E. A. Kraut

Measurement of ZnSe–GaAs(110) and ZnSe–Ge(110) heterojunction band discontinuities by x‐ray photoelectron spectroscopy (XPS)

X‐ray photoelectron spectroscopy was used to study the growth and energy‐band alignment of ZnSe–GaAs(110) and ZnSe–Ge(110) heterojunctions. The ZnSe–GaAs heterojunctions were formed by growing ZnSe on GaAs(110). Growth temperatures were varied to produce both epitaxial and nonepitaxial interfaces. For ZnSe grown at ∠300 °C on GaAs(110), the valence‐ band discontinuity ΔEv was 0.96 eV; for ZnSe deposited at room temperature and crystallized at ∠300 °C, ΔEv is 1.10 eV. The Ge–ZnSe(110) interfaces were formed by depositing Ge(ZnSe) on ZnSe(Ge)(110) at room temperature, followed by ∠300 °C crystallization. The corresponding ΔEv’s were 1.52 and 1.29 eV, respectively. Our measured ΔEv values for epitaxial heterojunctions are compared with the predictions of theoretical models. Our results demonstrate that substantial interface structure dependent contributions to ΔEv can occur at Ge–ZnSe(110) and GaAs– ZnSe(110) heterojunctions.

Measurement of semiconductor heterojunction band discontinuities by x‐ray photoemission spectroscopy

Accurate knowledge of the band discontinuities at a heterojunction interface and the factors that affect their magnitude are of both fundamental and practical interest. The application of x‐ray photoemission spectroscopy (XPS) to the direct, contactless, and quantitative measurement of the valence band discontinuity (ΔEv) at abrupt heterojunctions is discussed. The topics covered include a description of a method to achieve precise measurement, results of ΔEv measurements for many heterojunction pairs selected from the lattice‐matched semiconductor series Ge, GaAs, ZnSe, CuBr, and AlAs, and a comparison of experiment to models that predict ΔEv. Examples are given to illustrate the effect on ΔEv of such preparation‐dependent parameters as growth sequence and interface crystallographic orientation.

Determination of the InAs–GaAs(100) heterojunction band discontinuities by x‐ray photoelectron spectroscopy (XPS)

The valence‐ and conduction‐band discontinuities have been determined for the InAs–GaAs (100) heterojunction (HJ) by means of XPS measurements. The values are ΔEv = 0.17±0.07 eV and ΔEc = 0.90±0.07 eV, respectively. The InAs‐GaAs HJ has a rather large lattice mismatch (7%) but high quality epitaxial heterojunctions may be grown by molecular beam epitaxy (MBE). We have carried out XPS measurements on MBE grown InAs (100) films and on the InAs‐GaAs (100) HJ by using a recently developed arsenic cap transfer technique which permits XPS measurements on atomically clean and ordered heterojunctions. XPS is used to obtain the In4d and As3d to valence‐band maximum binding‐energy differences for bulk InAs by fitting the experimental XPS valence‐band density of states (VBDOS) to an instrumentally broadened theoretical VBDOS. The values are (EInAsIn4d−EInAsv) = 17.43 ±0.02 eV and (EInAsAs3d−EInAsv) = 40.77±0.02 eV, respectively. The core‐level separation ΔECL between the In4d and Ga3d core levels for a thin InAs–GaA...

Measurement of InP/In0.53Ga0.47As and In0.53Ga0.47As/In0.52Al0.48As heterojunction band offsets by x‐ray photoemission spectroscopy

X‐ray photoemission spectroscopy (XPS) has been used to measure the valence‐band offset ΔEv for the lattice‐matched InP/ In0.53Ga0.47As and In0.53Ga0.47As/ In0.52Al0.48As heterojunction interfaces. The heterojunctions were formed by molecular‐beam epitaxy. We obtain values of ΔEv (InP/In0.53Ga0.47As) =0.34 eV (ΔEc/ ΔEv=43/57) and ΔEv (In0.53Ga0.47As/ In0.52Al0.48As) =0.22 eV (ΔEc/ ΔEv =68/32) for the respective interfaces. By combining these measurements with available XPS ΔEv (InP/ In0.52Al0.48As) data we find that band offset transitivity is satisfied. Accordingly, the band offsets for heterojunction pairs formed from InP, In0.53Ga0.47As, and In0.52Al0.48As are not influenced by interface specific effects.

