Você está na página 1de 27

Accepted Manuscript

3+
Jahn-Teller distortion of Mn -occupied octahedra in red beryl from Utah indicated by
optical spectroscopy

Jana Fridrichová, Peter Bačík, Andreas Ertl, Manfred Wildner, Július Dekan, Marcel
Miglierini
PII: S0022-2860(17)31279-6
DOI: 10.1016/j.molstruc.2017.09.081
Reference: MOLSTR 24330

To appear in: Journal of Molecular Structure

Received Date: 1 August 2017


Revised Date: 21 September 2017
Accepted Date: 21 September 2017

Please cite this article as: J. Fridrichová, P. Bačík, A. Ertl, M. Wildner, Jú. Dekan, M. Miglierini, Jahn-
3+
Teller distortion of Mn -occupied octahedra in red beryl from Utah indicated by optical spectroscopy,
Journal of Molecular Structure (2017), doi: 10.1016/j.molstruc.2017.09.081.

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to
our customers we are providing this early version of the manuscript. The manuscript will undergo
copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please
note that during the production process errors may be discovered which could affect the content, and all
legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT
1 Jahn-Teller distortion of Mn3+-occupied octahedra in red beryl from
2 Utah indicated by optical spectroscopy

3 Jana Fridrichová a,*, Peter Bačíka, Andreas Ertlb, c, Manfred Wildner c, Július
4 Dekand, Marcel Miglierinid

PT
5
a
6 Comenius University in Bratislava, Faculty of Natural Sciences, Department of Mineralogy
7 and Petrology, Ilkovičova 6, 842 15 Bratislava, Slovakia

RI
b
8 Mineralogisch-Petrographische Abt., Naturhistorisches Museum, Burgring 7, 1010 Wien,
9 Austria

SC
c
10 Institut für Mineralogie und Kristallographie, Geozentrum, Universität Wien, Althanstrasse
11 14, 1090 Wien, Austria

U
d
12 Slovak University of Technology in Bratislava, Faculty of Electrical Engineering and
13 Information Technology, Institute of Nuclear and Physical Engineering, Ilkovičova 3, 812 19
AN
14 Bratislava, Slovak Republic
15
M

16 *Corresponding author. Tel: +421 2 60296 294


17 E-mail address: fridrichova@fns.uniba.sk
D

18
TE

19
C EP
AC

1
ACCEPTED MANUSCRIPT
20 Abstract

21 Red beryl from Utah is chemically homogeneous and contains only Fe <0.163, Mn
22 <0.018, and Mg <0.016 apfu. Channel sites contain only up to Cs 0.011, K 0.009, Rb 0.004,
23 and Na 0.004 apfu. This suggests only very slight tetrahedral (Cs,K,Rb)Li□-1Be-1 substitution,
24 octahedral Na(Fe2+,Mg)□-1Al-1 substitution can be excluded. Fe and Mn are trivalent as
25 documented by Mössbauer spectroscopy and optical absorption spectroscopy. Red beryl

PT
26 optimized formula is
27 ~[(Cs,Rb,K)0.020.98]Σ1.001.00(Al1.79Fe3+0.16Mn3+0.02Ti4+0.02Mg0.01)Σ2.00Be3(Si6O18). Location

RI
28 of Mn3+ was estimated to the octahedral Al3+ site, other choices are improbable due to the
29 bond-length requirements. No Mn3+-induced Jahn-Teller structural distortion was detected

SC
30 due to site symmetry restrictions and small Mn3+ content. However, optical spectroscopy
31 shows broad band at ~7190 cm-1 assigned to the excited level of the spin-allowed pseudo-

U
32 tetragonal split E ground state of elongated six-fold Mn3+ coordination. Crystal field
calculations indicate that the local Mn3+ environment complies well with crystal chemical
AN
33
34 expectations for Jahn-Teller distorted Mn3+O6 octahedra.
M

35 Keywords: red beryl; manganese; Jahn-Teller effect; structure refinement; optical absorption
36 spectroscopy; crystal field calculation
D

37
TE

38 1. Introduction

39 Beryl hexagonal structure was first determined by [1] and later refined by [2] and [3]. It
EP

40 consists of AlO6 octahedra and BeO4, SiO4 tetrahedra. Rings of SiO4 tetrahedra form channels
41 parallel to the c axis and are connected with Al octahedral and Be tetrahedral sites (Fig. 1). In
C

42 channels, there are two sites available – 2a site at 0 0 ¼ usually occupied by H2O in two
AC

43 orientations and larger alkali cations (Cs, Rb, and K) and 2b site at 0 0 0 which is occupied by
44 Na [4]. Alkali cations at channel sites balance charge deficiency due to heterovalent
45 substitutions at octahedral and tetrahedral sites e.g., [5-6]. The AlO6 octahedron is angularly
46 distorted and Al is often substituted by trivalent Fe, Cr, V and Mn as well as divalent Fe, Mn
47 and Mg cations. The substitution of larger cations results in increase of volume, Me–O bond
48 length and distortion of octahedron [6]. However, all Me–O bonds in AlO6 octahedron with
49 trigonal symmetry are of same length, no bond-length distortions have been observed e.g.,
50 [4,6-9].

2
ACCEPTED MANUSCRIPT
51 The colour of red beryl is produced by the Mn3+ [5]. Trivalent manganese in octahedral
52 ligand field has high-spin electronic configuration [Ar]3d4 : (t2g)3(eg)1 and a low-spin
53 configuration [Ar]3d4 : (t2g)4. Silicate minerals prefer high-spin configuration, although both
54 of these configurations are degenerated and according to Jahn-Tellerʼs theory [10], any non-
55 linear molecule with orbital degeneracy undergoes distortion and symmetry reduction to
56 remove the degeneracy. In the high spin configuration of trivalent manganese [Ar]3d4 :

PT
57 (t2g)3(eg)1, three electrons are located in t2g orbitals and fourth one must be placed in dz2 or dx2-
2
58 y orbital. If the four ligands arranged in a plane XY shift towards to a central Mn3+ ion, the eg
electron prefer dz2 orbital, because it is less repulsed by oxygen ligands than in dx2-y2 orbitals.

RI
59
60 Thus octahedral symmetry reduces to tetragonal dipyramide extended in the direction of c

SC
61 axis [11]. In octahedral complexes, the Jahn–Teller distortion is small when t2g orbitals are
62 involved; this distortion becomes larger when there is uneven electronic occupancy in the eg
63 orbitals [12].

