Residual Donor Concentration Research Articles

Type conversion of (Hg,Cd)Te induced by the redistribution of residual acceptor impurities

Bulk (Hg,Cd)Te grown by the quench-anneal process has been found to contain residual acceptor impurities which are normally depleted from the n-type surface by an internal gettering mechanism during low temperature postannealing in saturated Hg vapor. When the n-type slices are postannealed for much longer periods, the Te saturated core is annihilated, and the gettered residual acceptor impurities can redistribute into the surface region. If upon homogenization the residual acceptor concentration exceeds the residual donor concentration, the material will convert to p type. A new method for applying neutron activation analysis (NAA) to (Hg,Cd)Te has been used to identify this acceptor impurity as Cu at levels below 1 ppm. An extension of the NAA technique has employed quartz encapsulated silicon samples to demonstrate Cu may come from either the quartz or annealing environment.

Advances in LEC growth of InP crystals

By reducing the temperature gradients in the vicinity of the crystal-melt interface, 35-mm-diameter InP boules with much reduced dislocation densities have been grown by the liquid-encapsulated Czochralski technique. A reduction in the residual donor concentration of InP grown by this technique has been achieved by using In-rich charges prepared by adding elemental In to polycrystalline InP ingot material. Nominally undoped crystals with carrier concentrations as low as 1–2 x4 1015 Cm − 1 and 77 K mobilities as high as 7.0 × 10 cm2 V−1 s−1 have been obtained. By growing doped crystals at increased seed or crucible rotation rates, short-range longitudinal variations in dopant concentration have been reduced to a few per cent, as determined by optical absorption measurements with a scanning CO2 laser.

Epitaxial growth of GaAs by low-pressure MOCVD

The epitaxial growth of GaAs from trimethylgallium and arsine at 1–5 Torr was investigated. At deposition temperatures of 580–750°C the residual donor concentration decreased with increasing arsine excess, not increased as has been reported for higher pressures. Increases in growth rate also markedly reduced the rate of incorporation of the residual donor, and at high growth rates the films became p-type. Under optimum conditions the electron mobility and the residual impurity concentrations obtained were comparable to the best values reported for films grown using metalorganics.

Semi-insulating In 0.53Ga 0.47As by Fe doping

High resistivity epilayers of In 0.53Ga 0.47As have been prepared by liquid-phase epitaxy (LPE) using Fe as a deep level acceptor to compensate the residual donor background impurities. Intrinsic conduction of this 0.75 eV bandgap ternary limits the resistivity to values much lower than for InP or GaAs, however, room-temperature values as high as 1.7×10 3 ω cm have been observed. Growth of the epilayers is from In solutions at 640°C on (100) Fe-doped InP substrates. Electrical compensation of the Fe dopant is critically dependent on reduction of residual donor concentrations by extended purging of the growth chamber and bake-out of the growth solutions. Insulated-gate field-effect transistor (MISFET) structures fabricated on the semi-insulating (SI) In 0.53Ga 0.47As epilayers indicate that the surfaces are slightly accumulated, and conduction modulation is readily achieved by electron accumulation at the surface with positive gate bias.

SIMS profile analyses for Ge-, Co- and Fe-doped InP substrates used in epitaxial growth

Cockayne and co-workers [1, 2] have shown recently that germanium (Ge) or cobalt (Co) doping of liquid-encapsulation Czochralski (LEC) grown indium phosphide (InP) single crystals produces potentially useful properties in the application of this I I I -V compound as a substrate material for the manufacture of a number of electronic devices. At high concentrations (>1019 atoms cm-3), germanimn hardens the lattice sufficiently to inhibit dislocation generation during LEC growth and allows single crystals with dislocation densities close to zero to be obtained [1]. Cobalt, on the other hand, can be used at low doping levels (~3 x 101Satomscm-3) to induce resistivities in the region of 107ohmcm [2]. An important factor in determining applications for these materials is a knowledge of the extent to which cobalt and germanium migrate into epitaxial layers grown on to substrates doped with these elements. Here, we report the results of secondaryion mass spectrometric (SIMS) analysis of undoped layers grown on to such substrates by either vapour phase epitaxy (VPE) or molecular beam epitaxy (MBE). Comparison is made with layers grown similarly on to Fe-doped substrates. The single crystals used as substrates were all grown by the standard LEC technique using r.f. heating [3] from polycrystalline material supplied by MCP (Electronic Materials) Ltd. This yields undoped crystals with a residual donor concentration within the range 2 to 5 x l0 is cm -a at RT (room temperature) and with a corresponding mobility range of 4000 to 5000cm 2 V -1 sec -1. The highpurity dopants, germanium, cobalt or iron were added to the starting charge prior to melting and growth. The dopant levels, quoted subsequently, were determined either chemically by spark-source mass spectrometry or from electrical data obtained by the Van der Pauw technique. Epitaxial layers were grown on to (1 0 0) substrates from the crystals using either VPE or MBE techniques. Prior to growth polished substrates were degreased in propan-2-ol, dried, dipped in 0.3% bromine in methanol solution for 60 sec, rinsed in deionized water and then blow dried. The VPE process employed was the well established technique [4] based on the reaction between indium and PC13 in a stream of H2. In this process, the substrate is heated in H~ to the growth temperature of 650 ° C over a period of 30 min followed by growth for a further 40 rain to produce layers 2 to 3 #m thick. The MBE process used indium and P2 beams in ultra-high vacuum [5], wherein the substrates are heated for 5 min at 505 ° C in a P2 flux to facilitate cleaning without surface degradation followed by growth at 420 ° C for 3 h to produce layers 0.5 to 1 #m thick. SIMS analysis was performed with a Cameca 1MS3F instrument. All samples were analysed under standard conditions using an O~ primary beam to erode through the epitaxial layer into the substrate. The isotopes counted were S6Fe+, S9Co + and 72Ge+, all for a period of 30 sec at each sampling point. Each sample was analysed by rastering a beam of either 0.3/2A or 1 gA over an area of 500#m square. An aperture diaphragm was inserted into the secondary ion beam to restrict ions counted to a central area of 60gin diameter, to avoid crater edge effects. The energy discriminating slits between the electrostatic and magnetic analysers were opened to allow a 10 eV bandpass centred on the maximum energy of an aluminium standard specimen, no offset being used. All other apertures and slits were set for maximum trans-

