High carrier mobility in a series of new semiconducting PPV-type polymers

Article abstract of DOI:10.1021/ma025893p

The synthesis and characterization of a series of semiconducting polymers of the polyphenylenevinylene (PPV) type is presented. These new polymers have every second phenylene ring substituted by two alkyl groups of equal length which are hexyl, heptyl, oc

Full text of DOI:10.1021/ma025893p

4374
Macromolecules 2003, 36, 4374-4384
High Carrier Mobility in a Series of New Semiconducting PPV-Type
Polymers
F r ed er ik C. Kr ebs* a n d Mik k el J ør gen sen
The Danish Polymer Centre, RISØ National Laboratory, P.O. Box 49, DK-4000 Roskilde, Denmark
Received November 29, 2002; Revised Manuscript Received March 24, 2003
ABSTRACT: The synthesis and characterization of a series of semiconducting polymers of the polyphe-
nylenevinylene (PPV) type is presented. These new polymers have every second phenylene ring substituted
by two alkyl groups of equal length which are hexyl, heptyl, octyl, nonyl, decyl, undecyl, and dodecyl.
The polymers were synthesized from the corresponding dialkyldialdehydes by a Horner-Wadsworth-
Emmons condensation reaction with 1,4-bis(diethylphosphonomethyl)benzene in a high boiling solvent
mixture. The purified polymers were found to have charge carrier mobilities in the range 0.01-0.1 cm2
V-1 s-1 and carrier lifetimes of the order of 0.1-5 ms measured by the PR-TRMC technique. A possible
explanation for the unexpectedly large magnitudes of both the charge carrier mobility and lifetimes was
arrived at by powder X-ray diffraction studies which provided a putative structural model. Finally the
electronic structure was studied by ultraviolet photoelectron spectroscopy and the photodegradative
properties of thin films of the polymers characterized by illumination through a suitable photopositive
mask allowing for subsequent doping giving highly anisotropic conducting patterns.
Sch em e 1. Alk ) Lin ea r Alk yl Ch a in s C6 to C12
In tr od u ction
High carrier mobility is an important goal for new
organic polymer electronic materials. Intense research
effort has allowed for the successful demonstration of
semiconductor applications such as field effect tran-
sistors1-3 (FET’s), light emitting diodes4,5 (LED’s), pho-
tovoltaic devices,6-9 etc. based on the semiconducting
polymer materials currently available. Most of them
exhibit very low carrier mobilities and carrier lifetimes
when compared to conventional silicon and gallium
arsenide counterparts.10 While it is not expected that
semiconducting polymer materials will attain the per-
formance of the inorganic counterparts in terms of
carrier mobilities, lifetimes and purity it is of large im-
portance to achieve the highest possible values for these
key properties in the semiconducting polymer materials
of interest. The fundamental understanding that allows
one to achieve high carrier mobility is thus a highly
attractive goal. The three most important requirements
can be identified as the structure, an extended π-system,
and readily accessible carrier states. The latter two
factors are intimately linked and can to a large extent
be solved by synthetic organic chemistry. The organiza-
tion and structure of polymer materials, however, are
currently very difficult to control, and a large effort
within this area is needed. It is the disordered molecular
structure in semiconducting polymer materials that is
believed to increase the barrier to interchain hopping,
which is considered to be the rate-limiting step to carrier
transport in molecular and polymer materials. In the
present work we have chosen to measure the carrier
mobilities by the pulse-radiolysis time-resolved micro-
wave conductivity (PR-TRMC) technique.11 In this
method the semiconducting polymer sample is placed
in a microwave reflection cavity and subjected to a short
(10-100 ns) electron pulse (from a 10 MeV linear
electron accelerator) that create electron/hole pairs. The
conductivity change is the measured as the transient
absorption of the microwaves using a GHz sampling
rate. The PR-TRMC method is contactless, thus avoid-
ing injection barrier problems between electrodes and
sample inherent in other methods for measuring carrier
mobilities, such as TOF measurements. The PR-TRMC
technique measures the maximum trap-free mobilities
of carriers in domains of the sample. Bulk mobilities
measured by other techniques can be considerably
smaller due to domain boundaries and injection barri-
ers.12 Semiconducting polymers such as the substituted
poly(p-phenylenevinylene)s (PPV’s) and PPV itself have
been known for a long time and have proved valuable
as materials for the development of polymer electronic
devices.13 It is surprising that most materials studied
aside from the parent compound have had alkoxy groups
as substituents.14,15 Few attempts have been made to
study the relationship between the molecular substitu-
tion pattern and the resulting influence this has on the
structure and on the mobility of charge carriers.
In this paper we present the synthesis and charac-
terization of a series of poly(dialkylstilbenevinylene)s
abbreviated dialkyl-PSV’s as shown in Scheme 1.
We then explored the material properties of these
semiconducting polymers with respect to the charge
carrier mobility, charge carrier lifetime, and the elec-
tronic energy levels and present a putative structural
model based on X-ray powder diffraction data.
Resu lts a n d Discu ssion
Syn th esis. The synthesis of the monomers was
performed as shown in Scheme 2. All of the compounds
1a -g were known16 and most of the compounds 2a -g
were known16 except 2d -f which were prepared by a
Corriu-Kumada coupling of the corresponding alkyl-
magnesium bromide on 1,4-dichlorobenzene using a
NiCl2-dppp catalyst followed by bromination according
to the literature.16 While compound 7c has been re-
ported in the literature17 no reaction details were given.
The selective metal-halogen exchange on compounds
* Corresponding author. E-mail: frederik.krebs@risoe.dk.
10.1021/ma025893p CCC: $25.00 © 2003 American Chemical Society
Published on Web 05/14/2003

Macromolecules, Vol. 36, No. 12, 2003
High Carrier Mobility 4375
Sch em e 2
Sch em e 3
2a -g where just one bromine is replaced was not
possible without the formation of some impurities that
were not easily removed. On the other hand, dilithiation
was impossible if a pure product was desired. We found
that the reaction sequence where compounds 3a -g, 4a -
g, and 5a -g were used directly in subsequent steps
without purification gave 6a -g which was easily puri-
fied. This could then be deprotected to give 7a -g which
was of a high purity suitable for condensation polym-
erization reactions.
P olym er iza tion . The polymerization reactions were
performed using sodium hydride as base in a high
boiling solvent mixture as shown in Scheme 3. This was
found to give the highest molecular weights.
The raw polymer products were then subjected to
Soxhlet extraction with THF for 24 h to remove low-
molecular weight material and finally recrystallized/
reprecipitated from boiling 1,2-dichlorobenzene. The
polymers 8c and 8d were of the highest molecular
weight, but common to all the polymers obtained was a
very broad distribution of molecular weights. In our
hands it was not possible to control the dispersity of the
polymer product using this mode of polymerization. We
initially thought that the co-crystallized solvent water
molecules in the compounds 7a -g was responsible for
the variation in dispersities observed but removal of this
water either by azeotropic distillation before adding the
sodium hydride to the polymerization mixture or by
dissolving in chloroform and drying the solution gave
the same results. We ascribe the variation in molecular
weight distributions observed to be governed by small
errors in the mixing of stoichiometric amounts of the
monomer components and the solubility of the final
polymer product under the reaction conditions.
lifetimes as the method is contactless and thus inde-
pendent of experimental parameters such as sample
preparation (i.e., thin films), sample orientation/orga-
nization, domain boundaries, and contact problems such
as electrode contact resistance. We have applied this
method on a general basis for the screening of newly
synthesized materials and selecting materials with high
mobilities and long lifetimes. The desirable charge
carrier properties of the particular materials presented
here were thus discovered by the standard use of this
technique. The technique relies on the measurement of
a change in reflected microwave power from a polymer
sample upon passage of high-energy electrons. From the
transient changes in conductivity the lifetime of the
carriers can be determined as well as the sum of the
mobility of the carriers. The setup used in this work has
the sample in a reflection cell measuring the reflected
microwave power Pr and the difference in reflected
microwave power ∆P (before and after irradiation with
an electron pulse). Warman et al. found an expression
that relates the maximum of the ratio ∆P/Pr to the
change in conductance (∆σ) of the sample due to the
radiation-induced carriers according to eq 1 under the
assumption that the medium is isotropic.
∆P
Pr
)
(
)
max
2λvacfmax
ꢀoc2
d
2
2 1/2
c
2n + 1
2d
1
a
-∆σ
,
fmax
)
+
)
(1)
(2)
(
[
)
( )
]
2 ꢀ
x
r
∆σ
D
Ep
µ
)
,
N+ + N-
∑
min
e(N+ + N-)
Ch a r ge Ca r r ier Mobilities. The minimum sum of
carrier mobilities was determined using the pulse
radiolysis time-resolved microwave conductivity tech-
nique (PR-TRMC) developed by Warman et al.18,19 The
technique is very useful to the polymer scientist looking
for new materials with high mobilities and long carrier
∆σ is thus related to the length of the sample, d, and
the frequency at which the maximum response is
observed, fmax (i.e., the largest ratio of ∆P/Pr). From this
the sum of carrier mobilities can be obtained since the
conductance is the product between the former and the
number of carriers (N+ + N-), which is known from

4376 Krebs and J ørgensen
Macromolecules, Vol. 36, No. 12, 2003
Ta ble 1. Su m of Ca r r ier Mobilities Mea su r ed by
P R-TRMCa
compound
Σµmin (m2 V-1 s-1
)
τ1 (ms)
τ2 (ms)
Mp
8a
8b
8c
8d
8e
8f
1.28 × 10-6
5.78 × 10-6
9.14 × 10-6
8.90 × 10-6
6.12 × 10-6
1.87 × 10-6
4.18 × 10-6
0.1517
0.3994
0.3926
0.5284
0.2586
0.1873
0.2209
2.628
4.843
4.295
6.017
3.245
2.623
2.576
13 500
3280
26 400
129 000
29 100
9470
8g
49 500
a
The Σµmin values are based on a carrier pair formation energy
of 25 eV. The first and second half-lives τ1 and τ2 for the materials
studied are also shown. The Mp values obtained from the SEC
measurements are also shown.
dosimetry according to eq 2. (see Experimental Section).
