APPLIED PHYSICS LETTERS 93, 193310 共2008兲
Band alignment at metal/organic and metal/oxide/organic interfaces
M. G. Helander,a兲 Z. B. Wang, J. Qiu, and Z. H. Lub兲
Department of Materials Science and Engineering, University of Toronto, 184 College St., Toronto,
Ontario M5S 3E4, Canada
共Received 26 August 2008; accepted 30 October 2008; published online 14 November 2008兲
Charge injection at metal/organic interfaces dictates the performance, lifetime, and stability of
organic electronic devices. We demonstrate that interface dipole theory, originally developed to
describe Schottky contacts at metal/semiconductor interfaces, can also accurately describe the
injection barriers in real organic electronic devices. It is found that theoretically predicted hole
injection barriers for various archetype metal/organic and metal/oxide/organic structures are in
excellent agreement with values extracted from experimental transport measurements. Injection
barriers at metal/organic and metal/oxide/organic interfaces can therefore be accurately predicted
based on the knowledge of only a few fundamental material properties of the oxide and organic
layers. © 2008 American Institute of Physics. 关DOI: 10.1063/1.3030979兴
Thin film organic semiconductors have attracted considerable interest due to their potential application in low-cost
and high performance organic electronic devices. Since all
charge carriers within an organic semiconductor device must
be injected from the anode and cathode,1 the performance,
lifetime, and stability of all organic electronics are typically
governed by the metal/organic interfaces at the electrode
contacts.2 However, despite the critical importance of these
interfaces, the physics that governs charge injection at metal/
organic interfaces has yet to be determined. In the case of
traditional semiconductors, such as Si or GaAs, it is widely
accepted that interface dipole theory most accurately predicts
the band alignment for Schottky contacts formed at metal/
semiconductor interfaces.3–6 The theory describes the formation of an interfacial dipole due to the charging of intrinsic
interface states, which tends to drive the band alignment toward a zero net dipole charge.7 Recently, Mönch8 argued that
this theory can also be used to describe metal/organic interfaces. However, despite the surprising accuracy of interface
dipole theory in describing traditional metal/semiconductor
contacts, there has been no direct experimental evidence to
demonstrate its relevance to predicting injection barrier
heights at metal/organic interfaces in real organic electronic
devices. In this letter we demonstrate that the hole injection
barriers predicted by the interface dipole theory are in excellent agreement with the barrier heights extracted from carrier
transport measurements performed on single carrier holeonly devices for a variety of metal/organic and metal/oxide/
organic archetype systems.
Single carrier hole-only devices were fabricated in a
Kurt J. Lesker LUMINOS® cluster tool on Corning® 1737
glass substrates 共50⫻ 50 mm2兲. The standard device structure is as follows: anode 共20 nm兲/oxide 共0 and 1 nm兲/organic
共500 nm兲/Au 共20 nm兲. The organic thickness was chosen as
500 nm in order to ensure that any built-in potential is negligible compared to the applied external potential 共⬃20 V兲;
given the interface dipoles at the metal/organic and metal/
oxide/organic interfaces discussed in this work, the value of
a兲
Author to whom correspondence should be addressed. Electronic mail:
michael.helander@utoronto.ca.
b兲
Electronic mail: zhenghong.lu@utoronto.ca.
0003-6951/2008/93共19兲/193310/3/$23.00
any built-in potential is unclear. Au was used as the cathode
for its high work function 共i.e., 5.1 eV兲 in order to create a
sufficiently large electron injection barrier to minimize the
electron current in the devices. Au and Ag were used as the
different anode materials. Sputter deposited Al2O3 and
HfO2 were used as the different oxide layers.
N , N⬘-diphenyl-N , N⬘-bis-共1-naphthyl兲 -1-1⬘-biphenyl-4 , 4⬘diamine 共␣-NPD兲 and 4 , 4⬘ , 4⬙-tris共N-3-methylphenylN-phenyl-amino兲triphenylamine 共m-MTDATA兲 were used as
the different organic semiconductor layers. The Al2O3 films
were deposited using radio frequency 共rf兲 magnetron sputtering as described elsewhere.9 The HfO2 films were deposited
using a similar procedure from a 99.95% pure HfO2 target
with a rf power of 140 W. Additional details of the device
fabrication have been discussed elsewhere.10 The thickness
of the organic layers was verified 共for each device兲 using
both a stylus profilometer 共KLA Tencor P-16+兲 and
capacitance-voltage 共CV兲 measurements 共Agilent 4294A兲.
The active area for all devices was 2 mm2. Current-voltage
共IV兲 characteristics were measured at room temperature 共300
K兲 in a vacuum cryostat with a base pressure of ⬃10−6 Torr
using an HP4140B picoammeter.
