Interstitial bismuth dimers and single atoms as possible centres of broadband near-IR luminescence in bismuth-doped glasses
aa r X i v : . [ c ond - m a t . m t r l - s c i ] J a n Interstitial bismuth dimers and single atomsas possible centres of broadband near-IRluminescence in bismuth-doped glasses
V.O.Sokolov , V.G.Plotnichenko, E.M.Dianov Fiber Optics Research Center of the Russian Academy of Sciences38 Vavilov Street, Moscow 119333, Russia
Abstract
Absorption, luminescence and Raman spectra of interstitial bismuth atoms, Bi , andnegatively charged dimers, Bi − , in alumosilicate, germanosilicate, phosphosilicateand phosphogermanate glasses networks are calculated by time-dependent densityfunctional method. On grounds of this calculation an extension of our previouslysuggested model of broadband near-IR luminescence in bismuth-doped glasses isput forward. Key words:
Computer simulation; Glasses; Bismuth; Luminescence
PACS:
Introduction
Near-IR broadband (1100 – 1400 nm) luminescence in bismuth-doped glassesdiscovered in Refs. [1, 2] is being studied intensively. By now the lumines-cence has been observed in many bismuth-doped glasses, such as alumosilicate(e.g. [1–7]), alumogermanate (e.g. [4, 8–10]), alumoborate (e.g. [10–12]), alu-mophosphosilicate, alumophosphate, alumophosphoborate [4, 12–14], chalco-genide [15,16] and in several bismuth-doped crystals (RbPb Cl [17], FAU-typezeolites [18], BaB O [19]). The bismuth-related IR luminescence is used suc-cessfully in laser amplification and generation (see, e.g. [20, 21]). The present Corresponding author.Tel.: +7 499 135 8093; Fax: +7 499 135 8139; E-mail: [email protected],[email protected]
Preprint submitted to Elsevier November 16, 2018 tate of research and applications of near-IR luminescence in bismuth-dopedglasses are examined in the recent reviews, Ref. [22, 23].However there is no commonly accepted model of the IR luminescence center.Several models are suggested, such as electronic transitions in Bi + [11, 13–15],Bi [15] and Bi [2–4, 8] interstitial ions, in BiO interstitial molecules [6],Bi , Bi − and Bi − interstitial dimers [5,9,16,24,25], other bismuth clusters [9],BiO complexes with tetrahedral coordination of the central bismuth ion [26].In our opinion there is good evidence for negative bismuth dimers being theluminescence centres (see e.g. [7, 16, 25]). However, up to now, there are noconvincing explanation of changes in absorption and luminescence spectrawith glass composition (see e.g. Refs. [21, 23]). In the context of the Bi − andBi − dimers as centres of the IR luminescence it would appear reasonable thatsuch changes are caused by Stark shift and splitting together with certainrearrangement of dimer states in glass network electric field. The latter doesobviously depend on glass composition.Alternatively, there is an evidence that single Bi atoms and ions do occur aswell in the bismuth-doped glass host. We studied Raman scattering of Ar laser457.9 nm-light in optical fibres with bismuth-doped alumosilicate glass coreand observed a narrow intensive band in the low-frequency part of the Ramanspectra, near 110 cm − (Ref. [25], Fig. 4, ”a” band). Our calculation [25]of neutral single bismuth atom, Bi , and negative single bismuth ion, Bi − ,incorporated in alumosilicate host in the sixfold ring interstitial sites provedboth Bi and Bi − have three vibrational modes practically not mixed withany vibrations of the rings. The frequency of these vibrational modes arefound to fall in the 75 – 105 and 85 – 130 cm − ranges, for Bi − and Bi ,respectively (see Fig. 5 in Ref. [25]). One mode corresponds to displacement ofthe interstitial bismuth atom along the rings axis and the other two correspondto transverse displacement of the atom. All three vibrational modes are Ramanactive, scattering in the transverse vibrations being an order as intensive asin the longitudinal ones. Hence the Raman band near 110 cm − in bismuth-doped alumosilicate glass may be indicative of single bismuth atoms, Bi , ornegative ions, Bi − in sixfold ring interstitial sites of the glass network. Recentlyelectronic transitions in the Bi atom has been supposed to be responsible forthe near-IR luminescence in bismuth-doped glasses [27].The aim of this work were to verify the assumptions regarding both the net-work electric field and the interstitial bismuth atoms by immediate calculationof optical properties of interstitial Bi atoms and ions in networks of silica- andgermania-based glasses. 