Photoluminescence study of (CaO), -x( ZnO), powder solids in air Loukia A. Loukatzikou*, Antonios T. Sdoukos and Philip J. Pomonis Department of Chemistry, University of Ioannina, Ioannina 45110, Greece Luminescence from (CaO), -,(ZnO), (x =0-1) powder solids, prepared by thermal treatment in air at 1423 K, was observed at room temperature by means of high-energy UV excitation (5.2-6.2 eV). In general, broad emission bands centred in the visible region of the spectrum were recorded, with features dependent upon the excitation wavelength as well as on the sample nature. The presence of small quantities of zinc within the CaO results in a dramatic enhancement of the photoluminescence intensity, reaching a maximum at 1YOaddition of Zn atoms/Ca +Zn atoms. In this case, a nearly twenty-fold increment in photoluminescence intensity was observed.The emission band showed a peak at 435 nm (2.85 eV) and was excited to a maximum by 210 nm (5.9 eV) light. Photoinduced electron transfer, followed by recombination processes involving oxygen anion vacancies as well as photogenerated F+ and F centres, are possibly the cause of the observed luminescence. In recent work based on collected literature data, Blasse' showed that complexes [M(d10)],[02-]y may show efficient luminescence at room temperature when excited with high- energy UV radiation. This report also drew attention to the fact that the spectral features of these complexes are similar to those of several oxides which do not contain any d10 ions. Also, Kunkely and Vogler2 reported highly efficient photo- luminescence of the molecular species Zn,O(CH, COO), in solution at room temperature.The reported spectra were similar to those of Zn,0(B02), crystal^.^ This similarity was attributed to the presence of a particular cluster of atoms, since both compounds were characterized by a tetrahedral arrangement of four Zn atoms around the central oxygen., Molecular cluster calculation^^-^ indicated the above arrange- ment to be a very good model of the oxygen chemical environment in ZnO. Photoluminescence from ZnO powders is well known.6-'' A narrow ultraviolet band and a broad green band are observed usually, with an intensity ratio dependent on the preparation as well as the excitation The ultraviolet emission is of an excitonic nat~re,~'~ while the green emission appears to be associated with the presence of oxygen anion vacancies near the surface of the ample.^.^?'' On the other hand, cathodoluminescence12 and X-ray l~minescence'~ have been observed from CaO powders in air although the well known photoluminescence of outgassed CaO powders in UUCUO~~-'~is reported to be 'quenched' by oxygen or air.Cathodolumin- escence has also been observed from CaO :Zn by Lehmann,I2 but no photoluminescence has been recorded so far, although Blasse reported that it might be possible.' In this paper we present for the first time the photoluminescence spectra of (CaO), -,(ZnO), polycrystalline powders in air. These mate- rials show an intense luminescence at room temperature, which was previously overlooked probably because of the high- energy UV radiation required for their excitation.Experimental Polycrystalline powders of (CaO), -,(ZnO), (x=0.000, 0.005, 0.010, 0.050, 0.100 0.300, 0.500, 0.700, 0.900, 0.950, 0.995 and 1.000) were prepared by firing intimate mixtures of CaO (Fluka, p.a) and ZnO (Merck, p.a). The trace element content of the starting materials was mainly Fe (~0.05%) and Pb (<0.01YO),with some Cu and Zn (<O.O05%) for the CaO and Pb (0.005Y0) for the ZnO. The starting materials were mixed and ground thoroughly before firing in air for 5 h at 1423 K. This procedure was repeated three times. Samples were quickly removed from the 673 K furnace after the high-temperature treatment, and were ground and stored in glass vials.These samples will be hereafter designated as CZ:y, where y is defined as the percentage of Zn atoms over the total metal (Zn +Ca) atoms in the sample [y =100Zn/(Ca+Zn); Table 1). We emphasise that neither Ca(OH), nor CaCO, were detect- able by thermal analysis examination of the solids treated as described above. X-Ray diffraction (XRD) patterns of the oxides were meas- ured at room temperature in a Philips system (PW 2253 lamp, PW 