Pt and Pd hydrogen defects

PtH and PdH have donor and acceptor levels lying deeper in the gap than those of Pt and Pd respectively. Adding a second H atom, however, results in a rapid drop in the t₂ manifold and the donor levels are pushed into the valence band. The acceptor levels are, however, hardly affected. The (-/-) level of PdH₂ appears to be very shallow. PtH₃ and PdH₃ defects are particularly interesting as the donor levels are absent and the acceptor levels are very deep. It seems that there is strong affinity to fill the t₂ manifold. These (-/0) levels are calculated to lie 1.4 eV and 1.0 eV below E_c although of course the error in the estimates increases with the depth. PtH₄ and PdH₄ also possess deep acceptor levels probably related to the 5s and 6s levels being brought into the gap.

Experimentally, in Pt doped hydrogenated p-Si, a level at E_v+.4 eV (H210) is formed below 400K, which upon annealing is converted into a deep acceptor, at E_c-.9 eV (H150) [#!sachse2-97!#]. This defect dissociates into Pt above 530K. In n-Si, two levels at E_c -.18 eV (E90) and E_c-.5 eV (E250) are formed around 300K but these dissociate above 600K. At lower annealing temperatures around 300K, they convert into H(150). However, H(150) converts into H(210) in p-Si and E(250) in n-Si above about 500K. The emission rate of E(90) increases with electrical field suggesting it is a donor. The H(150) acceptor is formed where the H concentration is maximum and a rough estimate is that this centre contains 2 to 3 times as much H as E(90) and E(250). The latter anneal out at slightly different temperature indicating that they are distinct defects -as are E(250) and H(210).

The results in Table 1 suggest that the deep acceptor, H(150), is likely to be PtH₃ or PtH₄. There are no associated donor levels. Only PtH has a deep donor and we assign this to H(210). This suggests that E(250) is the (-/0) level of PtH₂ as this defect is believed to be distinct from H(210). This leaves the (-/0) level of PtH to be detected. The calculations give no support to the idea that the donor level around E_c-.18 eV (E90) is a Pt-hydrogen defect. Nor could this level be the (-/-) level in this region detected in EPR [#!uftring-95!#]. However, its character and location suggests a (0/+) level associated with bond centred H trapped near an unknown neutral Pt related complex.

EPR experiments [#!uftring-95!#] demonstrate that the PtH₂ defect has a (-/0) level between 0.23 and 0.87 eV below E_c. We place it at E_c-.45 eV and identify it with E(250). However, disturbingly there are no DLTS levels reported for the (-/-) level of PtH₂.

More levels have been detected for Pd-H_n defects [#!sachse3-97!#]. A suggested correspondence between the Pt and Pd levels allows us to tentatively assign the H(280) and E(200) levels to PdH and PdH₂ respectively, and H(140) to the deep acceptor level of PdH₃. The presence of an additional DLTS level, namely E(160) at E_c-.29 eV, which is close to the calculated (-/0) level of PdH suggests that this is the acceptor level missing in PtH. Thus the two levels E(160) and H(280) should be correlated. Two hole traps H(45) and H(55) around 0.08 eV may be due to double donor levels of PdH where the H is separated by different distances from Pd. Similarly, if E(60) is a donor level analogous with the Pt related E(90) level, then it must be related to H$_{\rm BC}$ close to a neutral defect. Apparently Pd can be passivated in contrast with what has been reported for Pt. The calculations do not show that substitutional Pd or Pt can be passivated but we now describe a defect which is inactive.