of the Less-Common
FUTURE TECHNOLOGICAL DOPED MATERIALS*
93 (1983) 243-251
Department of Znorganic and Analytical 91904 (Israel)
Chemistry, The Hebrew University of Jerusalem, Jerusalem
(Received March 21,1983)
Summary The possible uses of rare-earth-doped glasses or crystals are discussed in connection with (a) luminescent solar concentrators (LSCs), (b) light sources for fibre optics and (c) glass lasers. LSCs can be made from glasses which absorb the major part of the solar spectrum and concentrate the emitted fluorescence at wavelengths to which the photovoltaic cells are most sensitive. The ways in which such glasses can be prepared using a combination of UOZ” + or Cr3 + with Nd3 + and Yb3 + are discussed. An economical analysis of the new type of collector is given. The use of Er3+ in fluoride glasses for fibre waveguide sources is outlined and recent results are presented. References are made to new laser materials based on Er3 + ahd Ho3 + .
1. Introduction The industrial uses of rare earths have been summarized in an excellent book edited by Gschneidner Cl]. In the present paper future possible applications based on the unique properties of rare earth ions arising from their optical spectra, which resemble narrow transitions in atomic spectroscopy, are discussed. The two most important fields for the future needs of society are renewable energy sources and communication media. We shall see that rare earth ions can contribute significantly in these two fields. The topics discussed in this paper have not yet been applied in practice but the research results are quite encouraging. We shall therefore concentrate on the scientific approach and the laboratory results obtained to date including investigations of luminescent solar concentrators (LSCs), light sources for fibre optics and glass lasers. *Paper presented at the Sixteenth Rare Earth Research Conference, The Florida State University, Tallahassee, FL, U.S.A., April l&21,1983. cr Elsevier Sequoia/Printed
in The Netherlands
The world’s conventional energy supplies, which are based mainly on readily available fossil fuel sources, are diminishing rapidly. The main approach to the energy crisis, nuclear fusion, is raising a great deal of hope but its practicability has still to be demonstrated. There is no doubt that solar energy, which is clean and non-hazardous, could contribute considerably to a solution of the energy problem if appropriate methods were developed to collect, concentrate, store and convert solar radiation which is diffuse and intrinsically intermittent . Owing to the original efforts of the National Aeronautics and Space Administration to supply electric current from silicon photovoltaic (PV) cells to space vehicles, such devices are now available at a cost of about $8 per watt of power. At present large-scale solar cell arrays are operating in inaccessible locations distant from conventional electricity plants. Previous estimates of price decreases to $l-$2 W- ’ in 1984, which were obtained by making comparisons with the aluminium or electronic computer industries, may be slightly optimistic as the difficulties of preparing inexpensive silicon with a high photoelectric yield cannot easily be removed by increased production. One way of lowering the price of PV electricity is to concentrate the solar radiation, particularly that part which is most efficient in PV energy conversion. It is hoped that this can be achieved with LSCs [Z]. The operation of an LSC is based on the absorption of solar radiation in a collector containing a fluorescent species in which the emission bands have little or no overlap with the absorption bands. The fluorescence emission is trapped by total internal reflection and concentrated at the edges of the collector which is usually a thin plate . LSCs have the following advantages over conventional solar concentrators: they collect both direct and diffuse light; there is good heat dissipation of non-utilized energy by the large area of the collector plate in contact with air so that essentially “cold light” reaches the PV cells; tracking the sun is unnecessary; the luminescent species can be chosen to allow matching of the concentrated light to the maximum sensitivity of the PV cell. 3. Factors governing the performance tors
The performance of an LSC can be compared with that of a conventional concentrator by using the effective concentration ratio. This is given by the geometric concentration ratio multiplied by the optical conversion efficiency of the collector. The geometrical factor is the ratio of the surface area A, to the area A, of the plate edges. The optical plate efficiency qop, is defined as the ratio of light delivered from all the edges to the light incident on the collector plate: IIopt= Pcl”JPi” where Pi, is the total solar power in watts incident on the collector total power emitted from the edges.
