X-ray excited luminescence of oxides doped with D10 ions

X-ray excited luminescence of oxides doped with D10 ions

~~ate~als Chemistry and Physics, 28 (1991)275-279 275 X-RAY EXCITED LUMINESCENCE OF OXIDES DOPED WITH d" IONS G. BLASSE Debye Research Institute...

257KB Sizes 5 Downloads 18 Views

~~ate~als

Chemistry

and Physics,

28 (1991)275-279

275

X-RAY EXCITED LUMINESCENCE OF OXIDES DOPED WITH d" IONS

G. BLASSE Debye Research Institute, University of Utrecht, P.O. Box 80.000, 3508 TA Utrecht (The Netherlands) L.H. BRIXNER E.I. du Pont de Nemours and Company, Central Research and Development Department, Experimental Station, P.O. Box 80.356, Wilmington, Delaware 19880-0356 (USA)

Received January 8, 1991; accepted February 18, 1991

ABSTRACT The X-ray excited luminescence of CaO-Ga(III), CaO-fn(IIIf and InBOs is reported. The metal ions with dl" configuration are responsible for the luminescence, but the nature of the optical transitions remains unclear.

INTRODUCTION Recently one of us has discussed the question whether metal ions with d" configuration luminesce [l]. In that publication only literature data were used. It was shown that complexes of the type [M(d")), - [O(-II)],may show efficient luminescence at room temperature. Sometimes this has been overlooked because the excitation wavelength may be outside the conventional ultraviolet, viz. < 220 nm. The nature of the optical transitions involved is complicated and not yet understood. In order to check this further we used X-ray luminescence spectroscopy of some d1°-ion doped compositions, By using X-rays the problem of the short excitation wavelength is overcome. The compositions investigated are CaO:Ga(IX) and CaO: InfIII) and InBO,. From the literature survey in [l] the dr" ions ZnffI) and Cd(I1) were shown to luminesce in CaO. Also the Ga(IIIf and In(II1) ions luminesce in some hosts, although this luminescence was not directly correlated with the d"O ions involved [l]. The present results confirm the expectation that the dlo ions can act as luminescent centres. 0254-0584/91/$3.50

OElsevierSequoia/Printed in The Netherlands

EXPERIMENTAL The InBOs sample was prepared by a spray-drying technique described in [2] and made available to us by Dr. H.T. Hintzen. Samples of doped and undoped CaO were prepared by firing CaCO, with or without GazOs or InzOs at 1200°C. Samples were checked by X-ray powder diffraction and found to be in agreement with reported data. The equipment used for X-ray excitation has been described before [3]. All measurements were performed at room temperature.

RESULTS CaO: Ga(III), In(II1) Since the literature contains several reports on a luminescence from undoped CaO [4-71, undoped as well as doped CaO samples were prepared and the results compared. Figure 1 shows the X-ray excited emission spectra obtained. Whereas undoped CaO shows an ultraviolet emission which will not be discussed any futher, the Ga(II1) and In (III) doped samples show a predominantly visible emission band with a maximum at 435 and 485 run,respectively. The undoped-CaO-emission is also observable (see Fig. 1).

350

500 nm

Fig. 1. X-ray excited emission spectra of CaO (a), CaO:Ga(III) (b) and CaO:In(III) (c) at 300 K. The ultraviolet emission of undoped CaO has been mentioned by Lehmann [4] for cathode-ray excitation. This author reports for CaO: In(III) a weak emission band with a maximum at 496 nm which is obviously the same emission as that from our sample. We conclude that the trivalent dl' ions Ga(II1) and In(II1) show visible luminescence in CaO. The divalent dl' ions Zn(I1) and Cd(II) show ultraviolet luminescence in CaO with an emission maximum at 330 and 305 nm, respectively [1,4].

277

A nominally emission

pure sample

under

The visible impurity.

of InBO, shows ultraviolet

X-ray excitation

emission

is clearly

Its intensity

with an intensity due to Tb(II1)

corresponds

very efficient

activator

green phosphor

in television

as well as visible

ratio of about

to the fact that Tb(II1)

in InbOa. This material projection

tubes

1:2 (Fig. 2)

which must be present

as an

is known

has strong potential

to be a as a

[2].

I

A #/Ii_ 10x

300

450

Fig. 2. X-ray

excited

The ultraviolet aware.

emission

emission

It consists

600nm

spectrum

of nominally

has not been reported

of a broad band with a maximum

pure InBOa at 300 K.

before

as far as we are

at 300 nm.

