American Mineralogisl, Volume 60, pages 335-337, 1975 On the Color Centersof lron in Amethystand SyntheticQuartz: A Discussion GrnHlno LnHuaNN Institutfur PhysikalischeChemie der UniuersitdtMunster, D-44 Milnster. Germanv Two recentpapers published by Cohen and Hassan and Chentsova(1959), both cited but misinterpreted (1974)in this journal presentresults of optical absorp- by Hassanand Cohen with respectto the site occupa- tion, biaxiality,opticaland thermal bleachingstudies. tion of Fe3+. The conclusions drawn from their results about the (2) The absorption spectrumof Fe'+ in green por- structuresof the various centersinvolved completely tions of synthetic quattz proues (not merely suggests) neglect recent results obtained by a number ofauthors this ion to occupy interstitial sites of distorted oc- (Matarrese, Wells, and Peterson,19691, Dennen and tahedral symmetry (1u): The bands at 10100 and '' 5E Puckett, 1972:Lehmann, 1967, 1970, 197l; Lehmann 13800cm are due to the'T" - transition split by and Bambauer, 1973). Well establishedresults of the lower symmetry (Lehmann, 1967). Cohen and electron paramagnetic measurements (Een) and Hassan take a similar band presentin this region in ligand field interpretations of optical spectra are amethyst as evidencefor the presenceofsuch a center either termed unlikely or not considered(and cited) in amethyst. However, there are marked differences at all. As a result the proposed models partly con- in position, shape, and dichroism of the two band tradict these results (coordination and site occupa- systems,and for an identical center the ligand field tion of Fe3+and Fe2+)and partly do not (at leastin spectrum must also be identical, not just similar' A our view) represent physically meaningful pictures detailedligand field interpretationassigns the band in uZz 5E (proposed structures of amethyst color centers). amethystto the - transition of Fea+on silicon While the results presentedby Cohen and Hassan sites (Lehmann, 1967). support the models basedon these Epn and optical (3) Contrary to the statementof Hassanand Cohen data (Lehmann and Bambauer, 1973), the con- the band at about 6.2 eY or 51000cm-l has been clusionsdrawn disturb rather than clarify the picture. found earlier in synthetic, iron-doped quartz together Although it is not possibleto repeatall the evidence with a shoulder near 41000 cm-' (Lehmann, 1970)' in detail (which the reader will be able to extract from The position of the first band as well as the weaker the above cited literature himself), we want to stress ligand field spectrumare consistentonly with Fe3+in in- the following points: terstitial sitesof distorted tetrahedralsymmetry (1.). (1) Cohen and Hassan attribute the intenseband While the ErR spectrumof this centerwould also be at about 5.6 eV or 46000 cm-' present in amethyst consistentwith a substitutional Fe3+ without short and under the rhombohedral facesin syntheticiron- range charge compensation (Matarrese et al, 1969), doped qvarlz to Fe3+ in interstitial sites of approxi- this assignmenthas been confirmed by the charge mately octahedralsymmetry. Comparative studiesof transfer spectra of Fe3+ in alkali silicate glasses optical spectra of Fe3+ with oxygen ions as nearest (Lehmann and Steinmann, to be published). The neighbors (Lehmann, 1970) clearly establishedthat assignmentsto substitutional and interstitial centers octahedral symmetry places the first charge trans- given by Cohen and Hassanfor the bandsat 6.2 and fer band of Fe3+at or below 41000cm-'or 5.1 eV 5.6 eV are seento be reversedfrom the ones estab- (decreasingwith distortion from the ideal 06 sym- lished by correlation of optical spectra and predic- metry). On the other hand Fe3+ in tetrahedral lat- tions of ligand field theory. tice sites of both AIPO4 and GaPOr shows a charge (a) The Epn work of Barry et al (1965)clearly es- transferband at 46000cm-1 or 5.6eV. Thus the band tablishedsubstitutional Fe8+ (S' centers)as a precur- at this position in quartz must be due to Fe3+on lat- sor of the amethystcolor center.This was confirmed tice sites (S centers). This conclusion is in complete in later EpR work and in addition it was found that agreementwith Epn work of Barry, McNamara, and interstitial Fe3+(1. centers)also acts as a precursor in Moore (1965) and with earlier findings of Tsinober natural and synthetic amethyst, both centers being 335 336 GERHARD LEHMANN necessaryfor the generationof amethystcolor centers accompanying the precipitation of FezOgparticles (Lehmann and Moore, 1966).The maximum concen- that occurs in highly colored amethystsin the same tration of amethyst color centers produced by temperaturerange (Lehmann, l97l) prove migration irradiation is limited by the centerpresent in smaller of substitutional Fe8+. We