286
J. Phys. Chem. C 2007, 111, 286-293
Pentacene Grown on Self-Assembled Monolayer: Adsorption Energy, Interface Dipole, and Electronic Properties Aloke Kanjilal,*,† Luca Ottaviano,‡ Valeria Di Castro,§ Marco Beccari,§ Maria Grazia Betti,|,⊥ and Carlo Mariani†,| INFM-CNR Center on nanoStructures and bioSystems at Surfaces (S3), Via G. Campi 231/A, I-41100 Modena, Italy, Dipartimento di Fisica, UniVersita` degli Studi dell’Aquila, Via Vetoio, I-67010 Coppito, L’Aquila, Italy, Dipartimento di Chimica, UniVersita` di Roma “La Sapienza”, Piazzale Aldo Moro 2, I-00185 Roma, Italy, Dipartimento di Fisica, UniVersita` di Roma “La Sapienza”, Piazzale Aldo Moro 2, I-00185 Roma, Italy, and INFM-CNR CRS-SOFT, Piazzale Aldo Moro 2, I-00185 Roma, Italy ReceiVed: August 4, 2006; In Final Form: October 5, 2006
The energetic, the morphology, the interface dipole formation, and the electronic states of pentacene grown on a self-assembled monolayer (SAM) of benzenethiolate on the Cu(100) surface are investigated by complementary structural and spectroscopic techniques. The growth morphology of the heterostructure has been investigated by atomic force microscopy (AFM), low-energy electron diffraction (LEED), Auger electron spectroscopy (AES), and photoemission spectroscopy, inferring tightly packed arrangement of grains constituted of pentacene molecules with nearly perpendicular orientation, when a single-layer is deposited on top of the benzenethiolate-SAM. The adsorption energy of the pentacene single layer on the benzenethiolate buffer layer (Ea ) 1.16 eV ≡ 112 kJ/mol) is found much weaker than for pentacene on the copper surface, as estimated by thermal desorption spectroscopy (TDS). The evolution of the spectral density of electronic states in the valence band, obtained by high-resolution ultraviolet photoelectron spectroscopy (HR-UPS), confirms the weak interaction of the pentacene molecules with the benzenethiolate-SAM and the formation of a semiconducting heterostructure, with a hole injection barrier reduced to 0.95 eV with respect to the pentacene/ Cu interface.
Introduction Aromatic π-conjugated molecules have recently been employed in prototype hybrid devices and are found quite efficient for applying in molecular electronics. Pentacene (C22H14), having π-conjugated electrons delocalized at five benzene rings, has widely been used in organic thin film transistors (OTFTs) for exhibiting the highest field-effect mobility (µ) at room temperature (300 K).1-5 Although the mobility of the organic systems is lower with respect to the inorganic semiconductors, the efficiency can be controlled by adding or removing specific functional groups during creation of new molecular architectures with suitable transport properties. The structure prepared by a self-assembling process, with large long-range ordered domains, gives promising advantages and suitable alternatives to siliconbased electronics.1,2,6,7 Generally, OTFTs require high-operating voltages because of the coupling with the gate dielectric layers. Recently, a high-mobility and low-voltage organic-TFT has been designed, with a heterostructure constituted of pentacene deposited on a buffer layer of a phenyl-terminated molecule,8 where the self-assembled monolayer has been used successfully as a dielectric layer instead of thin oxide for reducing the operating voltages.8 Since charge transport in OTFTs occurs within the first few layers of molecules in contact with the dielectric,7,9,10 it is crucial * Author to whom correspondence should be addressed. Phone: +39 06 49914281; fax: +39 06 4957697; e-mail:
[email protected]. † INFM-CNR Center on nanoStructures and bioSystems at Surfaces (S3). ‡ Universita ` degli Studi dell’Aquila. § Dipartimento di Chimica, Universita ` di Roma “La Sapienza”. | Dipartimento di Fisica, Universita ` di Roma “La Sapienza”. ⊥ INFM-CNR CRS-SOFT.
to understand the structural and the electronic properties at the interface.11-13 Furthermore, the Schottky barrier can be tuned by incorporating self-assembled monolayers (SAMs) with orientated dipoles between the metal and the organic layer.14 However, the thermodynamic equilibrium for the SAM with alkyl chains remains doubtful because of the strong insulating nature of the alkyl chains.14,15 On the other hand, benzenethiol (C6H5SH, see Figure 1a), the simplest aromatic thiol consisting of a single phenyl group, can be used for preparing a suitable buffer layer: it presents more delocalized charge compared to other aromatic thiols,16 and it can be firmly anchored at the metal surface with the S-head group after deprotonation.17,18 Benzenethiolate (C6H5S, Bt in the following)-SAM can be used as a buffer layer for preventing chemical interactions between the organic material and the conducting substrate, balancing the carrier (holes or electrons) injection for achieving improved recombination efficiency.9 The device properties can be optimized by improving the morphology and the crystal structure of the first few monolayers (MLs).5,10-13 Furthermore, the charge carrier mobility in pentacene thin films is presumed to be severely affected by defects/traps at grain boundaries,3,5,8 which demands fabrication of good-quality epitaxial thin films.3,19 Recent electronic structure calculation12 for the growth of the pentacene molecules on the benzenethiol-covered Au surface has revealed that the formation of shallow charge traps with energies e100 meV above (below) the valence band maximum (conduction band minimum) depends on sliding of the pentacene molecules along their long molecular axis. Pentacene molecular ordering at the inorganic substrate depends on a subtle balance between the molecule-molecule
10.1021/jp065026c CCC: $37.00 © 2007 American Chemical Society Published on Web 12/02/2006
Pentacene Grown on Self-Assembled Monolayer
J. Phys. Chem. C, Vol. 111, No. 1, 2007 287 Experimental Section
Figure 1. Schematic sketch of (a) S-H bond breaking for benzenethiol (C6H5SH) or S-S bond breaking for difenyl-disulphide molecule (C6H5S-C6H5S), during growth of the benzenethiolate (C6H5S)-SAM through S bonding to the substrate; (b) pentacene (C22H14) adsorption on the benzenethiolate-SAM.
