Temperature and Composition Dependence of Exciton Peak Positions and Band Gap Energies of Zn1?xMgx(?0.19Se Epitaxial Films

Abstract

The temperature dependence of the 1s exciton energy has been measured in Zn1—xMgxSe epitaxial films at compositions x = 0, 0.07, 0.12, and 0.19 from 2 K up to 280 K. Detailed numerical fits of the individual temperature dependences are provided on the basis of an analytical four‐parameter representation developed recently by one of the authors. These are compared with previously used three‐parameter models of Viña et al. and Varshni. The x‐dependence of the exciton energy, E1s(T, x), and of the fundamental band gap energy, Eg(T, x), is given to very good approximation by linear functions of the composition x for any T from absolute zero up to room temperature. A comparison with recent room temperature band gap energy data by Jobst et al. shows that this linear dependence holds up to x ≈︂ 0.7. The magnitudes of the model‐dependent empirical parameters, which control the temperature dependence of the band gap energy in different compounds, are found to change significantly with increasing magnesium content. From the magnitude of the effective phonon temperature, particularly in the case of ZnSe, we conclude that the main contributions to the band gap shrinkage effect are due to acoustic phonons.

Parameter Sets Due to Fittings of the Temperature Dependencies of Fundamental Bandgaps in Semiconductors

“…Instead, according to [50], one has to endow the data points in question with weighting factors which are inversely proportional to the square of the corresponding error bars (cf. [23] and [32]). This means, for the data constellation in consideration, a reduction of the weight for the E g (T !…”

“…In analogy to most gallium-and indium-based III±V compounds, a considerable variety of E g (T) and E gx (T) data sets has been published during the past three decades also for II±VI compounds (for a list of references see [31,32]). The majority of these data sets, however, can be readily seen to be rather scarce, rather restricted (usually to T 300 K) and/or relatively inaccurate, so that they appear more or less insufficient for a trustworthy determination of empirical parameter sets (independently of the analytical model used).…”

“…Special experimental efforts were necessary for the wide bandgap materials zinc sulphide (ZnS) and zinc selenide (ZnSe), since the corresponding Debye temperatures are relatively high (Q D % 440 K and 340 K, respectively). Two qualitatively different methods have been combined in [31] in order to determine the 1s exciton peak energy positions from cryogenic temperatures up to about 4Q D /3 [31], namely: absorption measurements performed within the ranges (2 to 387) K for ZnS [31] or (4 to 280) K for ZnSe [32] and reflectance measurements performed within the ranges (303 to 541) K for ZnS or (300 to 500) K for ZnSe. In contrast to both wide bandgap materials, the Debye temperature in the case of zinc telluride (ZnTe) is relatively low (Q D % 260 K).…”

“…(1)) and a series of reflectance data points covering the range between 300 and 500 K. Let us compare the present parameter values obtained by fitting the whole data set for ZnSe with eq. (1) (given in Table 1) with the earlier values given in [32] (which resulted from a fit of merely the E 1s (T 280 K) data set [32] by the same equation). From this comparison we see that the parameter shifts caused by the inclusion of the 21 additional points [31] (filled circles in Fig.…”

Temperature dependence of excitonic emission in cubic CdSe thin film

“…It is very common that Varshni parameters fall in a broad range for the same material. This could be attributed to the empirical origin of Varshni relation [18,19]. Manoogian and Woolley [15] have suggested that after removing the contribution of the thermal expansion of lattice, the b parameter follows the…”