A site-selective spectroscopy study of Ag nanoclusters dispersed in oxyfluoride glass hosts has been carried out. The nano- to millisecond, essentially non-exponential, luminescence kinetics of Ag nanoclusters has been detected in the spectral range from 450 to 1000 nm, when excited at discrete wavelengths in the range 250 to 450 nm. Based on these experimental observations, the energy level configuration coordinate diagram for the involved ground and excited singlet/triplet states of the Ag nanoclusters has been proposed and confirmed by the density functional theory (DFT). The sites for the Ag nanoclusters are argued to be multiple. The structure/geometry of the involved Ag nanoclusters has been suggested to involve spin-paired dimers Ag2+, or tetramers Ag42+, with a varying elongation/distortion along the tetramer diagonals.
©2012 Optical Society of America
Recently, luminescent Ag nanoclusters have attracted increased interest owing to a variety of potential novel applications, such as in nanolabels, nanosensors, and nano-scale light sources, white light generation, screen monitor phosphors, down-shifting of the solar spectrum, e.g. in [1–7] and refs therein. These nanoclusters offer unique optical properties that are intermediate between the properties of the bulk Ag metal and single Ag atoms, and are reminiscent of the properties of simple inorganic molecules. In particular, whereas bulk silver is not normally luminescent and single Ag atoms show a narrow emission band in the UV/blue part of the spectrum, the Ag nanoclusters do emit a broad luminescence, which covers the whole visible range of the spectrum ([1–6] and refs therein).
While the majority of the hosts/scaffolds employed for the Ag nanoclusters are liquid, polymer or organics based materials (see, e.g., the recent review ), only very recently it has been found that some bulk, oxyfluoride, glasses can also host Ag nanoclusters . The obvious advantages of the glassy state, such as the possibility for fiber, film, and arbitrary bulk shape manufacturing, open novel application perspectives [5,6]. On the other hand, the mechanism for luminescence of Ag nanoclusters, in particular when dispersed in a solid hosts, i.e. in glassy host, has not yet been addressed in depth, and even the energy level diagram of Ag nanoclusters has not been proposed to date (e.g. in review ).
In this work, the kinetics of the luminescence of Ag nanoclusters dispersed in an oxyfluoride glass host has been studied. The kinetics has been modeled and explained using the density functional theory (DFT). The energy level and configuration coordinate diagrams for the Ag nanoclusters have been proposed, within the experimental testing view that the Ag nanoclusters are mostly paired dimers/tetramers with a varying shape. A model for the mechanism of luminescence of Ag nanoclusters has been suggested. The luminescence of Ag nanoclusters dispersed in glass hosts may be used in applications, such as color flexible monitors driven by UV light, white light generation under UV pump, tunable light sources/lasers across the whole visible range, and down-conversion of the solar spectrum, e.g. in [1–7] and refs therein.
The glass melting method used was similar to the previously reported procedure for preparation of similar oxyfluoride glasses containing lanthanide dopants, but no silver dopant . A mixture containing SiO2, Al2O3, CdF2, PbF2, ZnF2, and AgNO3 powders was batched and melted in a Pt-crucible at about 1000°C for 1 hour. The chemical formula of the typical prepared oxyfluoride glass was 33(SiO2)9.5(AlO1.5)32.5(CdF2)19.5(PbF2)5.5(ZnF2), in mol%, which was doped with an identified optimum amount of 5 wt% of AgNO3 [5,6]. A glass with such chemical composition will be called further in the text as a base glass. The glass melt was cast into preheated mold and then allowed to cool down to the room temperature. The resulting pieces of glass, typically of 4 × 1 × 0.3 cm3 in size, were then polished and cut in smaller pieces for optical and structural characterization.
The steady state luminescence spectra were obtained by exciting the samples with light from a 300 W Xe arc lamp transmitted through a 0.25 m double-grating monochromator and detecting with a 0.25 m monochromator and an extended range Hamamatsu photomultiplier. The spectral response function of the set-up has been taken into account .
