Identifying the magnetoconductance

responses by the induced charge transfer

complex states in pentacene-based diodes
Cite as: Appl. Phys. Lett. 101, 053307 (2012); https://doi.org/10.1063/1.4742135
Submitted: 22 June 2012 • Accepted: 17 July 2012 • Published Online: 01 August 2012

Wei-Shun Huang, Tsung-Hsun Lee, Tzung-Fang Guo, et al.

ARTICLES YOU MAY BE INTERESTED IN

Magnetoconductance responses of triplet polaron pair charge reaction in hyperfine coupling

regime
Applied Physics Letters 103, 253304 (2013); https://doi.org/10.1063/1.4852755

Magnetoconductance responses in organic charge-transfer-complex molecules

Applied Physics Letters 99, 073307 (2011); https://doi.org/10.1063/1.3627170

Detailed Balance Limit of Efficiency of p-n Junction Solar Cells

Journal of Applied Physics 32, 510 (1961); https://doi.org/10.1063/1.1736034

Appl. Phys. Lett. 101, 053307 (2012); https://doi.org/10.1063/1.4742135 101, 053307

© 2012 American Institute of Physics.

 APPLIED PHYSICS LETTERS 101, 053307 (2012)

Identifying the magnetoconductance responses by the induced charge

transfer complex states in pentacene-based diodes
Wei-Shun Huang,1 Tsung-Hsun Lee,1,2 Tzung-Fang Guo,1,3,a) J. C. A. Huang,2,3
and Ten-Chin Wen4
1
Department of Photonics, National Cheng Kung University, 70101 Tainan, Taiwan
2
Department of Physics, National Cheng Kung University, 70101 Tainan, Taiwan
3
Advanced Optoelectronic Technology Center, National Cheng Kung University, 70101 Tainan, Taiwan
4
Department of Chemical Engineering, National Cheng Kung University, 70101 Tainan, Taiwan

(Received 22 June 2012; accepted 17 July 2012; published online 1 August 2012)
We investigate the magnetoconductance (MC) responses in photocurrent, unipolar injection,
and bipolar injection regimes in pentacene-based diodes. Both photocurrent and bipolar
injection contributed MC responses show large difference in MC line shape, which are
attributed to triplet-polaron interaction modulated by the magnetic field dependent singlet
fission and the intersystem crossing of the polaron pair, respectively. By blending
2,3,5,6-tetrafluoro-7,7,8,8-tetracyanoquinodimethane into pentacene, all the MC responses are
suppressed but the MC response at unipolar injection regime is enhanced, which is attributed to the
induced charge transfer complex states (CT complex states). This work identify the MC responses
between single carrier contributed MC and exciton related MC by the induced CT complex states.
C 2012 American Institute of Physics. [http://dx.doi.org/10.1063/1.4742135]
V