1050 goniometer, PW 1?65/50 analogue detector) by using Cu-Ka radiation (2. =1.542 A). Photoluminescence measurements were carried out in air at room temperature, using a Perkin Elmer LS-3 fluorescence spectrometer equipped with a 150 W xenon lamp and a side- window photomultiplier tube (EM1 9781 RA), with reflection grating monochromators with fixed slits of 10nm.Thus, the wavelength accuracy was +2 nm. The scanning rate was 120nm min-'. The powders were held in a front surface accessory and a 1% neutral density filter was placed in front of the emission monochromator in order to record the emitted light. All excitation spectra were corrected for the xenon lamp intensity as well as for the excitation monochromator efficiency. The emission spectra were not corrected for the photomultiplier response. The maxima of spectral bands were identified by point-to-point measurements of the luminescence intensity us. wavelength in the range of interest. Successive measurements of a maximum were found to lie within the instrument's repeatability (+ 1nm).Table 1 Designation of the prepared samples designation sample [y = 100 Zn/(Ca +Zn)] CaO cz:0.00 Ca0.995Zn0.0050 Ca0.99Zn0.010 Ca0.95Zn0.050 Ca0.90Zn0. loo CZ :0.5 cz:1 cz:5 cz:10 Cao.7oZno.3oO CZ: 30 Ca0.50Zn0.500 CZ: 50 Cao.3oZno.7oO CZ :70 Cao.1oZno.900 CZ : 90 OsZnO. 95 cz :95 Ca0.01Zn0.990 cz :99 Ca0.005Zn0.9950 ZnO cz : 99.5 cz:100 J. Muter. Chem., 1996, 6(5), 887-893 887 Results The XRD patterns of the CZ :y solids are shown in Fig. 1. The patterns are indicative of well crystallized solids and show the crystal phases of one or both of the parent oxides in varying ratios. CaO was the only phase detected in the case of CZ :0-1 solids, while the ZnO phase alone was detected in the CZ:99-100 solids.However, the patterns of both CaO and ZnO were detected in all the other solids, e.g. CZ: 5-95, indicating that they are mixtures of the parent oxides in varying ratios. The CaO phase dominates for the CZ: 5-30 samples, while that of ZnO is dominant in the CZ :50-95 ones. The emission spectra of (CaO), -x(ZnO), powders excited by 200, 215 and 240 nm light are shown in Fig. 2. Upon excitation in the 200-240nm range, an intense visible lumi- nescence visible to the naked eye was observed from all the samples and broad emission bands in the 300-650nm region were recorded. Thus, the Stokes' shifts of the various emission bands are very large (2.5-4 eV).The emission bands may consist of one, two or three components peaking in the visible range of the spectra, although ultraviolet emission is also recorded, either as a shoulder (A =360 nm for sample CZ :1) or as a tail of the main emission. A very weak red emission band, with a peak at ca. 726 nm, can also be distinguished for the samples CZ :0.5-5. A similar red emission is also apparent in the case of pure CaO, but not in that of pure ZnO. All the emission spectra recorded for excitation at 200nm consist of broad bands with at least three overlapping compo- nents [Fig. 2(a)]. Although the overall and the relative compo- nent intensities vary across the samples, the emission band maxima are similar. The main emission band has a maximum either in the blue (Lax=435-448 nm for CaO and CZ :0.5-95 solids) or in the blue-green (A,,, =473 nm for CZ :99-100) region of the spectrum.CZ:30 i I 0cz:10 I I0 0 cz:5 I 1 cz:1 i I 0 CaO I 1 1.5 2.0 2.5 Excitation of CZ: 0.5-70 solids at 215 nm [Fig. 2(b)] results in a broad emission band with a maximum in the blue region of the spectrum (Amax =435-448 nm). Nevertheless, the corre- sponding spectra of the other CZ :y solids are similar to that obtained from excitation at 200 nm. The most striking result of the Zn addition to CaO (0.5-90%) is a dramatic enhance- ment of the photoluminescence intensity, which is maximized for 1% Zn addition. In this case, a nearly twenty-fold increase of the luminescence intensity was observed with respect to that of CaO [Fig.2( b)]. The emission now reaches a maximum at 435 nm (2.85 eV) and is excited to its maximum by 210 nm (5.9 eV) light. The emission spectra of CZ :0.5-70 recorded upon excitation at 240 nm are similar to the corresponding ones obtained from 200 nm excitation (Fig. 2). However, excitation of CaO and CZ: 90-100 