2 alculations In our previous works [24, 25] we used calculated data for free negativelycharged Bi dimers to explain absorption and luminescence spectra of bismuth-doped glass. Experimental data available for free Bi atom were used to thesame purposes in Ref. [27]. However an influence of glass network (interstitialelectrostatic field, in essential) must be taken into account to discuss opticalspectra of interstitial dimers and atoms.We performed quantum-chemical calculation of interstitial electric field in net-work of glasses of several compositions (silica, germania, alumosilicate, ger-manosilicate, phosphosilicate, phosphogermanate). Cluster approach was usedto model the atomic environment of interstitial sites. Two coaxial sixfold ringseach formed by SiO , GeO (for germanosilicate glasses with low germaniacontent), AlO or O = PO tetrahedra or GeO octahedra (for germanosilicateglasses with high germania content) were incorporated in proper numbersin the clusters. The distance between the ring centres was optimized duringthe calculations. To ensure aluminum atoms being fourfold coordinated, extraelectrons were placed in the aluminum-containing clusters, one electron peraluminum atom. In GeO -containing clusters two extra electrons were addedper sixfold coordinated germanium atom to make it be stable in this coordina-tion. Dangling bonds of the outer oxygen atoms in the clusters were saturatedwith hydrogen atoms. All calculations were carried out with GAMESS (US)quantum-chemical code [28] by DFT method using BLYP functional which isknown to provide nice agreement of calculated geometrical parameters withexperimental data. We used the bases and effective core potentials developedin Ref. [29]. One d -type polarization function with ζ = 0 . , Bi − and Bi − dimers and of single Bi atom and Bi − ion in alumosilicatenetwork was performed in our previous works [24,25] using the above-describedcluster approach. Calculated configuration for Bi − interstitial dimer was pre-sented in figure 3 of Ref. [25] for alumosilicate glass. In the present work suchcalculations were repeated for silica, germania, germanosilicate, phosphosili-cate, and phosphogermanate clusters. The equilibrium configurations of neg-atively charged dimers, both Bi − and Bi − , in the interstitial site formed bytwo sixfold rings are found to be quite similar in all these networks: the dimersare aligned along the common axis of the rings, the first bismuth atom beingpractically in the center of one ring and the second atom being between tworings. Again, equilibrium position of Bi atom and Bi − ion are found to bebetween two six-rings in all these networks. As a representative example, cal-culated configuration of alumosilicate glass cluster with Bi interstitial atom3s shown in figure 1. Such configurations turns out to be highly stable for bothdimers and atom or ion: bismuth atoms do not form bond with any atom ofthe rings and returned to the equilibrium positions even after dimer or singleatom is displaced considerably from that position. Vibrational properties ofinterstitial Bi atom and Bi − dimer in all these networks are found to be quitesimilar to those calculated in Ref. [25] for alumosilicate host (see above). Bi SiAl O
Figure 1. Calculated configuration of Bi atom in the sixfold-rings interstitial siteof alumosilicate glass Electric field strength was calculated in the point of the Bi atom position (or,what is the same, near the bismuth atom of the Bi dimer which is situatedbetween two sixfold rings) and in several (8 – 10) points in its vicinity. Thevalue averaged over all such points was used in further calculations of opticalspectra. The mean electric field strengths in ring interstitial sites calculatedfor several glass compositions are given in table 1. As evident from these data,the interstitial electric field is found to be approximately