. We entirely agree with Burns et al. (1983)that todoro- Again independently of both Wadsley and Yoshimura, kite is a valid mineral species. Unfortunately, all our Feitknecht and Marti (1945) synthesized a samples of todorokite from various origins investigated oxide phasewhich later turned out to be the dehydration up to 1978were not monomineralic. From 1978to 1980 product of Wadsley'sphase. we have, however, received two samplesfrom Japanand Buser and GrUtter (1956) were the first to realize the one from the French Eastern Pyr6n6eswhich turned out close relation between Wadsley's phase and one of the to be almost monomineralic (for the latter sample cf. constituents occurring in deep sea manganesenodules. Perseil and Giovanoli, 1982).I, therefore, withdraw my Levinson (1960) pointed out that Wadsley's phase pro- previous opinion (Giovanoli and Btirki, 1975)that todoro- ducesa powder diffraction pattern which shows a striking kite is not a valid mineral species. resemblanceto that of todorokite. Up to that time period The comparison of todorokite with psilomelane the metal scavenging constituent in marine manganese nodules was usually called l0A-manganatel after Buser, Burns et al. (1983)devote much of their report to the inferred structure of todorokite as first published (Burns and Burns, 1977) based on morphological similarities. t Feitknecht (1945) had, following Sarkar and Dhar (1921), fibrous as is psilome- unfortunately introduced the chemicalname "" which Some occurrencesof todorokite are (1977) created confusion with the mineral name manganite for y lane (romanechite). Burns and Burns make the MnOOH. Buser followed this nomenclature.Feitknecht correct- incorrect statement that psilomelane has Mn2+ ions in ed this in Giovanoli et al.. 1970. particular structure positions. This has been emphasized 0003-oM)vEs/oI 02-0202$02.00 202 GIOVANOLI : TODOROKITE-BUSERITE PROBLEM : DISCUSSION 203

by Burns in a discussionremark (Burns, 1979).Picking up todorokiteplates and eliminatesthe possibilitythat this particularpoint madein 1978and printed in 1979we it is a mass of needlesthat appear morphologically have synthesizedpsilomelane and verified that no dis- as a plate." cernibleamount of Mn2+ can be found in its strucrure. The contentof suchlarge tunnels as proposedby Burns The lower-valencemanganese ions are Mn3*. This ex- and Burns has also never been established.All known plains the coordination of these sites as due to Jahn- tunnels in Mn-O phases contain tightly fitting large Teller deformation.Mn2+ ions are much too large to fit cations;e.g., Ba2*, K* and others.Until thesequestions into thesepositions while Mn3+ fits very nicely (Giovan- areanswered, and as long as no diffractionevidence is put oli and Balmer, 1983).It is also noteworthy that in the forward, the structure drawingsas shown in Figures 2 and case of cryptomelaneand hollandite the valenceshave 3 in Burns et al. (1983) must be regarded as being routinely beenindicated as Mn2* and Mna* until Post et speculative. al. (1982)showed that the presenceof Mn2* ions in these compoundsis most unlikely. We have, as early as 1969, The reduction of 7A-manganate to 7-MnOOH rejected the presenceof Mn2* as most improbablein On p. 975 ff. Burns et al. discuss at great length a syntheticgroutite a-MnOOH (Giovanoliand Leuenberger, reaction we have undertaken(Giovanoli et al., l97l) in 1969).Glaunsinger et al. (1979)settled an old controversy the context of other topotactical transformations.No by confirming that all Mn ions in 7-MnOOH also are simulationof naturalgeochemical environments has been trivalent. intended(as Burns et al. suggest);only the processas exclusivelytaking place under a particularredox poten- Lattice imaging tial is implied. Burns et al. claim that the structural interpretation of this reaction was wrong. They refer to When Turner and Buseck(1979) produced high resolu- Figure I (in Burns et al., 1983)where we have indeed tion electron micrographsof natural samplesof psilome- drawnthe left-handside incorrectly. The figureis alsonot lane it was obviousthat the contrastlineages correspond to scale.Since its purposewas to show that startingand to the Ba2+in the tunnelsof this structure.Occasionally end products have virtually identical Mn-Mn distances intergrowth with hollandite was observed. We could only in one structure direction, this error in a purely reproducethese results on a large variety of synthetic schematicdrawing-although deplorable-is of little im- psilomelaneand hollandite phasesand also found fre- portancein the context. quent intergrowth (Giovanoli and Balmer, 1983). We Burns et al. (1983),however, go still further in their have, however, also found similar transmissionmicro- deductionsfrom the error in Figure l. They questionthat scopeimages in 7A-phyllomanganate(Giovanoli and Ar- the reductionproduct was I completelyreject rhenius,1985) and the high resolutionelectron micro- 7-MnOOH. this interpretationsince it has been showirbeyond doubt graphs are therefore not unambiguous. Turner and through X-ray diffraction patterns (in Giovanoli et al., Busecklater investigatedtodorokite and also interpreted 1971)that is the end product which can easily the lineagesfound