at 240 nm gives rise to some special spectral features [Fig. 2(c)]. In particular, the blue emission maximum of CaO (A=444-446 nm) is now replaced by a violet one at 409 nm. At the same time, blue-green (Amax =479 nm) and green (Lax=527 nm) emissions are recorded as shoulders on the main violet bond. The emission spectra of CZ:90-100 consist now of two partly overlapping bands with maxima at 508-523 nm and 406-416 nm.The green emission clearly dominates over the violet one. The photoexcitation spectra of some CZ :y solids are shown in Fig. 3. The corresponding excitation spectra of the other solids are quite similar to that of CZ:O.5. The CaO and CZ :99-100 excitation spectra recorded for an emission wave- length of 440nm show a broad band with a maximum at 200-208 nm, as well as a second weak band at 230-238 nm. However, the same emission (440 nm) is the result of only one excitation band peaking at 210-218nm in the case of CZ :0.5-95 solids. The above excitation bands also result in a green emission (520 nm), although such an emission arises also . , 1.....ll..l.t ' ..L .II CZ95 I I cz99 1.. ..11T r... I....I .. ~ 1.5 2.0 2.5 dlA Fig. 1 XRD patterns of CZ :y solids: 0,CaO; ,ZnO 888 J. Muter. Chem., 1996, 6(5),887-893 0.00 0.00 I I I I I 250 350 450 550 650 750 250 350 450 550 650 750 300 400 500 600 700 A/nm Fig. 2 Emission spectra of CZ :y solids ( y =0.00-100): (a),Iex200 nm, (b) ,Iex=215 nm, (c) I,,, =240 nm= from a new excitation band (380-390 nm) for CZ :70-100 (Fig. 3). Excitation of the CZ: 99-100 samples is also apparent in the whole A= 200-400 nm region. We emphasise that although the emission of CZ :0.5-70(4 solids peaks at nearly the double the excitation wavelength, the increased luminescence intensity is not a result of a ‘second- -100 order’ effect. In fact, the emission maximum shifts only slightly -99s with the excitation wavelength in the whole 200-240 nm range, -99 while the overall emission also shows an increased intensity. So, the increased luminescence intensity is an intrinsic property of the solids and is not enhanced by secondary effects.We have also noticed that the emission spectra were not corrected for the photomultiplier response. This could actually affect the spectral shapes, especially the relative band intensities recorded. However, all the spectra should be uniformly affected and the picture obtained would not be modified drastically. Discussion CaO The excitation spectrum of CaO powders in air (Fig. 3) is rather similar to the absorption spectrum of deformed CaO single crystals,” which exhibit maxima at 214 and 270 nm, although the main absorption peak is that at 270nm. A similarity has also been recognized between the absorption bands of deformed crystals and those of high-surface-area 0.5 alkaline-earth-metal oxide powders.20-22 The former was attri- -buted to vacancy clusters formed during deformation,” and0.00 the latter to transitions from low-coordination oxygen 200 300 200 300 400 anions.21,22 Excitation of the alkaline-earth-metal oxide powders inA/nm DUCUU, with energies in the region of their optical absorptions, Fig.3 Excitation spectra of CZ:y solids: (a) ,Ie,=440 nm, (b) ,Iem= produce lumines~ence~~-~~~~~ that has also been attributed to 520 nm transitions from low-coordinated surface ions.14-16 Such ions J.Muter. Chem., 1996, 6(5),887-893 889 have been assumed to be present in the outgassed alkaline- earth-metal oxide powders as a result of their high-temperature treatment zn ~UCUU,~'22 but their involvement in photoprocesses over powder oxides has been questioned24 26 However, the excitation energies of the used CaO powders in air are rather different from those of CaO powders zn z)ac~u,'~'' with the exception of an excitation near 240 nm The other excitation bands of outgassed CaO powders are all centred on smaller energies Luminescence from outgassed CaO powders is reported to be 'quenched' by molecular oxygenI4 So, the absence of luminescence when CaO powders were excited in air in the 260-380nm range may be attributed to the action of oxygen The emission spectra recorded for CaO powders are not identical to any spectrum of the oxide reported in the literature, although similar emissions have been observed Excitation of