Many studies have experimentally demonstrated a strong
correlation between the metal work function and the injection barrier for holes 共␾Bp兲 or electrons at metal/organic interfaces. In the Schottky–Mott limit the vacuum level of the
organic and metal aligned, forming a region of net space
charge at the interface, as depicted in Fig. 1共a兲. However, the
Schottky–Mott limit is rarely observed at real metal/organic
interfaces due to the formation of a strong interfacial dipole
⌬ between the metal surface and organic molecules, which
tends to pin the Fermi level. An effective metal work function can be defined as the difference between the pristine
metal work function ␾m and the dipole,
␾m,eff = ␾m − ⌬,
共1兲
where ␾m,eff is the effective metal work function.
In the case of traditional inorganic semiconductors, deviation from the Schottky–Mott limit was first described by
Bardeen11 as the result of a large density of surface states,
which act to pin the Fermi level. Heine12 later demonstrated
that surface states do not exist in the forbidden gap of most
93, 193310-1
© 2008 American Institute of Physics
Downloaded 13 May 2009 to 131.175.136.39. Redistribution subject to AIP license or copyright; see http://apl.aip.org/apl/copyright.jsp
193310-2
Appl. Phys. Lett. 93, 193310 共2008兲
Helander et al.
FIG. 1. Schematic energy level diagram of the band alignment at a metal
organic interface, 共a兲 Schottky–Mott limit 共vacuum level alignment兲, and 共b兲
pinning of the Fermi level by induced interface states 共interfacial dipole兲.
metal/semiconductor interfaces but rather that gap states are
induced in the semiconductor as a result of the rapidly decaying tail of the electronic wave function from the metal.
These metal induced gap states, later discussed by Tersoff,13
are independent of the energy levels in the semiconductor
and hence can either be donorlike or acceptorlike when close
to the valence or conduction bands, respectively. A charge
neutrality level ECNL = e␾CNL is defined as the point at which
the interface states are equally donorlike and acceptorlike.
Therefore, at metal/semiconductor interfaces the charge neutrality level of the semiconductor will tend to align with the
Fermi level of the metal as a result of charge transfer between the metal and the induced gap states, forming an interfacial dipole,14
␾m,eff = ␾CNL + S␾共␾m − ␾CNL兲.
共2兲
Here, the interface slope parameter S␾ 共for weakly interacting interfaces兲 was empirically found by Mönch8,15 to depend on the optical dielectric constant,
S␾ =
1
,
1 + 0.1共␧⬁ − 1兲2
共3兲
where ␧⬁ is the optical dielectric constant 共high frequency
limit of the dielectric function兲. Mönch8 also argued that Eq.
共2兲 should be applicable to organic semiconductors, same as
for Schottky contacts with solid xenon. A theoretical method
for estimating ␾CNL for molecules has been proposed.16,17
The band alignment for a weakly interacting metal/
organic interface based on interface dipole theory is schematically depicted in Fig. 1共b兲. In the case of the metal/
oxide/organic interfaces considered in this work, the oxide
layer is thin enough that we can treat it as a surface modification layer instead of a discrete layer. As a result the dipoles
at the metal/oxide and oxide/organic interfaces are interdependent and hence can be treated as a single combined dipole, which is schematically depicted in Fig. 2.
Figure 3 shows typical IV characteristics for single carrier devices utilizing Au/ Al2O3 bilayer as anodes and
␣-NPD and m-MTDATA hole transport layers. The dashed
curves represent the estimated space charge limited current
共SCLC兲 for ␣-NPD and m-MTDATA. The solid curves are
the corresponding calculation results based on the well
known carrier injection model proposed by Scott and
Malliaras,1 and are used to extract the experimental injection
barrier heights, ␾Bp, by fitting the room temperature 共300 K兲
IV data. This model describes the equilibrium contributions
FIG. 2. Schematic energy level diagram of the band alignment at a metal/
oxide/organic interface where the oxide interlayer is treated as a surface
modification layer on the metal. Here, S䊐 is taken as the dielectric value
given by Eq. 共3兲 since the oxide interlayer ensures a weakly interacting
interface.