as tunnels. They inferred, therefore, 7-MnOOH be reproduced. that todorokite is a tektomanganatewith large tunnels, an assumption that fitted nicely into the speculations of Conclusions Burns and Burns. It is possiblethat this is indeedthe case (l) and that-except for hollandite and psilomelane tun- Naturally occurring todorokite is a valid mineral nels-larger tunnelsalso exist, at leastas isolateddefects. specles. (2) Potterand Rossman(1979) who have thoroughlystudied The structure of todorokite is unknown and Figures the infrared spectra of Mn-O phasesdiscuss this point at 2 and 3 in Burns et al. (1983)are speculative. (3) great length on the grounds of WeissenbergX-ray diffrac- Psilomelanecontains Mn3* ions, not Mn2*. (4) tion patterns they have also taken of todorokite. Their Larger tunnelsthan those in psilomelanecould not text which is only partly quoted by Burns et al. (1983) be producedhitherto and the possibilityof their existence says: as a structuralunit is remote.However, they may occur ". . . No chain or channelmanganese oxides com- as isolatedlattice defects. posed of such large units are known to exist, References although isolated quadruple chains have been ob- servedin hollanditein electronmicroscope images Burns,R. G. and Burns,V. M. (1977)The mineralogyand crystalchemistry polyme- (Turner and Buseck, 1979). It appears that for of deep-seamanganese nodules, a tallic resourceof the twenty-firstcentury. In: Mineralogy: highly polymerized structures a layer structure is Towardsthe 2lst Century(Centenary Volume, Mineralogical preferred. . .]Weissenberg [. single-crystalX-ray Society,London, 49-67) PhilosophicalTransactions of the diffraction indicates that the sample is ordered in RoyalSociety of London,A286,283-301. the same plane as the plates but disordered in Burns,R. G. (1979)Discussion remark in: ClaudeLalou, Ed., La planes perpendicularto them. This suggeststhat it gendsedes nodules de mangandse.Colloque International du consistsof a randomsuperposition of single-crystal CentreNational de la RechercheScientifique no. 289,p. 314. GI OVAN OLI : TODO RO KITE_B US ERITE PRO B LEM : DI SC US S I ON

Burns,R. G., Burns,V. M., andStockman, H. (1983)A review Preparationand Magnetic Propertiesof Manganite. Joumal of of the todorokite-buseriteproblem: Implications to the miner- Solid StateChemistry, 29, ll7-120. alogy of marine manganesenodules. American Mineralogist, Levinson,A. A. (1960)Second occurence oftodorokite' Ameri- 68, 972-980. can Mineralogist, 45, 802-807. Buser,W. and Griitter, A. (1956)Ueber die Natur der Mangan- Perseil,E. A. and Giovanoli,R. (1982)Etude comparativede la knollen. SchweizerischeMineralogisch-Petrographische Mit- todorokite d'Ambollas (Pyr6n€esOrientales), des manganates teilungen,36,49-62. d l0A rencontr6sdans les nodules polymetalliquesdes oc6ans Feitknecht, W. and Marti, W. (1945) Ueber Manganite und et des produitsde synthEse.Comptes Rendus Acad6mie des ki.instlichenBraunstein. Helvetica Chimica Acta, 28, 149-157. SciencesParis.294. S€rie II. 199J02. Giovanoli,R. and Leuenberger,U. (1969).Ueberdie Oxydation Post, J. E., von Dreele, R. B., and Buseck,P. R. (1982) von Manganoxidhydroxid.Helvetica Chimica Acta, 52, 2333- Symmetry and cation displacementsin : Structure 2347. refinements of hollandite, and priderite. Acta Giovanoli,R., Stah[, E., and Feitknecht,W. (1970)Ueber die Crystallographica,B38, 1056-1065. Oxidhydroxidedes vierwertigen Mangans mit Schichtengitter. Potter, R. M. and Rossman, G. R. (1979) The tetravalent l. Mitteilung:Natriummangan(ll,III)manganat(IV). Helvetica manganeseoxides: Identification, hydration, and structural ChimicaActa, 53, 209-220. relationships by infrared spectroscopy. American Mineral- Giovanoli,R., W. FeitknechtW., and Fischer,F. (1971)Ueber ogist,64,ll99-1218. Oxidhydroxide des vierwertigen Mangansmit Schichtengitter. Straczek,J. A., Horen, A., Ross, M., and Warshaw, C' M. 3. Mitteilung:Reduktion von Mangan(lll)manganat(lV)mit (1960) Studies of the manganeseoxides. IV. Todorokite. Zimtalkohol.Helvetica Chimica Acta. 54. 1112-1124. American Mineralogist, 45, ll7 4-l 181. Giovanoli,R., Biirki, P., Giuffredi,M., and Stumm,W. (1975) Turner, S. and Buseck, P. R. (1979) tunnel Layei Structured ManganeseOxide Hydroxides. IV. The structures and their intergrowths. Science, 203,456-458. Buserite Group; Structure Stabilizationby Transition Ele- Wadsley,A. D. (1950)A hydrous manganeseoxide with ex- ments.Chimia. 29. 517-520. changeproperties. Journal ofthe American Chemical Society, Giovanoli,R. and Balmer, B. (1983)Darstellung und Reaktion 72, 1782-1784. von Psilomelan(Romandchite) Ba2Mn15O36 . 4H2O. Chimia, Yoshimura, T. (1934)Todorokite, a new manganesemineral 77 4)4_4)7 from the Todoroki Mine, Hokkaido, Japan. Journal of the Giovanoli,R. and Arrhenius,G. ( 1985)Morphology and Compo- Faculty of Science,Hokkaido University, Sapporo,Series 4, sition of the ManganeseNodules: Mineralogy. In P. Halbach Yol.2,289*297. and G. Friedrich, Eds., The ManganeseNodule Belt of the Pacific.Friedrich Enke PublishingHouse; Stuttgart, in press. Manuscriptreceived, May 27, 1983; Glaunsinger,W. S., Horowitz, H. S., and Longo, J. M. (1979) acceptedfor publication,September 4, 1984.