The millisecond kinetics of luminescence has been recorded using an optical parametrical oscillator (OPO) pumped by 3rd harmonic of a YAG:Nd laser as the excitation source. The signal was detected with a digital storage oscilloscope controlled by a personal computer. The samples were excited by 10 ns pulses with 20 Hz repetition rate. The time resolution of this setup was 20 ns.
Pico- and nanosecond luminescence kinetics has been recorded using a LifeSpecII spectrometer of Edinburgh Instruments. The samples were excited by 70 ps pulses with 1 MHz repetition rate produced by a semiconductor laser operating at 406 nm. A multi-channel plate photomultiplier tube was used for the detection. The system was operated in time correlated single photon counting mode and had an instrumental response function (IRF) with a half maximum width of about 70 ps.
The results of structural studies of this Ag-nanoclusters doped glass have been presented elsewhere [5,6,9]; they indicate the size of Ag nanoclusters in as prepared glasses studied in this work is less than 1 nm, and extra heat-treatment results in condensation of these tiny Ag nanoclusters to about 20 nm diameter Ag nanoparticles. No any crystalline phases have been detected in these as-prepared glasses either by X-ray diffraction pattern or by transmission electron microscopy.
3.1. Steady state luminescence of Ag nanoclusters
Figure 1 shows the typically observed emission and excitation spectra of Ag nanoclusters dispersed in the base oxyfluoride glass, when excited and detected at the indicated wavelengths. The excitation spectra in Fig. 1 were found to correspond to the respective absorption spectra of the Ag nanoclusters. The emission spectra in Fig. 1 show a blue shifting when the excitation wavelength is changed to shorter wavelengths from, i.e. from 457 to 320 nm, and the excitation spectra also show a blue shift when the detection wavelength shifted from 700 nm to the shorter value 450 nm. These shifts point out an unhomogeneous broadening due to a size/site distribution of the Ag nanoclusters [5,9], similar to the case of semiconductor quantum dots, e.g., in . The emission and excitation spectra shown in Fig. 1 correspond to the “border” excitation and detection wavelengths, respectively. That is, excitation below 320 and above 457 nm did result in essentially weaker luminescence spectra, and detection below 450 and above 700 nm did in fact result in weaker excitation spectra.
3.2 Time resolved luminescence of Ag nanoclusters
3.2.1 Microsecond kinetics of luminescence
Microsecond emission decays of Ag nanoclusters were obtained under excitation at 355 and 420 nm wavelengths, which are close to the short and long wavelength excitations borders of Ag nanoclusters discussed around Fig. 1. The emission was detected at 450, 500 and 800 nm; the results are presented in Fig. 2 . The decay kinetics was found to be well fit by a double exponential function given by Eq. (1):Table 1 . The rise kinetics in Fig. 2 was shorter than the time resolution limit of the setup at 20 ns; and therefore it is not shown.
As it is seen from Fig. 2 and Table 1, the decay kinetics becomes faster with detection at shorter wavelengths for the both excitation wavelengths. Also the kinetics becomes faster with shortening the excitation wavelength from 420 to 355 nm. These results will be explained in the Discussion Section in the terms of spin-allowed singlet-singlet and spin-forbidden triplet-singlet transitions of Ag nanoclusters. The current experimental data agree with those obtained for emitting Ag clusters dispersed in zeolite matrix, e.g. in .
3.2.2 Nanosecond kinetics of luminescence
The long lifetime microseconds kinetics presented in the previous chapter have both singlet-singlet and triplet-singlet components contributions and was detected for all the excitation wavelength range, which is seen in Fig. 1. On the other hand, the short lifetime nanosecond range kinetics presented in this chapter can be only due to spin allowed singlet-singlet transitions of Ag nanoclusters, e.g. in [12,13]. The nanosecond decays kinetics of Ag nanoclusters is shown, for example, in Fig. 3 , when detected at several indicated wavelengths. The excitation wavelength for these decays was 406 nm, which falls near to the middle between the excitation borders of Ag nanoclusters discussed around Fig. 1. The decays have been found to be well fit by a double exponential function of Eq. (1); the obtained τfast and τslow lifetimes are summarized in Table 2 . The rise kinetics in Fig. 3 was shorter than the time resolution limit of the setup at 70 ps; therefore it is not shown.
