Effects of covalency,p−dcoupling, and epitaxial strain on the band offsets of II-VI semiconductors

Abstract

Using the first-principles all-electrons method, we have systematically studied the natural band offsets among zinc blende BeX, MgX, and ZnX (XϭS, Se, Te͒. We show that ZnX, which has large anion p-cation d repulsion, always has higher natural valence band maximum ͑VBM͒ than BeX and MgX, whereas BeX, which shows strong covalency, has higher natural VBM than MgX due to kinetic-energy-induced valence band broadening. However, epitaxial strain could reverse these trends. We found that for these isovalent semiconductors, the band offset is not sensitive to interface atomic compositions.Extensive experimental and theoretical studies have been carried out to understand the nature of the band offsets between two semiconductor compounds forming heterostructure. 1,2 The band offset is one of the most important parameters in device modeling 1 and in determining the dopability of semiconductor compounds. 3 For conventional II-VI as well as III-V semiconductors, it is shown 2 that ͑i͒ for common-cation systems ͑e.g., ZnS, ZnSe, and ZnTe͒, the valence band maximum ͑VBM͒ increases as the anion atomic number increases, whereas the valence band offset ⌬E v decreases as the cation atomic number increases, e.g., from ZnX to CdX to HgX (XϭS, Se, and Te͒. ͑ii͒ For commonanion system, the compound with shallow occupied cation d orbitals ͑e.g., ZnSe͒ always has higher VBM than the one without it ͑e.g., MgSe͒. The valence band offset ⌬E v decreases when the anion atomic number increases, e.g., from sulphides to selenides to tellurides. It has been shown that coupling between the anion p and cation d states plays a decisive role in determining these chemical trends of valence band offsets. 2 Furthermore, previous studies 4,5 have shown that, for isovalent semiconductors, the band offset is not sensitive to the interface atomic structures.Recently, considerable interest has arisen for using BeX, MgX, and their alloys with the conventional II-VI semiconductor ZnX as materials for light-emitting and laser diodes in the blue/green region. 6 In addition to the technological importance, these unconventional II-VI compounds also show some interesting, but unexplained, physical phenomena. For example, recent photoelectron spectroscopy measurements of Nagelstrasser et al. 7 show that ⌬E v at the BeTe/ZnSe interface is highly dependent on the interface composition and can vary by as much as 0.8 eV. Subsequent ab initio pseudopotential calculations of Bernardini et al. 8 show that the experimentally proposed valence band offset cannot occur in the nominal BeTe/ZnSe interface. Instead, other types of interfaces, such as BeSe/ZnSe or ZnTe/ZnSe may form, which could explain the scattering of the experimental data. However, their theoretical study does not include the spin-orbital coupling. Furthermore, they only presented the band offset for strained interfaces and show that BeSe, which does not have an active cation d orbital, has a higher VBM than that for ZnSe, which does have an active Zn 3d orbital, in contrast to the general understanding. 2 It...

THEORETICAL CHARACTERS OF MgX (X = Te, Se, S AND O) CLUSTERS

“…18 Synchronously, there are many other papers reporting characters of MgX materials in theory. [19][20][21][22][23][24][25][26][27][28] Lately, MgTe and its mixed ternary alloys have been extensively studies, containing electronic, lattice dynamical and mechanical properties in zinc-blende structure. 29 However, the Raman and absorption spectra, bonding characters and molecule orbital of MgTe, MgSe, MgS and MgO have not been reported to this day.…”

Structural properties and spatial ordering in multilayered ZnMgTe/ZnSe type-II quantum dot structures

“…To further explore the possibility to enhance p-type doping in this material system, we have attempted to modify the QD bandgap by incorporating submonolayer quantities of Mg along with Te to form ZnMgTe instead of pure ZnTe QDs. The choice of Mg was made due to the absence of cation core d-electrons in MgTe, which increases the bandgap and lowers the valence band maximum relative to that of ZnTe, 5,6 reducing the valence band offset with ZnSe. 6,7 This in turn is expected to reduce the hole confinement energy, and hence to enhance the p-type conductivity.…”

“…The choice of Mg was made due to the absence of cation core d-electrons in MgTe, which increases the bandgap and lowers the valence band maximum relative to that of ZnTe, 5,6 reducing the valence band offset with ZnSe. 6,7 This in turn is expected to reduce the hole confinement energy, and hence to enhance the p-type conductivity. A hole free-carrier concentration of the order of mid-10 15 cm À3 has been measured, 8 for the first time, in such samples.…”

Radiative transitions in stacked type-II ZnMgTe quantum dots embedded in ZnSe

“…The carrier confinement in type-II systems, such as Zn-Se-Te, is independent of the bandgaps of the underlying materials but rather depends on the band offsets. The choice of Mg is due to the absence of cation core d-electrons in MgTe, which while increases the bandgap, more importantly, lowers the valence band maximum relative to that of ZnTe, 10,11 reducing the valence band offset (VBO) with ZnSe. 11,12 Thus, incorporation of Mg in ZnTe is expected to reduce the hole confinement energy and enhance p-type conductivity.…”

“…The choice of Mg is due to the absence of cation core d-electrons in MgTe, which while increases the bandgap, more importantly, lowers the valence band maximum relative to that of ZnTe, 10,11 reducing the valence band offset (VBO) with ZnSe. 11,12 Thus, incorporation of Mg in ZnTe is expected to reduce the hole confinement energy and enhance p-type conductivity. A hole free carrier concentration in the order of mid-10 15 cm À3 has been measured for the first time, 13 in such samples.…”