to the current density 共J兲 in analogy with the Langevin recombination of an electron-hole pair in the bulk,
J = Jinj − Jrec = 4N0␺2e␮F exp共− e␾Bp/kBT兲exp共f 1/2兲, 共4兲
where ␺ is a function of reduced electric field f
= e3F / 4␲␧kB2 T2, namely,
␺ = f −1 + f −1/2 − f −1共1 + 2f 1/2兲1/2 ,
共5兲
where N0 is the density of states, ␮ is the field-dependent
mobility, F is the electric field, ␧ is the dielectric constant of
the organic, kB is the Boltzmann constant, and T is temperature. The measured field-dependent mobilities of the organic
molecules used in this study have been reported elsewhere.18
In order to calculate the net effect of the oxide layer on
the injection barrier, we treat the thin oxide as a surface
modification layer on the metal. The interaction between the
oxide and the metal creates an interfacial dipole layer that
modifies the work function of the pristine metal surface. This
results in the organic molecules deposited on top of the oxide
modified metal experiencing an effective metal work function as given by Eq. 共2兲. In this case the morphology of the
oxide film on the metal surface plays a minor extrinsic role
as compared to the intrinsic interface states that tend to
FIG. 3. 共Color online兲 Current density as a function of electric field for
single carrier devices utilizing Au/ Al2O3 bilayer anodes with ␣-NPD and
m-MTDATA hole transport layers. Experimental results 共symbols兲 along
with theoretical calculations 共solid curves兲 are shown. The estimated SCLC
共dashed curves兲 is also shown as the upper limit. The electric field 共F兲 is
taken as F = V / d, where V is the voltage and d is the device thickness
共⬃500 nm兲.
Downloaded 13 May 2009 to 131.175.136.39. Redistribution subject to AIP license or copyright; see http://apl.aip.org/apl/copyright.jsp
193310-3
Appl. Phys. Lett. 93, 193310 共2008兲
Helander et al.
TABLE I. Material properties for the oxides and organics used in this study:
charge neutrality level 共ECNL兲, optical dielectric constant 共␧⬁兲, and interface
slope parameter 共S␾兲. All energies are reported with respect to vacuum level
in units of eV.
Material
ECNL
a
Al2O3
HfO2
␣-NPD
m-MTDATA
5.2
4.8b
4.2c
4.1f
␧⬁
b
3.4
4.0b
3.0d
3.0d
S␾
0.69b
0.52b
0.33e
0.10f
a
Taken from Ref. 3.
Taken from Ref. 7.
c
Taken from Ref. 21.
d
Taken from Ref. 20.
e
Taken from Ref. 24.
f
Taken from Ref. 23.
b
dictate the band alignment.15,19 Treating the metal/oxide as a
layered anode structure, the effective metal work function
can be plugged back into Eq. 共2兲 to extract the effective work
function of the metal/oxide after consideration of the effect
of the organic layer. The hole injection barrier is then taken
as the difference between this effective work function and
the highest occupied molecular orbital 共HOMO兲 in the organic. The HOMOs of ␣-NPD and m-MTDATA are taken as
5.4 and 5.1 eV, respectively.20 The required material parameters to extract the theoretical barrier heights—namely, the
charge neutrality level, optical dielectric constant, and interface slope parameter—are summarized in Table I. The theoretical barrier heights along with the extracted experimental
values for the different metal and metal/oxide bilayer anodes
are summarized in Table II; the calculated effective metal
work functions for the various anode structures are also
shown. For all of the metal/organic interfaces considered in
this study, the insertion of a thin oxide layer dramatically
reduces the hole injection barrier due to Fermi level pinning
by the oxide 共see Fig. 2兲. Remarkably the hole injection barriers predicted by interface dipole theory are in excellent
agreement with the experimentally determined values; for
nearly all of the structures, the agreement is better than 100
meV. Similar agreement between theory and experiment was
TABLE II. Comparison of theoretical and experimental hole injection
barriers 共␾Bp兲 at metal/organic and metal/oxide/organic interfaces. The
calculated effective metal work function 共␾m,eff兲 of the anodes are also
shown. All barriers are given in units of V.
Anode
␾m,eff
Organic
␾Bp Theor.
␾Bp Expt.
Au
Au/ HfO2
Au/ Al2O3
Ag/ Al2O3
Au
Au/ HfO2
Au/ Al2O3
Ag/ Al2O3
5.1
4.96
5.13
4.79
5.1
4.96
5.13
4.79
m-MTDATA
m-MTDATAb
m-MTDATAb
m-MTDATAb
␣-NPD
␣-NPDb
␣-NPDb
␣-NPDb
0.90
0.41
0.29
0.53
0.90
0.68
0.56
0.80
0.86a
0.46
0.44
0.47
0.95c
0.60
0.55
0.77
a
Taken from Ref. 25.
The interface slope parameter 共S␾兲 is taken as the dielectric value given by
Eq. 共3兲 since the oxide interlayer ensures a weakly interacting interface,
whereas the metal/organic interfaces are better represented by the experimental value that includes other effects.
c
Taken from Ref. 26.
b
also found based on photoemission,21 albeit with a larger
error range due to the limited resolution of photoemission.