It is reasonable to assume that the isotropic condition
is valid since the powdered and randomly oriented
sample completely fills the waveguide and the high
energy electrons interact weakly with the material
(large penetration depth). The number of carriers is
derived from D/Ep where D is the measured dose and
Ep is the electron/hole pair formation energy. A value
of 25 eV for Ep has been used extensively in previous
PR-TRMC studies.20-25 Alternatively the value of Ep
can be estimated from a weighted sum of the electronic
band gap for the constituents (i.e., backbone and side
chains) of the molecules based on a model by Alig et
al.26 and used in the interpretation of PR-TRMC
results.27-29 For our results presented in Table 1 an Ep
value of 25 eV has been used. When deriving an Ep value
based on the Alig-model for our systems a lower value
for Ep is obtained (i.e., Ep ∼ 15 eV for 8c) thus
corresponding to a reduction of the mobilities to 60% of
the values given in Table 1. The major advantage of the
technique is that the magnitude of the carrier mobility
obtained is the minimum average carrier mobility
within a domain and it is independent of sample
preparation, i.e., no contact resistance.
F igu r e 1. PR-TRMC dose normalized transient (below)
showing the end-of-pulse conductivity value, σEOP, that was
used to extract the minimum sum of carrier mobilities, µmin
,
in Table 1 for polymer 8d . Above, the 100 ns electron pulse
that creates the carriers is shown (recorded with a Faraday
cup after passage through the sample).
This means that even higher mobilities within indi-
vidual domains along certain directions is possible as
documented for columnar discotic liquid crystals where
the mobility along the columnar stack is significantly
larger than the minimum sum of mobilities. A disad-
vantage is that it is the sum of carrier mobilities that
is obtained and the technique does thus not allow for
distinguishing between hole and electron mobilities
except for one specific example.27 The general results
for the polymers 8a -g are presented in Table 1 for
thermally annealed samples.
It is noteworthy that the observed magnitudes of the
carrier mobilities observed are much larger, by 1-3
orders of magnitude, than for other PPV derivatives20,28
and larger than for polythiophenes by 1 order of
magnitude.21,29 Values similar to ours have been re-
ported recently however.11 While the values in Table 1
are very high in general and particularly high for
polymeric materials, the mobility in the direction of the
stacks could be higher by up to a factor of 3 as shown
for discotic liquid crystalline materials where the carrier
transport is known to be 1-dimensional. The lifetime of
the carriers was observed to be in the millisecond range
that is much longer when compared to lifetimes ob-
served for alkoxysubstituted PPV derivatives.20
F igu r e 2. Very long conductivity transient for the polymer
8d recorded using a slow sampling rate of 2.5 Ms s-1 to monitor
the conductivity decay in the millisecond range. The biexpo-
nential fits according to the carrier lifetime values given in
Table 1 are shown. The difference between the data and the
fitting function is shown at the top as the percentage deviation.
had little influence on the magnitude of the carrier
mobilities. However, the lifetimes of the carriers was
increased by a factor of 103. These very large values of
carrier mobilities and the long lifetimes led us to
attempt studying the structure and organization of the
recrystallized/reprecipitated materials by powder X-ray
diffraction.
Str u ctu r a l Mod el. It seemed puzzling that these
materials should exhibit such high values for the
mobility when compared to the known PPV derivatives
which almost exclusively has alkoxy substituents and
The decays were all nonmonoexponetial and a fit was
obtained by using two half-lives as shown in Figure 2.
These data are also presented in Table 1. The purifica-
tion step whereby the polymer materials were recrystal-
lized/reprecipitated from boiling 1,2-dichlorobenzene

Macromolecules, Vol. 36, No. 12, 2003
High Carrier Mobility 4377
F igu r e 4. Stereoview of the diacetal 6c showing how the alkyl
groups form a torsion angle with the benzene ring close to 90
°.
F igu r e 3. Number of occurrences in the CSD of structures
in a given 10° torsion angle span for, respectively, alkyl- and
alkoxy-aryl compounds with search fragments shown as
insets. There were 365 alkyl occurrences and 457 alkoxy
occurrences.
a substitution pattern that has every phenylene unit
substituted. We started by considering the typically
observed behavior of the substituents in the solid state
by use of the Cambridge Structural Database30 (CSD)
which contains a wealth of information on the structure
of molecular organic materials. By searching the data-
base for the geometries of selected molecular fragments,
we succeeded in identifying striking differences in the
conformational behavior of substituents generally em-
ployed in semiconducting polymers.
Most significantly the torsion angle span observed
between an aromatic system and a substituent revealed
that alkyl and alkoxy groups are orthogonal as shown
in Figure 3.
F igu r e 5. Observed (blue), calculated (red), and difference
(green) plot for 8c. The powder data were acquired on samples
annealed at 150 °C for 24 h under argon using Cu KR
radiation.
This essentially implies that the methylene group
directly attached to a benzene ring prefers a torsion
angle of about 90° while it is about 0° for similar alkoxy
substituents as shown in Figure 3. The total number of
occurrences in the CSD for respectively alkyl- and
alkoxy-substituted aryl compounds of the type indicated
by the search fragments were similar. It is evident that
the torsion angles for alkyl groups are more relaxed and
a larger conformational range is available. These prefer-
ences are undoubtedly responsible for the differences
in packing properties of dialkyl- vs dialkoxy-substituted
polymers. We used this information to obtain a reason-
able starting model for our powder diffraction studies
and also solved the structure of compound 6c to further
substantiate these findings. The molecular structure for
6c is shown in Figure 4 where a torsion angle between
the benzene ring and the innermost methylene group
is close to 90° as expected.
Recent studies of model compounds for alkyl- and
alkoxy-substituted polyparaphenylenevinylenes pre-
sented single-crystal X-ray data and molecular struc-
tures observed are consistent with our results.31,32
X-ray powder diffraction on the polymers 8a -g (see
also Supporting Information) performed in this study
showed that the alkyl substituted PPV’s had a reason-
ably degree of crystallinity and many Bragg peaks were
observed. Considering the very rigid nature of the
polymer this led us to try and model the crystal
structure from powder data on annealed samples of the
material by rigid body refinement and finding support
for our initial structural model through the CSD-search
shown in Figure 3 but also a search on trans-stilbenes.
Several initial models were attempted, and all con-
verged on a solution where the experimental powder
diffractogram and the calculated powder diffractogram
were in good agreement as shown in Figure 5. Our
putative model obtained in this way show that the
polymer PPV backbone of alternating benzene rings and
vinylene groups run along the space diagonal of the
triclinic unit cell. This arrangement maximizes the
overlap between the polymer chains, which in turn
facilitates high carrier mobility. The main finding is that
the stacking of the stilbene units is eclipsed giving rise
to near maximum π-overlap as shown in Figure 6. The
alkyl chains gives rise to an interdigitating pattern that
cause the structure to become layered. The extended
sheets of the PPV backbone that are isolated by the
alkyl chains probably hold the explanation to the long
carrier lifetime that is only surpassed by a few columnar
discotic liquid crystal materials where such large carrier
lifetimes and mobilities have been observed.25 For our
materials there is no significant correlation between the
magnitude of the observed lifetimes and the alkyl chain
length. We ascribe this to a more 2-dimensional trans-
port mechanism giving conduction paths both along the
polymer chains, but also between the chains due to the
good overlap. In the case of poly(3-alkylthiophene)s a
correlation between alkyl chain length and carrier
lifetime has been observed.29 One aspect of the values
in Table 1 is the lack of correlation between the alkyl
chain length and the carrier mobility and lifetime. For

4378 Krebs and J ørgensen
Macromolecules, Vol. 36, No. 12, 2003
F igu r e 6. Stereoview of the putative structural model ob-
tained from the powder X-ray diffraction data for 8c. The
polymer strands are stacked in sheets with the alkyl chains
surrounding them. Every second vinylene group forms torsion
angles significantly different from 0° forcing the intermolecular
π-π interactions to be eclipsed. One alkyl chain extends above
the plane of the polymer strand and the other alkyl chain
extends below the plane of the polymer strand.
the discotic liquid crystal materials studied using the
PR-TRMC technique there is a strong correlation
between alkyl chain length and the carrier lifetimes.
This observation has been rationalized through the
insulating nature of the alkyl chain core for these highly
1-dimensional systems. Aside from the possibility of a
more 2-dimensional nature of the conduction paths in
our materials as mentioned above, there are at least
four important issues. First, the molecular weights vary
considerably between the samples studied. Second, the
most crystalline samples (see Supporting Information)
generally have the highest values for both carrier
mobility and lifetime (i.e., 8a and 8f are the most weakly
scattering samples and they both have the shortest
lifetimes and smallest carrier mobilities). Third, there
seems to be a maximum around 8c and 8d where the
number of side chain atoms approximate the number
of backbone atoms and this might be important for the
efficient packing with respect to carrier transport.