U
64 The aim of this investigation is to consider the influence of major chromophore – Mn3+
AN
65 on octahedral distortion and the potential impact of the Jahn-Teller effect. Beryl as a silicate
66 mineral prefers high spin configuration and thus the distortion can be larger when there is
uneven electronic occupancy in the eg orbital, although Mn3+ content is not represented in
M

67
68 such a large extent. We used single-crystal X-ray diffraction for structure investigation that
D

69 helps us to check for symmetry reduction and any subsequent distortion of the octahedral site.
70 We applied spectroscopic methods, such as optical absorption spectroscopy to identify the
TE

71 chromophores responsible for the colouration of red beryl and their potential impact on
72 polyhedral Jahn-Teller distortion. On the other hand, we are expecting presence of Fe.
EP

73 Trivalent iron will not cause octahedral distortion, but Fe2+ has electronic configuration [Ar]
74 3d6 which can undergo Jahn-Teller distortion and thus Mössbauer spectroscopy was applied to
determine iron valence. Finally, we performed crystal field calculations to quantify the extent
C

75
76 of such local distortions.
AC

77 2. Material and methods

78 Beryl composition of the studied sample (Table 1) was established with a CAMECA
79 SX100 electron microprobe in wavelength-dispersive mode at the State Geological Institute of
80 Dionýz Štúr, Bratislava, under the following conditions: accelerating voltage 15 kV, beam
81 current 20 nA, and beam diameter 5 µm. The samples were analyzed with the following
82 standards: wollastonite (Si Kα, Ca Kα), TiO2 (Ti Kα), Al2O3 (Al Kα), pure Cr (Cr Kα), fayalite
83 (Fe Kα), rhodonite (Mn Kα), MgO (Mg Kα), pure Ni (Ni Kα), pure Zn (Zn Kα), albite (Na

3
ACCEPTED MANUSCRIPT
84 Kα), orthoclase (K Kα), Rb2ZnSi5O12 glass (Rb Lα), pollucite (Cs Lα). Detection limits of the
85 measured elements varied between 0.01 and 0.05 wt. %. The beryl crystal-chemical formula
86 was calculated on the basis of 8 octahedral and tetrahedral cations per formula unit; Be was
87 assumed with 3.00 apfu (atoms per formula unit).
57
88 Room temperature Fe Mössbauer spectrometry was performed in transmission
57
89 geometry, and the standard constant-acceleration spectrometer was equipped with a Co/Rh

PT
90 radioactive source. Calibration of the velocity scale was accomplished with a 12.5 µm α-Fe
91 foil. Isomer shift values (Table 2) are quoted with respect to a room temperature Mössbauer

RI
92 spectrum of the calibration α-Fe foil.

93 Raman analyses (Table S1) were performed on the LabRAM-HR Evolution (Horiba

SC
94 Jobin-Yvon) spectrometer system with a Peltier cooled CCD detector and Olympus BX-41
95 microscope (Masaryk University, Department of Geological Sciences). Raman spectra in the

U
96 range of 50-4000 cm-1 were excited by blue diode laser (473 nm) and then processed in
97 Seasolve PeakFit 4.1.12 software. Absorption bands were fitted by a Gauss function with
AN
98 automatic background correction and Savitzky-Golay smoothing.
99 Single-crystal X-ray diffraction. After the quality of the hand selected beryl crystal
M

100 was checked, X-ray data were collected with a Bruker APEX II diffractometer equipped with
101 a CCD area detector and an Incoatec Microfocus Source IµS (30 W, multilayer mirror, Mo-
D

102 Kα). The data were integrated and corrected for Lorentz and polarization factors with an
TE

103 absorption correction by evaluation of partial multiscans. The structure was refined with
104 SHELXL97 [13] using scattering factors for neutral atoms and a beryl starting model (X-ray
105 data) from [9]. Small amounts of Cs (with some K and Rb), which were already measured by
EP

106 chemical analysis, were detected at 0 0 ¼ and subsequently refined. No H of a water molecule
107 could be detected. The electron density at the octahedral Al site was modelled by refining Al
C

108 vs. Fe (assuming full site occupancy). The refinement was performed with anisotropic
AC

109 displacement parameters for all atoms. Table S2 provides crystal data and details of the
110 structure refinement. The high-quality refinement converged at a R1(F) value of 1.47% (Table
111 S2). In Table S3, we list the atomic parameters, anisotropic and equivalent isotropic
112 displacement parameters, and in Table S4 we present selected interatomic distances.

113 Polarised optical absorption spectra (Table 3) in the near ultraviolet, visible and near
114 infrared spectral regions were measured at room temperature in the range 32000-5000 cm-1 on
115 a mirror-optics microscope IR-scopeII, attached to a Bruker IFS66v/S FTIR spectrometer
116 (University of Vienna). The spectra were acquired with the electric light vector parallel and

4
ACCEPTED MANUSCRIPT
117 perpendicular to the crystallographic c-axis on a crystal slab containing the (10 1 0) face,
118 which was prepared with a thickness of 1480 µm from a prismatic single crystal of red beryl.
119 The measuring spot was 200 µm in diameter. A quartz beam splitter and appropriate
120 combinations of light sources (Xenon or Tungsten lamp) and detectors (GaP-, Si- or Ge-
121 diodes) were used to cover the desired spectral range. Hence, each full spectrum is combined
122 from three partial spectra (32000-20000 cm-1: spectral resolution 40 cm-1, averaged from 1024

PT
123 scans; 20000-10000 cm-1: spectral resolution 20 cm-1, averaged from 1024 scans; and 10000-
124 5000 cm-1: spectral resolution 10 cm-1, averaged from 512 scans), which were aligned in

RI
125 absorbance for perfect match, if necessary. For all measurements a calcite Glan-prism was
126 used as polariser. The final polarized absorption spectra are displayed as wavenumber vs.

SC
127 linear absorption coefficient. For a detailed description of procedures and programs used for
128 the present crystal field calculations the reader is referred to the paper by [14] on Al-replacing
129 chromophores in varicoloured kyanites.

130 3. Results
U
AN
131 3.1 Chemical composition and structural refinement
M

132 The sample of beryl from Utah was a single crystal of vibrant red colour. It contained
133 black inclusions of bixbyite. Red beryl was chemically homogeneous and along with the main
D

134 elements in the formula it contained only 0.156-0.163 apfu Fe, 0.016-0.018 apfu Mn, 0.010-
135 0.016 apfu Mg (Table 1). Channel sites were vacant in vast majority, it contained only up to
TE

136 0.011 apfu Cs, 0.009 apfu K and 0.004 apfu Rb. Interestingly, Na content was very low,
137 below 0.004 apfu. This suggests that only very slight tetrahedral (Cs,K,Rb)Li□-1Be-1
EP

138 substitution influences the beryl composition and octahedral Na(Fe2+,Mg)□-1Al-1 substitution
139 can be excluded.
C

140 Moreover, the extremely low alkali content and increased Fe and Mn content indicate that
AC

141 both this cations are in trivalent form because channel sites did not contain enough positive
142 charge to balance substitution of divalent cations in octahedra. This was also supported by
143 Mössbauer spectroscopy which revealed that all Fe really is in trivalent form (Fig. 2, Table 2).
144 Mössbauer spectrum of red beryl was fitted by two doublets with isomer shift typical for Fe3+
145 in octahedral coordination. Doublet 1 displays relatively large quadrupole splitting for typical
146 octahedral Fe3+ which could be attributed to different local environment but also can result
147 from lower quality of spectrum due to relatively low Fe content for this method. Nevertheless,

5
ACCEPTED MANUSCRIPT
148 no indication of Fe2+ doublets was observed in the spectrum and consequently, the presence of
149 Fe2+ can be excluded.