Organic semiconductor devices without any ferromag- straightforward. In most studies of the OMAR reveal that the
netic elements have large magnetoresistance in room temper- MC responses resulting only from one magnetic mechanism.
ature.1–3 Besides, the performances such as photocurrent, However, when several MC components are observed in or-
electroluminescence, and photoluminescence intensity could ganic semiconductor device, obviously, one magnetic mech-
be changed by applied magnetic field.4 These magnetic field anism is not enough to elucidate the results properly.
effects in organic semiconductor devices are called organic Wang et al. have shown that several components of the
magnetoresistance (OMAR). By studying the OMAR, one MC response could coexist in polymer-fullerene blends.11 In
could realize the role of spin transport and reaction in or- addition, they could identify the underlying mechanisms by
ganic semiconductor devices;4–7 consequently, it is helpful changing the blending concentration, operating under different
to understand the device’s performance in organic electron- applied bias, and fabricating unipolar devices. However, the
ics and to develop spintronic devices with additional results are more complicated owing to the blending of the full-
functionality.8,9 erene creates additional MC components. Lee et al. have dem-
With the development of the OMAR, heretofore, several onstrated that, by blending a high electron affinity material
mechanisms have been provided.2–5,10 Gillin et al. provided into the hole transport material, the induced charge transfer
an excitonic model based on the magnetic field dependent complex states (CT complex states) could contribute a nega-
intersystem crossing (ISC).5–7 The entire magnetoconduc- tive MC response satisfying the bipolaron model.12 The study
tance (MC) responses are related to the effect of triplet exci- shows that the MC response could be induced by blending the
ton on charge transport, the so called triplet-polaron high electron affinity material even through there is no MC
interaction (TPI) model, and some independent processes response in neat hole transport material. We think this could
could be interacted as following. One is site blocking effect be a trick to simplify and identify the MC responses in hole
in which the triplet exciton is considered as blocking site for transport materials. In this work, we investigate the MC
polaron with the same spin state. Another one is triplet exci- responses in pentacene-based devices in which the MC com-
ton could be considered as a shallow trap for carrier with op- ponents could be identified by the applied bias voltage. A
posite spin state. The other one is polaron which could high electron affinity material, 2,3,5,6-tetrafluoro-7,7,8,8-tet-
transport into triplet state resulting quenching of triplet. In racyanoquinodimethane (F4-TCNQ), is used to blend into
contrast to the excitonic models, for explaining the MC pentacene for increasing the unipolar MC component. By
responses in unipolar devices, Bobbert et al. demonstrated comparing both the current-voltage characteristics and MC
that magnetic field dependent bipolaron formation could responses between pure pentacene and blend devices, we pro-
result in a negative MC response in single carrier devices.10 pose that part of the induced trap states could contribute to
This bipolaron model refers to the spin transport process MC response at unipolar injection regime, which could be
between two sites in bipolaron states, which is tunable by well explained by bipolaron formation. Additionally, the MC
applying magnetic field. From this simulation result, one responses in photocurrent and bipolar injection regimes are
could explain the MC response in single carrier devices attributed to TPI modulated by the magnetic field dependent
singlet fission and the ISC of polaron pairs, individually.
a)
Author to whom correspondence should be addressed. Electronic mail: The organic devices were fabricated in a standard
guotf@mail.ncku.edu.tw. arrangement by sandwiching active layer between a

0003-6951/2012/101(5)/053307/5/$30.00 101, 053307-1 C 2012 American Institute of Physics

V
053307-2 Huang et al. Appl. Phys. Lett. 101, 053307 (2012)