CaO at 200-215 nm causes a major blue emission with a maximum at 444-446 nm [Fig 2(a),(b)] A broad emission band which peaks at 426nm has also been observed from outgassed powders in uucuu by excitation in the 235-300nm region An emission peaking at 454nm has also been observed from CaO single crystals upon excitation at 270 nm 27 CaO spectra in air also show emission bands with maxima at 479-480 and 525-527 nm (Fig 2) The emission band from deformed single crystals has a maximum at 477 nm and a shoulder at cu 510nm is also clear," but in this case no distinct emission peak was observed at shorter wavelengths However, the emission spectra of deformed CaO crystals reported in the literature seem to be more similar to the present results Although broad emission bands centred at 495 or 477nm have also been recorded from outgassed CaO powders zn U~CUUby excitation at 330 nm,I4 a green emission has never been observed from the oxide zn U~CUU However, the green emission is the result of air-annealing of deformed single ~rystals,,~ although it has also been observed for undeformed crystals 28 Excitation of CaO at 240nm results mainly in a violet emission band peaking at 409 nm [Fig 2(c)] An emission band with a maximum at 405 nm has also been obtained from outgassed powders zn uucuu, by excitation at 282 or 310 nm Note that the violet emission is the major one observed from these powders at low temperatures for excitation over the whole 235-300 nm range l5 In conclusion, the photoluminescence spectra of CaO pow- ders in the present work are not quite identical to some spectra reported in the literature, although similar emissions have been observed However, the material treatment as well as the experimental procedure followed here are quite different from those reported in the literature Thus, the observed spectral differences may be attributable to the different preparation procedure and, in particullar the quick quenching of the solids Note that the luminescence from CaO powders reported here is not a result of grinding A bright blue luminescence quite similar to that of pressurized single crystals has been observed previously from thermally treated CaO powders after grinding,29 but a second firing step was found to remove this luminescence completely In contrast, the luminescence from CaO recorded here is permanent On the other hand, the observed luminescence is not due to the presence of small quantities of Ca(OH),, as a result of the CaO sensitivity to the ambient atmosphere Although no luminescence was observed previously from Ca(OH), after 254 or 365 nm excitation,'2 more re~ently'~ some spectral features of partially dehydroxylated CaO powders were ascribed to the presence of Ca(OH), These features consist of excitations at ca 270 and 300nm, far from the present CaO excitation region, and of emissions at 353, 405 and 443 nm l7 The recorded luminescence does not seem to be a result of the trace elements present in the starting material In fact, Fe is known to act as an efficient fluorescence trap The well 890 J Muter Chem , 1996, 6(5),887-893 known luminescence of Pb2+ impurity ions (<O 01) in CaO,,' 31 which shows vibrational structure at low tempera- tures, is centred on the ultraviolet region (A=348-368 nm), although 0 1% or more Pb in CaO was reported to emit almost exclusively in a broad distribution in the visible region The CaO Cu emission spectrum consists of two overlapping bands with maxima at 390 and 448 nm ZnO The luminescence spectra of the ZnO powders used, excited by high-energy UV light (A =200-215 nm) are different from those reported in the literature, owing to the different energy used for their excitation Thus, the spectra obtained here show very large Stokes' shifts, in contrast to the small Stokes' shift of the well known luminescence of ZnO obtained by excitation near its energy gap The excitation spectrum of ZnO for the green (520 nm) emission [Fig 3(b)] is quite similar to the absorption spectrum of a ZnO thin layer However, the blue (A,,, =435-445 nm) and blue-green (nmax=473-484 nm) emis- sion components recorded here [Fig 2(a)(b)] seem to be only a result of high-energy UV excitation, especially near 208 nm [Fig 3(a)] Excitation of ZnO near its energy