150 The optimized crystal-chemical formula of red beryl, considering the single-crystal
151 structural refinement (Table S2-S4), the electron-microprobe analyses (Table 1) and
152 Mössbauer spectroscopy data (Table 2), is ~[(Cs, Rb, K)0.020.98]Σ1.00 1.00
153 (Al1.79Fe3+0.16Mn3+0.02Ti4+0.02Mg0.01)Σ2.00 Be3(Si6O18). We cannot exclude that very small

PT
154 amounts of H2O also occupy the channel site, but there was no clear evidence for such
155 occupancy neither by structural refinement nor by Raman spectroscopy and optical

RI
156 spectroscopy in NIR region (Figs. 3-4; Table S1).

SC
157 3.2 Optical spectroscopy

158 Polarized optical spectra of red beryl were measured in two orientations (Fig. 4; Table 3).

U
159 The basic assignment of the spectra is as follows: the broad band at ~7190 cm-1 is assigned to
the excited level of the spin-allowed tetragonal split 5E(D) ground state of a (presumably)
AN
160
161 elongated six-fold Mn3+ coordination. The two intense bands in the visible region (maxima at
162 ~17720 and ~18560 cm-1) are clearly attributable to the spin-allowed 5T2(D) state of
M

163 octahedral Mn3+, in a tetragonal distortion split into B2 and E levels. It is quite typical that this
164 splitting is rather small; hence the sequence of B2 and E is not definitely fixed. Within the
D

165 high energy shoulder of these two strong bands, some weak to very weak features (indicated
TE

166 by arrows in Fig. 4) at 18440, 19780, 20270 and 21080 cm-1 can be assigned to spin-forbidden
167 transitions of Mn3+ as given in Table 3. The underlying broader shoulders centred at ~20850
168 and ~21300 cm-1 might hint to a small contribution from a second type of Mn3+ site (“type
EP

169 II”); a band in the NIR region at ~11350 cm-1 might also arise from this Mn3+ “type II”. Quite
170 clear-cut is the assignment of the bands around 23370 and 26700-26990 cm-1 to characteristic
C

171 spin-forbidden levels of octahedral Fe3+ (Table 3).


AC

172

173 4. Discussion and Conclusions

174 4.1 Chemical composition of red beryl

175 Red beryls are derived from topaz rhyolites in Utah, USA; these rhyolites are enriched in
176 fluorine and also contain elements such as Li, Rb, Cs, U, Th and Be [15-16]. The chemical
177 composition of hosting rhyolite is reflected in the red-beryl composition with an increased Cs

6
ACCEPTED MANUSCRIPT
178 and Rb content. The significant presence of Li, although it is not possible to be measured by
179 EMPA, was not confirmed by SREF. However, the presence of Cs and Rb in the channel site
180 suggests that small amount of Li could substitute for Be [4].

181 The main substitution in the octahedra involves Fe3+ and Mn3+ in exchange for Al3+. The
182 presence of Fe3+ and absence of Fe2+ was confirmed by the Mössbauer spectrum. Although
183 the Fe content is relatively low and, therefore, the error of measurement is relatively high,

PT
184 there is no indication of Fe doublets with isomer shift corresponding to Fe2+. It is in
185 agreement with the structural data, which display relatively short Me–O bond length of

RI
186 1.9166 Å. Beryl with similar proportion of Al substituted dominantly by Fe2+ has Me–O bond
187 length increased up to 1.920 Å [6]. Moreover, Fe2+ would induce absorption in visible to IR

SC
188 region, typically at ~610, ~760 and ~810 nm [5, 17-18]. Similarly, the presence of Mn2+ was
189 excluded by optical absorption spectroscopy as discussed later. High oxidation state of these

U
190 cations could result from the beryl high-temperature and highly oxidative genetic
191 environment.
AN
192 Beryl displayed only small proportion of channel-sites constituents. Water can be
incorporated into structure only at the 2a site located between the stacked six-membered rings
M

193
194 together with K, Rb and Cs, whereas Na enters the 2b site in the centre of each six-membered
ring [19-21]. The chemical compositional study shows very small amounts of Na, Rb (up to
D

195
196 0.004 apfu), and red beryl is relatively H2O poor compared to other beryls e.g., [22] and
TE

197 references therein, it was not detectable with Raman or NIR spectroscopy. This can be
198 attributed to high-temperature (300-600°C, [23]) and low-pressure conditions during beryl
EP

199 formation in rhyolites. Low to atmospheric pressure caused devolatilization of rhyolite


200 magma [24] and therefore, there was not sufficient H2O pressure to incorporate it into the
201 channel site of beryl.
C
AC

202 4.2 Structural assignment of Mn in red beryl

203 A detailed analysis of the Mn3+ contribution dominating the optical absorption spectra is
204 necessarily based on its structural site in beryl. Judging from the refinement data, there are
205 four sites where 0.018 Mn apfu can be fully or partly accommodated:

206 1) the 4c octahedral Al3+ site, symmetry 32: this is the crystal-chemically obvious option;
207 since the Al-Fe occupancy refinement gives even a slightly higher averaged electron density
208 (14.27 e–) than EMPA (13.89 e–) it is very likely that the octahedron in the major Mn-host.

7
ACCEPTED MANUSCRIPT
209 2) the 6f tetrahedral Be site, symmetry 222: according to refinement tests, only a minor
210 part of the Mn content could be accommodated at this site; it may be considered for the
211 proposed minor Mn “type II”.

212 3) channel 2a site at 0 0 ¼, symmetry 622: the electron density at 2a is enough to


213 accommodate Cs and a major portion of Mn, but since there is also Rb and K (plus perhaps
214 marginal water) allocated, it may be only considered for the proposed minor Mn “type II”.

PT
215 4) channel 2b site at 0 0 0, symmetry 6/m: the electron density at 2b has to be considered

RI
216 with caution, since it can only be refined with fixed (or otherwise very high) isotropic
217 displacement parameter (besides, centric high symmetry sites tend to collect “fake density”).

SC
218 Therefore it was completely excluded in the final structure refinement. Anyway, its electron
219 density surpasses the low Na+Ca content; hence it may also be considered for the proposed
220 minor Mn “type II”.

