We demonstrate a self-assembly based simple method to prepare organic-organic heterobilayers on a metal substrate. By either sequential- or co-deposition of para-sexiphenyl (p-6P) and pentacene molecules onto the Cu(110) surface in ultrahigh vacuum, p-6P/pentacene/Cu(110) heterobilayer is synthesized at room temperature. The layer sequence of the heterostructure is independent of the growth scenario indicating the p-6P/pentacene/Cu(110) is a self-assembled structure with lowest energy. Besides, the bilayer shows a very high orientational ordering and is thermally stable up to 430K.

Organic-organic heterostructures are elementary building block of many optoelectronic devices, e.g., organic field effect transistors (OFETs), organic light emitting diodes (OLEDs), and organic photovoltaics devices.1–4 Consequently, the fabrication of such structures is of both fundamental and technical relevance.5–7 Molecular self-assembly is a widely used strategy for fabrication of organic nanostructures.8 It accounts for many of the most difficult steps in nanofabrication and leads to target structures which are the thermodynamically most stable ones. For instance, via strong chemical bonds of certain molecular species to surfaces, ordered monomolecular structures, so called self-assembled monolayers (SAMs) can be created.9 In this letter, we report the fabrication of a self-assembled heterobilayer, namely para-sexiphenyl/pentacene (p-6P/pentacene) on the Cu(110) surface using either sequential or co-deposition of the corresponding organic molecules at room temperature. The simple self-assembly process allows the synthesization of well ordered and robust organic-organic heterobilayers on metal substrates and thus can be applied in the fabrication of optoelectronic devices based on ultrathin organic heterojunction.

The experiments have been carried out in a UHV chamber with a base pressure of 1 × 10−10 mbar. The substrate can be heated by electron beam bombardment from the backside and cooled down to 15K with a continuous flow liquid He cryostat. The Cu(110) surface was cleaned by repeated cycles of Ar+ ion sputtering and subsequent annealing at 800K. The organic molecules, namely, para-sexiphenyl and pentacene were evaporated from a thoroughly degassed organic molecular beam epitaxy source equipped with three Knudsen cells. The distance between the evaporator and substrate is about 150mm. In all the experiments, the temperatures of the Knudsen cells containing pentacene and p-6P were kept at 450K and 530K, respectively. Under these conditions, a constant deposition rate of 1ML/1600s was achieved for both molecules. The growth is monitored in-situ using a photoelastic modulator (PEM) based RD spectrometer.10,11 The normalized difference in reflectivity defined as:12,13

\begin{eqnarray}\frac{\Delta R}{R}=2\, \frac{R_{[1\overline{1}0]}-R_{[001]}}{R_{[1\overline{1}0]}+R_{[001]}},\end{eqnarray}
ΔRR=2R[11¯0]R[001]R[11¯0]+R[001],
(1)

is recorded either in spectroscopic mode over a photon energy range between 1.5 and 5.5eV or by recording transients where the ΔR/R signal at a selected photon energy is measured as a function of time. In addition, the structure and morphology of the deposited bilayer was investigated at 15K using a home-built variable temperature scanning tunneling microscope (VT-STM).