transparent electrode and a metal electrode. The transparent regimes with/without illumination. Figs. 1(a) and 1(b) are
electrode was comprised of the cleaned indium-tin-oxide both the current and MC ratio (at applied magnetic field
(ITO) covered glass substrate (RITEK Corp., 15 X/w), and B ¼ 100 mT) as a function of bias voltage in ITO/PEDOT:
coated with poly(3,4-ethylenedioxythiophene): poly(styrene- PSS/pentacene/LiF/Al (pentacene device) with and without
sulfonate) (PEDOT:PSS; Baytron P, Bayer AG, Germany) as illumination, respectively. As shown in Fig. 1(a), the thresh-
hole transport layer. Afterward, the substrate was capped old point is around 0.9 V. While the bias voltage is above
with either pure pentacene or a blend of pentacene and 0.9 V, a negative MC response could be observed. When
F4-TCNQ (with blending ratio 1:0.28) which was deposited keep increasing the applied bias ( > 1.3 V), the MC response
by thermal evaporation in high vacuum condition would be inversed from negative to positive, and the maxi-
(4.0  106 hPa) as a active layer. The total thickness of the mum positive MC response is located on 1.5 V (MC  0.2%).
active layer is 80 nm in both devices. LiF (0.5 nm) and Al This result shows that the MC has strong bias dependence
(100 nm) were thermally deposited on the surface of active and could be relevant to the unbalanced and balanced injec-
layer as cathode in a vacuum chamber (4.0  106 hPa). The tion between electrons and holes. Therefore, for simplicity,
active area of device is 6 mm.2 All the fabricating processes we divide the applied bias regime into bipolar (0.9–1.3 V)
were carried out in a nitrogen environment without exposing and unipolar (>1.3 V) injection regime. Fig. 1(b) shows,
to atmosphere. In MC measurement, all the devices under illumination, the open-circuit voltage (Voc) is around
were mounted in a vacuum tube (vacuum condition  0.9 V in pentacene device. There are positive MC responses
7.0  103 hPa), then placed between two magnet poles with which could be observed when the applied bias is below Voc.
the direction of current flow perpendicular to applied mag- With increasing the applied bias voltage approach to Voc, the
netic field. In order to remove the influence of drifting stress positive MC ratio would be increased rapidly (MC  22% at
at constant bias, the MC curves were averaged through two- 0.9 V). In addition, when increasing the bias voltage above
times measurement.13 Here, the MC ratio is defined as Voc, the MC response would be reversed from positive to
negative in pentacene device (around the regime at 0.9 V).
DIðBÞ IðBÞ  Ið0Þ Subsequently, keep increasing the bias voltage, the sign of
MC ¼ ¼ ;
Ið0Þ Ið0Þ the MC would be reversed again from negative to positive.
From Fig. 1(b), the photocurrent shows stronger MC
where I(B) and I(0) are the device current with/without response than the injection current indicates that there exists
applied magnetic field, respectively. All the measurements a more magnetic field responsible process during the creation
were performed at room temperature, and the MC responses of photocurrent. By comparing Fig. 1(a) with (b), it is notice-
were found to be independent with the direction of applied able that both the MC ratios tend to the same when the
magnetic field. The MC measurements under illumination applied bias is above the threshold voltage because the injec-
were carried by HeNe laser (632 nm, 20 mW) as light tion current increases rapidly and becomes dominant.
source. Besides, at bipolar injection regime, the positive MC
The MC responses with different underlying mecha- response would be enhanced by illumination. This suggests
nisms have different dependences with applied bias.4,6 In that under illumination the formation of the excited states
this work, we investigate the behavior of the MC compo- could contribute to the positive MC response which could be
nents in pentacene-based devices at different applied bias explained by excitonic models. Fig. 1(c) shows both current
and MC ratio as a function of bias voltage without illumi-
nation in ITO/PEDOT: PSS/pentacene: F4-TCNQ/LiF/Al
(pentacene: F4-TCNQ device). It is apparently that blending
F4-TCNQ into pentacene film would lower the injection cur-
rent by comparing the results with Fig. 1(a). Furthermore,
there is only a negative MC response can be observed when
applying bias above the threshold point (0.9 V). Besides,
the magnitude of the negative MC is increased with increas-
ing bias voltage. It seems that there are some correlations
between the lower conductivity and the observed negative
MC response.
F4-TCNQ is a high electron affinity material, which is
commonly used as the blending material of the hole-
transport layers in small molecular light emitting diodes.14–16
Through F4-TCNQ blend into the hole transport layer, the CT
complex states could be induced, and the increase of conduc-
tivity could be expected. Recently, we study the MC response
of the CT complex states,12 and discover that there is
an enhancement in both conductivity and negative MC in
FIG. 1. Both the current and MC ratio (B ¼ 100 mT) as a function of bias N,N0 -bis-(1-naphyl)-N,N0 -diphenyl-1,10 -biphenyl0 -4,40 -diamine
voltage in pentacene device (a) with and (b) without illumination. (c) Both
the current and MC ratio (B ¼ 100 mT) as a function of bias voltage in pen-
(NPB): F4-TCNQ devices in comparison with both pure
tacene: F4-TCNQ device. Solid lines are I-V characteristics, and dash lines NPB and pure F4-TCNQ devices. However, there is only
with empty circle are MC ratio with bias voltage. the elimination of the conductivity observed in pentacene:
053307-3 Huang et al. Appl. Phys. Lett. 101, 053307 (2012)