221
U
If we choose the most obvious option and assign Mn3+ to the octahedral Al3+ site with
AN
222 symmetry 32, an obvious problem is that the 5E(D) ground state of (ideal) octahedral Mn3+
223 does not split in strict trigonal symmetry, i.e., the band at 7190 cm-1 should be absent (Fig.
M

224 5b), and static Jahn-Teller distortion is formally forbidden. Hence, we have to assume a local
225 violation of the trigonal symmetry around the Mn3+ dopant cations with different local site
D

226 symmetry. The logical choice would be to assume a local (pseudo-)tetragonal distortion to an
elongated octahedron, as expected for Jahn-Teller distorted Mn3+, with the elongated O–Mn–
TE

227
228 O axis corresponding to one of the three (trigonal equivalent) O–Al–O axis in statistical
229 variation. In such a local tetragonal case, the 5E(D) ground state splits into 5B1 (= the new
EP

230 ground state in case of elongation) and 5A1 (7190 cm-1), the excited spin-allowed octahedral
5
231 T2(D) state into 5B2 and 5E (17720 and 18560 cm-1) (in Fig. 5c shown with labels for the
C

232 remaining symmetry subgroup 2). This would be the preferred choice; however, the O–Al–O
AC

233 axes form φ-angles of ~60° and ~30° with the c-axis and with a direction perpendicular to c,
234 respectively. In turn, this means that for such a (pseudo-)tetragonal approach a ratio of
235 polarisation intensities for each individual transition of around 25:75% is expected (cos2φ
236 relation); this is quite obviously not the case, as it is clearly seen for the non-degenerate
5
237 A1(D) band at 7190 cm-1 with its strict polarisation perpendicular to the c-axis. If we ignore
238 this specific problem and assume a (pseudo-)tetragonal octahedron with the elongated O–Mn–
239 O axis parallel to c (Fig. 5d), then the Vis band polarisation could be explained

8
ACCEPTED MANUSCRIPT
240 straightforward (B1→B2 is allowed ∥c, B1→E perpendicular), but the NIR band would be
241 symmetry forbidden in any case (i.e., both B→A and A→B are forbidden).

242 A comparison of the observed polarization behaviour with selection rules for the above-
243 mentioned octahedral cases (Fig. 5a-d) shows that they are not able to explain all the details of
244 the optical spectra consistently. Hence, we also paid special attention to the Be-tetrahedron
245 (Fig. 5e), since the Jahn-Teller-affected Cu2+ is reported to partly occupy the Be site in beryl

PT
246 e.g. [25], thus bypassing the Jahn-Teller problem. Surprisingly, energy calculations for Mn3+
247 at this site yield level energies which tolerably comply with the measured data, i.e., at least in

RI
248 the rough order of the observed bands (Fig. 5e). However, as noted already above, according
249 to the refinement only a minor part (~1/4 to 1/3) of the total Mn could be allocated there;

SC
250 furthermore, as for the octahedron, the selections rules cannot explain the observed
251 polarization behaviour in a satisfying way (irrespective of the specific ground state, which is

U
252 difficult to derive in that particular case); besides, there would exist a (minor) valence
253 problem. However, the Be tetrahedron is to be favoured as host for Mn3+ “type II”.
AN
254 From the crystallographic perspective, Mn3+ assignment to the octahedral site is the most
likely because its ionic radius of 0.64 Å is only slightly bigger than that of Al3+ – 0.53 Å [26].
M

255
256 Second possible assignment to Be2+-bearing tetrahedron is less likely because it is occupied
by very small Be2+ (0.27 Å, [26]). Moreover, ionic radius of Mn3+ was not determined in the
D

257
258 tetrahedral coordination [26]. However, in Mn-doped spinel, which was presumed to contain
TE

259 tetrahedrally coordinated Mn3+, the Mn3+–O bond length was estimated at 1.88 Å [27] which
260 is significantly more than Be–O bond length of 1.6527(3) Å in red beryl.
EP

261 In contrast, assignments of Mn3+ in channel sites are unlikely due to exactly opposite
262 reason. Smaller 2b site inside the tetrahedral ring is usually occupied by Na. Sodium is larger
C

263 cation (1.12 Å, [26]) which in beryl channel forms 2.30 Å long bonds with ring oxygens [9].
AC

264 This is unrealistically long bond for Mn which as trivalent cation usually do not form
265 complexes with coordination 8 and therefore, its ionic radius for this coordination is not
266 published [26]. However, we could consider Mn2+ in 8-fold hexagonal bipyramidal
267 coordination which was later oxidized to trivalent form. Nevertheless, this is also very
268 improbable scenario, because Mn2+ in 8-fold coordination has ionic radius of 0.96 Å [26]
269 which is 0.16 Å shorter than Na and its oxidation would decrease its ionic radius even more.
270 For the comparison, in the octahedral coordination Mn2+ has ionic radius of 0.83 Å and Mn3+
271 0.645 Å [26]. Consequently, if we consider similar shrinkage of 0.2 Å in ionic radius during

9
ACCEPTED MANUSCRIPT
272 the Mn2+ to Mn3+ oxidation, ion with ionic radius of 0.76 Å is far too small to form bonds
273 with ring oxygens. The Mn assignment to the 2a channel site is even more unrealistic,
274 because this site is occupied by large K, Rb and Cs cations in 12-fold coordination [4].
275 Maximal published coordination of Mn is, as mentioned, 8-fold coordination of Mn2+ with
276 ionic radius of 0.96 Å, while ionic radii of K, Rb and Cs in 12-fold coordination are 1.64, 1.72
277 and 1.88 Å, respectively [26]. Consequently, Mn would not be able to form stable bonds at

PT
278 this site. In conclusion, we definitely could assign all Mn3+ into the octahedral site.

279 4.3 Jahn-Teller effect on MnO6 octahedron in red beryl

RI
280 From the octahedral cations analysed in red beryl from Utah, a Jahn-Teller effect could
be attributed only to d elements which include Fe3+, Mn2+, and Mn3+ (Fe2+ presence was

SC
281
282 excluded by Mössbauer spectroscopy). In the case of Fe3+ and Mn2+, electronic configuration
283 [Ar] 3d5, with five unpaired electrons in the 3d orbitals, there are two possible configurations

U
284 - a high-spin and a low-spin. If Fe3+ and Mn2+ form the low-spin configuration, theoretically
AN
285 these would be subjected to the Jahn-Teller octahedral distortion, since the low-spin t2g5eg0
286 configuration is degenerated [12]. However, the low-spin configuration occurs when repulsive
forces of ligands are stronger than pairing energy of non-axial orbitals. In the spectrochemical
M

287
288 series O2- is regarded as weaker ligand [11]. Consequently, in silicate minerals with O2- as
dominant ligand, d elements more preferably form the high-spin configuration due to the
D