The p-6P/pentacene heterobilayer on Cu(110) can be readily prepared by subsequent deposition of first pentacene and then p-6P on Cu(110). However, to demonstrate the self-assembly of the p-6P/pentacene heterobilayer on Cu(110), here we only show the results of deposition in the inverse sequence. To this end, one monolayer of p-6P was first deposited on clean Cu(110) and the pentacene molecules were then deposited on top of it. In order to obtain exactly one monolayer of p-6P, about 1.2 monolayer of material were deposited, initially, and the sample was then heated to 500K in order to desorb the excess p-6P molecules.14 The desorption of p-6P has been monitored by recording the change of the RD signal at 3.5eV during annealing. The p-6P monolayer prepared in this way shows a well ordered c(22×2) super structure with respect to the Cu(110) substrate.15 Because of the strong electronic interaction with the metallic substrate, the optical absorption related to the intramolecular HOMO-LUMO transition of the p-6P molecules is quenched.16 Consequently, no characteristic absorption peak of p-6P can be observed in the RD spectrum recorded at this stage (see Fig. 1(b)). We attribute the broad feature with a positive sign located at 3.5eV to transitions between “interface” states resulting from the hybridization of the Cu electronic states and the molecular orbitals of the molecules.16 In fact, this is a quite general phenomenon for organic molecules in direct contact with a metal substrate. On the other hand, it has been demonstrated that with just one monolayer of molecule underneath, the organic molecule can be electronically decoupled from the substrate and thus becomes optically active.17 On the so prepared p-6P monolayer on Cu(110), pentacene molecules were deposited and the RD signals at 3.5eV and 3.9eV were monitored during deposition (Fig. 1(a)). The photon energy of 3.5 eV corresponds to the excitation energy of the HOMO-LUMO transition in p-6P, whereas 3.9 eV is the HOMO-LUMO+1 transition energy in pentacene, the RD signals at 3.5eV and 3.9eV can thus be used as fingerprints of p-6P and pentacene molecules, respectively.16,18,19 Interestingly, as soon as the deposition of pentacene is started, the negative RD signal at 3.5 eV, which is the signature of p-6P, spontaneously increases in amplitude, whereas the RD signal at 3.9eV, i.e., the finger-print of pentacene, remains unchanged. According to the previous discussion, the absence of an RD signature at 3.9eV indicates that the deposited pentacene molecules are in direct contact with the Cu(110) substrate, whereas the increasing amplitude of the RD signal at 3.5eV suggests that the p-6P molecules originally lying right on top of the substrate have been lifted up into the layer above. From this observation we conclude that the pentacene molecules exchange with the p-6P molecules and lie directly on top of the Cu(110) surface, whereas the p-6P molecules are promoted into the second monolayer and thus contribute to the RD signal at 3.5eV. Another possible reason for the absence of an RD signal at 3.9eV, namely, that the deposited pentacene molecules would be randomly oriented on the surface, can be ruled out by the STM results discussed in detail later. This situation continues until the RD amplitude at 3.9eV starts to increase after about 1600 seconds of pentacene deposition, indicating the completion of the first pentacene monolayer. The corresponding RD spectra recorded after different steps of pentacene deposition are plotted in Fig. 1(b). Indeed, up to the completion of the first pentacene monolayer, only the RD signature of p-6P at 3.5eV grows with pentacene coverage. When the pentacene coverage exceeds one monolayer, a shoulder at 3.9eV appears which is contributed by the pentacene in excess of one monolayer. The extra pentacene molecules can be desorbed by annealing at 400K. As a result, a well ordered heterobilayer structure of p-6P/pentacene/Cu(110) is created. Due to the fact that the dipole moment of the HOMO-LUMO transition for p-6P and HOMO-LUMO+1 transition for pentacene are orientated along the long molecular axis, the negative sign of the corresponding RD signals shows that both, the p-6P and the pentacene molecules orientated preferentially aligned along the [

$1\bar{1}0$
11¯0] direction of the Cu(110) substrate. The layer inversion clearly demonstrates that the p-6P/pentacene structure is energetically favored with respect to pentacene/p-6P on the Cu(110) surface. Therefore, the p-6P/pentacene on Cu(110) is, indeed, a self-assembled heterobilayer. In fact, this bilayer structure is thermally stable up to 430K. Above this temperature, the structure will be destroyed due to the desorption of the p-6P layer.

The morphology of the p-6P/pentacene/Cu(110) bilayer was investigated by STM at a temperature of 15K (Fig. 2). The length of molecules in the top layer fits very nicely to that of p-6P (∼2.87nm) but is much longer than expected for pentacene (∼1.64nm). Therefore, the topmost layer exclusively contains p-6P molecules, confirming the inversion between pentacene and p-6P. Moreover, the p-6P molecules are well ordered with their long molecular axis inclined by ±22° out of the [1

$\bar{1}$
1¯0] direction of the substrate, leading to the two mirror domains in Fig. 2. The molecular orientation as resolved by STM is fully consistent with the negative sign of the RD signal at 3.5eV (see Fig. 1). Most interestingly, the p-6P molecules are packed in the rows which are tilted by ±10° out of the [001] direction. In fact, Müller et al.,20 have shown that a dense monolayer of pentacene on Cu(110) forms a
$\left( {6\atop \pm 1}\, {1\atop \pm 4}\right)$
6±11±4
superstructure with two molecules per unit cell. In this structure, the pentacene molecules are packed in rows inclined by exactly
$\arctan (1/4\sqrt{2})=10^{\circ }$
arctan(1/42)=10
away from the [001] direction. This straight epitaxial relationship suggests that the pentacene monolayer underneath the p-6P layer exhibits the same structure as that of a dense, uncovered monolayer of pentacene on Cu(110). In particular, the pentacene molecules should still be aligned along the [1
$\bar{1}$
1¯
0] direction of the substrate. This also confirms that the absence of a pentacene related signal in the RD-spectra is due to the actual quenching of the optical absorption rather than a loss of (orientational) anisotropy in the buried pentacene monolayer.

FIG. 1.