Fourth, we only present the minimum sum of carrier
mobilities and have not for instance corrected for the
dimensionality, the differential dose distribution (be-
tween side chain and backbone), and the carrier pair
survival probability, which has been assumed to equal
1 (i.e., 100% carrier pair survival probability). With
respect to this last point, an attempted correction using
any of these terms would only increase our observed
value for the sum of carrier mobilities. We have thus
provided a conservative set of values for the carrier
mobilities for the compounds 8a -g.
The model obtained this way should be considered
with some caution. In particular, the difference between
calculated density (0.833 g cm-3) and the density found
experimentally (0.95 g cm-3) for polymer 8c is quite
large. The magnitude of the density is however not
unreasonably low for a polymer material. We cannot
provide a solid explanation for this discrepancy but
tentatively ascribe it to the few degrees of freedom in
our simple model that however does fit the experimental
diffraction data quite nicely. While our putative model
explains the observed materials properties (high carrier
mobility and long carrier lifetime) the information
contained in the powder diffractogram is limited. The
crystallinity observed is high for a polymer material but
is very poor when compared to molecular crystals. This
puts severe constraints on our ability to draw specific
conclusions regarding the detailed geometry of the
system. In this manner, the packing arrangement
F igu r e 7. Multiple powder diffractograms collected for com-
pound 8a upon heating from 30 (top) to 250 °C (second from
the bottom). Upon cooling back down to 65 °C, the fine
structure of the peaks have disappeared.
illustrated in Figure 6 is very likely to be a good
representation. The detailed packing of the alkyl chains,
however, is most likely very disordered, and the
illustration in Figure 6 should be considered as an
effective geometry obtained from the Rietveld refine-
ment.
Th er m a l Beh a vior . We chose to obtain further
information on the behavior of the alkyl side chains by
thermal studies. The DSC data generally showed a
transition around 110-130 °C, which is ascribed to side
chain melting.33 The transition was not easily observed
for all the compounds and was found to be irreversible
if heated above the melting point of the polymers which
were generally observed at temperatures in the interval
200-300 °C. In the case of 8a the side chain melting in
the temperature interval 70-150 °C was very difficult
to quantitate from the DSC trace. A powder diffraction
study using synchrotron radiation however more clearly
showed the changes observed in the 70-150 °C interval
and the irreversibility when heating above melting (see
Figure 7). The use of X-ray powder diffraction for
thermal studies turned out to be an alternative method
for the establishment of phase changes and confirmed
the irreversible melting and subsequent crystallization
also observed in the DSC measurements. The fact that
the side chain melting or mesophase transition is not
reversible (on the time scale of the experiment) if the
samples are melted indicates that the desirable high
mobility/long lifetime phases are obtained only by
recrystallization/reprecipitation of the polymers.
This information is very valuable if devices or studies
where one makes use of the desirable materials proper-
ties are to be made. We then performed an annealing
study also using synchrotron powder X-ray diffraction
for 8c as shown in Figure 8. Here the annealing is
shown to alter the diffraction pattern irreversibly (on
the time scale of the experiment) while maintaining a
high degree of crystallinity when heating above the first
transition (to 150 °C) and cooling back down. Upon

Macromolecules, Vol. 36, No. 12, 2003
High Carrier Mobility 4379
Ta ble 2. Con d u ctivities Obta in ed for th e P r istin e a n d
P h otoblea ch ed F ilm s Neglectin g P ossible Differ en ces in
th e Mech a n ism s Con d u ction (i.e. in jection of ca r r ier s in
th e u n d op ed film s)
undoped (Ω1- m-1
)
doped (Ω1- m-1
)
pristine
photobleached
4.0 × 10-6
1
1.6 × 10-7
5.1 × 10-3
Ta ble 3. Sin gle Cr ysta l a n d P ow d er X-r a y
Cr ysta llogr a p h ic Da ta for 6c a n d 8c Resp ectively
study
formula
6c
8c
C
34H58O4
C
32H44
formula wt
cryst syst
530.80
triclinic
P1h
428.69
triclinic
P1
space group
Z
2
1
a, Å
5.714(2)
16.601(7)
17.899(7)
108.577(7)
98.008(8)
97.567(8)
1565.3(11)
1.126
5.1900
11.0153
16.8393
63.7905
96.7334
97.6577
854.16287
0.833
b, Å
c, Å
R, deg
â, deg
γ, deg
V, Å3
F, g cm3 (calcd)
F, g cm3 (obsd)
cryst dimens, mm
type of radiation
wavelength, Å
µ, cm-1
0.95
0.25 × 0.25 × 0.10 powder
Mo KR
Cu KR
1.5418
0.710 73
0.071
F igu r e 8. Thermal annealing study for compound 8c using
synchrotron X-ray diffraction. The pristine purified sample
(trace shown above) was heated to 150 °C (second trace from
the top) and cooled back to 40 °C (middle trace) where some
of the peaks have become more pronounced. Finally, heating
to 200 and 220 °C (lower traces) led to a merging of the peaks
in the 10-20° in 2θ interval into a broad feature.
T, K
120(2)
293(1)
0.0512, 0.0665
0.22
1121
780
R-P, R-WP (model)
R-WP (w/over background)
no. of profile points
no. of reflcns
no. of reflcns
no. of unique reflcns
(with I > 2σ)
Rint
-
-
-
-
19 922
2839
0.0949
0.0689, 0.2092
R(F), Rw(F2) all data
Ta ble 4. Resu lts fr om th e P h otoelectr on Sp ectr oscop ic
Da ta on Th in F ilm s of th e P olym er s 8a -g on
P olycr ysta llin e Silver
VB
VAC
EF
EF
cutoff
∆
IP
8a
8b
8c
8d
8e
8f
0.90
1.05
0.70
0.80
0.90
1.10
0.90
3.80
45.30
45.50
45.35
45.35
45.35
45.40
45.30
-0.40
-0.75
-0.25
-0.35
-0.45
-0.70
-0.40
4.70
4.50
4.65
4.65
4.65
4.60
4.70
3.45
3.95
3.85
3.75
3.50
3.80
8g
an electrode material, the optical band gap, and the
ionization potential are important materials properties.
The ionization potential, IP, the distance from the
Fermi level of the electrode material to the position of
the valence band, EFVB, and distance from the Fermi
level of the electrode material to the vacuum level,
EFVAC, were determined using ultraviolet photoelectron
spectroscopy (UPS) on thin films of the polymers spin-
coated on a silver substrate by procedures described in
the literature.34-36 The ionization potential of the
F igu r e 9. Ultraviolet photoelectron spectrum of the polymer
8c spin coated on a polycrystalline silver surface using a
photon energy of 50 eV. The large signal toward the cutoff at
around -46 eV is due to secondary electrons. The reference
level is the vacuum level with respect to silver.
heating further (to above 200 °C), the diffraction pattern
merged to broad features and the distinct pattern was
no longer observed.
Thermogravimetric analysis (TGA) showed that all
the polymers decomposed with onset temperatures at
around 325 °C and derivative peak decomposition
temperatures at around 480 °C. The polymers can thus
be considered as quite stable thermally.
polymers ranged from 4.5 to 4.7 eV (see Table 4). The
VB
position of the valence band edge for the polymers, EF
,
was lower than the Fermi level of silver by about 0.7-
1.1 eV. In Figure 9, a photoelectron spectrum of the
dioctyl derivative is shown with reference to the Fermi
VB
level of silver. The values obtained for EF represents
the barrier to hole injection into the valence band and
is a reasonable value for electron rich materials such
as the materials studied here.
P h otod egr a d a tion a n d th e F or m a tion of P a t-
ter n s. We found that irradiation with UV light in air
led to the rapid disappearance of the characteristic
yellow color of the polymer films. We ascribe this to the
Electr on ic Str u ctu r e. While high carrier mobilities
and lifetimes are suitable for constructing devices the
actual application of the material may depend on many
other factors relating to the physical construction of the
device. Aside from the preparation of films or samples
with materials properties identical to the bulk sample
the position of the filled energy levels in the context of

4380 Krebs and J ørgensen
Macromolecules, Vol. 36, No. 12, 2003
F igu r e 10. Photoresist properties of dioctyl-substituted polymer. A cast film was irradiated with a 150 W xenon lamp (100 mW
cm-2) for 5 min through a positive mask. The result is the image to the left where the exposed part of the film has been
photobleached. When the film was then exposed to ICl vapor, the doping gave rise to a dark coloration of the unexposed part of
the films.
well-known oxygen dependent photodegradation (a 2 +
2 cycloaddition between the polymer vinylene groups
and oxygen).37 When performing the same experiment
in a vacuum of P < 10-7 mbar the yellow color persisted
with undetectable change in absorbance of the film upon
exposure for 24 h. We applied this property as an in-
built photoresist.
used to offer a possible explanation for the high values
of carrier mobilities and lifetimes the variation in
crystallinity along with the variation in the molecular
weights was also used to explain the variation in the
observed values. The large values for the carrier mobil-
ity and lifetime for the materials presented here was
established using the PR-TRMC technique that does
not provide insight into the individual contributions
toward the carrier mobility and lifetime from ordered
and disordered domains, both present in the sample.
Our results does however provide a firm basis for
further study in terms of applications. We have further
characterized the thermal behavior, the electronic struc-
ture and the photodegradative properties of the materi-
als and applied the ladder properties to make patterns
with high conduction anisotropy.