gap (Aex= 320-380 nm) results exclusively in a broad green emission band, quite similar to that referred to in the literature67 No ultraviolet emission component was observed at the same time, although such an emission was obtained together with the main green one, after 240 nm excitation [Fig 2(c)] This is probably due to the sample treatment method The emission spectra of CZ 0 5-70 (iex=215 nm) exhibit close similarities to that of Zn,O(BO,), Excitation of these crystals at 250 nm results in a broad emission band with a maximum at 435 nm There is also good agreement between the exci- tation wavelengths recorded from CZ 0 5-90 solids (Aex= 210-218 nm) and the absorption of the Zn,O(CH,COO), solutions at 216nm,2 although the emission from the latter reaches a maximum at shorter wavelengths (A=372 nm) Thus, there is a clear similarity in the luminescence behaviour of Cz 0 5-70 and Zn,O(BO,), solids as well as Zn,O(CH,COO), solutions On the other hand, the lumi- nescence spectra of CZ 0 5-70 solids are quite similar to that of deformed MgO crystals which do not contain any Zn The latter absorb at 2175nm (57eV) and emit at 4275nm (2 9 eV) l9 The similarity between the Zn,O(CH,COO), and Zn40( BO,), luminescences has been recognized and has been related to the common tetrahedral arrangement of four Zn atoms around the central oxygen in these compounds A similarity has also been recognized between the luminescences of the above compounds with those of several oxides which do not contain any d'' ions In general, an analogy between the luminescences of 0x0-d" and 0x0-do complexes has been established,' 32 while the optical transitions involved seem to be of a very complicated nature ' 32 Thus, a question arises about the possible role of either Zn2+ ions (d") or a particular cluster of atoms, in the determinatim of the luminescence behaviour of CZ y solids, in particular CZ 0 5-70 No other crystal phases except for CaO and ZnO were detected in the solids, so the photo-effects obtained should be attributable to a synergistic action between the two oxides The luminescence spectra recorded for the CZ y solids show that such an effect exists This is especially clear in the emission spectra obtained by 240 nm excitation [Fig 2(c)] These spec- tra are characterized by a progressive 'red' shifting across the CZ y samples, as the Zn addition to CaO increases It can be anticipated that the above spectra arise from excitation in the range of a second, minor excitation state of pure CaO and ZnO oxides, which is absent in the case of CZ 05-95 solids However, the oxide’s XRD phase prevalence across the CZ :y samples does not affect their spectral features in an expected way.The influence of the ZnO phase prevalence becomes apparent only when the Zn concentration reaches 90%. Thus, although the luminescence behaviour of single-phase CZ:90-99.5 solids is quite similar to that of pure ZnO, the behaviour of CZ :50-70 differs. It rather resembles the behav- iour of CZ: 5-30 as well as that of single-phase CZ: 0.5-1 solids, which in turn is rather different to that of pure CaO.On the other hand, a synergistic action is also clear in the emission spectra obtained by 200 nm excitation [Fig. 2(a)]. A similarity between the spectral features of CZ:y solids as a whole is now clear. The Zn addition to CaO affects the overall luminescence intensity but leaves emission peak positions virtually unchanged. Although such a similarity is not clear in the emission spectra recorded by 215 nm excitation, it seems that the CZ :0.5-70 solids do not exhibit some special kind of luminescence, e.g. different from that of the other samples. The luminescence is now increased dramatically, probably owing to a synergistic action between the two oxides.The emission spectra seem to be rather structureless, possibly because of the relatively increased blue component emission (Lax= 435-448 nm), overlapping with components at longer wavelengths. In conclusion, the emission spectra of CZ :y solids recorded by excitation at 200-215 nm show that the percentage of Zn atoms present in CaO does not play any significant role in the determination of the emission energies, although it determines the overall and the relative intensities of the