(1) and PO,,is the
The optical plate efficiency depends on the following factors: (1) the fraction ylabsof light absorbed; (2) the quantum efficiency ye,of the fluorescent species; (3) the Stokes efficiency qs which is the ratio of the average energy of emitted photons to the average energy of the absorbed photon and is given by
(4) the fraction q, of the light trapped in the collector
where n is the refractive index of the light-emitting medium; (5) the transport efficiency qtr which takes the transport losses due to matrix absorption and scattering into account; (6) the efficiency qsel due to losses arising from selfabsorption of the colourants. The expression for the optical efficiency, including reflection, is given by rlopt= (I-
where R is the Fresnel reflection coefficient and is about 4%. The ultimate requirements of the collector plates are their long-term stability towards photodegradation and corrosion and the absence of selfabsorption of the emitted light by the material of the plate. These last two factors can apriori be achieved by using highly fluorescent inorganic species of which the rare earth ions are the best representatives. Separation between absorption and emission can be overcome by utilizing the fact that some of the rare earth ions emit to electronic levels positioned at energies above the ground state so that self-absorption is prevented. The drawback of rare earth ions lies in their low absorption coefficients originating from the parity-forbidden f-f transitions. In order that these ions can be utilized in LSCs the energy should be absorbed by species with higher transition probabilities and transferred to the rare earth ions, or alternatively they have to be incorporated into transparent media in which the transition probabilities increase. The first attempt to use neodymium ions for solar collectors was made by Levitt and Weber  who measured devices consisting of Owen-Illinois ED-2 neodymium-doped laser glass. The efficiency of this glass for solar collection is rather low because of the low absorption of Nd3 + in this glass. This problem can be circumvented by incorporating Nd3 + into tellurite glasses. Calculated plate efficiencies of Nd3+ in tellurite glasses based on optical measurements of the absorption and emission spectra and the quantum efficiency of Nd3+ in these glasses are 12%, which is three times higher than in conventional silicate glasses [4,5]. Reisfeld and Kalisky [S] have shown that a, combination of UO,*+ with Nd3+ or Ho3+ in glasses extends the spectral sensitivity range of LSCs as a result of the efficient energy transfer from the 20 500 cm- 1 excited level of the uranyl ion. Both energy transfer and direct excitation to the Nd3 + levels provide the well-known laser transition  4F3,2 + 4I11,2 (1060 nm), with a branching ratio around 0.5, as well as the transition 4F3,2 + 419,2 (880 nm) to the ground
state. The latter transition is resonant and repopulates the 4F3,2 level, and thus little energy is lost by reabsorption. Both transitions lie in the spectral range in which the silicon solar cells have high spectral sensitivity. A different method of sensitizing Nd3+ and Yb3+ emissions by the Cr3 + ion has been described recently [&lo]. The absorption spectrum of Cr3 + in glasses (and crystals) consists of two broad bands covering most of the solar spectrum with peaks at 450 nm (4A, + 4T,) and at 650 nm (4A, + 4T,). Glasses provide low field Cr3 + sites in which the zero-phonon 4T, level lies below the zero-phonon 2E state. In these low field cases the room temperature Cr3 + emission from 4T2 consists of a broad unstructured band centred in the near IR. Quantum yields for luminescence are low in glass hosts; for example the highest quantum yield of Cr3+ found to date in glass is 11°?&170~(silicate glass)
The best efficiencies of 23% for 4T2 + 4A are obtained for lithium lanthanum phosphate (LLP) glasses [lo]. The explanation for the noydiative transitions of Cr3+ in glasses is still being investigated by Lempicki’s group at General Telephones and Electronics Laboratories and by our group. At this point it is absolutely clear that the theory of multiphonon relaxation which holds for rare earth ions cannot be applied to Cr3+. Our present studies [lo, 121 show that the probability of energy transfer from Cr3 + to Nd3 + in LLP may be higher than that of the non-radiative transfer within the Cr3+ system. This is probably the reason for the good laser performance of such glasses as reported recently by Hlrig et al. r131. The integrated lifetime at room temperature is 24.6 ps which is indicative of the proximity of 4T2 and ‘E states in the LLP glass . Figure 1 shows the absorption spectrum of 0.31 mol.% Cr,O, in LLP glass. Figure 2 shows the absorption spectra of 0.15 mol.% Cr203, 2 mol.% Nd203 and 0.15 mol.% Cr,O, + 2 mol.% Nd,O, in LLP glass. It can be seen from Fig. 2 that the absorption is additive. The same is true for LLP glass with Cr,O, and Yb,O,
Fig. 1. Absorption
600 h (nm)
spectrum of 0.31 mol.% Cr3+ in LLP (8.4 x lOI Cr3* ions cmw3).