As it is seen from Fig. 3 and Table 2, the decay kinetics becomes faster with shortening the detection wavelength; the results will be explained in the Discussion Section in the terms of spin allowed singlet-singlet transitions of Ag nanoclusters. The lifetimes collected in the Table 2 vary from 0.4 to 6.4 ns, which agree with the lifetime data of Ag nanoclusters dispersed in other hosts, e.g. in [1,14,15] and refs therein.
Despite the luminescence of Ag nanoclusters has been extensively investigated in recent years, e.g. in [1–6] and refs therein, its mechanism has not yet been understood. Based on the above experimental kinetics data, we carry out here quantum chemistry calculations of the low-lying electronic energy surfaces of Ag nanoclusters and propose the mechanism for their luminescence. The used approach can be applied to the Ag nanoclusters dispersed in other hosts.
It was argued in [5,6] that the Ag nanoclusters in this oxyfluoride glass host are dispersed in the fluorite-type lattice and their growth is nucleated by the F- vacancies. These F- vacancies/F-centers/color centers are the intrinsic vacancies/defects in the fluorite lattice matrix, e.g. in review . The arguments were based, in particular, on the temperature dependence for the nucleation and growth of Ag nanoclusters/nanoparticles in this oxyfluoride glass and inevitability of the fluorite component for the dispersion of the Ag nanoclusters in the glass. It was also argued in other works, that the luminescent Ag nanoclusters comprise small amounts of only 2-4 Ag atoms, e.g. in [14,15] and refs therein. Therefore, we have undertaken a quantum chemistry investigation of the electronic structure of Ag nanoclusters consisting of only 2-4 Ag atoms using a density functional theory (DFT). As an experimental proof for the structure of Ag nanoclusters, electron spin resonance (ESR) study has been carried out by means of the methods and setups developed in , which has shown that the Ag nanoclusters in these oxyfluoride glasses remain diamagnetic from T = 4 K up to the room temperature. Thus, a conclusion was made , that the most plausible candidates for Ag nanoclusters in this glass host are the Ag42+ diamagnetic tetramers (or spin-paired Ag2+ dimers), because other Ag nanoclusters consisting of 2 to 4 silver atoms either do not enter the fluorite lattice due to the charge compensation criterion , or are paramagnetic, or their calculated excitation and emission spectra locate far in the UV or IR part of the spectrum , i.e. far from the experimental data of Fig. 1.
Figure 4(a) shows the Ag42+ tetramer embedded in the fluorite-type lattice of MF2, where M stands for two-valence cations such as Pb and/or Cd , which form fluorite lattice component in this glass as indicated by Raman scattering  and optical spectroscopy  data. Due to disorder in the glassy state, only the short-range order of the crystalline fluorite lattice will be preserved in this counterpart glass, as argued, e.g. in . Figure 4(b) shows a closer view of the Ag42+ tetramer surrounded by the nearest F- atoms and encompassing two F- vacancies resulting in the nanocluster [Ag4M4(HF)34]10+, which is further investigated by means of DFT. As it was mentioned above, the F- vacancies are required for nucleation and charge compensation of the Ag42+ . Due to F- vacancies, the Ag42+ becomes elongated as shown schematically in Figs. 4(a) and 4(b). In the DFT calculation, the external bonds on the terminating fluorine ions of this nanocluster were saturated by attaching one H+ ion to each F-. Ag nanoclusters consisting of larger amounts of silver atoms have been also computed by DFT in this work, however the Ag42+ based nanoclusters depicted in Fig. 4(b) were found to provide a best qualitative fit to the steady state spectra and kinetic reported in this paper.