emission bands. Considerable differences are observed in the emission spectra of the solids recorded for excitation at 240 nm. These are related to the present Zn percentage and probably arise from the influence of a new excitation state of pure ZnO near 240 nm (Fig.3). Thus, a similarity really exists across the luminescence spectra of CZ:y solids as a whole, which also suggests a similarity in the nature of the photoprocesses involved. Proposed photoluminescence model In order to describe the photoluminescence processes for CZ :y solids, a simplified model is proposed based on the experimen- tal data. Indeed, a general energy diagram can be drawn as shown in Fig. 4. This is based on the following experimental observations: EemJ where Eex2denotes the energy corresponding to the weaker excitation band maximum at 230-238 nm (5.2-5.4 eV), recorded in the excitation spectra of some samples (CaO, CZ:99-loo), while Eeml, Eem2, Eem3 and Eem4denote the energies corresponding to the emission bands maxima for the blue-green (460-490 nm), green (500-530 nm) blue (415-450 nm) and violet (400-412 nm) components respectively.A general mechanism explaining the photoprocesses involved as well as the various energy levels shown in Fig. 4 could be as follows. High-energy UV light (Eexl=5.7-6.2 eV) can induce the formation of levels of electrons and holes in the solids, possibly via excitation of lattice 02-ions, Then, level 1 in Fig. 4 corresponds to the top of the valence band of the solid, while levels 2 and 3 correspond to the electron and hole levels generated, respectively. We propose that the elec- trons released from the valence band could arrive initially in the vacuum level.In fact, the vacuum level lies about 6eV above the top of the valence band of most semiconductor oxides,33e.g. near to the main excitation energy EeX1recorded for our solids. The main excitation energy of CaO is smaller 0 0 Fig. 4 General energy diagram describing the photoluminescence pro- cesses over CZ :y solids: 0,electron; 0,hole that its bandgap, but in this case the vacuum level lies inside the bandgap, as a result of its negative electron affinit~.~~.~~ Then, emission could be a result of recombination processes between photogenerated electrons and holes, possibly proceed- ing through oxygen vacancies present in the solids as a result of their treatment. The following general emission mechanism is possible: during the trapping of an electron from level 2 (Fig.4) by an oxygen vacancy, blue-green light (460-490 nm) is produced. This process results in the formation of an F+ centre. Next, a second electron from level 2 could be trapped by an Ff centre to form an F centre and produce the green emission (500-530 nm). The above processes are both accompanied by changes in the oxidation state of the ions of the electrons level. Then, the F+ and F centres electrons recombine with holes at level 3. The recombination of an electron coming from either an Ff or an F centre with a hole is accompanied by either the blue (415-450nm) or the violet (400-412 nm) emission respectiveiy. At the same time, an oxygen vacancy or an F+ centre is reformed. If so, levels 4 and 5 probably correspond to F+ and F centres levels, respectively.It must be stressed here that no distinct levels, but energy bands should be formed in the CZ:y solids during the irradiation, as the observation of broad excitation and emission bands suggests. However, although the problem of optical transitions involved in the above luminescence processes is probably of a very complicated nature, the proposed simplified model aims to provide a first approach to the matter. A complete answer to this problem requires at least a more detailed energy-level calculation. By applying the above general model in the cases of CaO, CZ :0.5 and ZnO solids, we obtain the energy diagrams shown in Fig. 5 (a), (b) and (c), respectively. These diagrams were designed by first fitting the experimental energies correspond- ing to the maxima of the emission bands recorded by 200 nm excitation, and then those corresponding to the maxima of the main excitation bands, recorded for the emission at 440 (CaO, CZ :0.5) or at 520 nm (ZnO).These energy values are indicated by bold characters in