DISCUSSION In view of the arguments Ga(II1)

and In(II1)

of the optical has been

transitions

reported

Fig. 2 the Stokes

presented

ions luminesce

in [l] there is little doubt

is unknown.

at about 45,500

that the

in CaO and in InBO,, even though

cm“

shift is estimated

For InBOs an optical [2]. Using

the nature

absorption

the emission

spectrum

edge of

to be 2 13,000 cm-', in line with data in

111. The Stokes

shift of the emission

since no excitation

spectra

reflection

of CaO: In(II1)

connection is 20,000 Ga(II1)

spectrum

are available.

and In(II1)

The present

cannot be determined,

from the diffuse

by Lehmann

[4] it follows

shift of the emission

in

of CaO: In(III)

agrees with data for the luminescence

of

in Sr,(PO,), [l].

data, together of d"

the dopant

However,

reported

with Fig. 1 that the Stokes cm-' or more. This value

of the emission whether

in the CaO samples

with those in [l], suggest

ions is not only large, but depends

fits the lattice

electrically

or not:

that the Stokes

shift

also on the fact if no charge

278

compensation is required the Stokes shift amounts to 2 eV or less (CaO:Zn(II),Cd(II);YaOs:In(III);p-GasOs); if the dl' ion carries an effective charge the Stokes shift amounts to 2.5 eV or more (Sr,(P0,)2:Ga(III),In(III); CaO:Ga(III),In(III)). It has been shown recently that effective charges may drastically influence the amount of relaxation in excited charge-transfer states: in CaS04:Eu(III) the charge-transfer relaxation is so large that the quantum efficiency of the Eu(II1) emission is low [8]; in A1a03:Ti(IV)the same holds true for the excited charge-transfer state of the titanate octahedron [9]. The results on the Stokes shift of the emission of d" ions suggest, therefore, that the optical transition involved has a certain charge-transfer character. This effect was already predicted in [l]. It is also a guideline where to find the most efficient d"-ion luminescence,viz. in lattices where no chargecompensation is required. Actually this is the case for the most efficient luminescent materials in this class: Zn40(B02)s,CaO:Zn(II), CaO:Cd(II) [l]. The results for InBOa have still another meaning. Low amounts of Tb(II1) are very effective in capturing the excitation energy. The ultraviolet emission of It-&O,shows no spectral overlap with allowed transitions on the Tb(II1) ion as can be found from a comparison of our Fig. 2 and the diffuse reflection spectrum of InBO,:Tb(III)in [2]. There is only overlap with the spin-forbidden 4f-5d transition. Following methods described in [lo] the critical distance for In(III)-Tb(II1)energy transfer is estimated to be 9A. With an estimated amount of 100 ppm Tb this would only yield a few percent of Tb(II1) emission. This shows that the Tb(II1) ion is populated by direct capture of charge carriers, since the excited In(II1) state cannot be very mobile in view of the large Stokes shift. It is interesting to compare this result with those for Cd3GazGe301a: Nd(III), Tb(II1) [ll]. This host contains only metal ions with d" configuration and is in a sense isoelectronicwith InBOs. The unactivated host shows broad-band emission which we assign to the same type of transition as discussed in [l] and in this paper. Energy transfer to Nd(II1) is very inefficient. This ion is not able to trap efficiently the charge carriers. The few percent of Nd(II1) emission correspond to the result of the estimation performed above. However, 10e3 Tb(II1) is able to trap the larger amount of the excitation energy, in CasGaaGe s0 r2 as well as in CdsGasGesOrs[ll]. Obviously charge carriers are directly trapped by Tb(II1). In the mixed crystals Cds_,CaxGazGe301s: Tb the activator ions are considerably less efficient in trapping the excitation energy which may be due to the disorder in the lattice which will restrict the charge carrier mobility. The authors of

279

[ll] propose that the holes are trapped by Tb(III) and the electrons by defects. The difference with our interpretation is that we do not invoke defect centres, but consider the intrinsic dr" ions as luminescent centres.

CONCLUSION In conclusion we have given further evidence that metal ions with d"" configuration may show luminescence if high energy excitation such as X-rays is employed.

REFERENCES G. Blasse, Chem. Phys. Letters, 175 (1990) 237. H.T. Hintzen, Ph. d thesis, University of Utrecht, 1990, chap.& L.H. Brixner, Inorg. Chim. Acta, 140 (1987) 97; L.H. Brixner, C.C. Torardi and G. Blasse, J. Solid State Chem., 89 (1990) 138. 4 W. Lehmann, J. Luminescence, 6 (1973) 455. 5 Y. Chen, M.M. Abraham, T.J. Turner and C.M. Nelson, Phil. Mag., 32 (1975) 99. 6 J.A. Garcia, A. Remon and J. Piqueras, Phys. Stat. Sol. (b), 89 (1985) 237. 7 H. Donker, W.M.A. Smit and G. Blasse, Phys. Stat. Sol. (b), 145 (1988) 333. 8 D. van der Voort and G. Blasse, J. Solid State Chem., 87 (1990) 350. 9 G. Blasse and J.W.M. Verweij, Mater. Chem. Phys., 26 (1990) 131. 10 G. Blasse, Mater. Chem. Phys., 16 (1987) 201. 11 R. Hirrle, J. Wiehl, W. Wischert and S. Kemmler-Sack, Phys. Stat. Sol. (a), 120 (1990) 643.