289
290 weak ligand-field strength; therefore neither Mn2+, nor Fe3+ in beryl is subjected to the Jahn-
TE

291 Teller distortion. Besides, optical spectroscopy reveals no hints for the presence of manganese
292 in the divalent state.
EP

293 The Jahn-Teller effect is manifested in many minerals by a measurable distortion of


294 octahedra or tetrahedra. The positive correlation between Cu content and the tetragonal
C

295 distortion of octahedra has been found in different Cu-bearing inorganic compounds and
AC

296 minerals including spinels, garnets, gadolinites and tourmalines [28-33]; a correlation of the
297 distortion with optical spectra has been established as well [34]. Similarly, measurable Jahn-
298 Teller distortion of octahedra or tetrahedra due to Fe2+ was observed, e.g., in spinels,
299 gadolinites-hingganites, chlorites and micas [35-36,28, 37-38]. However, since no symmetry
300 reduction was detected in our beryl sample, no bond-length or unexpected bond-angle
301 distortion was observed at the Al site with trigonal site symmetry. Therefore, a Mn3+ Jahn-
302 Teller elongation of one of the O–Me–O axes cannot be detected, due to the site-symmetry
303 restrictions. Since the X-ray data were measured in triclinic symmetry, we also tried structure

10
ACCEPTED MANUSCRIPT
304 refinements in monoclinic and triclinic symmetry (thus lifting the symmetry restriction at the
305 Al site), but these low-symmetry refinements in fact confirmed the standard high symmetry of
306 beryl: the strongest differences of shortest to longest individual Al–O bond lengths were only
307 around 0.0014 Å, which is clearly within a 1-2 σ limit and therefore not significant. However,
308 it is important to keep in mind that the Mn-content is very low compared to Al and Fe at the
309 same structural site and therefore, the Mn content is too low for inducing Jahn-Teller

PT
310 distortion detectable by SREF.

311 The Jahn-Teller effect on Mn3+-bearing octahedral distortion in silicates such as

RI
312 piemontites [39-40], viridines and kanonaite [41], pyroxenes and amphiboles [42],
313 montmorillonite [43], and kyanites [14] was described from their optical spectra. In the

SC
314 literature, several papers on red beryl focused on – or at least include – optical absorption
315 spectroscopy of red beryls, mostly from the Utah localities, and/or of morganites and

U
316 sometimes of synthetic Mn-bearing beryls [44-50,23,51,5,52-53]. The spectra from Utah
317 samples found in the literature are more or less similar to our present spectra.
AN
318 However, there is one very important difference to published spectra – the NIR-region
has either not been measured by earlier authors, or not in a suitable quality to notice the rather
M

319
320 prominent band at ~7200 cm-1. This band is decisive for the spectral interpretation since it
very likely represents the excited part of the tetragonal split ground state 5E(D) of the Jahn-
D

321
322 Teller distorted Mn3+ cation. Although this band could be also due to the water presence, there
TE

323 was no evidence in Raman spectra (Fig. 3). Therefore, water contribution to this band is
324 negligible. Similarly, the weaker NIR band at ~11350 cm-1 has not been interpreted before. Its
EP

325 general interpretation is not straightforward – it might represent a Fe2+ feature only occurring
326 in red Utah beryls, or it might belong to a second, minor occupied type of Mn3+ site in the
327 beryl structure (“type II”) which is, however, structurally improbable. Since Mössbauer data
C

328 show no Fe2+ present (and a contribution of marginal Ni2+ or deoxidized Ti3+ can be ignored),
AC

329 we can attribute all spectral features to Mn3+ and Fe3+.

330 Apart from the various problems discussed above, the Jahn-Teller-elongated pseudo-
331 tetragonal octahedron with elongation ∥ to the c-axis is the most probable case for Mn3+,
332 especially since crystal field energy calculations (in a tetragonal approximation) give an
333 extraordinary good agreement of observed vs. calculated energy levels (Table 4). The crystal-
334 field and free-ions parameters corresponding to this calculation (Table 5) are within expected
335 ranges, indicating a comparatively weak crystal field strength expected for beryl (Dqcub =

11
ACCEPTED MANUSCRIPT
336 1470 cm-1 vs. e.g. 1680 cm-1 for Mn3+ in kyanite; [14]) with a typical Jahn-Teller elongation
337 of the octahedron. In fact, comparing the elongation of 13.3 % predicted from our crystal field
338 calculations (Table 5) with structural data compiled by [54] – who found average Mn3+–O
339 distances of 1.936 Å for the four short bonds and 2.237 Å for the two elongated bonds,
340 equalling an elongation of 15.5 % – the agreement with this expectation is excellent. Finally,
341 it should be mentioned that calculations for a pseudo-tetragonal compressed octahedron also

PT
342 give a very good agreement of energies, but the respective selection rules deviate still more
343 from observation than for the elongated case.

RI
344

SC
345 ACKNOWLEDGMENTS

346 This work was supported in part by Ministry of Education of Slovak Republic grant

U
347 agency under the contracts VEGA-1/0079/15 and VEGA-1/0499/16 (JF, PB) and Austrian
348 Science Fund (FWF) project no. P26903-N19 to AE.
AN
349
M

350 References

351 [1] W.L. Bragg, J. West, The structure of beryl, Proc. R. Soc. 3A (1926) 691–714.
D

352 [2] G.V. Gibbs, D.W. Breck, E.P. Meagher, Structural refinement of hydrous and anhydrous
TE

353 synthetic beryl, Be3Al2Si6O18 and emerald Be3Al1.9Cr0.1Si6O18, Lithos 1 (1968) 275–285.

354 [3] B. Morosin, Structure and thermal expansion of beryl, Acta Crystallogr. B28 (1972) 1899–
EP

355 1903.

356 [4] F.C. Hawthorne, P. Černý, The alkali-metal positions in Cs-Li beryl, Can. Mineral. 15
C

357 (1977) 414–421.


AC

358 [5] D.L. Wood, K. Nassau, The characterization of beryl and emerald by visible and infrared
359 spectroscopy, Am. Mineral. 53 (1968) 777–800.

360 [6] C. Aurisicchio, G. Fioravanti, O. Grubessi, P.F. Zanazzi, Reappraisal of the crystal
361 chemistry of beryl, Am. Mineral. 73 (1988) 826–837.

362 [7] G Artioli, R Rinaldi, K Ståhl, PF Zanazzi, Structure refinements of beryl by single-crystal
363 neutron and X-ray diffraction, Am Mineral 78 (1993) 762–768

12
ACCEPTED MANUSCRIPT
364 [8] G. Artioli, R. Rinaldi, C.C. Wilson, P.F. Zanazzi, Single-crystal pulsed neutron diffraction
365 of a highly hydrous beryl, Acta Crystallogr. B51 (1995) 733–737.