the Figures. The bandgaps of the pure oxides used are based on the literature data.34,35 In the case of CZ :0.5 we propose that E, >5.7 eV. The agreement of this model with the experimental data is very good in the case of CaO and satisfactory enough in the other cases. In fact, the estimated distance from Fig. 5(a) between electron and hole levels (5.38 eV) corresponds to the maximum of the second weak excitation band recorded from CaO (Fig.3) at 230 nm (5.4 eV). In addition, the estimated energy between F and hole levels (3.02 eV) corresponds to the violet emission at 409 nm (3.03 eV), recorded upon excitation of the oxide at 240 nm [Fig. 2(c)]. In contrast, neither a distinct emission peak near 3.08 eV (402 nm), nor a second excitation peak near 5.5 eV (225 nm) were recorded in the spectra of the CZ:O.5 solid. Finally, the emission at 406 nm (3.05 eV), J. Muter. Chem., 1996, 6(5),887-893 891 conduction band ,, level EL=? 5.47 cv, (5 I* m (473 nm) I ' . , 023 eV F+ Fig. 5 Energy diagrams describing the photoluminescence processes over CaO (a), CZ : 0.5 (b) and ZnO (c) recorded by excitation of ZnO at 240 nm, corresponds satisfac- torily to the F and holes levels distance, estimated to be 3.02 eV [Fig.5(c)], although the estimated distance between electron and hole levels (5.41 eV) does not correspond very well to the maximum of the second excitation band of the oxide at 240 nm (5.2 eV). According to the proposed energy diagrams, the F and F+ levels are estimated to be 3.16 and 3.39 eV below the conduc- tion band of CaO, as well as 0.03 eV and 0.26 eV below the conduction band of ZnO, respectively. The F-centre ground level has been found to be 3.11 eV below the conduction band of CaO single crystals,36 while the Ft centre absorbs at 3.6-3.7 eV,37-39 which corresponds to the F' distance from the valence band rather than the conduction band of CaO in Fig.5(a). In the case of ZnO, F and F' levels have been estimated to be 0.02-0.05 eV40-42 and 0.19 eV4' or 0.30-0.45 eV4' below the conduction band respectively, in satisfactory agreement to the corresponding energies estimated here from Fig. 5(c). Zn2+ ions, although containing closed d shells, facilitate the photoprocesses over CZ : y solids as described above, because they can really trap electrons and be converted to unstable Zn+ ions, which in turn can easily return electrons to the oxygen vacancies. Thus, an intense luminescence is produced. The blue emission (i,,,= 435-448 nm) dominates for the CZ : 0.5-70 solids, possibly because the recombination pro- cesses proceed mainly through the F+ centres. Bearing in mind that the energy gap of ZnO is 3.34 eV at room temperat~re,~' the F centres position (3.31 eV above the valence band) estimated from Fig.5(b) suggests that Zn2'-F pairs should be formed in the CZ : 0.5 solid after high-energy excitation. If so, the recombination processes should proceed niore easily cia Ff rather than via F centres, because F electrons could then easily be trapped by Zn2' ions. Similar effects are possible in the case of CZ: 1-70 solids. In contrast, the green emission (Amax = 508-523 nm) dominates for the CZ : 90-99.5 solids, as a result of the relatively high conductivity of ZnO. Now, although the F electrons are close to the conduction band [Fig. 5(c)], they prefer to recombine with holes in the hole level. The observation that their excitation by light of energy E < 5.2 eV (i240 nm) results exclusively in green (i,,,= 508-523 nm) and violet (i,,,= 406-416 nm) emission possibly indicates that F centres are formed exclusively under these conditions.The involvement of FC as well as F centres in the photo- luminescence processes over the alkaline-earth-metal oxide powders in mcuo has also been considered in the pa~t.'~,'*,~~ Charge-transfer transitions'6 and relaxation processes have also been con~idered.~~ The model developed here is in good agreement with the interpretation of the phosphorescence as well as the thermoluminescence of UV-irradiated CaO powders in cacu~.'