agree with Cohen and concentration. This fact is consistent only with a Hassan that this result is highly surprising for a model in which one of them acts as the donor, the trivalent ion bound by partly covalent M-O bonds, other one as the acceptorof the transferredelectron: but neverthelessit has to be accepted as an ex- perimentally proven fact. If the equilibration occurs uia jumps from one lattice site to a neighboring one, Fe3+(S,)* Fe3+(1) 'heat Fea+(S)* Fe,+(Ia) #i3T5or light cronor acceptor each singlediffusion jump will result in a net change of the occupation ratio of the three sites.In this way (The reverseprocess with S, centers as acceptors and diffusion coefficientsas small as 10-18cmz/sec are 1ocenters as donors would be extremelyunlikely as a sufficientto accomplishan equilibration.If, however, result of ionizing irradiation). the equilibration occurs uiapreferential conversion of This model is in complete agreement with the Fe3+from the more highly substitutedsites (i.e., by predictionsof ligand field theory. The visible colora- an overall decreaseof the S, center concentration), tion of amethystis dominated by Fe'+ whereasFe2+ longer diffusion paths and consequently larger diffu- just contributes a weak band near 6000 cm-' and a sion coefficientsare necessary.tWe are at present in- charge transfer band above 50000 cm-1 (Lehmann, vestigating these processesto determine the relative 1967).This model is (exceptfor the coordination of importance of thesetwo mechanisms.In any casethe the interstitial ion) practically equivalent to the one brown particles of FerOr are formed mainly from proposed by Cohen and Hassan. However, in con- substitutionalFe3+, not from interstitialions aloneas trast to the smoky quartz center and other color postulated by Cohen and Hassan, in amethyst. centersassociated with main group metal ions, a hole In conclusionwe may say that the method of elec- associatedwith a transition metal ion is (at leastin its tron paramagnetic resonance (Ern) gave final electronicground state)highly localizedon this tran- answersfor a number of very specificproblems. In sition metal ion. Therefore the only reasonable contrast, ligand field theory can only serve to decide physical description for centers of this kind seemsto whether a model is consistentwith the optical spec- us to be one with the valency increasedby one for the trum, but cannot definitelyexclude additional models electrondonor and decreasedby one for a transition that were not specificallyconsidered. However, in metal ion as electron acceptor. The existenceof these combination with the results of Ern, it places our ions then must result in ligand field and charge model far beyond the mere likelihood of one among transfer spectra specific for both valency and coor- severalequally plausiblealternatives. This model has dination of these ions. The positions and intensities been completely confirmed in a recent investigation of absorption bands are predictable within rather of natural amethystsfrom different locations (H. D. narrow limits, but do not necessarilybear any resem- Stock and G. Lehmann, to be published). blanceto those of hole or electron centersassociated ' Note added in proof: Recent results (H. D. Stock and G. Leh- with main group ions as implied by Schlesingerand mann, to be published) show the second mechanism to dominate. Cohen (1966)in an earlierpublication. The occurrenceof amethyststhat rapidly bleachin References daylight (H. Harder, private communication) in- Blnnv, T. I., P. McNltrlln.l, AND W. J. MooRE (1965) dicates that other defectsbesides interstitial Fe3+may Paramagnetic resonance and optical properties of amethyst. ,I. act as electron acceptors in amethyst. Chem. Phys. 42, 2599-2606. (1974) (5) The work of Barry et al (1965)clearly demon- CoHrN. A. J.. eNo F. HlsslN Ferrous and ferric ions in synthetic oFqvartz and natural amethyst. Am. Mineral. 59, strated the optical biaxiality of amethyst to be 719-728. causedby the unequal substitutionof the threelattice DENNEN,W. H., luo A. M. Pucrrm (1972) On the chemistry sites by Fe3+(and furnished an explanation for the and color of amethyst. Can. Mineral. ll, 448-456. different ciegreeof biaxiality in different crystals).The HasslN, F., AND A. J. ConrN (1974) Biaxial color centers in thermal equilibration among these sites at tem- amethyst quartz.,4t1t Mineral. 59, 709-7 18. LruueNw, G. (1967) Farbzentren des Eisensals Ursache der Farbe peratures near 600oC has also been established by von Amethyst. Z. Naturf. 22s, 2080-2085. theseauthors and was confirmed in later work. This - (1970) Ligand field and charge transfer spectra ofFe(lII)-O equilibration as well asthe changesin the Ern spectra complexes. Z Physik Chem. (Frankfurt), 72, 279-297. COLOR CENTERS