DFT calculations were carried out using the Gaussian 09 package . In these calculations, the 2-double-zeta basis set LANL2DZ  has been used in conjunction with the long range corrected Perdew–Burke–Ernzerhof functional (LC-wPBE) . The time-dependent DFT approach (TD-DFT)  has been used to model the observed vertical optical transitions. The molecular orbitals involved in the electronic transitions have been also analyzed and depicted using the Gaussview program (not shown here) .
Figure 5(a) shows the calculated lowest electronic energy surfaces, or configuration coordinate diagram (CCD), of the embedded Ag42+ tetramer as a function of its two diagonals, longer, d1, and shorter, d2. Figure 5(b) shows the cross-section of the energy surfaces from Fig. 5(a) viewed along d1. In Fig. 5, we show only the lowest electronic states which only can contribute to the experimental excitation and emission spectra of the visible range, as shown in Fig. 1. S0 is the ground and S1 and S2 are excited singlet states, while and T1 and T2 are the excited triplet states, respectively. The S2 state will also not be considered in further discussion because it lies far in UV, i.e. far from the experimental conditions of Fig. 1.
Excitation of Ag nanoclusters emission occurs by the fast S0→S1 (spin-allowed) transition, as depicted by up-headed arrow in Fig. 5(b). This is because the rise kinetics of Ag nanoclusters luminescence presented in Figs. 2 and 3, have typical times substantially shorter than 1 ns, indicating spin-allowed process for the excitation.
Five possible emission transitions for the excited states S1, T2, and T1 in Fig. 5 are the following transitions: S1→S0 (spin-allowed), S1→T1 (spin-forbidden), T2→S0 (spin-forbidden), T2→T1, (spin-allowed), and T1→S0.(spin-forbidden); they are shown by down-headed arrows in Fig. 5(b). The singlet-singlet and triplet-triplet transitions are spin-allowed, therefore they have relatively short lifetimes; while the triplet-singlet transitions are spin-forbidden, i.e., they have relatively longer lifetimes, e.g. in [12,13] and refs therein; this difference in lifetimes will explain the nanosecond to microsecond range of experimental lifetimes reported in Figs. 2 and 3 and Tables 1 and 2.
These emission transitions, as seen in the energy scale of Fig. 5(b), occur either around the blue range of the spectrum, S1→S0 (fast transition); the green-yellow range, T2→S0 (slow transition); the yellow-red range, T1→S0 (slow transition); infrared range, S1→T1 (slow transition); and far infrared range, T2→T1 (fast transition). The last two transitions have relatively small energy/quantum therefore they are quenched non-radiatively, and thus do not contribute to the observed experimental emission spectra, e.g. in Fig. 1.
Also, the near-infrared transitions S1→T1 and T2→T1 can be quenched by cross-relaxation energy transfer processes (S1,S0→T1,T1) and (T2,S0→T1,T1), respectively, resulting in increase of yellow-red emission band T1→S0. These energy transfer processes are caused by multipole interaction between Ag nanoclusters in the ground, S0, and excited, S1 and T2, states, following the mechanism of Inokuti-Hiroyama as in the case of rare earth dopants ( and refs therein).
Moreover, ultrafast intersystem crossing is known to occur in luminescence of molecules, e.g. in [12,13] and refs therein, whilst the Ag nanocluster can be treated as a molecule. In the inter-system crossing, the excitation energy transfers from the singlet to the triplet state due to intersection of the singlet, S1, and triplet, T2, states in the CCD, as seen in Fig. 5. The intersystem crossing may be an ultrafast process of the order of 10−12 s, which is defined by the attempt frequency for the hops from the singlet to the triplet state [12,13].
