J. MATER. CHEM., 1994, 4(8), 1349-1350 MATERIALS CHEMISTRY COMMUNICATIONS Red Bismuth Emission in Alkaline-earth-metal Sulfates Mariska A. Hamstra, Hilda F. Folkerts and George Blasse Debye Institute, Utrecht University, Postbox 80.000, 3508 TA Utrechf, The Netherlands The red luminescence of bismuth-doped alkaline-earth-metal sulfates is ascribed to the presence of divalent bismuth. The optical transitions are due to transitions within the 6p shell. The Bi3+ ion is known to show an ultraviolet, blue or sometimes even green luminescence.' In 1886 Lecoq de Boisbaudran reported a red luminescence for bismuth in the alkaline-earth-metal sulfates,, which was confirmed later by Kroger et aL3 The latter authors suggested that this lumi- nescence may be due to trivalent or pentavalent bismuth.Here we propose a very different explanation. We were led to this via two ways. First, our research on the luminescence of (s), ions in sulfates led us to a reinvestig- ation of the red emission mentioned above. Secondly, an orange emission for bismuth-doped SrB,O, was reported recently from this laboratory, and shown to be due to divalent bismuth with electron configuration (6s),( 7s)'. Our present results show that the red emission in the sulfates has the same origin. MSO, :Bi (M =Ca, Sr, Ba) samples were prepared by conventional solid-state techniques with firing up to 900 "C. The bismuth concentration was ca. 0.1 mol%. The samples were shown to be single phase by X-ray powder diffraction using Cu-Kcr radiation.The optical measurements were per- formed as described before., The alkaline-earth-metal sulfates show a red photolumin- escence of medium intensity for Ba and Sr, whereas for Ca it is very weak and disappears with time. Fig. 1 shows as an example the emission and excitation spectra of BaSO, :Bi. The decay time of this emission is practically temperature independent and amounts to 4 ps. The data for CaSO,: Bi were very hard to measure and are considered to be inaccurate. The emission spectra agree with those reported (half) a century ag0.~3~ Table 1 summarizes our results. Some samples also showed a bluish emission that could only be excited in the shorter-wavelength UV region, the presence of which depends on the preparation conditions and on the bismuth concentration.At room temperature this 300 400 500 600 hlnm Fig. 1 Emission (---) and excitation (-j spectra of the luminescence of BaSO, : Bi2+ at 4.2 K. Excitation wavelength 455 nm, monitored emission wavelength 625 nm. 0gives the radiant power per constant wavelength interval, and qr the relative quantum output, both in arbitrary units. emission is usually quenched. Kroger et aL3 reported similar results. There is a striking analogy between the results for MSO, :Bi M=Sr, Ba and SrB,O,:Bi (Table 1). This relates also to the values of the decay times of the emissions involved. This analogy is even more pronounced if one considers the fact that the Stokes shift in the SrB,O, host lattice is a1 ways very small., For this reason we assume that the red emission of bismuth in the alkaline-earth-metal sulfates is due to divalent bismuth.Its electron configuration is (SS)~(6p)', vielding a 6P1/2ground state and a crystal-field-split 6P3/2excited state. In Table 1 the two crystal-field levels are denoted (I) and (2) in order of increasing energy. Since the emission is a 6p intraconfigurational transition C2P3I2(l)-+2Pl/2],it is formally parity forbidden. Tht. observed decay time (4ps) points indeed to a partly forbidden rransition. Since the uneven crystal-field terms mix with the (6~)~(7s)', ,SI1, and the 2P3/2and 2Pljzstates,, the parity selection rule becomes partly lifted. The emission and the lowest excitaticbn band [2P1/2-2P3/2(1)]show a side band at ca.1050 cm-' lower and higher energy, respectively (see Fig. 1). This can bc. ascribed to a vibronic transition involving the asymmetric a1 sulfate stretching vibration which is at ca. 1100 cm-' in the infrared spe~trum.~This transition has also been observed in SrB,07: Bi2+ where it is at 1200 cm-' lower or higher energy, corresponding to the borate stretching vibrations. In CaSO, :Bi the divalent state of bismuth is obviously less stable. The reason for this can be the smaller size available, or the hydration of CaSO, yielding oxidation of Bi2+. Although the data for CaSO, :Bi are intriguing in comparison with the others in Table 1, we do not discuss these further in view of their inaccuracy. The much larger Stokes shift is not unusual if one compares with data observed for isoelectronic Pb+.In CaF,: Pb+ a Stokes shift of ca. 5000 cm-' has been reported.6 The observed blue emissions are ascribed to Bi", which may be present in the host lattice or in a second phase. Unfortunately the corresponding absorption band is situated in the shorter-wavelength UV region where it is expected to overlap with the allowed (6s),( 6p)l+( 6s),( 7s)' transition on the Bi2+ ion. As in SrB,07:Bi2+, we observed the red Bi2+ emission also upon excitation of the sulfates in this region: see, e.g., the excitation band at 260 nm in Fig. 1. This shows that the (6s),(6p)' --+(SS)~(~S)' transition is in fact situated in this region (ca. 40000 cm-'). Finally, we note that Kroger et aL3 reported a similar red luminescence for bismuth-activated CaP,O,, Ca2P207, Sr,P,07, SrP206 and Ca3(P04)2.The emission maxima vary from 650 to 700 nm. We assume that this emission is also due to divalent bismuth, This shows that Bi2+ is less rare than suggested by inorganic chemistry textbooks. Kroger et aL3 J. MATER. CHEM., 1994, VOL. 4 Table 1 Positions of some optical transitions of Biz+ in several host lattices (all values in lo3 cm-') composition absorption 2p1,2+zp3,2(11 zpl/2 +2p3/2 SrB,O, :Biz+ 17.4 21.2 BaSO, :Biz+ 17.0 22.0 SrSO, :Biz+ 17.4 21.7 CaS04:Bi2+ 20.0 23.6 This work; this work; uncertain values. obtained efficient cathodoluminescence for the red-emitting sulfates and phosphates.No doubt the Bi2+ ion can very efficiently capture a hole. If an electron recombines with this captured hole, the red bismuth emission occurs. In conclusion, there is ample evidence that Bi2+ can be substituted up to amounts of ca. 0.1 mol% in several alkaline- earth-metal compounds. References 1 G. Blasse, Prog. Solid State Chem., 1988,18, 79. 2 Locoq de Boisbaudran, C.R. Acad. Sci. Paris, 1886,103,629. emission Stokes ref. shift (2) 17.1 0.3 16.0 1.o 16.4 1.o 16.2 3.8 3 F. A. Kroger, J. Th. G. Overbeek, J. Goorissen and J. van den Boomgaard, Trans. Electrochem. Soc., 1949,96, 132. 4 G. Blasse, A. Meijerink, M. Nomes and J. Zuidema, J. Phys. Chem. Solids, 1994,55, 171. 5 L. H. Brixner, M. K. Crawford and G.Blasse, J.Solid State Chem., 1990, 85, 1. 6 M. Fockele, F. Lohse, J. M. Spaeth and R. H. Bartram, J. Phys. Condens. Matter, 1989, 1, 13; M. Fockele, F. J. Ahlers, F. Lohse, J. M. Spaeth and R. H. Bartram, J. Phys. C, 1985,18,1963. Communication 4/03131H; Received 25th May, 1994