HAL Id: jpa-00225117
https://hal.archives-ouvertes.fr/jpa-00225117
Submitted on 1 Jan 1985
HAL is a multi-disciplinary open access
archive for the deposit and dissemination of
sci-entific research documents, whether they are
pub-lished or not. The documents may come from
teaching and research institutions in France or
abroad, or from public or private research centers.
L’archive ouverte pluridisciplinaire HAL, est
destinée au dépôt et à la diffusion de documents
scientifiques de niveau recherche, publiés ou non,
émanant des établissements d’enseignement et de
recherche français ou étrangers, des laboratoires
publics ou privés.
LUMINESCENCE INDUCED BY DEFECTS IN
CsCdBr3
Chantal Andraud, F. Pellé, O. Pilla
To cite this version:
JOURNAL DE PHYSIQUE
Colloque C7, supplément au nOIO, Tome 46, octobre 1985 page C7-489
LUMINESCENCE INDUCED BY DEFECTS
I N
CsCdBr3C. Andraud, F. Pellé and 0. pilla+
C.N.R.S. Ecole CentraLe des A r t s e t Manufactures, 92295 Chatmay-MaZabry Cedex, France
Résumé
Les propriétés optiques du composé monocristallin CsCdBr ont été étudiées. Les effets des impuretés, de la temperature: des traitements thermiques et de l'irradiation X sont discutés. Les bandes observees sont attribuées à des pseudomol~cules Br;-
localisées sur des défauts d'empilement.
-
Abstract
Optical spectra of pure CsCdBr3 crystals are reported. The effects of impurities, temperature, thermal annealing and X
irradiation are discussed. The main features are assigned to a B r 2 pseudomolecule on a lattice defect.
CsCdBr3- crytallizes in the hexagonal P6 /mmc space group. The 3
important ieature of this structure is its one-dimensionality. It consists of chains of face sharing (CdBr6)4-octaedra parallel to the C axis. The CS+ ions form an hexagonal sublattice between the chains (Fig. 1 a).
Hexagonal cubic
-
@.O
Cs+cd++
Br-
Figure 1: Perovslcite-lilce structures. Al1 ions illustrated lie in the (110) plane.(a) hexagonal structure, ( c ) cubic perovskite, (b) a possible stacking fault of CsCdBr in presence of PL++ or Sn++
impurities. 3
Crystals were grown by the Bridgman method under bromine atmosphere. The crystal structure and the impurities concentration were checked by standard methods; the level of unwanted impurities is found to be less than 10ppm. As grown pure crystals are transparent and easily cleave along the C axis.
The absorption edge lies at about 5eV and no luminescence is detected upon excitation at energy higher than 5eV, whereas several emissions are observed when exciting in the band gap.
+permanent address : Dipartimento di Fisica, Università di Trento, Italy
C7-490 JOURNAL DE PHYSIQUE
Al1 the studied crystals, excited at 3.9-4eV, show an emission band at about 3.29eV at IOK. At temperatures higher than IOOK, the 3.29eV emission disappears and an emission band grows up at 1.9eV (T=80K) and shifts with temperature (2.leV at 300K). Both the 3.29eV and the 2eV emissions have the same excitation spectrum which consists, at low temperature, of a well defined doublet at 3.91 and 3.97eV1 the halfwidth of each component being about 50meV (Fig.2). i4oreover another weak emission is observed in the range 50 - 150K at 2.9eV whose excitation spectrum consists of a broad band at 4.leV (
halfwidth
-
0.6eV). It should be mentioned that the intensity of this latter band increases with respect to the other emissions in iodine doped samples, where an absorption band is also detected at 4.2eV. t"10reover similar excitation and emission bands have been observed in iodine doped CsCdClg (1). Assignement of this band to centers related to iodine impurities is tempting but would require further investigations.1.5 2 2.5 3 3:5 L
Energy ( e V 1
Figure 2: Excitation and emission spectra of pure CsCdBr3. Full line T-IOK, dashed line T=170R. The intensities are not on scale.
Both 3.9eV excitation and related emissions are strongly dependent on nature and concentration of impurities. In crystals doped with trivalent and some divalent impurities such as Hn++the bands are wealc or not observed at all. On the contrary, these are more intense in lead and tin doped crystals (2). In these samples the 3.9eV band is directly detectable in the absorption spectrum, while in pure crystals, it is not.