the same in silica,germanate and germanosilicate (with low germania content) hosts but consid-erably higher both in glasses containing aluminate or phosphate componentand in germanosilicate glasses with high germania content. This is attributable4o charged (AlO and GeO ) or strongly dipole-polarized (O = PO ) structuralunits in the rings around interstitial bismuth sites in such glasses. Table 1Calculated electric field in sixfold ring interstitial sites of glass networkRing composition Electric field, 0.001 a.u.6SiO – 2GeO – 2AlO – 2GeO – 2O = PO – 2O = PO Optical spectra calculations were performed by time-dependent density func-tional theory (TDDFT) method using Octopus program [30] and Hartwigsen-Goedecker-Hutter pseudopotentials [31] with spin polarization and spin-orbitinteraction taken into account. PBE density functional [32] was used in theground-state calculation. To obtain the linear optical absorption spectrum ofthe system, the Octopus code excites all frequencies of the system by givingcertain (small enough) momentum to the electrons and then evolves the time-dependent Kohn-Sham equations in real space for a certain real time [33].The dipole-strength function (or the photo-absorption cross section) is thenobtained by a Fourier transform of the time-dependent dipole moment. Adia-batic LDA approximation is used in these calculations to describe exchange-correlation effects. The Octopus code uses real-space uniform grid inside thesum of spheres around each atom of the system (a single one in our case). Thesphere radius and the grid spacing were taken to be was 8.0 and 0.25 ˚A, re-spectively, in our calculations. The real-time propagation was performed with2 · time steps with the total simulation time of about 20 fs. The Fouriertransform was performed using third-order polynomial damping (see [30] fordetails of the code).Shown in figure 2 are the calculated optical spectra of free Bi − dimer and ofinterstitial Bi − dimers in sixfold ring interstitial sites of networks of silica (orgermania or germanosilicate), alumosilicate, phosphosilicate and phosphoger-manate glasses. Figures 3 and 4 display similar spectra for free and interstitialBi atoms and for free and interstitial Bi − ion, respectively. Compairing the5esults for free Bi − dimers with previous configuration interaction calcula-tions [24, 25, 34] and those for free Bi with experimental and theoretical dataavailable [35–38] (see figure 5), one readily sees that both transition wave-lengths and transition intensities are described adequately in our TDDFTcalculations. (cid:0) (cid:1) (cid:2) (cid:3) (cid:4)(cid:5)(cid:6)(cid:4)(cid:5)(cid:6) (cid:3)(cid:5)(cid:4) (cid:3)(cid:5)(cid:6) (cid:2)(cid:5)(cid:4) (cid:2)(cid:5)(cid:6) (cid:1)(cid:5)(cid:4) (cid:1)(cid:5)(cid:6)(cid:4)(cid:5)(cid:3)(cid:3)(cid:3)(cid:4)(cid:3)(cid:4)(cid:4) (cid:7) (cid:8) (cid:9) (cid:10) (cid:11) (cid:12) (cid:13) (cid:14) (cid:15) (cid:11)(cid:16) (cid:7) (cid:17) (cid:12) (cid:11) (cid:10)(cid:10) (cid:18) (cid:10) (cid:19) (cid:17) (cid:14) (cid:15) (cid:11)(cid:16) (cid:20)(cid:7) (cid:21) (cid:12) (cid:9) (cid:15) (cid:14) (cid:12) (cid:21) (cid:12) (cid:22) (cid:7) (cid:23)(cid:16) (cid:15) (cid:14) (cid:10) (cid:24)(cid:16)(cid:19)(cid:12)(cid:25)(cid:22)(cid:20)(cid:7)(cid:19)(cid:26)(cid:7)(cid:27)(cid:21)(cid:28)(cid:19)(cid:29)(cid:19)(cid:16)(cid:25)(cid:14)(cid:30)(cid:20)(cid:7) (cid:0) (cid:31)(cid:7) Figure 2. Optical cross-section spectra of free Bi − dimer (dots) and of interstitialBi − dimers in silica, germania or germanosilicate (with low germania content) host(short dashes), alumosilicate host (dashes), in phosphosilicate or germanosilicate(with high germania content) host (dashes plus dots) and in phosphogermanatehost (solid) Figure 2 shows calculated optical spectra of interstitial Bi − dimer in compar-ison with those of a free dimer. Similar spectra for interstitial Bi atom arepresented in figure 3. From figure 2 it will be noticed how optical spectrumof an interstitial Bi − dimer is modified in comparison with free dimer. Ac-cording to our previous calculations performed by configuration interaction6CI) method [25], absorption at 860, 720, 460 nm and .
