As in GaN, p-type conductivity in InN has proven to be difficult to
achieve. Even though the valence band edge lies 1.6 eV below the Fermi
level stabilization energy [236], the low position of the
conduction band edge makes efficient p-type doping very
difficult. Another issue is the pinning of the surface Fermi level above
the conduction band edge, due to native donor defects, which leads to a
n-type accumulation layer at the surface [237,238]. Any
study of p-type bulk material has to isolate the effects of this
accumulation layer. This was achieved by Jones et al., who
provided the first indirect evidence of a net concentration of
acceptors, but who were however unable to verify the presence of free
holes [239]. Later works were not yet able to demonstrate net
p-type conductivity [240], but an activation energy for the
Mg acceptor of about 61 meV was extracted by photoluminescence
measurements [241]. Using the same value for the activation
energy Wang et al. [242] calculated a hole mobility in the
range of 1736 cm
/Vs for a hole concentration of about
(1.4
3.0)
10
cm
. However, they used a suggested
effective hole mass value (0.42m
) [243], which was not
experimentally confirmed. The same value was adopted by Fujiwara et
al. [244], who reported mobilities of
25
70 cm
/Vs. Recent works agree [245], that free holes
can be detected only for moderate Mg contents. Most of
the evidences of electrical conductance related to free electrons are
yet to be confirmed.