Thin films of the yellow PPV polymers were spin-
coated onto a Pyrex glass slide. A 150 W xenon lamp
was used to illuminate the film through a positive mask
for 5 min. The exposed areas became colorless, leaving
a yellow image of the mask as shown in Figure 10.
Doping was then affected with vapors of iodine or more
effectively with iodine monochloride. Only the nonirra-
diated parts changed color to the characteristic blue-
black of the charge-transfer complex also shown in
Figure 10. The conductivity increase upon doping was
very high for the unexposed part of the film. dc-
conductivity measurements with and without doping
was performed for both pristine and photobleached
films. The films were cast from 1,2-dichlorobenzene onto
an etched pattern on 10 Ω square-1 ITO slides. In Table
2, the results from the experiment are shown.
Exp er im en ta l Section
Syn th esis. All reagents were commercially available. Some
signals in 13C NMR spectra were missing. This is ascribed to
accidental isochrony and was observed for the compounds 2e,
2f, 6g, and 7d . Compounds 2a -c and 2g were prepared as
described in ref 16.
Gen er a l P r oced u r e for th e P r ep a r a tion of Com p ou n d s
2d )f. The dialkylbenzenes 1d -f were prepared by a Corriu-
Kumada coupling as described (also ref 16). Compounds 1d -f
(0.2 mol) were each dissolved in dichloromethane (100 mL) and
bromine (0.4 mol) was added. The resulting mixtures were kept
in the dark to avoid radical bromination of the alkyl side
chains. After stirring for 48 h at room temperature, the
products were recovered and recrystallized from ethanol.
In the undoped films, the dc conductivity should be
treated with caution as the material in this state is
semiconducting with very few intrinsic carriers. Upon
doping, the behavior becomes ohmic. The conductivity
anisotropy is very high between doped and undoped
material and is close to a factor of 106. The doping
anisotropy between bleached and unbleached material
is of the order of 103. This property could be particularly
useful for the simple design of electronic circuitry where
a homogeneous film can be patterned after application.
It should however be kept in mind that there are many
uncertainties associated with these observations as the
degree and homogeneity of the doping is unknown. The
results show the relative differences between doped/
undoped and pristine/photobleached films, whereas the
absolute measures are very dependent on the experi-
mental procedure.
1
2d : Colorless needles in 65% yield; mp 44-45 °C. H NMR
(250 MHz, CDCl3, 300 K, TMS): δ ) 0.9 (t, 6H, CH3), 1.1-1.4
(m, 24H, CH2), 1.5-1.6 (m, 4H, CH2), 2.7 (t, 4H, CH2), 7.4 (s,
2H, ArH). 13C NMR (63 MHz, CDCl3, 300 K, TMS): δ ) 14.8,
23.4, 30.03, 30.05, 30.1, 30.2, 30.5, 32.6, 36.2, 123.8, 134.4,
142.0. Anal. Calcd for C24H40Br2‚C2H5OH: C, 58.43; H, 8.68.
Found: C, 58.46; H, 8.35.
1
2e: Colorless microcrystals in 60% yield; mp 55-56 °C. H
NMR (250 MHz, CDCl3, 300 K, TMS): δ ) 0.9 (t, 6H, CH3),
1.1-1.4 (m, 28H, CH2), 1.5-1.6 (m, 4H, CH2), 2.7 (t, 4H, CH2),
7.4 (s, 2H, ArH). 13C NMR (63 MHz, CDCl3, 300 K, TMS): δ
) 14.8, 23.4, 30.0, 30.1, 30.2, 30.3, 30.5, 32.6, 36.2, 123.8, 134.4,
142.0. Anal. Calcd for C26H44Br2: C, 60.47; H, 8.59. Found:
C, 60.53; H, 8.77.
Con clu sion s
In this work, we have presented the synthesis and
characterization of a new series of semiconducting
polymers based on polyphenylenevinylene which exhibit
high carrier mobilities and long lifetimes. This is mainly
ascribed to the high degree of crystallinity for these
materials and the structural model obtained from
powder diffraction experiments. While crystallinity was
1
2f: Colorless microcrystals in 70% yield; mp 58-59 °C. H
NMR (250 MHz, CDCl3, 300 K, TMS): δ ) 0.9 (t, 6H, CH3),
1.1-1.4 (m, 32H, CH2), 1.5-1.6 (m, 4H, CH2), 2.7 (t, 4H, CH2),
7.4 (s, 2H, ArH). 13C NMR (63 MHz, CDCl3, 300 K, TMS): δ
) 14.8, 23.4, 30.0, 30.1, 30.2, 30.3, 30.5, 32.6, 36.2, 123.8, 134.4,
142.0. Anal. Calcd for C28H48Br2: C, 61.76; H, 8.89. Found:
C, 61.31; H, 8.95.

Macromolecules, Vol. 36, No. 12, 2003
High Carrier Mobility 4381
Gen er a l P r oced u r e for th e P r ep a r a tion of 6a )g. The
dibromdialkyl compounds 2a -g (0.165 mol) were each dis-
solved in dry THF (500 mL) and cooled. At -10 °C, the
compound crystallizes into a thick mass. The use of a large
magnetic bar made stirring of the mixture become possible at
lower temperatures. At -60 °C, nBuLi in hexanes (110 mL,
1.6 M) was added and the mixture stirred for 15 min. The
mixture did not become entirely clear even though all solids
dissolved. Dry DMF (25 mL) was added and the mixture
allowed to reach room temperature. After 1 h, HCl(aq) (37%,
100 mL) was added and the mixture evaporated until the THF
had been removed. The aqueous phase was extracted with
ether (3 × 300 mL) and the light yellow ether phase washed
with water (500 mL). Drying over MgSO4 and evaporation gave
an oil. A typical yield of 94% was obtained (on occasion the
crude product could be crystallized from 90% ethanol (600
mL)). The crude product was refluxed in benzene (500 mL)
containing neopentylglycol (16 g, 0.28 mol, excess) and p-
toluenesulfonic acid (100 mg) with a water separator. After 5
h, the mixture was cooled and washed with NaHCO3(aq)
(1 M, 300 mL) and water (300 mL). Drying with MgSO4 and
evaporation gave an oil in a typical yield of 83%. The product
from above was dissolved in dry THF (500 mL) and cooled to
-70 °C, and nBuLi in hexanes (1.6 M, 1.1 equiv) was added.
Stirring for 10 min and addition of dry DMF (2 equiv) was
performed. The mixture was allowed to reach room tempera-
ture. HCl(aq) (37%, 100 mL) was added and the mixture
stirred for 2 min. The phases were separated and the organic
phase washed with water (200 mL). The organic phase was
evaporated without drying to give a yellow oil in a typical yield
of 76%. The crude product was refluxed in benzene (500 mL)
containing p-toluenesulfonic acid and neopentylglycol (2 equiv)
with a water separator. After 2 h, the reaction was stopped
and the organic phase washed with NaHCO3(aq) (1 M, 300
mL) and water (300 mL) and finally dried and evaporated to
give an oil that crystallizes. The oil was dissolved in absolute
ethanol (300 mL) and crystallized.
(d, 4H, CH2), 5.5 (s, 2H, CH), 7.4 (s, 2H, ArH). 13C NMR (63
MHz, CDCl3, 300 K, TMS): δ ) 14.6, 22.3, 23.1, 23.7, 29.8,
29.9, 30.0, 30.1, 30.3, 30.6, 32.1, 32.4, 32.6, 78.3, 100.4, 127.6,
136.6, 138.4. Anal. Calcd for C38H66O4: C, 77.76; H, 11.33.
Found: C, 77.57; H, 11.37.
6f: Colorless microcrystals in 63% yield based on 2f; mp 93-
94 °C. 1H NMR (250 MHz, CDCl3, 300 K, TMS): δ ) 0.8 (s,
6H, CH3), 0.9 (t, 6H, CH3), 1.2-1.4 (m, 38H, CH2, CH3), 1.5-
1.6 (m, 4H, CH2), 2.7 (t, 4H, CH2), 3.6 (d, 4H, CH2), 3.8 (d, 4H,
CH2), 5.5 (s, 2H, CH), 7.4 (s, 2H, ArH). 13C NMR (63 MHz,
CDCl3, 300 K, TMS): δ ) 14.8, 22.6, 23.4, 23.9, 30.0, 30.2,
30.3, 30.36, 30.4, 30.5, 30.9, 32.4, 32.6, 32.9, 78.6, 100.6, 127.8,
136.8, 138.7. Anal. Calcd for C40H70O4: C, 78.12; H, 11.47.
Found: C, 78.08; H, 11.57.
6g: Colorless thin needles in 67% yield based on 2g; mp 85-
86 °C. 1H NMR (250 MHz, CDCl3, 300 K, TMS): δ ) 0.8 (s,
6H, CH3), 0.9 (t, 6H, CH3), 1.2-1.4 (m, 42H, CH2, CH3), 1.5-
1.6 (m, 4H, CH2), 2.7 (t, 4H, CH2), 3.6 (d, 4H, CH2), 3.8 (d, 4H,
CH2), 5.5 (s, 2H, CH), 7.4 (s, 2H, ArH). 13C NMR (63 MHz,
CDCl3, 300 K, TMS): δ ) 14.8, 22.6, 23.4, 23.9, 30.1, 30.2,
30.3, 30.35, 30.4, 30.5, 30.9, 32.4, 32.6, 32.9, 78.6, 100.6, 127.8,
136.8, 138.7. Anal. Calcd for C42H74O4: C, 78.45; H, 11.60.