400 nm wavelengthsin the free Bi − dimer is caused by transitions from the ground state to theexcited ones. The IR luminescence at 1450, 1300 and 1050 nm wavelengthscorresponds to spin-forbidden transitions from three lowest excited states tothe ground state and the visible luminescence at 750 nm is caused to a spin-allowed transition with low oscillator strength from one of the other excitedstates to the ground state. Notice that compairing results of CI calculationswith optical spectra calculated in TDDFT with Octopus package (such asshown in figures 2, 3, 4) one should realize that the latters include all possibletransitions in the system under consideration. (cid:0) (cid:1) (cid:2) (cid:3) (cid:4)(cid:5)(cid:6)(cid:4)(cid:5)(cid:6) (cid:3)(cid:5)(cid:4) (cid:3)(cid:5)(cid:6) (cid:2)(cid:5)(cid:4) (cid:2)(cid:5)(cid:6) (cid:1)(cid:5)(cid:4) (cid:1)(cid:5)(cid:6)(cid:3)(cid:3)(cid:4) (cid:7) (cid:13) (cid:14) (cid:15) (cid:17) (cid:21) (cid:29) (cid:7) (cid:17) (cid:12) (cid:11) (cid:10)(cid:10) (cid:18) (cid:10) (cid:19) (cid:17) (cid:14) (cid:15) (cid:11)(cid:16) (cid:20)(cid:7) (cid:21) (cid:12) (cid:9) (cid:15) (cid:14) (cid:12) (cid:21) (cid:12) (cid:22) (cid:7) (cid:23)(cid:16) (cid:15) (cid:14) (cid:10) (cid:24)(cid:16)(cid:19)(cid:12)(cid:25)(cid:22)(cid:20)(cid:7)(cid:19)(cid:26)(cid:7)(cid:27)(cid:21)(cid:28)(cid:19)(cid:29)(cid:19)(cid:16)(cid:25)(cid:14)(cid:30)(cid:20)(cid:7) (cid:0) (cid:31)(cid:7) Figure 3. Optical cross-section spectra of free Bi atom (dots; the part for energy < . atom in silica, germaniaor germanosilicate (with low germania content) host (short dashes), alumosilicatehost (dashes), in phosphosilicate or germanosilicate (with high germania content)host (dashes plus dots) and in phosphogermanate host (solid) (cid:1) (cid:2) (cid:3) (cid:4)(cid:5)(cid:6)(cid:4)(cid:5)(cid:6) (cid:3)(cid:5)(cid:4) (cid:3)(cid:5)(cid:6) (cid:2)(cid:5)(cid:4) (cid:2)(cid:5)(cid:6) (cid:1)(cid:5)(cid:4) (cid:1)(cid:5)(cid:6)(cid:3)(cid:4)(cid:3)(cid:4)(cid:4)(cid:3)(cid:4)(cid:4)(cid:4)(cid:3)(cid:4)(cid:4)(cid:4)(cid:4) (cid:7) (cid:13) (cid:14) (cid:15) (cid:17) (cid:21) (cid:29) (cid:7) (cid:17) (cid:12) (cid:11) (cid:10)(cid:10) (cid:18) (cid:10) (cid:19) (cid:17) (cid:14) (cid:15) (cid:11)(cid:16) (cid:20)(cid:7) (cid:21) (cid:12) (cid:9) (cid:15) (cid:14) (cid:12) (cid:21) (cid:12) (cid:22) (cid:7) (cid:23)(cid:16) (cid:15) (cid:14) (cid:10) (cid:24)(cid:16)(cid:19)(cid:12)(cid:25)(cid:22)(cid:20)(cid:7)(cid:19)(cid:26)(cid:7)(cid:27)(cid:21)(cid:28)(cid:19)(cid:29)(cid:19)(cid:16)(cid:25)(cid:14)(cid:30)(cid:20)(cid:7) (cid:0) (cid:31)(cid:7) Figure 4. Optical cross-section spectra of free Bi − ion (dotted line) and of interstitialBi − ion in alumosilicate host (solid line). For the interstitial Bi − dimers, the absorption in the .