Thus, the microsecond emission kinetics and their wavelength detection dependence reported in Fig. 2 and Table 1 is accounted for by the CCD diagrams in Fig. 5 when this kinetics correspond to the above mentioned forbidden triplet-singlet transitions T2→S0 and T1→S0, and perhaps to singlet-triplet transition S1→T1; the relative amplitudes of these transitions can be evaluated from the steady state emission spectra of Fig. 1. As for the observed shortening of lifetime with shortening of excitation wavelength, Fig. 2 and Table 1, this can be also explained within the CCD of Fig. 5. Indeed, the shorter-wavelength excitation at 355 nm will transfer the system into a high excited vibration state of S1 where the probability of intersystem crossing to T2 will be low due to mismatch of their energy. As a result, the luminescence will occur predominantly by an allowed transition S1→S0 into excited vibration states of S0 giving rise to a broad luminescence band with shorter nanosecond lifetime. By contrast, a longer-wavelength excitation at 420 nm will excite the system in a low-lying vibration state of S1, which will be closer to the crossing point of S1 and T2, resulting in more efficient inter-system crossing and longer lifetime for the respective forbidden transitions T2→S0 (green-yellow) and T1→S0 (red). These transitions are responsible for the observed slow kinetic of luminescence in the microsecond range.
The nanosecond emission kinetics and their wavelength detection dependence reported in Fig. 3 and Table 2 is also accounted for by the CCD diagrams in Fig. 5. This kinetics corresponds to the allowed S1→S0 transition, which occurs without inter-system crossing, as argues above for the case of 355 nm excitation wavelength.
The double exponent decay of luminescence in nanosecond, Fig. 2 and Table 1, and microsecond, Fig. 3 and Table 2, ranges may indicate several, or maybe only two, non-equivalent Ag42+ sites in the fluorite lattice; which may be distinguished for instance by elongation along the diagonal of the Ag42+ tetramer. The fast decays components presented in Tables 1 and 2 account for about 80% and the slow components for about 20%, respectively. The slow component may also originate from the energy hops between Ag nanoclusters as has been proposed earlier . Alternatively, intrinsic defects/traps of the glass host may trap photoexcited electrons and return them back to the Ag nanoclusters. Such mechanism for luminescence accompanied by intermittency/blinking of the luminescence is well known for semiconductor quantum dots and some molecules when the luminescence is detected from the single dot/molecule, e.g. in [26–28] and refs therein. The characteristic lifetime of such luminescence varies in the broad range accounting for the slow decay components presented in Tables 1 and 2.
Finally, in this calculation and interpretation of luminescence lifetimes, we considered the Ag42+ nanoclusters because the previous work  based on DFT and complete active space perturbation theory of second order (CASPT2) showed that namely these nanoclusters provide the best fit to the experimental excitation spectra of Ag nanoclusters luminescence and absence of electron spin resonance (ESR) signal (diamagnetism of Ag nanoclusters) in this glass host; some preliminary data on this have been reported also in .
Time resolved study of the luminescence of Ag nanoclusters dispersed in the oxyfluoride glass host has been undertaken from the nanosecond to microsecond ranges. The observed luminescence lifetimes have been interpreted in terms of a newly proposed energy level configuration coordinate diagram for Ag nanoclusters. The diagram involves the overlapping singlet and triplet states and intersystem crossing between these states, which when combined, accounts for an extremely broad distribution range of luminescence lifetime and the breadth of emission band of the Ag nanoclusters.
We are grateful to the support from the Methusalem Funding of the Flemish Government. We also acknowledge the support of Agencia Canaria de Investigación, Innovación y Sociedad de la Información, Gobierno de Canarias (SolSubC200801000088, SolSubC200801000286 and FPI grant PI97718301), and Ministerio de Ciencia e Innovacion (MAT2009-12079). JJV acknowledges support from INPAC, KU Leuven. NTC thanks the Government of Vietnam for a doctoral scholarship.