366 [9] L.A. Groat, G.R. Rossman, M.D. Dyar, D. Turner, P.M.B. Piccoli, A.J. Schultz, L.
367 Ottolini, Crystal chemistry of dark blue aquamarine from the True Blue showing,
368 Yukon Territory, Canada, Can. Mineral. 48 (2010) 597–613.

PT
369 [10] H. Jahn, E. Teller, Stability of Polyatomic Molecules in Degenerate Electronic States. I.
370 Orbital Degeneracy, Proc. R. Soc. A Mat 161 (1937) 220–235.

RI
371 [11] R.G. Burns, Mineralogical applications of crystal field theory, Cambridge University
372 Press, Cambridge (2005).

SC
373 [12] W. Li, G. Zhou, T.C. Wai Mak, Advanced Structural Inorganic Chemistry, Oxford
374 Science Publications, Oxford (2008).

U
375 [13] G.M. Sheldrick, A short history of SHELX, Acta Crystallogr. A64 (2008) 112–122.
AN
376 [14] M. Wildner, A. Beran, F. Koller, Spectroscopic characterisation and crystal field
377 calculations of varicoloured kyanites from Loliondo, Tanzania, Miner. Petrol. 107
M

378 (2013) 289–310.

379 [15] D.M. Burt, M.F. Sheridan, J.V. Bikun, E.H. Christiansen, Topaz rhyolites—Distribution,
D

380 origin, and significance for exploration, Econ. Geol. 77 (1982) 1818–1836.
TE

381 [16] E.H. Christiansen, D.M. Burt, M.F. Sheridan, R.T. Wilson, The petrogenesis of topaz
382 rhyolites from the western United States, Contrib. Mineral. Petrol. 83 (1983) 16–30.
EP

383 [17] A.N. Platonov, M.N. Taran, O.E. Minko, E.V. Poeshin, Optical absorption spectra and
384 nature of color of iron-containing beryls, Phys. Chem. Miner. 3 (1978) 87–88.
C

385 [18] D.C. Price, E.R. Vance, G. Smith, A. Edgar, B.L. Dickson, Mössbauer effect studies of
AC

386 beryl, J. Phys. Colloq. C6 12 37 (1976) 811–817.

387 [19] C. Aurisicchio, O. Grubessi, P. Zecchini, Infrared spectroscopy and crystal chemistry of
388 the beryl group, Can. Mineral. 32 (1994) 55–68.

389 [20] B.A. Kolesov, C.A. Geiger, The orientation and vibrational states of H2O in synthetic
390 alkali-free beryl, Phys. Chem. Miner. 27 (2000) 557–564.

13
ACCEPTED MANUSCRIPT
391 [21] J. Fukuda, K. Shinoda, S. Nakashima, N. Miyoshi, N. Aikawa, Polarized infrared
392 spectroscopic study of diffusion of water molecules along structure channels in beryl,
393 Am. Mineral. 94 (2009) 981–985.

394 [22] J. Fridrichová, P. Bačík, V. Bizovská, E. Libowitzky, R. Škoda, P. Uher, D. Ozdín, M.


395 Števko, Spectroscopic and bond-topological investigation of interstitial volatiles in beryl
396 from Slovakia, Phys. Chem. Miner. 43 (2016) 419–437.

PT
397 [23] J.E. Shigley, T.J. Thompson, J.D. Keith, Red Beryl from Utah: A review and update,
398 Gems Gemol. 39 (2003) 302–313.

RI
399 [24] H. Shinohara, Excess degassing from volcanoes and its role on eruptive and intrusive

SC
400 activity, Rev. Geophys. 46 (2008) RG4005.

401 [25] M.N. Taran, V.A. Klyakin, Optical spectra of Cu2+ ions in synthetic beryl, J. Appl.

U
402 Spectrosc. 53 (1990) 1167–1169.
AN
403 [26] R.D. Shannon, Revised effective ionic radii and systematic studies of interatomic
404 distances in halides and chalcogenides, Acta Crystallogr. A32 (1976) 751–767.
M

405 [27] F. Bosi, U. Hålenius, G.B. Andreozzi, H. Skogby, S. Lucchesi, Structural refinement and
406 crystal chemistry of Mn-doped spinel: A case for tetrahedrally coordinated Mn3+ in an
D

407 oxygen-based structure, Am. Mineral. 92 (2007) 27–33.


TE

408 [28] J. Ito, S.S. Hafner, Synthesis and study of gadolinites, Am. Mineral. 59 (1974) 700–708.
409 [29] Z.A. Kazet, P. Novak, V.I. Soklov, Cooperative Jahn-Teller effect in the garnets, J. Exp.
EP

410 Theor. Phys.+ 83 (1982) 1483–1499.

411 [30] U. Henn, H. Bank, F.H. Bank, H. Von Platen, W. Hofmeister, Transparent bright blue
C

412 Cu-bearing tourmalines from Paraíba, Brazil, Mineral. Mag. 54 (1990) 553–557.
AC

413 [31] P.C. Burns, F.C. Hawthorne, Static and dynamic Jahn-Teller effects in Cu2+ oxysalt
414 minerals, Can. Mineral. 34 (1996) 1089–1105.

415 [32] M.C. Kemei, S.L. Moffitt, L.E. Dagaro, R. Seshadri, M.R. Suchomel, D.P. Shoemaker,
416 K. Page, J. Siewenie, Structural ground states of (A,A‘)Cr2O4 (A = Mg, Zn; A‘ = Co,
417 Cu) spinel solid solutions: Spin-Jahn-Teller and Jahn-Teller effects, Phys. Rev. B89
418 (2014) 174410.

14
ACCEPTED MANUSCRIPT
419 [33] A. Ertl, O.S. Vereshchagin, G. Giester, E. Tillmanns, H.-P. Meyer, T. Ludwig, I.V.
420 Rozhdestvenskaya, O.V. Frank-Kamenetskaya, Structural and chemical investigation of
421 a zoned synthetic Cu-rich tourmaline, Can. Mineral. 53 (2015) 209–220.

422 [34] M. Wildner, G. Giester, M. Kersten, K. Langer, Phys. Polarized electronic absorption
423 spectra of colourless chalcocyanite, CuSO4, with a survey on crystal fields in Cu2+
424 minerals, Chem. Miner. 41 (2014) 669–680.

PT
425 [35] J.B. Goodenough, Jahn-Teller distortions induced by tetrahedral-site Fe2+ ions, J. Phys.
426 Chem. Solids 25 (1964) 151–160.