For the case of Au/ Al2O3 / m-MTDATA, the high experimental current density near the SCLC 共see Fig. 3兲 combined
with the low predicted injection barrier 共0.29 eV兲 strongly
suggests that the contact is already in the quasi-Ohmic injection regime described by Wolf et al.,22 and therefore
injection-based models may no longer be applicable in extracting the injection barrier. As a result the extracted experimental injection barrier 共0.44 eV兲 is most likely an overestimate of the actual value.
In summary, we have demonstrated that interface dipole
theory can indeed accurately describe the injection barriers in
real organic electronic devices. It is found that theoretically
predicted hole injection barriers for various archetype metal/
organic and metal/oxide/organic contacts are in excellent
agreement with experimentally determined values based on
transport experiments. Therefore, once the material properties for a given organic molecule are known, namely, the
charge neutrality level and optical dielectric constant, the
injection barriers at metal/organic interfaces can be predicted. It is further found that interface dipole theory not
only explains why metal oxides act as universal hole injection layers but can also be used to accurately predict the
injection barriers.
We wish to acknowledge funding for this research from
Ontario Centres of Excellence and Natural Sciences and
Engineering Research Council 共NSERC兲 of Canada.
J. C. Scott and G. G. Malliaras, Chem. Phys. Lett. 299, 115 共1999兲.
H. Aziz, Z. D. Popovic, N. X. Hu, A. M. Hor, and G. Xu, Science 283,
1900 共1999兲.
3
A. A. Demkov, L. R. C. Fonseca, E. Verret, J. Tomfohr, and O. F. Sankey,
Phys. Rev. B 71, 195306 共2005兲.
4
J. Robertson, Appl. Surf. Sci. 190, 2 共2002兲.
5
W. Mönch, Surf. Sci. 299, 928 共1994兲.
6
F. Flores, R. Perez, R. Rincon, and R. Saiz-Pardo, Philos. Trans. R. Soc.
London, Ser. A 344, 567 共1993兲.
7
Y.-C. Yeo, T.-J. King, and C. Hu, J. Appl. Phys. 92, 7266 共2002兲.
8
W. Mönch, Appl. Phys. Lett. 88, 112116 共2006兲.
9
G. Yip, J. Qiu, W. T. Ng, and Z. H. Lu, Appl. Phys. Lett. 92, 122911
共2008兲.
10
M. G. Helander, Z. B. Wang, and Z. H. Lu, Appl. Phys. Lett. 93, 083311
共2008兲.
11
J. Bardeen, Phys. Rev. 71, 717 共1947兲.
12
V. Heine, Phys. Rev. 138, A1689 共1965兲.
13
J. Tersoff, Phys. Rev. Lett. 52, 465 共1984兲.
14
A. M. Cowley and S. M. Sze, J. Appl. Phys. 36, 3212 共1965兲.
15
W. Mönch, Phys. Rev. Lett. 58, 1260 共1987兲.
16
H. Vázquez, Y. J. Dappe, J. Ortega, and F. Flores, Appl. Surf. Sci. 254,
378 共2007兲.
17
H. Vázquez, Y. J. Dappe, J. Ortega, and F. Flores, J. Chem. Phys. 126,
144703 共2007兲.
18
S. W. Tsang, M. W. Denhoff, Y. Tao, and Z. H. Lu, Phys. Rev. B 78,
081301 共2008兲.
19
J. Robertson, J. Vac. Sci. Technol. B 18, 1785 共2000兲.
20
S. W. Tsang, Z. H. Lu, and Y. Tao, Appl. Phys. Lett. 90, 132115 共2007兲.
21
H. Vázquez, W. Gao, F. Flores, and A. Kahn, Phys. Rev. B 71, 041306
共2005兲.
22
U. Wolf, S. Barth, and H. Bassler, Appl. Phys. Lett. 75, 2035 共1999兲.
23
Z. B. Wang, M. G. Helander, and Z. H. Lu 共unpublished兲.
24
H. Vázquez, F. Flores, and A. Kahn, Org. Electron. 8, 241 共2007兲.
25
Z. B. Wang, M. G. Helander, J. Qiu, and Z. H. Lu, Appl. Phys. Lett.
共unpublished兲.
26
M. G. Helander, Z. B. Wang, and Z. H. Lu, Proc. SPIE 7051, 70510Z
共2008兲.
1
2
Downloaded 13 May 2009 to 131.175.136.39. Redistribution subject to AIP license or copyright; see http://apl.aip.org/apl/copyright.jsp