Fig. 2. Absorption spectra of Cr3’, Nd3+ and Cr3+ +Nd3+ in LLP (4.2 x 1OL9 Cr3+ ions cm-‘; 5.44 x 10” Nd3+ ions cm-3): curve a, 0.15 mol.% Cr +2mol.% Nd; curve b, 2 mol.% Nd; curve c, 0.15 mol.“A Cr.
additions. Figure 3 shows the absorption spectra of 0.31 mol.% Cr,O, and 3 mol.% Yb,O, in LLP glass. Figure 4 presents the emission spectra of Cr3+ in LLP glass (curve a is for 0.05 mol.% Cr,03 and curve b is for 0.31 mol.% Cr,O,) arising from the 4T, + 4A, transition. Figure 5 shows the emission spectra of Nd3+ excited directly at 585 nm to Nd3+ or indirectly at 647 nm via energy transfer from Cr3 ‘. This curve provides evidence for energy transfer from Cr3 + to Nd3 + . Figure 6 shows the emission spectra of Yb3+ excited directly at 915 nm into the 7F,,2 -+ 2F5,2 transition (Fig. 6(a)) and excited via Cr3+ into 4T2 (Fig. 6(b)). Figure 6(b) again provides evidence for energy transfer between Cr3’ and Yb3+.
8 E :: P 2
a, .? z f ,
Fig. 3. Absorption spectrum of 0.31 mol.% Cr3++3.0 cmm3+8.3 x 10ZoYb3+ ionscme3).
900 X (nm)
mol.% Yb3+ in LLP (8.4x 1Ol9 Cr3+ ions
Fig. 4. Emission spectra of Cr’+ (excited at 647 nm): curve a, 0.05 mol.% Cr; curve b, 0.31 mol.% Cr.
Fig. 5. Emission spectra of Nd3 + in LLP: curve a, 0.15 mol.% Cr +2 mol.% Nd (excited at 585 nm); curve b, 2 mol.% Nd (excited at 585 nm); curve c, 0.15 mol.% Cr + 2 mol.% Nd (excited at 647 nm). Fig. 6. Emission spectra of Yb3+ in LLP (0.5 mol.% Yb3+ = 1.36 x 10” Yb3+ ions cmm3; 3.0 mol.% Yb3+ ~8.31x1020Yb3fionscm~3):---,0.3mol.~Cr+0.5mol.~Yb;---,0.3mol.~Cr+3.0mol.~~ Yb; ., ytterbium absorption.
Energy transfer efficiencies of 92% between Cr3 + and Nd3 + and of 38% between Cr3 + and Yb3+ have been obtained in LLP glasses. It is not surprising that the energy transfer efficiencies between Cr3 + and Nd3 + are higher than the corresponding transfer in Yb ’ + because there is a much higher spectral overlap between Cr3+ and Nd3 +.