Measurement of potential at semiconductor interfaces by electron spectroscopy

Electron spectroscopy performed in ultrahigh vacuum can be used to measure potential and heterojunction band discontinuities at abrupt semiconductor interfaces. The technique provides a direct contactless and nondestructive means to determine and correlate interface chemistry and potential. This article discusses some of the factors which affect applications of Auger electron spectroscopy, ultraviolet photoelectron spectroscopy, soft x‐ray photoelectron spectroscopy, and x‐ray photoelectron spectroscopy for semiconductor interface potential measurements.

Effect of growth sequence on the band discontinuities at AlAs/GaAs (100) and (110) heterojunction interfaces

The valence band discontinuity ΔEv for the AlAs grown on GaAs (AlAs–GaAs) heterojunction interface is compared to the corresponding ΔEv for the GaAs grown on AlAs interface (GaAs–AlAs) for both (100) and (110) crystallographic orientations by using x‐ray photoemission spectroscopy (XPS). Molecular‐beam epitaxy within the XPS system was used to prepare the samples. The XPS method has the feature that ΔEv differences are measured more accurately than the corresponding absolute magnitudes. For the (100) interface, the value of ΔEv(AlAs–GaAs) is 0.10±0.02 eV larger than ΔEv(GaAs–AlAs). For the (110) interface, ΔEv(AlAs–GaAs) is 0.13±0.04 eV larger than ΔEv(GaAs–AlAs). The average discontinuity magnitudes are: ΔEv(AlAs–GaAs)(100)=0.46 eV; ΔEv(GaAs–AlAs)(100)=0.36 eV; ΔEv(AlAs–GaAs)(110)=0.55 eV; and ΔEv(GaAs–AlAs)(110)=0.42 eV. Thus, both a growth sequence effect and an orientation dependence have been found to be intrinsic characteristics of the (100) and (110) AlAs/GaAs heterojunction interfaces. As a conseq...

Measurement of GaAs/InP and InAs/InP heterojunction band offsets by x-ray photoemission spectroscopy

The unstrained valence‐band offset ΔEv for the x=0 and x=1 end points of the InxGa1−xAs/InP (100) heterojunction system has been measured by x‐ray photoemission spectroscopy (XPS). Although the GaAs/InP and InAs/InP interfaces are strained because of lattice mismatch, the ΔEv values obtained by the XPS measurement method used are characteristic of an unstrained interface. Values of ΔEv (GaAs/InP)=0.19 eV and ΔEv (InAs/InP)=0.31 eV are observed. A linear interpolation between the x=0 and x=1 values gives ΔEv (In0.53 Ga0.47As/ InP)=0.25 eV for the x=0.53 lattice‐matched interface (ΔEc /ΔEv =58/42).

Valence-band discontinuities for abrupt (110), (100), and (111) oriented Ge-GaAs heterojunctions

Abstract Valence-band discontinuities, Δ E v , have been derived from XPS data on abrupt Ge-GaAs interfaces for five different crystallographic orientations. The Δ E v values for epitaxial Ge layers grown on (111)Ga, (100)Ga, (110), (100)As, and (111)As GaAs substrates are 0.48, 0.55, 0.56, 0.60 and 0.66 eV, respectively.

Effect of interface composition on the band offsets at InAs/AlSb (001) heterojunctions

X‐ray photoemission spectroscopy has been used to directly measure the dependence of the valence band offset ΔEv on interface composition (AlAs‐like versus InSb‐like) for InAs/AlSb (001) heterojunctions grown by molecular‐beam epitaxy. A staggered band alignment with a consistent average value of ΔEv = 0.16 eV is measured for three samples with the InSb‐like interface. A difference of 0.07 eV is observed between the average ΔEv values for two AlAs‐like interface samples (0.15 and 0.22 eV, respectively). The growth‐to‐growth variation in ΔEv value for the AlAs‐like interface, and the 0.06 eV difference between the InSb‐like ΔEv value and the larger AlAs‐like ΔEv value are both attributed to interface composition differences. We interpret the ΔEv = 0.22 eV value as being most characteristic of the AlAs‐like interface. A tight binding model calculation of the effect of local strain on InAs/AlSb band offsets is compared to experiment.

Diffraction of Electromagnetic Waves by a Right‐Angle Dielectric Wedge

The problem of diffraction of a plane‐polarized electromagnetic wave incident on a right‐angle dielectric wedge is formulated as a singular integral equation in k space. A solution of the singular integral equation is constructed as a power series in the index of refraction. This series converges when the index of refraction is near unity. Using this solution, the electric‐field amplitude at the tip of the wedge is examined. We also prove as incorrect a closed‐form analytic expression claimed in the literature to be a global solution of the problem considered here.