RI
427 [36] G.H. Faye, The optical absorption spectra of iron in six-coordinate sites in chlorite,
428 biotite, phlogopite and vivianite: Some aspects of pleochroism in the sheet silicates,

SC
429 Can. Mineral. 9 (1968) 403–425.
430 [37] A. Kyono, S.A. Gramsch, T. Yamanaka, D. Ikuta, M. Ahart, B.M. Mysen, H.-k. Mao,

U
431 R.J. Hemley, The influence of the Jahn–Teller effect at Fe2+ on the structure of chromite
432 at high pressure, Phys. Chem. Miner. 39 (2012) 131–141.
AN
433 [38] P. Bačík, J. Fridrichová, P. Uher, J. Pršek, M. Ondrejka, Crystal chemistry of gadolinite-
434 datolite group silicates, Can. Mineral. 52 (2014) 625–642.
M

435 [39] R.G. Burns, R.G.J. Strens, Structural interpretation of polarized absorption spectra of the
436 Al–Fe–Mn–Cr–epidotes, Mineral. Mag. 36 (1967) 204–226.
D

437 [40] K. Langer, R.M. Abu-Eid, P. Anastasiou, Absorptionsspektren synthetischer Piemontite


TE

438 in den Bereichen 43,000-11,000 (232,6-909,1 nm) and 4,000-250 cm-1 (2.5-40 µm), Z.
439 Kristallogr. 144 (1976)434–436.
EP

440 [41] I. Abs-Wurmbach, K. Langer, F. Seifert, E. Tillmans, The crystal chemistry of (Mn3+,
441 Fe3+)-substituted andalusites (viridines and kanonaite), (All_x_yMn3+xFe3+y)2(O/SiO4):
442 crystal structure refinements, Mössbauer and polarized optical absorption spectra, Z.
C

443 Kristallogr. 155 (1981) 81–113.


AC

444 [42] S. Ghose, M. Kersten, K. Langer, G. Rossi, L. Ungaretti, Crystal field spectra and Jahn
445 Teller effect of Mn3+ in clinopyroxene and clinoamphiboles from India, Phys. Chem.
446 Miner. 13 (1986) 291–305.

447 [43] D.M. Sherman, N. Vergo, Optical-spectrum, site occupancy, and oxidation-state of Mn in
448 montmorillonite, Am. Mineral. 73 (1988) 140–144.

449 [44] H.A. Hänni, M.S. Krzemnicki, Caesium-rich morganite from Afghanistan and
450 Madagascar, J. Gemmol. 28 (2003) 417–429.

15
ACCEPTED MANUSCRIPT
451 [45] I. Adamo, G.D. Gatta, N. Rotiroti, V. Diella, A. Pavese, Gemmological investigation of a
452 synthetic blue beryl: A multi-methodological study, Mineral. Mag. 72 (2008) 799–808.
453 [46] V.P. Solntsev, E.I. Kharchenko, A.S. Lebedev, V.A. Klyakhin, A.G. Il'in, Nature of color
454 centers and EPR of a manganese-activated beryl, J. Appl. Spectrosc. 34 (1981) 111–
455 115.
456 [47] J.M. Gaite, V.V. Izotov, S.I. Nikitin, S.Y. Prosvirnin, EPR and optical spectroscopy of

PT
457 impurities in two synthetic beryls, Appl. Magn. Reson. 20 (2001) 307–315.
458 [48] J.E. Shigley, E.E. Foord, Gem-quality red beryl from the Wah-Wah Mountains, Utah,

RI
459 Gems. Gemol. 20 (1984) 208–221.
460 [49] J.E. Shigley, S.F. McClure, J.E. Cole, J.I. Koivula, T. Lu, S. Elen, L.N. Demianets,

SC
461 Hydrothermal synthetic red beryl from the Institute of Crystallography, Moscow, Gems
462 Gemol. 37 (2001) 42–55.
463 [50] B.M. Laurs, W.B. Simmons, G.R. Rossman, E.P. Quinn, S.F. McClure, A. Peretti, T.

U
464 Armbruster, F.C. Hawthorne, A.U. Falster, D. Günther, M.A. Cooper, B. Grobéty,
AN
465 Pezzottaite from Ambatovita, Madagascar: A New Gem Mineral, Gems Gemol. 39
466 (2003) 284–301.
M

467 [51] K. Nassau, D.L. Wood, An examination of red beryl from Utah, Am. Mineral. 53 (1968)
468 801–806.
D

469 [52] K. Schmetzer, H. Bank, W. Berdesinski, Eine seltene rote Varietät der Mineralart Beryll
TE

470 (früher Bixbit genannt), Z. Dtsch. Gemmol. Ges. 23 (1974) 139–141.

471 [53] A.N. Platonov, M.N. Taran, V.A. Klvakhin, On two colour types of Mn3+-bearing beryls,
EP

472 Z. Dtsch. Gemmol. Ges. 38 (1989) 147–151.

473 [54] R. Norrestam, The effective shapes and sizes of Cu2+ and Mn3+ ions in oxides and
C

474 fluorides, Z. Kristallogr. 209 (1994) 99–106.


AC

475 [55] D.M. Adams, I.R. Gardner, Single-crystal vibrational spectra of beryl and dioptase, J.
476 Chem. Soc. Dalton 1974 (1974) 1502–1505.

477 [56] B. Charoy, P. De Donato, O. Barres, C. Pinto-Coelho, Channel occupancy in an alkali-


478 poor beryl from Serra Branca (Goias, Brazil): Spectroscopic characterization, Am.
479 Mineral. 81 (1996) 395–403.

480 [57] A. Hagemann, A. Lucken, H. Bill, J. Gysler-Sanz, H.A. Stalder, Polarized Raman spectra
481 of beryl and bazzite, Phys. Chem. Miner. 17 (1990) 395–401.

16
ACCEPTED MANUSCRIPT
482 [58] R.X. Fischer, E. Tillmanns, The equivalent isotropic displacement factor, Acta
483 Crystallogr. C44 (1988) 775–776.

484 [59] P.H.M. Uylings, A.J.J. Raassen, J.F. Wyart, Energies of N-equivalent electrons
485 expressed in terms of two-electron energies and independent three-electron parameters:
486 a new complete set of orthogonal operators: II. Application of 3dN configurations, J.
487 Phys. B-At. Mol. Opt. 17 (1984) 4103–4126.

PT
488

489

RI
490 Figure captions

SC
491 Fig. 1. Projection of the crystal structure of red beryl from Utah parallel to the c-axis.

492 Fig. 2. Room temperature Mössbauer spectra of red beryl from Utah. Experimental points are

U
493 plotted by black solid symbols and the resulting fit is given by the red solid line. Individual
494 spectral components are also given.
AN
495 Fig. 3. Raman spectrum of red beryl from Utah. No Raman bands were observed in region
496 between 1400 and 3400 cm-1.
M

497 Fig. 4. Polarized optical absorption spectra of red beryl from Utah with band energies (cm-1),
D

498 wavelengths (nm, in parentheses), and assignments for tetragonal (spin-allowed Mn3+) or
TE

499 cubic symmetry. Uncertain features are labelled in grey.