The use of porous supports in which rare earths can be incorporated as ions or complexes at room temperature in glassy materials for LSCs has been reported recently by our group . Porous materials [lS, 171 can be used to produce transparent plates if the average pore size is much smaller than the wavelength of the incident solar light. We used the following supports: (1) Vycor porous glass with an average pore diameter of 40-70 A which is sufficiently large for the penetration of rare earth ions; (2) compressed pyrogenic silica gel (Cab-0-Sil, Aerosil) (when compressed under a load of 4-10 tf cm-’ the fine spherical particles (7-14 pm) yield porous transparent discs). To the best of our knowledge this work is the first investigation of the application of impregnated porous materials to solar collectors. Europium trisbenzoylacetonate adsorbed on Vycor porous glasses or compressed pyrogenic silica discs shows a fluorescence spectrum characteristic of Eu3+ when excited via the ligand. The quantum efficiency of the Eu3 + emission from this complex adsorbed on porous supports is higher than that from a similar complex of the same optical density in a solution of dichloromethane. The emission spectra of Pr3+, Sm3+, Eu3+ , Tb3+ ,Y,D 3+ Ho3+ , Er3+ and Tm3 + on porous Vycor glass or Cab-0-Sil exhibit strong luminosity owing to the fluorescence of the rare earth ions adsorbed on these supports. The increase in luminosity compared with the fluorescence efficiency of these ions in solution is due to a decrease in the non-radiative relaxation governed by OH stretching frequencies. As already mentioned these new types of material provide a new approach to LSCs in which fluorescent species can be incorporated at room temperature. 4. Light sources for fibre optics The development of an optical fibre communication system is extremely desirable since the transmission capacity per unit cross-sectional area of optical fibres is expected to be from ten to 100 times larger than that of conventional cables, As a result of highly developed techniques of glass fibre production for optical waveguides the limiting factor in the attenuation is the Rayleigh scattering, which decreases as the inverse fourth power of the wavelength, and the absorption peak near 950 nm due to the OH group. Impurities other than OH groups are usually absent because extremely pure raw materials are now available. It should be noted that the material dispersion in most glasses decreases significantly between 1300 and 1600 nm. Nd3+-doped materials emitting at 1060-1370 nm or Er 3 +-doped materials emitting at 1500 nm may have an advantage over GaAs lasers as they may form an integrated source for optical communications [lo]. Rare earth sulphides form glasses with Ga,S, and Al,S, if appropriate conditions are maintained. Such glasses have been prepared by Flahaut and his group who have studied the phase diagrams of these materials in detail [Ml. The glasses 3Al,S,.La,S, (ALS) and 3Ga,S,+La,S, (GLS) are transparent between 1500 and 25 000 cm-’ and between 1500 and 20 000 cm- ’ respectively and may play an important role in optical fibre waveguides.
Laser emission cross sections of the Nd3+ emissions at 1077 and 1370 nm in 3Ga,S3*0.85La,S3.0.15Nd,S, (GLS) and 3A1,S3~0.872La,S3~0.126Nd,S, (ALS) were obtained from the calculated matrix elements of Nd3+ in the glasses and the experimentally measured intensity parameters, emission half-widths and lifetimes. The laser threshold power for side pumping in the chalcogenide glasses is much lower than for ED-2 glass. The peak cross sections of Nd3+ are 7.95 x 10m2’ cm2 at 1077 nm and 3.60 x 10m2’ cm2 at 1320nm in GLS and 8.20 x 10e2’ cm2 at 1077 nm and 4.10 x 10e2’ cm2 at 1370 nm in ALS compared with 2.90 x 10m2’ cm2 at 1060 nm and 0.72 x 10e2’ cm2 at 1300 nm for ED-2 glass