Found: C, 78.41; H, 11.80.
Gen er a l P r oced u r e for th e P r ep a r a tion of 7a )g.
Compound 6a -g (20 mmol) was dissolved in trifluoroacetic
acid (200 mL), and water (15 mL) was added whereby the clear
solution becomes turbid. The mixture was stirred with heating
to the boiling point and then left to cool on the stirrer for 1 h.
The mixture was then cooled with an ice/acetone bath. Water
(100 mL) was then added with vigorous stirring. The product
crystallized. After filtering, the crude and wet product was
recrystallized from acetonitrile. It was found to be important
to use the crude and wet product for the recrystallization. If
recrystallized material was subjected to further crystallization
or the crude product was dried before recrystallization, this
led to degradation of the product. The traces of water/acid
present in the crude product were thus found to be imperative
for a high purity product. Further all the elemental analysis
of the products 7a -g indicated a small water content.
6a : Colorless needles in 56% yield based on 2a ; mp 98-99
1
°C. H NMR (250 MHz, CDCl3, 300 K, TMS): δ ) 0.8 (s, 6H,
1
CH3), 0.9 (t, 6H, CH3), 1.2-1.4 (m, 18H, CH2, CH3), 1.5-1.6
(m, 4H, CH2), 2.7 (t, 4H, CH2), 3.6 (d, 4H, CH2), 3.8 (d, 4H,
CH2), 5.5 (s, 2H, CH), 7.4 (s, 2H, ArH). 13C NMR (63 MHz,
CDCl3, 300 K, TMS): δ ) 14.8, 22.6, 23.3, 23.9, 30.2, 30.9,
32.3, 32.4, 32.9, 78.6, 100.6, 127.9, 136.8, 138.7. Anal. Calcd
for C30H50O4: C, 75.90; H, 10.62. Found: C, 75.90; H, 10.72.
6b: Colorless plates in 53% yield based on 2b; mp 105-106
7a : Colorless microcrystals in 51% yield; mp 48-49 °C. H
NMR (250 MHz, CDCl3, 300 K, TMS): δ ) 0.9 (t, 6H, CH3),
1.1-1.4 (m, 12H, CH2), 1.5-1.6 (m, 4H, CH2), 3.0 (t, 4H, ArH),
7.7 (s, 2H, ArH), 10.4 (s, 2H, CHO). 13C NMR (63 MHz, CDCl3,
300 K, TMS): δ ) 14.7, 23.2, 29.8, 32.3, 32.5, 33.0, 133.8, 137.4,
144.1, 192.4. Anal. Calcd for C20H30O2‚1/6H2O: C, 78.64; H,
10.01. Found: C, 78.72; H, 10.01.
1
1
°C. H NMR (250 MHz, CDCl3, 300 K, TMS): δ ) 0.8 (s, 6H,
7b: Colorless microcrystals in 81% yield; mp 61-62 °C. H
CH3), 0.9 (t, 6H, CH3), 1.2-1.4 (m, 22H, CH2, CH3), 1.5-1.6
(m, 4H, CH2), 2.7 (t, 4H, CH2), 3.6 (d, 4H, CH2), 3.8 (d, 4H,
CH2), 5.5 (s, 2H, CH), 7.4 (s, 2H, ArH). 13C NMR (63 MHz,
CDCl3, 300 K, TMS): δ ) 14.8, 22.6, 23.4, 23.9, 29.8, 30.5,
30.8, 32.3, 32.5, 32.9, 78.5, 100.6, 127.8, 136.8, 138.7. Anal.
Calcd for C32H54O4: C, 76.45; H, 10.83. Found: C, 76.33; H,
10.96.
NMR (250 MHz, CDCl3, 300 K, TMS): δ ) 0.9 (t, 6H, CH3),
1.1-1.4 (m, 16H, CH2), 1.6-1.7 (m, 4H, CH2), 3.0 (t, 4H, ArH),
7.7 (s, 2H, ArH), 10.4 (s, 2H, CHO). 13C NMR (63 MHz, CDCl3,
300 K, TMS): δ ) 14.7, 23.3, 29.7, 30.1, 32.4, 32.5, 33.1, 133.8,
137.4, 144.1, 192.4. Anal. Calcd for C22H34O2‚1/6H2O: C, 79.23;
H, 10.38. Found: C, 79.48; H, 10.39.
1
7c: Colorless needles in 91% yield; mp 67-68 °C. H NMR
6c: Colorless needles in 59% yield based on 2c; mp 97-98
(250 MHz, CDCl3, 300 K, TMS): δ ) 0.9 (t, 6H, CH3), 1.1-1.4
(m, 20H, CH2), 1.6-1.7 (m, 4H, CH2), 3.0 (t, 4H, ArH), 7.7 (s,
2H, ArH), 10.4 (s, 2H, CHO). 13C NMR (63 MHz, CDCl3, 300
K, TMS): δ ) 14.7, 23.3, 29.9, 30.1, 30.2, 30.4, 32.5, 33.1, 133.8,
137.4, 144.1, 192.4. Anal. Calcd for C24H38O2‚1/6H2O: C, 79.73;
H, 10.69. Found: C, 79.30; H, 10.69.
1
°C. H NMR (250 MHz, CDCl3, 300 K, TMS): δ ) 0.8 (s, 6H,
CH3), 0.9 (t, 6H, CH3), 1.2-1.4 (m, 26H, CH2, CH3), 1.5-1.6
(m, 4H, CH2), 2.7 (t, 4H, CH2), 3.6 (d, 4H, CH2), 3.8 (d, 4H,
CH2), 5.5 (s, 2H, CH), 7.4 (s, 2H, ArH). 13C NMR (63 MHz,
CDCl3, 300 K, TMS): δ ) 14.8, 22.6, 23.4, 23.9, 29.9, 30.1,
30.5, 30.9, 32.3, 32.6, 32.9, 78.5, 100.6, 127.8, 136.8, 138.7.
Anal. Calcd for C34H58O4: C, 76.93; H, 11.01. Found: C, 76.91;
H, 11.08.
1
7d : Colorless needles in 96% yield; mp 74-75 °C. H NMR
(250 MHz, CDCl3, 300 K, TMS): δ ) 0.8 (t, 6H, CH3), 1.2-1.4
(m, 24H, CH2), 1.5-1.6 (m, 4H, CH2), 3.0 (t, 4H, CH2), 7.7 (s,
2H, ArH), 10.4 (s, 2H, CHO). 13C NMR (63 MHz, CDCl3, 300
K, TMS): δ ) 14.7, 23.3, 29.9, 30.0, 30.1, 32.5, 33.0, 133.8,
137.2, 144.0, 192.3. Anal. Calcd for C26H42O2‚1/6H2O: C, 80.15;
H, 10.95. Found: C, 80.05; H, 10.98.
6d : Colorless needles in 62% yield based on 2d ; mp 99-100
1
°C. H NMR (250 MHz, CDCl3, 300 K, TMS): δ ) 0.8 (s, 6H,
CH3), 0.9 (t, 6H, CH3), 1.2-1.4 (m, 30H, CH2, CH3), 1.5-1.6
(m, 4H, CH2), 2.7 (t, 4H, CH2), 3.6 (d, 4H, CH2), 3.8 (d, 4H,
CH2), 5.5 (s, 2H, CH), 7.4 (s, 2H, ArH). 13C NMR (63 MHz,
CDCl3, 300 K, TMS): δ ) 14.8, 22.6, 23.3, 23.9, 30.0, 30.1,
30.2, 30.5, 30.8, 32.3, 32.6, 32.9, 78.5, 100.6, 127.8, 136.8, 138.7.
Anal. Calcd for C36H62O4: C, 77.37; H, 11.18. Found: C, 77.36;
H, 11.33.
1
7e: Colorless microcrystals in 97% yield; mp 77-78 °C. H
NMR (250 MHz, CDCl3, 300 K, TMS): δ ) 0.8 (t, 6H, CH3),
1.1-1.4 (m, 28H, CH2), 1.5-1.6 (m, 4H, CH2), 3.0 (t, 4H, CH2),
7.7 (s, 2H, ArH), 10.4 (s, 2H, CHO). 13C NMR (63 MHz, CDCl3,
300 K, TMS): δ ) 14.5, 23.0, 29.7, 29.8, 29.90, 29.92, 29.96,
32.2, 32.3, 32.7, 133.5, 137.1, 143.8, 192.1. Anal. Calcd for
C28H46O2‚1/6H2O: C, 80.52; H, 11.18. Found: C, 80.38; H, 11.18.
6e: Colorless hairfine crystals in 66% yield based on 2e; mp
1
92-93 °C. H NMR (250 MHz, CDCl3, 300 K, TMS): δ ) 0.8
1
(s, 6H, CH3), 0.9 (t, 6H, CH3), 1.2-1.4 (m, 34H, CH2, CH3),
1.5-1.6 (m, 4H, CH2), 2.7 (t, 4H, CH2), 3.6 (d, 4H, CH2), 3.8
7f: Colorless needles in 99% yield; mp 83-84 °C. H NMR
(250 MHz, CDCl3, 300 K, TMS): δ ) 0.8 (t, 6H, CH3), 1.1-1.4

4382 Krebs and J ørgensen
Macromolecules, Vol. 36, No. 12, 2003
(m, 32H, CH2), 1.5-1.6 (m, 4H, CH2), 3.0 (t, 4H, CH2), 7.7 (s,
2H, ArH), 10.4 (s, 2H, CHO). 13C NMR (63 MHz, CDCl3, 300
K, TMS): δ ) 14.8, 23.4, 30.0, 30.1, 30.2, 30.23, 30.3, 30.31,
32.5, 32.6, 33.1, 133.7, 137.4, 144.1, 192.4. Anal. Calcd for
C30H50O2‚1/6H2O: C, 80.84; H, 11.38. Found: C, 81.06; H, 11.45.