An analytical evaluation of GaAs grown with commercial and repurified trimethylgallium

An analytical study of the impurities in trimethylgallium (TMGa) and subsequent correlation of the effect of these impurities on resulting GaAs films grown by metalorganic chemical vapor deposition (MOCVD) is presented. The effects of using fractional distillation techniques to improve the quality of TMGa and to help isolate and identify major source impurities in TMGa is detailed. Photothermal ionization data are presented which show the residual donor species present and their relative concentrations in the epitaxial layers. Correlations of the residual donor concentrations with TMGa preparation are made. It is demonstrated that high purity GaAs with μ77 K ≈ 125,000 cm2/V-sec can be grown by MOCVD using repurified trimethylgallium and arsine source materials.

Liquid Phase Epitaxial Growth of High Purity GaAs by Sliding Boat Method

High purity LPE layers are grown by the sliding boat method, varying the experimental conditions, such as the pre-heat temperature, growth temperature and the type of carrier gas. Dominant residual impurity and its purification mechanism are systematically investigated. The results of the present work (residual net donor concentration>1×1014 cm-3) can be satisfactorily explained by assuming oxygen as a dominant residual impurity which behaves as a shallow donor. The purification mechanisms are explained by both deoxidization of Ga2O3 by the hydrogen gas and the desorption of H2O and O2 adsorbed by the porous graphite. It is possible to grow high purity undoped layers reproducibly with carrier concentrations in the (1∼3)×1014 cm-3 range, taking special care not to introduce oxygen in the grown layers.

Behaviors and Doping Kinetics of Residual Impurities in Epitaxial n-GaAs Layers

Electrical properties are measured at 300°K and 77°K of epitaxial n-GaAs layers grown by the Ga/AsCl3/H2 method under various growth conditions. To investigate doping behaviors of main residual impurities, donor and acceptor concentrations ND and NA are calculated as a function of the growth temperature and substrate orientation ((100), (110), (311) A and (111) B). For each orientation, the compensation ratio K(≡NA/ND) is found to be nearly constant and K is increased as the arsenic partial pressure is decreased, both suggesting that amphoteric impurities (probably Si) are main donors and acceptors. Reaction kinetics are developed based on a simple model of amphoteric impurity doping and compared with experiment. Deviations of K from the constant value and some other possible effects are also discussed.

Distribution of Impurities in Zn,O‐Doped GaP Liquid Phase Epitaxy Layers

A series of variously doped liquid phase epitaxy (LPE) layers were analyzed to establish the distribution of Zn,O and residual impurities contributing to the net impurity gradient previously observed in the p‐layer of high efficiency LPE diodes. From these experiments we conclude that the net impurity gradient is primarily a consequence of a decreasing Zn concentration along the growth direction, and, to a lesser degree, an increasing residual donor concentration. The distribution of intentional and unintentional impurities was found to be independent of substrate doping level.The effect of cooling rate on the various impurity distributions was also studied, covering the range 0.5°–18°C/min. Residual impurity levels and associated gradients were significantly reduced by decreasing the cooling rate which indicates that residual impurity incorporation is kinetically controlled, probably by slow diffusion of these impurities in front of the growing LPE layer. Similarly, the O level appears to be decreased by reduced cooling rate. In contrast, Zn doping was found to be independent of cooling rate which suggests that its incorporation was essentially under equilibrium conditions. Assuming this to be the case, we have shown that the observed Zn distribution is consistent with the temperature dependence of the Zn solid solubility. Using an equilibrium description for Zn incorporation, we have computed the near‐junction Zn and O concentrations in the previously reported high efficiency diodes to be , respectively. The corresponding level of residual impurities yields a net donor concentration of .