References and links
3. Z. Shen, H. Duan, and H. Frey, “Water-soluble fluorescent Ag nanoclusters obtained from multiarm star poly(akrilic) acid as molecular hydrogel templates,” Adv. Mater. 19(3), 349–352 (2007). [CrossRef]
4. I. Díez, M. Pusa, S. Kulmala, H. Jiang, A. Walther, A. S. Goldmann, A. H. E. Müller, O. Ikkala, and R. H. A. Ras, “Color tuneability and electrochemiluminescence of silver nanoclusters,” Angew. Chem. Int. Ed. 48(12), 2122–2125 (2009). [CrossRef]
5. V. K. Tikhomirov, V. D. Rodríguez, A. Kuznetsov, D. Kirilenko, G. Van Tendeloo, and V. V. Moshchalkov, “Preparation and luminescence of bulk oxyfluoride glasses doped with Ag nanoclusters,” Opt. Express 18(21), 22032–22040 (2010). [CrossRef] [PubMed]
6. A. S. Kuznetsov, N. T. Cuong, V. K. Tikhomirov, M. Jivanescu, A. Stesmans, L. F. Chibotaru, J. J. Velázquez, V. D. Rodríguez, D. Kirilenko, G. Van Tendeloo, and V. V. Moshchalkov, “Effect of heat-treatment on luminescence and structure of Ag nanoclusters doped oxyfluoride glasses and implication for fiber drawing,” Opt. Mater. 34(4), 616–621 (2012). [CrossRef]
7. V. D. Rodríguez, V. K. Tikhomirov, J. Mendez-Ramos, A. C. Yanes, and V. V. Moshchalkov, “Towards broad range and highly efficient down-conversion of solar spectrum by Er3+-Yb3+ co-doped nano-structured glass-ceramics,” Sol. Energy Mater. Sol. Cells 94(10), 1612–1617 (2010). [CrossRef]
8. V. K. Tikhomirov, D. Furniss, A. B. Seddon, I. M. Reaney, M. Beggiora, M. Ferrari, M. Montagna, and R. Rolli, “Fabrication and characterization of nanoscale, Er3+-doped, ultratransparent oxyfluoride glass-ceramics,” Appl. Phys. Lett. 81(11), 1937–1939 (2002). [CrossRef]
9. V. K. Tikhomirov, T. Vosch, E. Fron, V. D. Rodríguez, J. J. Velázquez, D. Kirilenko, G. Van Tendeloo, J. Hofkens, M. Van der Auweraer, and V. V. Moshchalkov, “Luminescence of oxyfluoride glasses co-doped with Ag nanoclusters and Yb3+ ions,” RCS Adv. 2(4), 1496–1501 (2012).
10. A. C. Yanes, J. Del Castillo, M. Torres, J. Peraza, V. D. Rodríguez, and J. Méndez-Ramos, “Nanocrystal-size selective spectroscopy in SnO2: Eu3+ semiconductor quantum dots,” Appl. Phys. Lett. 85(12), 2343–2345 (2004). [CrossRef]
11. W. Chen, A. G. Joly, and J. Roark, “Photostimulated luminescence and dynamics of AgI and Ag nanoclusters in zeolites,” Phys. Rev. B 65(24), 245404 (2002). [CrossRef]
12. D. A. McQuarrie and J. D. Simon, Physical Chemistry, a Molecular Approach (University Science Books, Sausalito, CA, 1997).
13. M. T. Colvin, A. L. Smeigh, E. M. Giacobbe, S. M. M. Conron, A. B. Ricks, and M. R. Wasielewski, “Ultrafast intersystem crossing and spin dynamics of zinc meso-tetraphenylporphyrin covalently bound to stable radicals,” J. Phys. Chem. A 115(26), 7538–7549 (2011). [CrossRef] [PubMed]
14. L. Maretti, P. S. Billone, Y. Liu, and J. C. Scaiano, “Facile photochemical synthesis and characterization of highly fluorescent silver nanoparticles,” J. Am. Chem. Soc. 131(39), 13972–13980 (2009). [CrossRef] [PubMed]
15. I. Rabin, W. Schulze, and G. J. Ertl, “Light emission during the agglomeration of silver clusters in noble gas matrices,” J. Chem. Phys. 108(12), 5137–5142 (1998). [CrossRef]
16. S. Hull, “Superionics: crystal structures and conduction processes,” Rep. Prog. Phys. 67(7), 1233–1314 (2004). [CrossRef]
17. A. Stesmans, “Structural relaxation of Pb defects at the (111)Si/SiO2 interface as a function of oxidation temperature: The Pb-generation-stress relationship,” Phys. Rev. B Condens. Matter 48(4), 2418–2435 (1993). [CrossRef] [PubMed]
18. N. T. Cuong, V. K. Tikhomirov, L. F. Chibotaru, A. Stesmans, V. D. Rodríguez, M. T. Nguyen, and V. V. Moshchalkov, “Experiment and theoretical modeling of the luminescence of silver nanoclusters dispersed in oxyfluoride glass,” J. Chem. Phys. (to be published).