8f (d iu n d ecyl-P SV): Bright orange polymer in 23% yield.
1H NMR (250 MHz, C6D4Cl2, 420 K, TMS): δ ) 0.93 (t, 6H, J
) 7 Hz), 1.34-1.6 (m, 32H), 1.84 (broad singlet, 4H), 2.93
(broad singlet, 4H), 7.15 (d, 2H, J ) 16 Hz), 7.5-7.6 (m, 8H).
SEC: Mp ) 9470; Mn ) 6470; Mw/Mn ) 3.060. TGA: onset )
390 °C, derivative peak ) 480 °C. DSC, mesophase ) 108 °C
(onset ) 100 °C), melting ) 167 °C (onset ) 156 °C, 3.2 KJ
mol-1).
8g (d id od ecyl-P SV): Bright orange polymer in 20% yield.
1H NMR (250 MHz, C6D4Cl2, 420 K, TMS): δ ) 0.93 (t, 6H, J
) 7 Hz), 1.3-1.55 (m, 36H), 1.85 (broad singlet, 4H), 2.93
(broad singlet, 4H), 7.15 (d, 2H, J ) 16 Hz), 7.5-7.6 (m, 8H).
SEC: Mp ) 49 500; Mn ) 31 600; Mw/Mn ) 2.997. TGA: onset
) 327 °C, derivative peak ) 480 °C; DSC, mesophase ) 110
°C (onset ) 100 °C), melting ) 197 °C (onset ) 175 °C, 3.8 KJ
mol-1).
Size Exclu sion Ch r om a togr a p h y (SEC). SEC was per-
formed in THF at 60 °C using polystyrene standards for
molecular weight determination. The polymers were generally
not very soluble in THF even at the boiling point. Therefore,
the polymer samples were dissolved in boiling 1,2-dichloroben-
zene and kept at 150 °C until injected into the SEC system.
At concentrations of 1 mg mL-1, the 1,2-dichlorobenzene
solutions could be diluted with five volumes of THF at 60 °C
without any precipitation taking place (dilution by a factor of
5 roughly corresponds to the dilution of the sample during the
SEC experiment).
Differ en tia l Sca n n in g Ca lor im etr y (DSC). DSC was
performed at 10° min-1 temperature gradients in the temper-
ature interval 50-350 °C. It was notable that all the samples
exhibited a mesophase transition (side chain melting) and a
melting point. When a sample was heated beyond the melting
point, subsequent cooling gave rise to a glass.
1
7g: Colorless microcrystals in 99% yield; mp 84-85 °C. H
NMR (250 MHz, CDCl3, 300 K, TMS): δ ) 0.8 (t, 6H, CH3),
1.1-1.4 (m, 36H, CH2), 1.5-1.6 (m, 4H, CH2), 3.0 (t, 4H, CH2),
7.7 (s, 2H, ArH), 10.4 (s, 2H, CHO). 13C NMR (63 MHz, CDCl3,
300 K, TMS): δ ) 14.8, 23.4, 30.0, 30.1, 30.2, 30.23, 30.3, 30.33,
30.36, 32.5, 32.6, 33.1, 133.8, 137.4, 144.1, 192.4. Anal. Calcd
for C32H54O2‚1/6H2O: C, 81.12; H, 11.56. Found: C, 81.16; H,
11.62.
Gen er a l P r oced u r e for th e P r ep a r a tion of th e P oly-
m er s 8a )g. The dialdehyde 7a -g (5 mmol) was dried at 40
°C in a vacuum oven for 24 h and then mixed with the
diphosphonic acid ester (5 mmol) in a 1:1 mixture of 1,2-
dichlorobenzene and diphenyl ether (150 mL), and the mixture
was degassed with argon. NaH (1 g) was added, and the
mixture was refluxed under argon with an air condenser for
24 h. The mixture rapidly acquires a cloudy yellow/green
coloration. The mixture was cooled and methanol (200 mL)
containing HCl(aq) (37%, 25 mL) was added and the mixture
was stirred for 30 min. An additional portion of methanol (1
L) was added and the crude product filtered and washed with
methanol. The crude yellow/orange product was then trans-
ferred to a Soxhlet extraction apparatus and extracted with
THF overnight. The extract contained low molecular weight
material. The contents of the extraction thimble was boiled
with stirring under argon in 1,2-dichlorobenzene (100 mL)
after 2 h the boiling solution was filtered through glass wool
to give a clear bright yellow solution that was left to cool slowly
under argon. The product forms a gel, and the contents of the
flask were poured into methanol (1 L) with vigorous stirring.
The yellow product was filtered and dried in the oven at 90
°C for 24 h.
Th er m ogr a vim etr ic An a lysis (TGA). TGA was per-
formed in the temperature interval 25-700 °C to establish the
thermal stability of the polymers that were all found to exhibit
Gaussian degradion peaks in the 440-520 °C temperature
interval with a maximum for all the polymers 8a -g at
480 °C.
1
8a (d ih exyl-P SV): Dark orange polymer in 56% yield. H
NMR (250 MHz, C6D4Cl2, 420 K, TMS): δ ) 0.94 (broad
singlet, 6H), 1.4-1.55 (m, 12H), 1.82 (broad singlet, 4H), 2.91
(broad singlet, 4H), 7.15 (d, 2H, J ) 16 Hz), 7.49-7.6 (m, 8H).
SEC: Mp ) 13 500; Mn ) 5960; Mw/Mn ) 3.507. TGA: onset
) 351 °C, derivative peak ) 480 °C; DSC, mesophase ) 125
°C (onset ) 100 °C), melting ) 266 °C (onset ) 242 °C, 2.4 KJ
mol-1).
P R-TRMC Ca r r ier Mobility Mea su r em en ts. The tech-
nique employed for the determination of carrier mobilities was
the well-known pulse radiolysis time-resolved microwave
conductivity technique (PR-TRMC). The system used the 10
MeV electron accelerator facility at Risø National Laboratory
where short pulses of electrons (10-100 ns pulse length) could
be passed through the sample placed in an R-band waveguide
(26.5-40GHz). Three microwave sources were employed. A
HP8690B Sweeper with a HP8697A 26.5-40GHz module in
conjunction with a broadband power amplifier. Two Gunn
diode sources in the frequency range 31-35 GHz were
employed when possible due to the high power (27 dbm) and
low noise obtained. The appropriate microwave source could
be switched into the system that consisted of a 26.5-40 GHz
HP R532A frequency meter and a 26-40 GHz HP R382A
attenuator. Hereafter the microwaves were lead into the first
port of a three-port circulator and from the second port to the
sample holder that was placed in the path of the electron beam.
The sample was evacuated to a pressure of 10-5-10-4 mbar
using a turbo pump during measurements. The reflected
microwave power was again entering the second port of the
circulator and taken from the third port through an isolator
and a circular waveguide cavity filter (TE111 band-pass filter
with a bandwidth of 500 MHz). At this point the reflected
power could be switched into a power meter (HP432A with a
HP R486A thermistor mount) for the absolute measurement
of microwave power or directly to a HP R422C Schottky barrier
diode with a time response better than 1 ns. At this point the
time response of the system is limited by the bandwidth of
the cavity filter (∼2 ns). The fast transients were recorded in
8b (d ih ep tyl-P SV): Bright orange polymer in 40% yield.
1H NMR (250 MHz, C6D4Cl2, 420 K, TMS): δ ) 0.93 (broad
singlet, 6H), 1.36-1.5 (m, 16H), 1.81 (broad singlet, 4H), 2.90
(broad singlet, 4H), 7.15 (d, 2H, J ) 16 Hz), 7.5-7.6 (m, 8H).
SEC: Mp ) 3280; Mn ) 3480; Mw/Mn ) 3.103. TGA: onset )
340 °C, derivative peak ) 480 °C; DSC, mesophase ) 123 °C
(onset ) 104 °C), melting ) 224 °C (onset ) 218 °C, 0.6 KJ
mol-1).
1
8c (d ioctyl-P SV): Bright orange polymer in 25% yield. H
NMR (250 MHz, C6D4Cl2, 420 K, TMS): δ ) 0.93 (broad
singlet, 6H), 1.34-1.55 (m, 20H), 1.84 (broad singlet, 4H), 2.92
(broad singlet, 4H), 7.15 (d, 2H, J ) 16 Hz), 7.5-7.6 (m, 8H).
SEC: Mp ) 26 400; Mn ) 14 000; Mw/Mn ) 3.179. TGA: onset
) 344 °C, derivative peak ) 480 °C; DSC, mesophase ) 113
°C (onset ) 96 °C), melting ) 198 °C (onset ) 180 °C, 1.6 KJ
mol-1).
8d (d im on yl-P SV): Bright orange polymer in 27% yield.
1H NMR (250 MHz, C6D4Cl2, 420 K, TMS): δ ) 0.93 (broad
singlet, 6H), 1.34-1.56 (m, 24H), 1.84 (broad singlet, 4H), 2.92
(broad singlet, 4H), 7.16 (d, 2H, J ) 16 Hz), 7.6-7.6 (m, 8H).
