Engineering topics

Yu, Kin Man

Publications and source records attributed to Yu, Kin Man.

6 records

DOE OSTI2024

Overcoming the doping limit in GaAs by ion implantation and pulsed laser melting

Most semiconductors exhibit a saturation of free carriers when heavily doped with extrinsic dopants. This carrier saturation or “doping limit” is known to be related to the formation of native compensating defects, which, in turn, depends on the energy positions of their conduction band minimum and valence band maximum. Here, we carried out a systematic study on the n-type doping limit of GaAs via ion implantation and showed that this doping limitation can be alleviated by the transient process of pulsed laser melting (PLM). For n-type doping, both group VI (S) and amphoteric group IV (Si and Ge) dopants were implanted in GaAs. For comparison, p-type doping was also studied using Zn as the acceptor. Implanted dopants were activated by the PLM method, and the results are compared to rapid thermal annealing (RTA). Our results reveal that for all n-type dopants, while implantation followed by the RTA results in a similar saturation electron concentration of 2–3 × 10 18 cm –3 , the transient PLM process is capable of trapping high concentration of dopants in the substitutional site, giving rise to a carrier concentration of >10 19 cm –3 , exceeding the doping limit of GaAs. However, due to scatterings from point defects generated during PLM, the mobility of n-type GaAs after PLM is low (~80–260 cm 2 /V s). Subsequent RTA after PLM (PLM + RTA) is able to remove these point defects and recover the mobility to ~1000–2000 cm2/V s. The carrier concentrations of these PLM + RTA samples are reduced but are still a factor of 3 higher than RTA only GaAs. This can be understood as the dopants are already incorporated in the substitutional site after PLM; they are less likely to be “deactivated” by subsequent RTA. This work is significant to the understanding of doping mechanisms in semiconductors and provides a means for device applications, which require materials with ultra-high doping.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

DOE OSTI2023

Transition metal elements as donor dopants in CdO

CdO has been shown to achieve a high electron concentration N (> 10 21 cm -3 ) and at the same time a high mobility μ (> 100cm 2 /V s) when doped with conventional shallow dopants (In or Ga), and consequently making it a transparent conducting oxide with very low resistivity ρ <10 -4 $\Omega$ cm. In this work, the properties of CdO thin films doped with a series of transition metal elements (CdO:TM) with partially filled 3d and 4d shells, including Sc, Ti, V, Cr, Fe, Y, Mo, and W, were investigated. We find that doping with these TM elements can effectively increase the N in CdO to a maximum N (N max ) of ~(7-12)×10 20 cm -3 with a dopant concentration x max of 4-7 %. However, unlike CdO:In, the μ of CdO:TM films drops rapidly from > 100 to < 10cm 2 /Vs as the dopant concentration x increases, so that they can only achieve a minimum ρ of ~ (1-2) × 10 -4 $\Omega$ cm, ~ a factor of 2-3 higher than that in CdO:In. As a result, free-carrier absorption and plasma reflection effects limit their optical transparency to <1200 nm. For most 3d TM dopants, a qualitatively higher d-donor level E d,donor gives rise to higher E F,max or a higher N max . Although at low x, the optical band gap E opt of CdO:TM follows the calculated values due to free-carrier effects, as x increases, E opt values are significantly higher than the calculated values. This is believed to be an effect of the anticrossing interaction of the localized d-levels and the extended CdO conduction-band (CB) states, giving rise to a lower occupied E - and an upper unoccupied E + subband. In conclusion, the restructured CBs have much flatter dispersion, which also results in a much higher effective mass $m^{\star}_{e}$, hence it can also explain the much lower μ of CdO:TM films with high N.

36 MATERIALS SCIENCE↗

DOE OSTI2020

Temperature-dependent growth of hexagonal and monoclinic gallium sulfide films by pulsed-laser deposition

We demonstrate the selective, pulsed-laser deposition of hexagonal GaS and monoclinic Ga2S3 films on sapphire substrates from a single Ga 2 S 3 target in high-vacuum conditions. Growth at substrate temperatures below 550 °C causes GaS film formation, which indicates non-stoichiometric transfer from the target to the film. Surprisingly, stoichiometric transfer occurs at substrate temperatures above 650 °C with monoclinic Ga 2 S 3 as the preferred, higher S-content phase. Through a series of growth and annealing experiments, we show that GaS nucleation under S-deficient conditions leads to the preferred growth of this layered, hexagonal phase below 550 °C. Furthermore, GaS films annealed above 650 °C under high vacuum are transformed to Ga 2 S 3 , reflecting the greater stability of the monoclinic phase. By first growing Ga 2 S 3 at a higher temperature and subsequently growing GaS at a lower temperature, we can fabricate GaS/Ga 2 S 3 heterostructures in a single growth process.