400 nm range andtransitions corresponding to the absorption and visible luminescence in the600 – 750 nm is found in our calculations to be practically unaltered in all thesurrounding glass network compositions studied. The absorption band near500 nm is found to be shifted towards shorter wavelengths (near 440 nm inphosphogermanate environment) and somewhat intensified. Besides, absorp-tion near 800 nm growths. In phosphate-bearing and in germanosilicate (withhigh GeO content) glasses absorption in the 900 – 1000 nm range becomessignificant. In glasses with no phosphate component present and in germanosil-icate glasses with low GeO content, the IR luminescence is not too differentfrom that in free dimers and occurring mainly in the 1000 – 1300 nm range.When passing to phosphate-bearing compositions and for high GeO contentone finds the IR luminescence bands to be shifted towards longer wavelengths,in the 1400 – 1600 nm range. 8 S / D / D / P / P / eV state a b c d e f g Figure 5. States, transition types and lifetimes in Bi atom corresponding to the6s configuration [35–38]: (a) 876 nm, M1, 32 ms; (b) 648 nm, M1, 156 ms;(c) 462 nm, M1, 18 ms and E2, 161 ms; (d) 302 nm, M1, 137 ms; (e) 976 nm, M1,833 ms and E2, 222 ms; (f) 460 nm, M1, 8 ms; (g) 564 nm, M1, 43 ms Again, optical spectra of Bi interstitial atoms are found in our calculations tochange compared with those of free atoms (figure 3). This change is relativelylow in alumosilicate glasses where Bi interstitial atoms give rise to absorptionbands at .
400 nm, near 500 nm and in 600 – 700 and 900 – 1100 nm ranges.9his absorption causes IR luminescence in 1000 – 1300 and 1400 – 1600 nmwavelength ranges. However the change turn out to be much more pronouncedin phosphosilicate glasses and in germanosilicate glass with high germaniacontent where the absorption should occur in 500 – 600 nm wavelength rangeand, especially, in 700 – 1000 nm range. The corresponding IR luminescencein these glasses is found to arise in 1000 – 1400 nm range and near 2000 nm.The above-described modification of optical spectra of bismuth interstitialdimers and atoms with composition of surrounding glass network gives aninsight into origin of variations of absorption and IR luminescence bands dis-covered in bismuth-doped glasses [21,23]. In particular, characteristic IR lumi-nescence band shift from ∼ − , is known to be ratherlow: e.g. in Ref. [38] the electron affinity of Bi is found to be 7600 cm − or 0 .
94 eV. Because of this, all transitions in . − ionization with neutral Bi atom formed andfree electron occurred in conduction band. Since optical cross-sections in Bi − ion are at least an order as high as those in Bi atom, one would expect theinterstitial Bi − ions are destroyed rapidly under both visible and IR irradia-tion. Thus interstitial Bi − ions in glass may promote formation of interstitialnegatively charged Bi dimers [25]. On the other hand, recombination lumi-nescence in 1500 – 1700 nm range is not improbable owing to conductionelectrons capturing in Bi interstitial atoms with Bi − ions formed. Discussion