F4-TCNQ device. We suggest that the induced CT complex

states are distributed on various energy levels, and could
be simply distinguished into polaron (shallow CT complex
states) and bipolaron states (deep CT complex states). In
addition, more induced polaron states would increase conduc-
tivity, but more induced bipolaron states would suppress
the conductivity. Consequently, the more induced bipolaron
states would be the reason why we observe the lower conduc-
tivity in pentacene: F4-TCNQ device. However, in absorption FIG. 3. The MC curves at several bias voltages under illumination in penta-
spectrum, there is no apparently additional absorption edge cene device. All the bias voltages are smaller than Voc, and the maximum
could be observed in F4-TCNQ blend into pentacene film, applied magnetic field is 100 mT.
which suggests that the amount of the induced CT complex
states are not much enough.13 Therefore, in addition to the which shows the same bias dependence and suggests both
formation of the deep CT complex states, there is another rea- components result from the same mechanism. Fig. 4 shows
son cause the lower conductivity. We already know that pen- the MC curves under different bias voltages in pentacene:
tacene has higher mobility due to the great crystallization. F4-TCNQ device without illumination. When applied bias
Blending F4-TCNQ into pentacene would disturb the crystal- voltage is above the threshold voltage (0.9 V), there is
lization and decrease grain size.17 Owing to the formation of only negative MC component could be observed. And this
the structural defects, more trap states induced in grain negative MC would be increased with increasing the bias
boundary could be expected. Accordingly, the lower conduc- voltage. It is worth noting that the shape of this negative
tivity in pentacene: F4-TCNQ device is partly attributed to MC in pentacene: F4-TCNQ device is almost the same
the structural defects in the grain boundary. with the shape of the negative MC in unipolar injection re-
The MC curves with the applied bias can help us to gime in pentacene device. Hence, we propose that the neg-
identify which MC component is dominant at specific ative MC responses in both devices are from the same
operation condition because different underlying mecha- mechanism which could be explained by bipolaron model.
nisms have different MC line shapes. Figs. 2 and 3 are the Furthermore, the enhancement of the negative MC in pen-
MC curves at various applied bias in pentacene device tacene: F4-TCNQ device is contributed by the induced
under dark and illumination, respectively. As the results bipolaron states. It is significant that in pentacene device
from Fig. 2 negative MC could be measured at unipolar there is a negative MC generated originally, and it means
injection regime, and positive MC is show up at bipolar some deep trap states (bipolaron) are formed essentially in
injection regime. In Fig. 2, the negative MC shows sharply pentacene device. Blending F4-TCNQ into pentacene could
increase with applied magnetic field at low field regime induce more bipolaron states to enlarge the negative MC
(10 mT), and positive MC would increase and become sat- response causes from the magnetic field dependent block-
urated at high applied magnetic field (100 mT). Although ing transport.
the sign inversion of the MC from negative to positive is In some organic semiconductor materials, the MC
common from paper reviews, however, the obvious differ- curves follow empirical function.1 By fitting the MC curves,
ence in MC line shape indicates that the MC responses the components of the MC could be defined precisely. Figs.
come from different magnetic field effects. From Fig. 3, 5(a)–5(c) demonstrate the MC curves with fitting line at dif-
under illumination, the MC is composed of one positive ferent bias regimes in pentacene device under illumination
and one negative MC components. In comparison with the and dark, individually. In Fig. 5(a), a linear combination of
positive MC component, the negative MC component is a one non-Lorentzian function and one Lorentzian function is
low field component which responses at a much lower used as the fitting line, that is,
applied magnetic field. Depending on the applied bias,
both positive and negative components would be enhanced B2 B2
MC ¼ c1 þ c 2 ;
rapidly when increasing the bias voltage approach to Voc B2 þ B210 ðjBj þ B20 Þ2
(0.9 V), and each component would be reversed when
the bias voltage is above Voc. It is important that both where c1 and c2 are the pre-factors, B is the applied magnetic
components increase with increasing bias voltage near Voc, field, and B10 and B20 are the saturation fields. Under