19. V. K. Tikhomirov, A. B. Seddon, M. Ferrari, M. Montagna, L. F. Santos, and R. M. Almeida, “The structure of Er3+-doped oxy-fluoride transparent glass-ceramics studied by Raman scattering,” Europhys. Lett. 64(4), 529–535 (2003). [CrossRef]
20. S. R. Elliott, “Medium-range structural order in covalent amorphous solids,” Nature 354(6353), 445–452 (1991). [CrossRef]
21. M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuseria, M. A. Robb, J. R. Cheeseman, G. Scalmani, V. Barone, B. Mennucci, G. A. Petersson, H. Nakatsuji, M. Caricato, X. Li, H. P. Hratchian, A. F. Izmaylov, J. Bloino, G. Zheng, J. L. Sonnenberg, M. Hada, M. Ehara, K. Toyota, R. Fukuda, J. Hasegawa, M. Ishida, T. Nakajima, Y. Honda, O. Kitao, H. Nakai, T. Vreven, J. A. Montgomery, Jr., J. E. Peralta, F. Ogliaro, M. Bearpark, J. J. Heyd, E. Brothers, K. N. Kudin, V. N. Staroverov, R. Kobayashi, J. Normand, K. Raghavachari, A. Rendell, J. C. Burant, S. S. Iyengar, J. Tomasi, M. Cossi, N. Rega, J. M. Millam, M. Klene, J. E. Knox, J. B. Cross, V. Bakken, C. Adamo, J. Jaramillo, R. Gomperts, R. E. Stratmann, O. Yazyev, A. J. Austin, R. Cammi, C. Pomelli, J. W. Ochterski, R. L. Martin, K. Morokuma, V. G. Zakrzewski, G. A. Voth, P. Salvador, J. J. Dannenberg, S. Dapprich, A. D. Daniels, O. Farkas, J. B. Foresman, J. V. Ortiz, J. Cioslowski, and D. J. Fox, Gaussian 09, Revision A.02 (Gaussian, Inc., Wallingford CT, 2009).
22. P. J. Hay and W. R. Wadt, “Ab initio effective core potentials for molecular calculations. Potentials for the transition metal atoms Sc to Hg,” J. Chem. Phys. 82(1), 270–283 (1985). [CrossRef]
24. R. Bauernschmitt and R. Ahlrichs, “Treatment of electronic excitations within the adiabatic approximation of time dependent density functional theory,” Chem. Phys. Lett. 256(4-5), 454–464 (1996). [CrossRef]
25. K. Driesen, V. K. Tikhomirov, C. Gorller-Walrand, V. D. Rodriguez, and A. B. Seddon, “Transparent Ho3+-doped nano-glass-ceramics for efficient infrared emission,” Appl. Phys. Lett. 88(7), 073111 (2006). [CrossRef]
26. M. Nirmal, B. O. Dabbousi, M. G. Bawendi, J. J. Macklin, J. K. Trautman, T. D. Harris, and L. E. Brus, “Fluorescence intermittency in single cadmium selenide nanocrystals,” Nature 383(6603), 802–804 (1996). [CrossRef]
27. P. Frantsuzov, M. Kuno, B. Janko, and R. A. Marcus, “Universal emission intermittency in quantum dots, nanorods, and nanowires,” Nat. Phys. 4(5), 519–522 (2008). [CrossRef]