We have established, on modelling grounds, that interstitial negatively chargedbismuth dimers, Bi − and Bi − , can actually occur in glass network and thatabsorption and luminescence spectra of the dimers correlate well with theexperimental spectra of bismuth-doped glasses. On the other hand, bismuthinterstitial atoms, Bi , and negative ions, Bi − , are demonstrated to occur inglass network as well. In particular, certain features observed in Raman spectraof bismuth-doped alumosilicate glasses are attributable to both Bi dimers andBi atoms (ions) present in glass together.So it would appear reasonable that both Bi interstitial atoms and Bi − (Bi − )negatively charged dimers occur in in bismuth-doped glasses as two types ofthe IR luminescence centres. Relation between interstitial Bi dimers and Biatoms (and ions) concentrations in the given glass is likely to depend predomi-nantly on bismuth dopant content. It should be stressed that in bismuth-dopedglasses considerable part of the dopant bismuth atoms are likely to be bonded10ith oxygen atoms of surrounding network. However such bismuth centresbear no relation to IR luminescence.It follows from the relation between calculated intensities of transitions corre-sponding to absorption and IR luminescence in Bi atom that the IR lumines-cence time constant is considerably less than in Bi − or Bi − dimers and that itdecreases rapidly with electrostatic field growth (figure 3). Similar calculationsof Bi dimers in electrostatic field prove the transition intensities to changeonly slightly and the transition wavelengths to change almost as rapidly as inBi atom with the field strength. Therefore, interstitial Bi atoms should giverise to much faster IR luminescence in comparison with Bi − and Bi − dimers.Unfortunately the transition intensities in atoms and dimers are hardly to becompared directly in our calculational approach. Rough estimate results inratio of the order of 10. Thus, interstitial Bi atoms may be responsible for”fast” component observed in IR luminescence.The calculations described in the present article and in Ref. [24, 25] are con-cerned with sixfold ring interstitial sites. The sixfold ring are known to bethe most abundant ring structure in silica- and germania-based glasses net-work. However there are other rings as well in such glasses, mainly five-, seven-and eightfold ones. We have performed some calculations for interstitial sitesformed by pairs of such rings in alumosilicate network to examine equilibriumstates of bismuth atoms and dimers and electric field strength. It is felt thatthe results and conclusions obtained for the sixfold rings still remain valid forsevenfold and, supposedly, in eightfold rings. The fivefold rings turn out to betoo narrow to hold the interstitial dimers and, supposedly, atoms. Conclusion
In summary, we calculated optical spectra of interstitial negatively chargedbismuth dimers, Bi − , neutral atoms, Bi , and negative ions, Bi − , in network ofsilica- and germania-based glasses with various composition. Considering ourprevious works [24, 25], the results of the present calculations allow definiteconclusions, as follows.— Both interstitial Bi − and Bi − dimers and interstitial Bi atoms and Bi − ions can occur in bismuth-doped silica- and germania-based glasses. IRluminescence in 1000 – 1600 nm in these glasses can be caused by all suchinterstitial centres.— IR luminescence decay time in Bi atoms is estimated to be an order asshort as in Bi − or Bi − dimers . Hence Bi − (Bi − ) dimers and Bi atomscan be assumed to cause ”slow” and ”fast” parts of the IR luminescence,respectively. 11 Interstitial electric field strength dependence of absorption and IR lumi-nescence in the interstitial Bi − (Bi − ) dimers is found to differ consid-erably from that in the interstitial Bi atoms. Pronounced dependenceof the IR luminescence spectra and its excitation spectra on glass com-position and bismuth content observed in several experiments may beattributable to the interstitial field variations.— interstitial Bi atom is found to capture under certain conditions an elec-tron from the glass conduction band with Bi − formed. Recombinationluminescence in the 1500 – 1700 nm is likely to occur owing to such cap-turing. On the other hand, the interstitial negative Bi − ions are readilydestroyed under . and Bi − (Bi − ) occurin these glasses as two types of the IR luminescence centres. Acknowledgements