(a) ΔR/R signals at 3.5eV and 3.9eV recorded at room temperature during the deposition of pentacene on one monolayer p-6P covered Cu(110) as a function of deposition time. (b) Corresponding RD spectra taken at different pentacene coverages indicated in monolayer (ML) units. The growth model is sketched in the right panel.

FIG. 1.

(a) ΔR/R signals at 3.5eV and 3.9eV recorded at room temperature during the deposition of pentacene on one monolayer p-6P covered Cu(110) as a function of deposition time. (b) Corresponding RD spectra taken at different pentacene coverages indicated in monolayer (ML) units. The growth model is sketched in the right panel.

Close modal
FIG. 2.

STM images of the p-6P/pentacene heterobilayer on Cu(110). The dashed lines in the inset indicate the orientation of the p-6P rows. The images were taken at 15K with a tunneling current of 50pA and a tip bias of 2V.

FIG. 2.

STM images of the p-6P/pentacene heterobilayer on Cu(110). The dashed lines in the inset indicate the orientation of the p-6P rows. The images were taken at 15K with a tunneling current of 50pA and a tip bias of 2V.

Close modal

The same self-assembled p-6P/pentacene heterobilayer can be prepared by co-deposition of pentacene and p-6P at room temperature on the Cu(110) surface. For this purpose, the deposition rates of p-6P and pentacene were adjusted to be close to each other and the RD signals at 2.1eV, 3.5eV and 3.9eV were recorded during deposition. The RD signal at 2.1eV stems from the electronic transition between surface states of Cu(110).21 Due to its surface state origin, the RD signal at 2.1eV is extremely sensitive to surface defects and adsorbates and thus lends itself as a probe of the surface coverage of the deposited molecules.21 As one can see from Fig. 3(a), the growth can be separated into three stages: (I) The growth of the first monolayer composed of both, pentacene and p-6P molecules. The RD signal at 2.1eV decreases monotonically as a function of deposition time, indicating the increasing coverage of adsorbed molecules on the surface. At this stage, the RD signals at 3.5eV and 3.9eV change only slightly due to the strong interaction of the adsorbed molecules with the Cu substrate. The absence of molecular optical absorption indicates that both pentacene and p-6P molecules adsorb directly on top of the Cu(110) surface. (II) The formation of the p-6P/pentacene/Cu(110) bilayer structure. As soon as the surface becomes fully covered (signalized by the completely quenching of the RD signal at 2.1eV), the amplitude of the RD signal at 3.5eV starts to increase while the RD signal at 3.9eV stays constant. This observation reveals three characteristics: (a) all the pentacene molecules deposited during stage (II) diffuse into the first monolayer by exchanging with p-6P molecules; (b) the p-6P molecules being freshly deposited together with those being expelled from the first monolayer, accumulate in the second monolayer; (c) the p-6P molecules are oriented preferentially along the [

$1\bar{1}0$
11¯0] direction of the substrate. (III) The growth of excess p-6P and pentacene after completion of the p-6P/pentacene bilayer. The onset of the increase of the RD signal at 3.9eV is a sign of the completion of the pentacene/Cu(110) interface layer. All subsequently deposited molecules, will be incorporated in the upper layers where they are optically active and thus give rise to a monotonous increase of the RD amplitudes at both, 3.5 and 3.9 eV. The RD spectra for coverages close to the bilayer completion are plotted in Fig. 3(b). The RD spectrum obtained after 1800 seconds of deposition corresponds to the completed p-6P/pentacene/Cu(110) bilayer and perfectly coincides with the characteristic spectrum shown in Fig. 1(b). The spectrum recorded after 2100s deposition shows some excess p-6P and pentacene on top of the heterobilayer, which can be desorbed by heating to 400K (dashed line in Fig. 3(b)).

FIG. 3.

(a) ΔR/R signals at 2.1eV, 3.5eV and 3.9eV recorded during the co-deposition at room temperature of p-6P and pentacene on Cu(110) as a function of deposition time. (b) ΔR/R spectra recorded from the bare Cu(110) surface and after successive co-deposition of p-6P and pentacene at room temperature (solid lines). The spectrum recorded after subsequent heating to 400K is shown by the dashed line. The three growth stages are sketched in the right panel.

FIG. 3.

(a) ΔR/R signals at 2.1eV, 3.5eV and 3.9eV recorded during the co-deposition at room temperature of p-6P and pentacene on Cu(110) as a function of deposition time. (b) ΔR/R spectra recorded from the bare Cu(110) surface and after successive co-deposition of p-6P and pentacene at room temperature (solid lines). The spectrum recorded after subsequent heating to 400K is shown by the dashed line. The three growth stages are sketched in the right panel.