FIG. 2. The MC curves at several bias voltages under dark in pentacene FIG. 4. The MC curves at several bias voltages in pentacene: F4-TCNQ
device. The maximum applied magnetic field is upon to 100 mT. device. The maximum applied magnetic field is upon to 100 mT.
053307-4 Huang et al. Appl. Phys. Lett. 101, 053307 (2012)

in high field by Zeeman splitting.21 As result, the applied

magnetic field could modulate the population of the triplet
exciton, consequently, causes a combination of one low
field and one high field MC component.
From TPI model, the Frenkel-triplet exciton could be a
trapping center or blocking site for polaron and, conse-
quently, cause mobility decrease.5–7 This process would hap-
pen originally. Under bipolar injection, the excited states
would be formed in long distance in which most of them are
polaron pairs. Applying magnetic field would suppress the
ISC rate of the polaron pair and, consequently, decrease the
triplet concentration, which would reduce the probability of
the polaron trapped by triplet exciton resulting in a positive
MC. Under illumination, the singlet exciton is formed, which
follows by the strong singlet fission. Applying magnetic field
would increase the singlet fission in low field and, conse-
quently, increase the triplet concentration and increase TPI
resulting in a negative MC. In contrast, in high field, the sin-
FIG. 5. The MC curves with fitting line (a) at 0.8 V under illumination, (b)
glet fission would be suppressed which resulting a positive
at 1.05 V under dark, and (c) at 1.8 V under dark in pentacene device. Solid MC. From above, we propose that the difference in MC line
lines are fitting lines, and all the empty circles are experimental data. The shape between photocurrent contributed MC and bipolar
maximum applied magnetic field is upon to 100 mT. injection contributed positive MC is resulted from TPI
modulated by the magnetic field dependent singlet fission
illumination and at applied bias below Voc (0.9 V) in penta- and the ISC of the polaron pair, respectively.
cene device, this linear combination could fit the MC curves In summary, we have demonstrated the MC responses in
perfectly. In this condition, B10 ¼ 63 mT and B20 ¼ 3.6 mT. pentacene-based devices. By blending F4-TCNQ into penta-
Under dark and at bias voltage above the threshold cene, the negative MC response in unipolar injection regime
(0.9 V) in unipolar injection regime, as shown in Fig. is enhanced. The result shows blending F4-TCNQ into pen-
5(b), the negative MC curves can fit with single non- tacene could help to simplify and identify the MC responses
Lorentzian function, and B0 ¼ 3.6 mT. When applying bias between excitonic and single carrier contributed MC
voltage in bipolar injection regime, as can be seen from response. On the other hand, the photocurrent and bipolar
Fig. 5(c), the positive MC curve can fit with single Lorent- injection contributed MC responses are attributed to TPI
zian function. Here, B0 ¼ 18 mT. We find that the saturation modified by the magnetic field dependent singlet fission and
field has different values under different operation condi- the intersystem crossing of the polaron pair, respectively.
tions. This strongly indicates multiple MC responses coexist
in pentacene device. By comparing Fig. 5(a) with (b), it is T.-F. Guo would like to thank the NSC of Taiwan
worth to note that both the saturation field and line shape of NSC99-2113 -M-006-008-MY3, the Asian Office of Aerospace
the negative MC component under illumination is the same Research and Development (AOARD-10-4054), and NCKU
with the negative MC response at bias voltage above the Landmark project for financially supporting this research.
threshold voltage (0.9 V). This result suggests that the
1€
negative MC component of photocurrent and the negative O. Mermer, G. Veeraraghavan, T. L. Francis, Y. Sheng, D. T. Nguyen,
MC of injection current could be contributed by the same M. Wohlgenannt, A. K€ ohler, M. K. Al-Suti, and M. S. Khan, Phys. Rev. B
72, 205202 (2005).
internal magnetic field. However, as discussed before, both 2
V. Prigodin, J. Bergeson, D. Lincoln, and A. Epstein, Synth. Met. 156, 757
positive and negative components in photocurrent regimes (2006).
3
have the same bias dependence; in other words, it seems J. Kalinowski, J. Szmytkowski, and W. Stampor, Chem. Phys. Lett. 378,
these two components are bound together. We should deal 380 (2003).
4
B. Hu, L. Yan, and M. Shao, Adv. Mater. 21, 1500 (2009).
with the MC components together. As proposed by Nguyen 5
P. Desai, P. Shakya, T. Kreouzis, and W. P. Gillin, Phys. Rev. B 76,
et al., the level-crossing in polaron pairs spin sublevel could 235202 (2007).
6
contribute to the combination of a negative MC component P. Desai, P. Shakya, T. Kreouzis, and W. P. Gillin, J. Appl. Phys. 102,
in low field and a positive MC component in high field.18 7
073710 (2007).
P. Desai, P. Shakya, T. Kreouzis, W. P. Gillin, N. A. Morley, and M. R. J.
The level-crossing could be a suitable model to satisfy the Gibbs, Phys. Rev. B 75, 094423 (2007).
MC response under photocurrent regime. However, in con- 8
T. Reichert and T. P. I. Saragi, Appl. Phys. Lett. 98, 063307 (2011).
9
trast with the dissociation of singlet exciton followed by the M. Prezioso, A. Riminucci, I. Bergenti, P. Graziosi, D. Brunel, and V. A.
intersystem crossing of polaron pair, the singlet fission is Dediu, Adv. Mater. 23, 1371 (2011).
10
P. A. Bobbert, T. D. Nguyen, F. W. A. van Oost, B. Koopmans, and M.
more dominant owing to the strong singlet fission in penta- Wohlgenannt, Phys. Rev. Lett. 99, 216801 (2007).
cene molecules.19,20 As proposed by Merrifield et al., 11
F. J. Wang, H. B€assler, and Z. V. Vardeny, Phys. Rev. Lett. 101, 236805
applying magnetic field could modulate the reaction rate of (2008).
12
T. H. Lee, J. H. Li, W. S. Huang, B. Hu, J. C. A. Huang, T. F. Guo, and
singlet fission by changing the amplitude of singlet compo-
T. C. Wen, App. Phys. Lett. 99, 073307 (2011).
nent of triplet pairs, in which the reaction rate is increased 13
See supplementary material at http://dx.doi.org/10.1063/1.4742135 for
in low field by increasing the spin states mix and decreased demonstration of the MC measurement and absorption spectra.
053307-5 Huang et al. Appl. Phys. Lett. 101, 053307 (2012)