A different type of material which is transparent in the IR region of the spectrum is found in a new family of fluoride glasses developed recently . These glasses exhibit a wide range of high transparencies from the UV to the midIR (about 0.2-7 pm) and have an ultrahigh transparency in the 2-5 pm region. In addition they have a relatively low refractive index of about 1.50 in the visible region and good chemical stability with high resistance to attack by water and acids. For example fluorozirconate glasses have low losses (less than 1 dB km- ‘) in the 1.5 pm range and show an improvement over conventional silicate glasses for use in long distance (more than 100 km) fibre waveguides for transoceanic or transcontinental communications. Fluorozirconate and lead gallium zinc fluoride glasses doped with ErF, may constitute light sources when integrated with fibre optic waveguides made from undoped materials. The stimulated peak cross section of Er3+ at 1.5 urn has been found to be 0.54 x 1O-2o [Zl]. Since the undoped materials are excellent candidates for fibre optic waveguides because of their high transmittance in the IR and their chemical stability, it has been suggested that erbium-doped materials with the desired laser emission wavelengths can be used as light sources for fibre optic systems. They could be integrated into one waveguide to provide integrated fibre optic systems. 5. Recent developments
in rare earth lasers
Research on neodymium lasers increased rapidly in the 1970s because of the requirement of large neodymium glass lasers for fusion research [7,22-241. Laser-driven fusion is one approach to a long-term solution of the world’s energy supply problems as it is based on the inexhaustible fuel deuterium which is obtained from water. The unique capability of lasers to produce a very high instantaneous power density over a very small area introduces the possibility of driving thermonuclear fuel to extremely high temperatures and densities at which fusion is expected to occur. The demonstration of the scientific feasibility of the initiation of a fusion burn with the energy produced from the pellet exceeding the absorbed beam energy can be attempted with laser energies of the order of 0.33-0.5 MJ and lifetimes of a few nanoseconds. Such a laser, the Nova neodymium glass laser, is now being completed in the Lawrence Livermore Laboratory.
Neodymium-doped yttrium aluminium garnet lasers have found extensive applications as range finders and have also become standard laboratory equipment for research in photochemistry and related fields. High efficiency laser emission has also been observed from yttrium lithium fluoride doped with Er3+ and Tm3+ [25,26], Ho3+ [27-291 and Nd3+ . All glass lasers developed to date have used a rare earth as the active ion and optical pumping for excitation . Gf these, flash-lamp-pumped neodymium glass lasers are the most frequently used and the most widely investigated. The spectroscopic data needed for estimation of the laser characteristics are usually obtained from small samples . The data include absorption, emission, non-radiative relaxation, energy transfer probabilities and laser cross sections. Laser operation predictions can be made from such data without actually demonstrating laser action. Stimulated emission cross sections of neodymium vary with glass composition. We have previously shown that the amount of covalency between the glassforming medium and the neodymium ion increases significantly with the emission cross section . This fact is demonstrated by the very high cross section of Nd3+ in chalcogenide glasses. The decision about the type of glass laser to be used for a specific application depends on the emission wavelength, pulse duration, signal output and optical configuration requirements. Additional work is still needed to establish the relative merits of various glasses for lasers. This work includes investigation of spectral inhomogeneities and their effects on large signal energy extraction, laser-induced damage threshold as a function of wavelength and pulse duration for a wider range of glass compositions .
In this paper we have concentrated on applications of rare earth ions in solar concentrators, as sources for fibre optics and in lasers. Additional applications in colour television and light sources may receive increased attention in the future.
Acknowledgment The author is extremely grateful to Mrs. E. Greenberg for assistance in the preparation of the manuscript. R.R. is “Enrique Berman” Professor of Solar Energy. References 1 2
K. A. Gschneidner, Jr. (ed.), Industrial Application of Rare Earth Elements, in ACS Symp. Ser. 164 (1981). R. ReiSfeld and C. K. J$rgensen, Struct. Bonding (Berlin), 49 (1982) 1.