~ The above luminescence, which also exhibits some similarity to the emission spectra of CaO in air.was ascribed to radiative tunnelling recombination processes between elec- tron and hole pairs in distant Ft and V- centres that were pho toformed.18 In conclusion, we have demonstrated in this work that an enhancement of the emission spectra of CaO-ZnO powders is observed by exciting them with i= 200-215 nm. The effect is particularly strong at 1% addition of ZnO into CaO where a 2000% increase of intensity is produced. A simplified photo- luminescence model is proposed, answering the question of the similarity observed across the CZ : y solids spectra as a whole and fitting satisfactory with the experimental data. We are grateful to Professor M. Marselos for the provision of the fluorescence spectrometer for the luminescence experi- ments, and to Dr.M. A. Demertzis for his help in the early stages of this study and for helpful discussions. References 1 G. Blasse, Chem. Phys. Lett., 1990,175,237. 2 H. Kunkely and A. Vogler, J. Cheni. Soc., Chem. Commun.: 1990, 1204. 3 A. Meijerink, G. Blasse and M. Glasbeek, J. Phys: Condens. Matter 2, 1990, 6303. 4 R. Bertoncello, M. Bettinelli, M. Casarin, A. Gulino, E. Tondello and A. Vittadini, Inorg. Chem., 1992,31,1558. 5 M. Casarin, E. Tondello, F. Calderazzo, A. Vittadini, M. Bettinelli and A. Gulino, J. Chem. Soc., Faradaj Trans., 199?,89,4363. 6 G. Heiland, E. Mollwo and F. Stockmann, Solid State Phys., 1959, 8, 191. 7 F. Van Craeynest, W. Maenhout-Van der Vorst and W. Dekeyser, Phys. Status Solidi, 1965,8, 841.8 W. Maenhout-Van der Vorst and F. V. Craeynest, Phys. Status Solidi, 1965, 9, 749. 9 M. Anpo and Y. Kubokawa, J. Phys. Chem., 1984: 88,5556. 10 M. Mienska, R. Leszczynski, S. Karolczak and S. Wysocki, Radiat. Phys. Chem., 1989,33,483. 11 H. Idriss and M. A. Barteau, J. Phys. Chem., 1992,96, 3382. 12 W. Lehmann, J. Luminescence, 1973,6,455. 13 G. Blasse and L. H. Brixner, Mater. Chem. Phys., 1991,28,275. 14 S. Coluccia, A. M. Deane and A. J. Tench, J. Chem. Soc., Faraday Trans. 1, 1978,74,2913. 892 J. Mater. Chem., 1996, 6(5), 887-893 15 16 17 18 19 20 21 22 23 24 25 26 27 28 S. G. MacLean and W. W. Duley, J. Phys. Chem. Solids, 1984, 45, 227. W. W. Duley, Philos. Mag. B., 1984,49, 159.W. W. Duley, High Temp. Sci., 1984, 17,409. Y. Yanagisawa, N. Inishi and A. Narumi, Phys. Rev. B, 1992, 46,11121. Y. Chen, M. M. Abraham, T. J. Turner and C. M. Nelson, Philos. Mag., 1975,32,99. R. L. Nelson and J. W. Hale, Discuss. Faraday Soc.,1971,52, 77. A. Zecchina, M. G. Lofthouse and F. S. Stone, J. Chem. SOC., Faraday Trans I, 1975,71,1476. E. Garrone, A. Zecchina and F. S. Stone, Philos. Mag., 1980, 42, 683. A. J. Tench and G. T. Pott, Chem. Phys. Lett., 1974,26,590. V. A. Shvets, A. V. Kuznetsov, V. A. Fenin and V. B. Kazansky, J. Chem. Soc., Faraday Trans 1,1985,81,2913. J. Cunningham, in Surface and Near-Surface Chemistry of Oxide Materials, ed. J. Nowotny and L. C. Dufour, Elsevier, Amsterdam, 1988,ch. 8.A. M. Stoneham and P. W. Tasker, in Surface and Near Surface Chemistry of Oxide Materials, ed. J. Nowotny and L. C. Dufour, Elsevier, Amsterdam, 1988,ch. 1. J. A. Carcia, A. Remon and J. Piqueras, Phys. Status Solidi. A, 1985,89,237. J. Llopis and J. Piqueras, J. Appl. Phys., 1983,54,4570. 29 H. Donker, W. M. A. Smit and G. Blasse, Phys. Status Solidi B, 1988,145,333. 30 A. F. Ellervee, Phys. Status Solidi B, 1977,82,91. 31 A. C. Van der Steen and L. T. F. Dijcks, Phys. Status Solidi B, 1981, 104, 283. 32 G. J. Dirksen, A. N. J. M. Hoffman, T. P. Vande Bout, M. P. G. Laudy and G. Blasse, J. Muter. Chem., 1991,1,1001. 33 P. A. Cox and A. A. Williams, Surf.Sci., 1986,175, L782. 34 0.V. Krylov, in Catalysis by Nonmetals, Academic Press, New York, 1970. 35 W. F. Wei, Phys. Rev. B, 1977,15,2250. 36 B. Henderson, S. E. Stokowski and T. C. Ensign, Phys. Rev., 1969, 183, 826. 37 A. E. Hughes and B. Henderson, in Point Defects in Crystals, ed. J. H. Grawford, Jr. and L. M. Slifkin, Plenum Press, New York, 1972, vol. 1. 38 L. S. Welch, A. E. Hughes and G. P. Summers, J. Phys. C: Solid State. Phys., 1980,13, 1791. 39 J. C. Kemp, W. M. Ziniker and E. B. Hensley, Phys. Lett. A, 1967, 25,43. 40 A. Poppl and G. Volkel, Phys. Status Solidi A, 1991, 125, 571. 41 W. Cope1 and U. Lampe, Phys. Rev. B, 1980,22,6447. 42 F. A. Kroger, in The Chemistry of Imperfect Crystals, North Holland, Amsterdam, 1964. Paper 5/06576C; Received 5th October, 1995 J. Muter. Chem., 1996,6(5), 887-893 893