36 MATERIALS SCIENCE↗

NASA NTRS2002

Origin of the Bandgap Bowing in ZNSe(1-x)Te(x) Alloys

Photomodulated reflection and optical absorption spectroscopies were used to measure the composition and hydrostatic pressure dependencies of interband optical transitions in ZnSe(1-x)Te(x) alloys. On the Te-rich side, the bandgap reduction and the reduction in the pressure dependence of the bandgap are well explained by the band anticrossing interaction between the ZnTe conduction band states and the Se localized state located at 2.85 eV above the valence band top of ZnTe. On the Se-rich side, an interaction between the degenerate Gamma valence bands of ZnSe and the localized Te state located at 0.1 eV above the valence band top of ZnSe is proposed to generate the bandgap reduction and the rapid increase of the spin-orbit splitting with increasing Te concentration. A linear interpolation between these two models accounts for the composition and pressure dependencies of the bandgap in the entire composition range.

Walukiewicz, W.↗

NASA NTRS2001

Band Anticrossing in Highly Mismatched Compound Semiconductor Alloys

Compound semiconductor alloys in which metallic anions are partially replaced with more electronegative isoelectronic atoms have recently attracted significant attention. Group IIIN(x)V(1-x), alloys with a small amount of the electronegative N substituting more metallic column V elements has been the most extensively studied class of such Highly Mismatched Alloys (HMAs). We have shown that many of the unusual properties of the IIIN(x),V(1-x) alloys can be well explained by the Band Anticrossing (BAC) model that describes the electronic structure in terms of an interaction between highly localized levels of substitutional N and the extended states of the host semiconductor matrix. Most recently the BAC model has been also used to explain similar modifications of the electronic band structure observed in Te-rich ZnS(x)Te(l-x) and ZnSe(Y)Te(1-y) alloys. To date studies of HMAs have been limited to materials with relatively small concentrations of highly electronegative atoms. Here we report investigations of the electronic structure of ZnSe(y)Te(1-y) alloys in the entire composition range, 0 less than or equal to y less than or equal to 1. The samples used in this study are bulk ZnSe(y)Te(1-y) crystals grown by either a modified Bridgman method or by physical vapor transport. Photomodulated reflection (PR) spectroscopy was used to measure the composition dependence of optical transitions from the valence band edge and from the spin-orbit split off band to the conduction band. The pressure dependence of the band gap was measured using optical absorption in a diamond anvil cell. We find that the energy of the spin-orbit split off valence band edge does not depend on composition and is located at about 3 eV below the conduction band edge of ZnSe. On the Te-rich side the pressure and the composition dependence of the optical transitions are well explained by the BAC model which describes the downward shift of the conduction band edge in terms of the interaction between localized Se states and the conduction band. On the other hand we show that the large band gap reduction observed on the Se-rich side of the alloy system is a result of an interaction between the localized Te level and the valence bands. This interaction leads to the formation of a Te-like valence band edge that strongly interacts with the light hole valence band. Calculations based on a modified k p model account for the reduction of the band gap and the large increase of the spin-orbit splitting observed in Se-rich ZnSe(y)Te(l-y) alloys. We will also discuss the importance of these new results for understanding of the electronic structure and band offsets in other highly mismatched alloy systems.

Yu, Kin Man↗

NASA NTRS2001

Band Anticrossing in Highly Mismatched Compound Semiconductor Alloys

Compound semiconductor alloys in which metallic anions are partially replaced with more electronegative isoelectronic atoms have recently attracted significant attention. Group IIIN(sub x)V(sub 1-x) alloys with a small amount of the electronegative N substituting more metallic column V elements has been the most extensively studied class of such Highly Mismatched Alloys (HMAs). We have shown that many of the unusual properties of the IIIN(sub x)V(sub 1-x) alloys can be well explained by the Band Anticrossing (BAC) model that describes the electronic structure in terms of an interaction between highly localized levels of substitutional N and the extended states of the host semiconductor matrix. Most recently the BAC model has been also used to explain similar modifications of the electronic band structure observed in Te-rich ZnS(sub x)Te(sub 1-x) and ZnSe(sub y)Te(sub 1-y) alloys. To date studies of HMAs have been limited to materials with relatively small concentrations of highly electronegative atoms. Here we report investigations of the electronic structure of ZnSe(sub y)Te(sub 1-y) alloys in the entire composition range, y between 0 and 1. The samples used in this study are bulk ZnSe(sub y)Te(sub 1-y) crystals grown by either a modified Bridgman method or by physical vapor transport. Photomodulated reflection (PR) spectroscopy was used to measure the composition dependence of optical transitions from the valence band edge and from the spin-orbit split off band to the conduction band. The pressure dependence of the band gap was measured using optical absorption in a diamond anvil cell. We find that the energy of the spin-orbit split off valence band edge does not depend on composition and is located at about 3 eV below the conduction band edge of ZnSe. On the Te-rich side the pressure and the composition dependence of the optical transitions are well explained by the BAC model which describes the downward shift of the conduction band edge in terms of the interaction between localized Se states and the conduction band. On the other hand we show that the large band gap reduction observed on the Se-rich side of the alloy system is a result of an interaction between the localized Te level and the valence bands. This interaction leads to the formation of a Te-like valence band edge that strongly interacts with the light hole valence band. Calculations based on a modified k(sup dot)p model account for the reduction of the band gap and the large increase of the spin-orbit splitting observed in Se-rich ZnSe(sub y)Te(sub 1-y) alloys. We will also discuss the importance of these new results for understanding of the electronic structure and band offsets in other highly mismatched alloy systems.

Yu, Kin Man↗