5 6 7 8
10 11 12 13 14 15 16
17 18 19 20 21 22 23 24 25 26 27 28 29 30 31
J. A. Levitt and W. H. Weber, Appl. Opt., 16(1979) 2684. R. Reisfeld, Development of planar fluorescent concentrators for photovoltaic systems, Annu. Rep. Joint Israel Ministry of Energy and U.S. Department of Energy, 1982 (U.S. Department of Energy, Washington, DC); U.S. Patent 4,367,267, January 4,1983. A. Kisilev and R. Reisfeld, to be published. R. Reisfeld and Y. Kalisky, Nature (London), 283 (1980) 281. R. Reisfeld and C. K. J$rgensen, Lasers and Excited States of Rare Earths, Springer, Berlin, 1977. R. Reisfeld, E. Greenberg, A. Kisilev and Y. Kalisky, in J. S. Connolly (ed.), Photochemical Conversion and Storage of Solar Energy, Academic Press, New York, 1981, p. 364. A. Lempicki, L. J. Andrews and B. C. McCollum, General Telephones and Electronics Annu. Prog. Rep., September 1, 1975December 31, 1980 (U.S. Department of Energy, Washington, DC). R. Reisfeld, in B. DiBartolo (ed.), Energy Transfer Processes in Condensed Matter, Plenum, New York, 1983. L. J. Andrews, A. Lempicki and B. C. McCollum, J. Chem. Phys., 74 (1981) 5526. R. Reisfeld, A. Kisilev and Y. Millstein, to be published. T. Harig, G. Huber and I. A. Shcherbakov, J. Appl. Phys., 52 (1981) 4450. A. Kisilev and R. Reisfeld, Sol. Energy, to be published. H. Mack, D. Avnir and R. Reisfeld, Extended Abstracts 161st Meet. Electrochemical Society, Montreal, May g-14,1982, Electrochemical Society, Princeton, NJ, p. 733. N. Manor, H. Mack, R. Reisfeld and D. Avnir, Abstracts 4th Int. Conf. on Photochemical Conversion and Storage of Solar Energy, Jerusalem, August 8-13, 1982, p. 296. R. Reisfeld, N. Manor and D. Avnir, Sol. Energy Mater., 8 (1983), in the press. J. Flahaut, in K. A. Gschneidner, Jr., and L. Eyring (eds.), Handbook on the Physics and Chemistry of Rare Earths, Vol. 4, North-Holland, Amsterdam, 1979, p. 31. A. Bornstein and R. Reisfeld, J. Non-Cryst. Solids, 50 (1982) 23. Abstracts 1st Int. Symp. on Halide and Non-oxide Glasses, Cambridge, Cambs., March 23-26,1982, p. 296. R. Reisfeld, G. Katz, C. Jacoboni, R. de Pape, M. G. Drexhage, R. N. Brown and C. K. Jtirgensen, J. Solid State Chem., to be published. M. J. Weber, in C. Stein (ed.), Critical Materials Problems in Energy Production, Academic Press, New York, 1976. D. C. Brown, High-Peak-Power Nd:Glass Laser Systems, Springer, Berlin, 1981. M. J. Weber, in K. A, Gschneidner, Jr., and L. Eyring, Handbook on the Physics and Chemistry of Rare Earths, Vol. 4, North-Holland, Amsterdam, 1979, p. 275. E. P. Chicklis, C. S. Naiman and A. Linz, Dig. Tech. Papers 7th Int. Quantum Electronics Conf., May 1972, p. 17. H. P. Jenssen, A. Linz, R. P. Leavitt, C. A. Morrison and D. E. Wortman, Phys. Rev. B, 1 I (1975) 92. N. Karayanis, D. E. Wortman and H. P. Jenssen, J. Phys. Chem. Solids, 37(1976) 675. Sh. N. Gefeisman, A. M. Tkachuk and V. V. Prizmak, Opt. Spectrosc. (U.S.S.R.), 44 (1978) 68. N. P. Barnes, D. J. Gettemy, N. J. Levinos and J. E. Griggs, Proc. LASL Optics Conf., in SPIE J., 190(1979) 297. M. G. Knights, W. F. Wing, J. W. Baer, E. P. Chicklis and H. P. Jenssen, IEEE J. Quantum Electron., 18(1982) 163. S. E. Stokowski, in M. J. Weber (ed.), Handbook of Laser Science and Technology, Vol. 1, Chemical Rubber Company, Boca Raton, FL, 1982, p. 215, and references cited therein. R. Reisfeld, Ann. Chim. Fr., 7(1982) 147. M. J. Weber, Int. Conf. on Lasers, New Orleans, December l&17,1982.