14 18
Y. Nosho, Y. Ohno, S. Kishimoto, and T. Mizutani, Nanotechnology 18, T. D. Nguyen, G. Hukic-Markosian, F. J. Wang, L. Wojcik, X. G. Li,
415202 (2007). E. Ehrenfreund, and Z. V. Vardeny, Nature Mater. 9, 345 (2010).
15 19
M. B. Khalifa, D. Vaufrey, and J. Tardy, Org. Electron. 5, 187 (2004). M. W. B. Wilson, A. Rao, J. Clark, R. S. S. Kumar, D. Brida, G. Cerullo,
16
J. H. Kaufman, N. Colaneri, J. C. Scott, and G. B. Street, Phys. Rev. Lett. and R. H. Friend, J. Am. Chem. Soc. 133, 11830 (2011).
20
53, 1005 (1984). R. H. Friend, M. Phillips, A. Rao, M. W. B. Wilson, Z. Li, and C. R.
17
H. Kleemann, C. Schuenemann, A. A. Zakhidov, M. Riede, B. L€ ussem, McNeill, Faraday Discuss. 155, 339 (2012).
21
and K. Leo, Org. Electron. 13, 58 (2012). R. E. Merrifield, P. Avakian, and R. P. Groff, Chem. Phys. Lett. 3, 155 (1969).