Ultrapure, high mobility organic photoconductors

164 W. Warta et al.
The investigations presented here are part of a larger
program designed to obtain reliable experimental data
over wide temperature ranges on charge-cartier trans-
port in organic photoconductors. Naphthalene and
perylene were selected, among others, to serve as model
substances. Naphthalene is one of the most simple
aromatic molecules; very extensive investigations are
available of both its molecular and its crystal pro-
perties. Its melting point, however, (m.p. = 80.5 ~ is
too low for any potential practical application. In
contrast, perylene, which melts at 278 ~ is an example
of a fairly stable, larger aromatic molecule. Both
materials can be zone-refined and obtained as fairly
perfect single crystals by the Bridgman method.
We have chosen not to measure the electrical conduc-
tivity as such, since it is in the most general case
composed of at least four independent material pro-
perties: the electron and hole concentrations, n- and
n +, and the electron and hole mobilities, #- and #+,
respectively
= e(n-#- + n+#+). (1)
Instead, we decided to focus on the basic transport
quantities, #- and/~+. These mobilities can, in prin-
ciple, be measured separately by the time-of-flight
technique, cf. [2], where the transit of electrons
or holes over macroscopic sample dimensions (of
typically 0.2-1 ram) is observed.
However, most chemical impurities can very efficiently
capture charge carriers during passage through the
crystal, because the wide band gaps which are typical
for this material class provide a wide energy range in
which impurity states can act as traps. It is therefore
crucial to study charge carrier transport properties at
extremely low impurity concentrations. In addition,
the concentration of physical (lattice) defects has to be
kept low by careful annealing and handling of the
crystals, since at the present level of chemical purity,
physical defects can no longer be considered to give rise
to only minor contributions for the total trapping
behaviour.
Traps which are energetically deep with respect to kT
can completely prohibit the observation of carrier
transits. Shallow traps can repeatedly hold charge
carriers for short time intervals, until they are
thermally released to the conduction states again. A
strongly reduced, thermally activated "effective" mo-
bility is therefore commonly observed in such cases,
cf. [3]. The intrinsic transport properties, the "micro-
scopic" mobilities (or "lattice mobilities") are ob-
scured in such situations. Early mobility results have
often suffered from these different trapping influences.
(This is at least and definitely known to be so in
cases where more efficient purification later led to
higher, non-thermally activated mobility values.)
Historically, the measured microscopic mobilities were
often on the order of 1 cm2/Vs, cf. [4, 5] which is very
small in comparison with the mobilities in standard
inorganic semiconductors, such as silicon or ger-
manium. An increase of the mobilities # with decreas-
ing temperature T, # oc T", (n < 0), was found with some
of the purer samples: This increase was usually re-
stricted to narrow temperature intervals near room
temperature (and therefore not very big). The reason is
that even very shallow trapping influences, not seri-
ously disturbing room-temperature transport, can
limit the observability of transits at lower tempera-
tures, since the time constants for thermal detrapping
increase exponentially on cooling. The reproducibility
of mobifity results was sometimes rather poor between
different laboratories, which may also have been due to
(unknown) impurities which were present in variable
amounts in the different samples.
There were only three cases reported in the literature
where crystal quality permitted reaching lower tem-
peratures: The electron-mobility tensor component
#~;~, in anthracene was measured down to liq. N2
temperature [4, 6]; (surprisingly it exhibited the ab-
normality of remaining nearly temperature-
independent between 373 and 83 K, amounting to
0.4cmZ/Vs). In naphthalene the observation of a
transition from a nearly temperature-independent
electron-mobility tensor component #~;c. = 0.4 cm2/Vs
to one rising exponentially below 120 K was accom-
plished in a measurement down to 31 K, where #~c.
=2cmZ/Vs was reached [7]. Durene was the only
example for which fairly high low-temperature mo-
bilities have been reported (#+ = 55 cmZ/Vs at 120 K)
[83.
The aim of the work presented here was to improve
purification procedures and to grow high-quality
single crystals of naphthalene and perylene in order to
find out if and to which maximum value intrinsic
charge carrier mobilities continue to increase with
decreasing temperature, and, if high charge carrier
mobilities might eventually constitute a common low-
temperature property of aromatic crystals.
On our way to gradually increased purities we found,
among other things, that temperature-dependent
charge carrier mobility measurements represent a very
sensitive analytical tool for the assessment of purity
of organic aromatic crystals and we wish to focus on
this aspect in this publication.
Purification Procedures
We started with the purest commercially available
material (naphthalene "for scintillation", Merck, pery-
lene "purum", Fluka). Naphthalene was prepurified by
liquid chromatography. After vacuum sublimation,