500 Fig. 5. (a) Observed polarization behaviour with transition energies (cm-1), and (b-e) selection
rules for spin-allowed states of Mn3+ in variable sites and local geometry of sites in red beryl
EP

501
502 from Utah.
C
AC

17
ACCEPTED MANUSCRIPT
TABLES:

Table 1

Representative compositions of red beryl from Utah.

1 2 3
SiO2 66.81 64.56 64.67
TiO2 0.23 0.21 0.26

PT
Al2O3 16.67 16.58 16.52
Fe2O3 2.39 2.35 2.23
Mn2O3 0.23 0.27 0.26

RI
BeO* 13.86 13.50 13.48
MgO 0.07 0.11 0.08
NiO 0.01 0.02 0.01

SC
ZnO 0.06 0.06 0.06
CaO 0.01 0.00 0.00
Na2O 0.02 0.01 0.00

U
K2O 0.08 0.06 0.06
Rb2O 0.07 0.06 0.07
AN
Cs2O 0.24 0.27 0.27
Total 100.74 98.07 97.96
Si4+ 6.022 5.974 5.990
M

Al3+T 0.000 0.026 0.010


T-sum. 6.022 6.000 6.000
Be2+ 3.000 3.000 3.000
D

Ti4+ 0.016 0.014 0.018


Al3+ 1.770 1.782 1.793
TE

Fe3+ 0.162 0.163 0.156


Mn3+ 0.016 0.019 0.018
Mg2+ 0.010 0.016 0.011
EP

Zn2+ 0.004 0.004 0.004


Ni2+ 0.000 0.001 0.001
O-sum. 1.978 2.000 2.000
C

Ca2+ 0.001 0.000 0.000


Na+ 0.004 0.002 0.000
AC

K+ 0.009 0.008 0.007


Rb+ 0.004 0.004 0.004
Cs+ 0.009 0.010 0.011
X
□ 1.973 1.975 1.978
C-sum. 0.027 0.025 0.022
The analytical data, acquired with an electron microprobe, is first reported in wt%, and then
converted to atoms per formula unit on the basis of 8 O+T cations per formula unit. All Fe
and Mn are treated as trivalent. * BeO was calculated assuming 3.00 apfu.
ACCEPTED MANUSCRIPT
Table 2

Mössbauer parameters of red beryl from Utah: relative area – A, isomer shift – IS, quadrupole
splitting – QS, and line width – LW.

component cation A (%) IS (mm/s) QS (mm/s) LW (mm/s)


D1 Fe3+ 11(1) 0.44(1) 2.53(2) 0.68

PT
D2 Fe3+ 89(1) 0.34(1) 0.79(2) 0.82

RI
Table 3 Energies and wavelengths of observed absorption bands in polarized optical spectra

SC
of red beryl from Utah and their probable assignments in tetragonal (spin-allowed Mn3+
quintet states) or cubic symmetry (other states).

U
∥ c assign. Mn3+, Mn3+, Fe3+ Mn3+ Mn3+ Mn3+ Mn3+ Mn3+ Fe3+ Fe3+
Fe3+ ? type II
AN
5 3 3 3 4
B2(D) E(H) T1(H) T2(F2) A1+4E(G) 4
E(D)
? ?
-1
cm 8540 11250 17720 18440 ~20850 21080 22240 23370 27800
nm 1170 890 565 542 480 474 450 428 360
M

eV 1.059 1.395 2.197 2.286 2.585 2.614 2.757 2.898 3.447


⊥ c assign. Mn3+ Mn3+ Mn3+ Mn3+ Mn3+ Mn3+ Fe3+ Fe3+ Fe3+
5 type II 5 3 3 type II 4
A1+4E(G) 4 4
D

A1(D) E(D) E(H) T1(H) T2(D) T2(D)


? ?
-1
cm 7190 11350 18560 19780 20270 ~21300 23370 26700 26990
TE

nm 1390 880 540 505 493 470 428 375 371


eV 0.891 1.407 2.301 2.452 2.513 2.641 2.898 3.310 3.346
EP

Table 4 Energies of Mn3+-assigned absorption bands observed in optical spectra of red beryl
from Utah and respective energy levels calculated for Mn3+ in a pseudo-tetragonal Jahn-
C

Teller-elongated octahedron. Spin-allowed energies are given in bold with tetragonal


AC

level assignments, observed spin-forbidden features are given with cubic assignments.
3
cm-1 5
B1(D) 5
A1(D) 5
B2(D) 5
E(D) 3
E(H) 3
E(H) 3
T1(H) 3
T1(H) T2(F2)
(part)
observed (0) 7190 17720 18560 18440 19780 20270 21080 22240
calculated 0 7175 17704 18579 18451 19831 20210 20888 23299
ACCEPTED MANUSCRIPT
Table 5 Selected parameters resulting from crystal field calculations for Mn3+ in red beryl
from Utah, corresponding to the level energies listed in Table 8. All values in cm-1,
except the predicted tetragonal elongation (%), the ratio C/B, and the nephelauxetic
ratio β. Racah B0 is taken from [59].

Mn3+

PT
tetragonal crystal field
Dqcub 1470
Dqeq (≡Dq) 1770

RI
Dqax 869
Dt 515
Ds 1150
tetragonal elongation 13.27

SC
Racah B 835
Racah C 3430
C/B 4.11

U
β (B0 = 1047 cm-1) 0.80

B4 8240
AN
B2 9800
s4 13672
s2 3600
Dqcub (from s4) 1492
M
D
TE
C EP
AC
ACCEPTED MANUSCRIPT

PT
RI
U SC
AN
M
D
TE
EP
C
AC
ACCEPTED MANUSCRIPT

PT
RI
U SC
AN
M
D
TE
EP
C
AC
ACCEPTED MANUSCRIPT

PT
RI
U SC
AN
M
D
TE
EP
C
AC
ACCEPTED MANUSCRIPT

PT
RI
U SC
AN
M
D
TE
EP
C
AC
ACCEPTED MANUSCRIPT

PT
RI
U SC
AN
M
D
TE
EP
C
AC
ACCEPTED MANUSCRIPT
Highlights

• Red beryl is chemically homogeneous, it contains only small Mn, Fe, Mg content
• Fe, Mn are trivalent as documented by Mössbauer and optical absorption spectroscopy
• No Mn3+-induced Jahn-Teller distortion was detected due to site symmetry restrictions
• Optical spectroscopy shows 7190 cm-1 band assigned to split E ground state of Mn3+
• Crystal field calculations indicate Jahn-Teller distorted Mn3+O6 octahedra

PT
RI
U SC
AN
M
D
TE
C EP
AC

Você também pode gostar