The 2eV emission it is much weaker than the 3.29eV one at 10K (Fig.3) and its width increases with temperature £rom 0.3 to 0.6eV.
We have fitted the thermal behavior of the widths of the two emission bands. In the single mode coupling approximation using the well-known relation:
6 =(hw/2m ) coth(hw/2k~)
we obtain hw of the order of some 60 cm-'for both bands. This value seems to be reasonable (3).
The lifetimes of the two emissions have been recorded at various temperatures, under excitation in the 3.9eV doublet using the second harmonic generation of a pulsed dye laser ( ~ = 4 n s ) . The lifetime of the high energy band is found to be 20ns at IOK, 15ns at 50K and at
T b 70K, it is faster than the exciting pulse. The 2eV emission
lifetime is much slower: ~ = 3 (1.5) vs at 70 (150)K
.
Figure 3: Emissions intensity as a function of temperature. ( + )
E=3.2eVr
.
2=2 eV. (note that the intensities of the 2eV emission band have been multiplied by a factor of 10).> t V)
5
so- + Z ' OThe two emissions and their excitation band could be related to the energy level scheme of a Br;- molecule located on a staclcing fault (Fig.lb). By referring us to the classical Self Trapped Exciton model (4), the 3.9eV doublet corresponds to the promotion of an electron to the 2p excited level whereas the 3.29 and 2eV emissions arise £rom the 2s,ls levels respectively.
Actually the local symmetry of the center is lower than Oh so that, for instance, the 2p level degeneracy is removed. This model can account for the observed polarization, widths, lifetimes of the three bands at least in a qualitative way. More precise assignments indeed would require the exact knowledge of the symmetry of the center, which is not yet available.
This hypothesis is supported by several observations. First of al1 is that CsPbBr3 (5) and CsSnBr (6) crystallize in the cubic Perovskite structure (octahedra d a r i n g corners) (Fig. 1 c) in addition in ~ b + + doped crystals
,
no emission has been recorded £rom the isolated lead ions. Even at low concentration, the lead impurities aggregate and we observe the characteristic emission lines of CsPbBr3 (7).The stacking faults, in this case, would be preferentially located on the border of the CsPbBr3 microcrystals. It is straightforward that the presence of cubic microdomains enhances the number of defects (see Figure 1). No 3.2eV emission is observed in Mn++ doped samples, but CsMnBr3 crystallizes in an hexagonal structure such as CsCdBr3. The introduction of trivalent~7-492 JOURNAL DE PHYSIQUE
impurities could also stabilize the chain structure, but because of charge compensation these ions enters in the sequence 142' -Vacancy-Me3+
.
These simple considerations could account for the observed dependence of the optical spectra on both nature and amount of impurities. It is worthwhile to note that another evidence of the relation between structural defects and the observed bands is the effect of thermal treatement. In annealed pure crystals (T = 50010 both emission and excitation bands definitively disappear, while no effect, apart from a slight narrowing of the 3.9eV doublet, is observed in lead doped crystals where the defects are not only due to the thermal hystory of the sample but are related to the impurity.Under X-ray excitation at low temperatures, we have also observed at about 2.leV the Vk-e- recombination luminescence. This band, which is in the proposed mode1 the analogous of the 2eV emission, arises £rom Br;- centers in a regular environment.
The 3.9, 3.2, 2eV bands have already been observed (2), but while for the 3.9 and 3.2eV bands a relation with the chain defects is proposed, the 2eV band is there attributed to the presence of lead. On the light of Our observations and the lead concentration in Our pure crystals being 0.5 ppm the relation of this latter emission with lead impurities is not probable. A more precise assignment of these bands require other experimental investigations which are in progress.
References
1. Pidzyrai10,N.S. and Khap1c0,Z.A.~ Ulcr. Fiz. Zh. 30 (1 985) 189. 2. PJolfert, A. and Blasse, G., J. Solid State Chem. 55 (1984) 344.
3. Orinchai, A.V., Lazarev, J.B., Peresh, E., oper ri os, B.M. and ~ ' O r d y a i , T.S., inorg. mat. 18 (1982) 1080.
4. Tanimura, K. and Itoh, n. Semiconductors and Insulators 5
(1983) 473 and references therein.
5. Haupt, H.J, Heidrich, K., Kunze1,H. and aversb berger, P., Z. Phys. Chem. BI10 (1978) 63.
6. Clark, S.j., Flint, C.D. and Donaldson, J.D., J. Phys. Solids 42 (1981) 133.