We are grateful to Prof. I.A.Bufetov for valuable discussions.12 eferences [1] K.Murata, Y.Fujimoto, T.Kanabe, H.Fujita, M.Nakatsuka, Fusion Engineeringand Design 44 (1999) 437.[2] Y.Fujimoto, M.Nakatsuka, Japan. J. Appl. Phys. 40 (2001) L279.[3] Y.Fujimoto, M.Nakatsuka, J. Non-Cryst. Solids 352 (2006) 2254.[4] V.V.Dvoyrin, V.M.Mashinsky, L.I.Bulatov, I.A.Bufetov, A.V.Shubin,M.A.Melkumov, E.F.Kustov, E.M.Dianov, A.A.Umnikov, V.F.Khopin,M.V.Yashkov, A.N.Guryanov, Opt. Lett. 31 (2006) 2966.[5] S.Khonthon, S.Morimoto, Y.Arai, Y.Ohishi, J. Ceramic Soc. Japan 115 (2007)259.[6] J.Ren, L.Yang, J.Qiu, D.Chen, X.Jiang, C.Zhu, Solid State Comm. 140 (2006)38.[7] B.Denker, B.Galagan, V.Osiko, I.Shulman, S.Sverchkov, E.Dianov, Appl.Phys. B 95 (2009) 801.[8] H.-P.Xia, X.-J.Wang, Appl. Phys. Lett. 89 (2006) 051917.[9] M.Peng, J.Qiu, D.Chen, X.Meng, C.Zhu, Optics Lett. 30 (2005) 2433.[10] M.Peng, X.Meng, J.Qiu, Q.Zhao, C.Zhu, Chem. Phys. Lett. 403 (2005) 410.[11] X.Meng, J.Qiu, M.Peng, D.Chen, Q.Zhao, X.Jiang, C.Zhu, Optics Express 13(2005) 1635.[12] B.Denker, B.Galagan, V.Osiko, S.Sverchkov, E.Dianov, Appl. Phys. B 87 (2007)135.[13] V.G.Truong, L.Bigot, A.Lerouge, M.Douay, I.Razdobreev, Appl. Phys. Lett. 92(2008) 041908.[14] X.Meng, J.Qiu, M.Peng, D.Chen, Q.Zhao, X.Jiang, C.Zhu, Optics Express 13(2005) 1628.[15] G.Yang, D.Chen, J.Ren, Y.Xu, H.Zeng, Y.Yang, G.Chen, J. Am. Ceram. Soc.90 (2007) 3670.[16] M.A.Hughes, T.Akada, T.Suzuki, Ya.Ohishi, D.W.Hewak, Optics Express 17(2009) 19345.[17] A.G.Okhrimchuk, L.N.Butvina, E.M.Dianov, N.V.Lichkova, V.N.Zagorodnev,K.N.Boldyrev, Optics Lett. 33 (2008) 2182.[18] H.-T.Sun, Y.Miwa, F.Shimaoka, M.Shimaoka, A.Hosokawa, M.Mizuhata,Sh.Hayashi, Sh.Deki, Optics Lett. 34 (2009) 1219.[19] L.Su, J.Yu, P.Zhou, H.Li, L.Zheng, Y.Yang, F.Wu, H.Xia, J.Xu, Optics Lett.34 (2009) 2504.20] V.V.Dvoyrin, V.M.Mashinsky, E.M.Dianov, Optics Lett. 32 (2007) 451.[21] E.M.Dianov, S.V.Firstov, S.V.Khopin, A.N.Guryanov, I.A.Bufetov QuantumElectronics 38 (2008) 615.[22] E.M.Dianov, Proc. SPIE 6890 (2008) 68900H.[23] I.A.Bufetov, E.M.Dianov, Laser Phys. Lett. 6 (2009) 487.[24] V.O.Sokolov, V.G.Plotnichenko, E.M.Dianov, Optics Lett. 33 (2008) 1488.[25] V.O.Sokolov, V.G.Plotnichenko, V.V.Koltashev, E.M.Dianov, J. Phys. D: Appl.Phys. 42 (2009) 095410.[26] E.F.Kustov, L.I.Bulatov, V.V.Dvoyrin, V.M.Mashinsky Optics Lett. 34 (2009)1549.[27] M.Peng, C.Zollfrank, L.Wondraczek, J. Phys.: Condens. Matter 21 (2009)285106.[28] M.W.Schmidt, K.K.Baldridge, J.A.Boatz, S.T.Elbert, M.S.Gordon, J.J.Jensen,S.Koseki, N.Matsunaga, K.A.Nguyen, S.Su, T.L.Windus, M.Dupuis,J.A.Montgomery, J. Comput. Chem. 14 (1993) 1347( ).[29] W.J.Stevens, H.Balsch, M.Krauss, J. Chem. Phys. 81 (1984) 6026; W.J.Stevens,M.Krauss, H.Basch, P.G.Jasien, Canadian J. Chem. 70 (1992) 612;T.R.Cundari, W.J.Stevens, J. Chem. Phys. 98 (1993) 5555.[30] A.Castro, H.Appel, M.Oliveira, C.A.Rozzi, X.Andrade, F.Lorenzen,M.A.L.Marques, E.K.U.Gross, A.Rubio, physica status solidi (b) 243 (2006)2465(