SEC: Mp ) 129000; Mn ) 57 100; Mw/Mn ) 11.335. TGA: onset
) 334 °C, derivative peak ) 480 °C; DSC, mesophase ) 191
°C (onset ) 173 °C), melting ) 326 °C (onset ) 309 °C, 1.5 KJ
mol-1).
8e (d id ecyl-P SV): Bright orange polymer in 25% yield. 1H
NMR (250 MHz, C6D4Cl2, 420 K, TMS): δ ) 0.93 (t, 6H, J )
7 Hz), 1.4-1.55 (m, 28H), 1.82 (broad singlet, 4H), 2.91 (broad
singlet, 4H), 7.15 (d, 2H, J ) 16 Hz), 7.49-7.6 (m, 8H). SEC:
Mp ) 29 100; Mn ) 20 600; Mw/Mn ) 2.413. TGA: onset ) 326
°C, derivative peak ) 480 °C; DSC, mesophase ) 186 °C (onset
) 169 °C), melting ) 319 °C (onset ) 297 °C, 2.3 KJ mol-1).
a 50 Ω system on a 4 Gs/s TDS784C oscilloscope from
Tektronix. The electron beam was after passage through the
sample picked up by a Faraday cup and the signal from the
Faraday cup used as the trigger source for the data collection.
Both the conductivity transient and the electron pulse were
stored for later analysis. The electron pulse was used to
monitor changes in beam current and pulse shape and to get

Macromolecules, Vol. 36, No. 12, 2003
High Carrier Mobility 4383
a measure of the absorbed radiation dose per measured beam
current. The conductivity was obtained by using a calibration
curve relating the reflected microwave power as a function of
detector diode output voltage. To extract the carrier mobility,
the number of carriers must be known. This was obtained from
dosimetry where radiochromic films were used to get the
distribution around the sample holder and inside the sample
holder in four depths using polyethylene spacers (this proce-
dure followed the ASTM E1275 standard “Practice for use of
a radiochromic film dosimetry system”). A better absolute
measure of the dose was obtained using alanine pellets where
the dose over the whole volume of the pellet is accurately
determined (this procedure followed the ASTM E1607 stan-
dard “Practice for use of the alanine-EPR dosimetry system”).
The typical dose for a 100 ns pulse was 25 Gy, and the
variation in dose over the sample volume was 10%. The value
for the end-of-pulse conductivity used in the calculation of the
minimum sum of carrier mobilities was obtained 100 ns after
the end of the electron pulse to ensure a conservative mea-
surement. An analysis of the conductivity transient gave the
carrier half-life, τ1/2. The lifetimes were determined using the
F900 program from Edinburg Instruments. A fast data acqui-
sition of 500 MS s-1 taking 5000 data points were used for
the determination of the first half-life that was taken as the
characteristic time constant. For the longer decay times 50 000
data points were recorded at 2.5 MS s-1 that was subsequently
thinned.
Sin gle-Cr ysta l X-r a y Diffr a ction . A crystals of 6c was
subjected to single-crystal X-ray diffraction. The crystals were
drawn directly from the mother liquor, coated with a thin layer
of protecting oil and mounted on glass fibers using Apiezon
grease and transferred quickly to the cold stream of nitrogen
(Oxford Cryostream) on the diffractometer (Siemens SMART
CCD Platform). The alkyl chains were found to be disordered
and this was treated as two mutually exclusive contributions
with a refinement of the sof. An almost complete sphere of
reciprocal space was covered by a combination of several sets
of exposure frames; each set with a different æ angle for the
crystal and each frame covering a scan of 0.3° in ω. Data
collection, integration of frame data and conversion to intensi-
ties corrected for Lorenz, polarization, and absorption effects
were performed using the programs SMART,38 SAINT,38 and
SADABS.39 Structure solution, refinement of the structures,
structure analysis and production of crystallographic illustra-
tions was carried out using the programs SHELXS97,40
SHELXL97,40 and SHELXTL.41 The structures were checked
for higher symmetry, and none was found.42 In all of the
structures H atoms were included in their calculated positions.
Crystallographic data (excluding structure factors) for the
structure reported in this paper have been deposited with the
Cambridge Crystallographic Data Centre as supplementary
publication no. CCDC-198729. Copies of the data can be
obtained free of charge on application to the Director, Cam-
bridge Crystallographic Data Centre, 12 Union Road, Cam-
bridge CB2 1EZ, U.K. (fax (+44)1223-336-033; e-mail deposit@
ccdc.cam.ac.uk).
polymers (8c and 8d ). A few grains of the polymer was covered
with a few mL of ethanol (δ ) 0.7893 g cm-3) weighed out
carefully. Acetic acid (δ ) 1.0492 g cm-3) was then added
dropwise while weighing the sample and ensuring good mixing,
until the polymer grains floated freely in the liquid. The
density could then be estimated to be 0.95 g cm-3 for 8c and
0.99 g cm-3 for 8d .
P h otoelectr on Sp ectr oscop y. The samples for ultraviolet
photoelectron spectroscopy measurements (UPS) were pre-
pared by spin-coating a 1 mg cm-3 solution of 8a -g in 1,2-
dichlorobenzene onto a freshly prepared polycrystalline silver
surface. The samples were then dried in a vacuum oven at 50
°C for 24 h with the exclusion of light. The samples were then
mounted in the load-lock and kept at a pressure of 10-9 mbar.
The photoelectron spectra were recorded at the ASTRID
storage ring at Aarhus University, Aarhus, Denmark. The
beamline consists of an SX-700 monochromator and a hemi-
spherical electron energy analyzer. An ESCA (electron spec-
troscopy for chemical analysis) of the samples was recorded
first to check the cleanliness of the area where the incident
photons illuminated the sample and to confirm the presence
of carbon and the absence of N, Cl, Na, and P. The photoelec-
trons were measured at an angle normal to the sample surface.
The ESCA scans were performed with 800 eV photons and a
resolution of 0.4 eV. The photoelectron spectra were recorded
using 50 eV incident photons and a resolution of 0.2 eV. The
experimental chamber was equipped with an ion gun, a mass
spectrometer and a gas dosing system. The sample holder was
electrically isolated from the chamber, and the sample was
kept at a potential of -9.5 V relative to the surrounding
instrument. This serves to eliminate the contribution from the
instrument work function at the low energy cutoff. A clean
silver substrate was first entered and sputtered with ionized
argon using an emission current of 22 mA and a potential of
3 kV. The sputtering was stopped after 1 h, and subsequent
measurements of the silver work function gave the reference
value for the work function (ΦAg, ) 4.2 eV). The samples 8a -g
on silver substrates were introduced and the photoelectron
spectra recorded. The onset and cutoff of the intensity normal-
ized photoelectron spectra were used to compute the work
function in the case of the pure silver substrate and the
ionization potential and valence band edge in the case of the
silver substrates containing the samples. The data are pre-
sented in Table 4.
The silver work function, ΦAg, was determined to be 4.2 eV.
The position of the valence band edge for 8a -g, EFVB, was
lower than the Fermi level of silver. The Fermi level in
compounds 8a -g are thus situated in the forbidden band gap
of the materials and the distance from the Fermi level of the
samples to the vacuum level is given by the difference between
the ionization potential, IP, and EFVB giving EFVAC. The vacuum
VAC
level shift, ∆, is obtained as: ∆ ) EF
- ΦAg. The vacuum
level shift reflects how the molecules orient at the surface
giving a dipole layer. The procedure employed in the analysis
of the data has been reported.34,36
P h otod egr a d a tion . The photodegradation was measured
by irradiation of a thin film of the appropriate polymer spin-
coated onto a Pyrex glass slide. The absorbance of the sample
was measured before and at intervals during irradiation. The
irradiation was done in the wavelength range 400-450 nm
with an incident power of 10 mW cm-2. In the atmosphere the
photodegradation seemed to take place at all wavelengths
below 550 nm and was a rapid process where the large
absorption of the polymer material at 430 nm disappeared in
less than 2 h as shown in Figure 11.
A similar experiment was performed in the absence of
atmospheric oxygen by irradiating the film in the same setup
with a vacuum chamber kept at a pressure of 10-7 mbar. Films
irradiated for 24 h only showed a small degradation of the
absorbance of less than 5%. The mechanism responsible for
the degradation in the absence of oxygen is likely to involve a
2 + 2 electrocyclic reaction mechanism known from topotactic
dimerizations or polymerizations of similar materials. The
photodegradative properties could be exploited by using the
conductive polymer material itself as a photoresist. Illumina-
tion of thin films of the material through a positive mask
P ow d er X-r a y d iffr a ction . The polymers diffracted X-rays
moderately well at scattering angles up to 45° in 2θ when
using Cu KR radiation (see Supporting Information) using
standard laboratory powder diffractometer from STOE. Pow-
der data was collected in from 1 to 60° in 2θ. Powder indexing
and structure solution was achieved for 8c using the Powder-
Solve module in ReflexPlus within the Materials Studio
package from Accelrys.43 The powder X-ray data collected for
the thermal studies were collected at the I711 beamline at the
MAXII synchrotron in Lund, Sweden.44,45 The setup employed
a HUBER G670 imaging-plate Guinier camera and an X-ray
wavelength of 1.302 Å, which was the longest possible wave-
length with the current beamline setup. Longer wavelengths
would have been desirable for these polymer samples as they
do not scatter to very high angle and they have large peaks at
low angle. A wavelength of 2 Å would thus have spread the
diffractogram over a larger span in 2θ.
