Low-Pressure CVD of GeE (E = Te, Se, S) Thin Films from Alkylgermanium Chalcogenolate Precursors and Effect of Deposition Temperature on the Thermoelectric Performance of GeTe

Abstract

The homologous series [Ge n Bu3(E n Bu)] (E = Te, Se, S; (1), (3) and (4)) and [Ge n Bu2(Te n Bu)2] (2) have been synthesized as mobile oils in excellent yield (72–93%) and evaluated as single-source precursors for the low-pressure chemical vapor deposition (LPCVD) of GeE thin films on silica. Compositional and structural characterizations of the deposits have been performed by grazing-incidence X-ray diffraction, scanning electron microscopy, energy-dispersive X-ray analysis, and Raman spectroscopy, confirming the phase purity and stoichiometry. Electrical characterization via variable-temperature Hall effect measurements is also reported. Given the strong interest in GeTe and its alloys for thermoelectric applications, variable-temperature Seebeck data were also investigated for a series of p-type GeTe films. The data show that it is possible to tune the thermoelectric response through intrinsic Ge vacancy regulation by varying the deposition temperature, with the highest power factor (40 μW/K2cm@629 K) and effective ZT values observed for the films deposited at higher temperatures.

Boosting Thermoelectric Performance in Epitaxial GeTe Film/Si by Domain Engineering and Point Defect Control

“…Because the extra amount of p was decreased, the L-sample exhibited higher S than polycrystalline GeTe bulks ,,,− and polycrystalline GeTe thin films (Figure c). S values of the L-sample decreased with the increase in p as well as those of the preceding studies. ,,,− We calculated theoretical p – S curves based on the Boltzmann transport theory under relaxation time and single parabolic band approximations. Then, the density-of-states effective mass m d is an important parameter for the analysis of TE properties .…”

“…To understand the origin of higher μ, we measured the μ of the L-sample as a function of T in the range of 200−400 K. Figure 4d is the T dependences of normalized μ values μ nor , which was defined by normalizing the experimental μ values by that measured at 200 K. For comparison, the μ nor values of polycrystalline GeTe thin films (the solid square) 60 were also plotted simultaneously. Intriguingly, the μ nor values of the Lsample decreased more largely than those of polycrystalline GeTe thin films when increasing T, implying that carrier transport mechanisms in both samples can be different.…”

“…Because the extra amount of p was decreased, the L-sample exhibited higher S than polycrystalline GeTe bulks 27,28,47,52−59 and polycrystalline GeTe thin films 60 (Figure 4c). S values of the L-sample decreased with the increase in p as well as those of the preceding studies.…”

“…27,28,47,52−60 The average p value of the L-sample was ∼3 × 10 20 cm −3 , which was lower than those of polycrystalline GeTe bulks (∼4−9 × 10 20 cm −3 ) 27,47,52−59 and polycrystalline GeTe thin films (∼1−3 × 10 21 cm −3 ). 60 Furthermore, this p value was lower than or comparable to those of polycrystalline GeTe bulks containing an extra number of Ge (∼3−5 × 10 20 cm −3 ). 28 The lower p value of the L-sample is likely attributed to the reduction of the V Ge , which comes from the film growth under the situation that Te is easily volatilized.…”

“…As shown in Figure 4a, the L-sample exhibited relatively high μ comparable to polycrystalline GeTe bulks 27,28,47,52−59 at almost the same p, and higher μ than that of the polycrystalline GeTe thin films. 60 Therefore, the Lsample exhibited almost the same σ tendency against p as polycrystalline GeTe bulks unlike polycrystalline GeTe films (Figure 4b).…”