Close modal

In conclusion, we have shown that the p-6P/pentacene heterobilayer on Cu(110) can be fabricated using either sequential or co-deposition of p-6P and pentacene at room temperature. The self-assembly of the heterobilayer is evidenced by the fact that the same structure can be fabricated using different deposition scenarios. The p-6P/pentacene heterobilayer is thermally stable up to a temperature of 430K which is the onset of desorption of the p-6P layer. Since the structure of the p-6P layer on top depends strongly on that of the underlying pentacene layer, which can actually be controlled by the pentacene coverage,20 the structure of this heterobilayer may be tuned by varying the coverage of pentacene within the first monolayer. The method proposed in this letter is perfectly suited to prepare organic-organic heterojunctions consisting of monolayer thick active components.

This work was supported by the Austrian Science Fund FWF through project P21422. One of the authors (C.Y.L.) was supported by a scholarship from the Eurasia-Pacific Uninet.

1.
S. R.
Forrest
,
Nature
428
,
911
(
2004
).
2.
A.
Dodabalapur
,
H. E.
Katz
,
L.
Torsi
,
R. C.
Haddon
,
Science
269
,
1560
(
1995
).
3.
P. K. H.
Ho
,
J. S.
Kim
,
J. H.
Burroughes
,
H.
Becker
,
S. F. Y.
Li
,
T. M.
Brown
,
F.
Cacialll
,
R. H.
Friend
,
Nature
404
,
481
(
2000
).
4.
C. W.
Tang
,
Appl. Phys. Lett.
48
,
183
(
1986
).
5.
W.
Chen
,
H.
Huang
,
S.
Chen
,
L.
Chen
,
H. L.
Zhang
,
X. Y.
Gao
,
A. T. S.
Wee
,
Appl. Phys. Lett.
91
,
114102
(
2007
).
6.
F. F.
So
,
F. R.
Forrest
,
Phys. Rev. Lett.
66
,
2649
(
1991
).
7.
N.
Koch
,
ChemPhysChem
8
,
1438
(
2007
).
8.
G. M.
Whitesides
,
B.
Grzybowski
,
Science
295
,
29
2418
(
2002
).
9.
F.
Schreiber
,
Prog. Surf. Sci.
65
,
151
(
2000
).
10.
D. E.
Aspnes
,
J. P.
Harbison
,
A. A.
Studna
,
L. T.
Florez
,
J. Vac. Sci. Technol. A
6
,
1327
(
1988
).
11.
O.
Acher
,
B.
Drévillon
,
Rev. Sci. Instrum.
63
,
5332
(
1992
).
12.
P.
Weightman
,
D. S.
Martin
,
R. J.
Cole
,
T.
Farrell
,
Rep. Prog. Phys.
68
,
1251
(
2005
).
13.
C.
Goletti
,
G.
Bussetti
,
P.
Chiaradia
,
A.
Sassella
and
A.
Borghesi
,
Org. Electron.
5
,
73
(
2004
).
14.
S.
Müllegger
,
A.
Winkler
,
Surf. Sci.
600
,
1290
(
2006
).
15.
S.
Berkebile
,
T.
Ules
,
P.
Puschnig
,
L.
Romaner
,
G.
Koller
,
A. J.
Fleming
,
K.
Emtsev
,
T.
Seyller
,
C.
Ambrosch-Draxl
,
F. P.
Netzer
,
M. G.
Ramsey
,
Phys.Chem.Chem.Phys.
DOI: (
2011
).
16.
Y.
Hu
,
K.
Maschek
,
L. D.
Sun
,
M.
Hohage
,
P.
Zeppenfeld
,
Surf. Sci.
600
,
762
(
2006
).
17.
R.
Forker
,
D.
Kasemann
,
T.
Dienel
,
C.
Wagner
,
R.
Franke
,
K.
Müllen
,
T.
Fritz
,
Adv. Mater.
20
,
4450
(
2008
).
18.
T. M.
Halasinski
,
D. M.
Hudgins
,
F.
Salama
,
L. J.
Allamandola
,
T.
Bally
,
J. Phys. Chem. A
104
,
7484
(
2000
).
19.
P.
Sony
,
A.
Shukla
,
Phys. Rev. B
75
,
155208
(
2007
).
20.
K.
Müller
,
A.
Kara
,
T. K.
Kim
,
R.
Bertschinger
,
A.
Scheybal
,
J.
Osterwalder
,
T. A.
Jung
,
Phys. Rev. B
79
,
245421
(
2009
).
21.
L. D.
Sun
,
M.
Hohage
,
P.
Zeppenfeld
,
RE.
Balderas-Navarro
,
Surf. Sci.
589
,
153
(
2005
).
This article is distributed under a Creative Commons Attribution 3.0 Unported License.