Den sity Mea su r em en ts. In view of the low-density found
in the modeling of the powder X-ray diffraction data (0.833 g
cm-3) a direct measurement was attempted for two of the

4384 Krebs and J ørgensen
Macromolecules, Vol. 36, No. 12, 2003
(9) Huynh, W. U.; Dittmer, J . J .; Alivisatos, A. P. Science 2002,
295, 2425-2427.
(10) Kraft, A. ChemPhysChem 2001, 2, 163-165.
(11) Warman, J . M.; Gelinck, G. H.; de Haas, M. P. J . Phys.:
Condens. Matter 2002, 14, 9935-9954.
(12) van de Craats, A. M.; Siebbles, L. D. A.; Bleyl, I.; Haarer, D.;
Berlin, Y. A.; Zharikov, A. A.; Warman, J . M. J . Phys. Chem.
B 1998, 102, 9625-9634.
(13) Hide, F.; D´ıas-Garc´ıa, M. A.; Schwartz, B. J .; Heeger, A. J .
Acc. Chem. Res. 1997, 30, 430-436.
(14) Wang, H.; Wang, H. H.; Urban, V. S.; Littrell, K. C.;
Thiyagarajan, P.; Yu, L. J . Am. Chem. Soc. 2000, 122, 6855-
6861.
(15) Staring, E. G. J .; Demandt, R. C. J . E.; Braun, D.; Rikken,
G. L. J .; Kessener, Y. A. R. R.; Venhuizen, T. H. J .; Wynberg,
H.; ten Hoeve, W.; Spoelstra, K. J . Adv. Mater. 1994, 6, 934-
937.
(16) Rehahn, M.; Schlueter, A.-D.; Feast, J . W. Synthesis 1988,
5, 386-388.
(17) Strehmel, B.; Sarker, A. M.; Malpert, J . H.; Strehmel, V.;
Seifert, H.; Neckers, D. C. J . Am. Chem. Soc. 1999, 121,
1226-1236.
F igu r e 11. Absorbance as a function of time of a thin film of
the polymer 8c upon irradiation in atmospheric oxygen. The
sample was irradiated with 10 mW cm-2 in the wavelength
range 400-450 nm.
(18) Infelta, P. P.; de Haas, M. P.; Warman, J . M. Radiat. Phys.
Chem. 1977, 10, 353-363.
(19) Warman, J . M.; de Haas, M. P. In Pulse Radiolysis; Tabata,
Y., Ed.; CRC Press: Boca Raton, FL, 1990; Vol. 6, pp 101-
132.
illustrated this. A thin yellow film of the polymer 8c was
irradiated, leading to decoloration of the illuminated area of
the film. The film could subsequently be doped with iodine or
iodine monochloride, leading to oxidation seen as a darkening
of the unexposed area of the film. The conductivity increase
upon doping was for the unexposed part of the film very high.
The dc conductivity with and without doping was performed
for both pristine and photobleached films. The films were cast
from 1,2-dichlorobenzene onto an etched pattern on 10 Ω
square-1 ITO slides. The thickness of the films was measured
using a DekTak 3030. Typical values were 150-200 nm. The
inter electrode gap was 1 mm.
(20) Gelinck, G. H.; Warman, J . M. J . Phys. Chem. 1996, 100,
20035-20042.
(21) de Haas, M. P.; van der Laan, G. P.; Wegewijs, B.; de Leeuw,
D. M.; Ba¨uerle, P.; Rep, D. B. A.; Fichou, D. Synth. Met. 1999,
101, 524-525.
(22) Schouten, P. G.; Warman, J . M.; de Haas, M. P. J . Phys.
Chem. 1993, 97, 9863-9870.
(23) Warman, J . M.; Schouten, P. G. J . Phys. Chem. 1995, 99,
17181-17185.
(24) Schouten, P. G.; Warman, J . M.; de Haas, M. P.; Fox, M. A.;
Pan, H.-L. Nature 1991, 353, 736-737.
(25) van de Craats, A.; Warman, J . M.; Fechtenko¨tter, A.; Brand,
J . D.; Harbison, M. A.; Mu¨llen, K. Adv. Mater. 1999, 11,
1469-1472.
Ack n ow led gm en t. This work was supported by the
Danish Technical Research Council (STVF). We would
like to express sincere gratitude to Ole Hagemann for
technical assistance with the synthetic organic chem-
istry, Torben J ohansen for technical assistance at the
Accelerator beamline, Arne Miller and Hanne Corfitzen
for the dosimetry measurements, J an Alstrup for the
DSC measurements, Sokol Ndoni and Walther B. Ped-
ersen for SEC measurements, and Zheshen Li, Søren
V. Hoffmann, and Philip Hofmann for technical support
at the ASTRID storage ring and at the beamline.
(26) Alig, R. C.; Bloom, S.; Struck, C. W. Phys. Rev. B 1980, 22,
5565-5582.
(27) Hoofman, R. O. M.; de Haas, M. P.; Siebbeles, L. D. A.;
Warman, J . M. Nature 1998, 392, 54-56.
(28) Remmers, M.; Neher, D.; Gru¨ner, J .; Friend, R. H.; Gelinck,
G. H.; Warman, J . M.; Quattrocchi, C.; dos Santos, D. A.;
Bre´das, J .-L. Macromolecules 1996, 29, 7432-7445.
(29) van der Laan, G. P.; de Haas, M. P.; Buik, A.; de Ruiter, B.
Synth. Met. 1993, 55-57, 4930-4935.
(30) Conquest Version 1.4 Cambridge Structural Database (CSD),
2002.
(31) Hohloch, M.; Maichle-Mo¨ssmer, C.; Hanack, M. Chem. Mater.
1998, 10, 1327-1332.
(32) Stalmach, U.; Schollmeyer, D.; Meier, H. Chem. Mater. 1999,
11, 2103-2106.
Su p p or tin g In for m a tion Ava ila ble: A CIF file contain-
ing single-crystal structure data for 6c, Figure S1, showing
the ORTEP structure of 6c, tables showing crystallographic
data, atomic coordinates and equivalent isotropic displacement
factors, bond lengths and angles, anisotropic displacement
factors, and hydrogen coordinates and isotropic displacement
factors for 6c, and Figure S2 showing powder diffraction data
for polymers 8a -g. This material is available free of charge
via the Internet at http//:pubs.acs.org.
(33) Bao, Z.; Chen, Y.; Cai, R.; Yu, L. Macromolecules 1993, 26,
5281-5286.
(34) Krebs, F. C.; J ørgensen, M. Macromolecules 2002, 35, 7200-
7206.
(35) Krebs, F. C.; J ørgensen, M. Macromolecules 2002, 35, 10233-
10237.
(36) Salaneck, W. R.; Lo¨gdlund, M.; Fahlman, M.; Greczynski, G.;
Kugler, T. Mater. Sci. Eng. 2001, R34, 121-146.
(37) Dam, N.; Scurlock, R. D.; Wang, B.; Ma, L.; Sundahl, M.;
Ogilby, P. R. Chem. Mater. 1999, 11, 1302-1305.
(38) SMART and SAINT. Area-Detector Control and Integration
Software; Siemens Analytical X-ray Instruments Inc.: Madi-
son WI, 1995.
(39) Empirical absorption program (SADABS) written by George
Sheldrick for the Siemens SMART platform.
(40) Sheldrick, George M. SHELX-97, Program for structure
solution and refinement, 1997.
Refer en ces a n d Notes
(1) Garnier, F.; Hajlaoui, R.; Yassar, A.; Srivastava, P. Science
1994, 265, 1684-1686.
(2) Wu¨rtner, F. Angew. Chem. 2001, 40, 1037-1039.
(3) Horowitz, G. Adv. Mater. 1998, 10, 365-377.
(4) Ciao, Y.; Parker, I. D.; Yu, G.; Zhang, C.; Heeger, A. J . Nature
(London) 1999, 397, 414-417.
(5) Friend, R. H.; Gymer, R. W.; Holmes, A. B.; Burroughes, J .
H.; Marks, R. N.; Taliani, C.; Bradley, D. D. C.; Dos Santos,
D. A.; Bre´das, J . L.; Lo¨gdlund, M.; Salaneck, W. R. Nature
(London) 1999, 397, 121-128.
(6) Granstro¨m, M.; Petritsch, K.; Arias, A. C.; Lux, A.; Andersson,
M. R.; Friend, R. H. Nature (London) 1998, 395, 257-260.
(7) Brabec, C. J .; Sariciftci, N. S.; Hummelen, J . C. Adv. Funct.
Mater. 2001, 11, 15-26.
(8) Schmidt-Mende, L.; Fechtenko¨tter, A.; Mu¨llen, K.; Moons, E.;
Friend, R. H.; MacKenzie, J . D. Science 2001, 293, 1119-
1122.
(41) Sheldrick, G. M. SHELXTL95; Siemens Analytical X-ray
Instruments Inc.: Madison WI, 1995.
(42) Spek, A. L. Acta Crystallogr. 1990, 46, C-34.
(43) Engel, G. E.; Wilke, S.; Ko¨nig, O.; Harris, K. D. M.; Leusen,
F. J . J . J . Appl. Crystallogr. 1999, 32, 1169-1179.
(44) Ståhl, K. J . Appl. Crystallogr. 2000, 33, 394-396.
(45) Cerenius, Y.; Ståhl, K.; Svensson, L. A.; Ursby, T.; Oskarsson,
Å.; Albertsson, J .; Liljas, A. J . Synchrotron Radiat. 2000, 7,
203-208.
MA025893P