and molecule-substrate interactions.8,12,20 A flat-lying orientation is favored by the interaction with the substrate,3,20-29 while a vertical molecular alignment is observed7,9,12,30 when the molecule-molecule interaction prevails. Standing-up pentacene has recently been obtained by using aromatic thiolate buffer layers on Au(111)4,12 as well as on Cu(100)31 surfaces. The Cu(100) surface is advantageous, as it provides a highly symmetric 4-fold chemisorption site for S-headgroup mediated molecular adsorption32,33 and constitutes a suitable template for the chemical binding site of the thiolate-SAM. In the present work, we investigate the growth mechanism, energetic, interface dipole layer, and the electronic properties of pentacene grown on a buffer layer of Bt-SAM, adsorbed on the Cu(100) substrate in ultrahigh vacuum (UHV), by means of combined structural and spectroscopic techniques. The heterostructure growth morphology and the surface topography are investigated by Auger electron spectroscopy (AES) and atomic force microscopy (AFM), respectively. The adsorption energy of the pentacene single layer to the benzenethiolate SAM is estimated by thermal desorption spectroscopy (TDS), and the evolution of the interface dipole and of the electronic spectral density in the valence band are investigated by high-resolution ultraviolet photoelectron spectroscopy (HR-UPS).
The photoemission experiments were performed in situ at the surface physics laboratory LOTUS, Dipartimento di Fisica, Universita` di Roma “La Sapienza”, in a UHV chamber equipped with angular-resolved high-resolution UPS apparatus, lowenergy electron diffraction (LEED), AES, a quadrupole mass spectrometer, and ancillary facilities for sample preparation. UPS data were acquired using a high-intensity He discharge lamp (He IR photons, hν ) 21.218 eV), with 45° incidence angle, and the photoemitted electrons were analyzed in the plane of incidence with a hemispherical SCIENTA SES-200 analyzer, used with energy resolution of 15 meV, by keeping the integration angle of (8° with respect to the direction of normal emission, as described elsewhere.34 All the spectra in the present study were collected along the normal emission geometry. Calibration of the binding energy (BE) scale with respect to the measured kinetic energy was made for the adsorbed systems, using the Cu Fermi energy edge. The high-binding-energy cutoff in the present UPS study was also registered while the sample was biased at -8.8 V for estimating the work function (WF) variation irrespective of the analyzer WF. Another set of samples was prepared in a second UHV chamber equipped with LEED, AES, ancillary facilities for sample preparation, and a TDS measurement setup. We did not observe significant electron beam damage of the organic layers during AES/LEED measurements with the typical exposure time (about 1 min) of a spectrum/pattern. To ensure reliability and reproducibility, we have dedicated several independent experimental runs to the different techniques. The photoemission data for gas-phase pentacene were taken at the gas-phase end station35 of the ELETTRA Synchrotron Radiation Laboratory (Trieste, Italy). We used a hemispherical electron spectrometer mounted at magic angle with respect to the polarization plane of the radiation (54.7°) and perpendicular to the directions of the photon beam and of the high-temperature molecular beam. The overall energy resolution was about 0.3 eV. The ionization potential energy was calibrated by using known gas sources. High-purity (99.9999%) Cu(100) single crystals were cleaned by repeated sputtering-annealing cycles consisting of sputtering with Ar+ ions (1 keV, 11.5 µA for 45 min, followed by 600 eV, 8 µA for 30 min) and annealing at 675 K for 10 min. In the present experiments, both liquid Bt and diphenyl-disulphide (C6H5S-SH5C6, DPDS) powder (purity 99%, supplied by Sigma Aldrich) were used, kept in two UHV-sealed pyrex glass ampoules. They were connected separately in the two UHV chambers through leak valves.32 The DPDS molecule was purified by heating-pumping cycles at about 345 K. Prior to the benzenethiol or DPDS exposure on the clean Cu(100) surface, the purity of the dosing vapor was verified by means of the quadrupole mass spectrometer. The vapor pressure of liquid benzenethiol was reported 1.3 mbar at 300 K, while for DPDS it is about 0.13 mbar at 365 K.36 The DPDS molecules (melting temperature is about 335 K)37 do not decompose in gas phase, but the S-S bond breaks when they interact with the Cu surface, building up a benzenethiolate layer with S attached to the substrate (schematically depicted in Figure 1a), as already observed for several disulfides adsorbed on Cu surfaces.38 In the case of thiols, the formation of the thiolateSAM involves spontaneous decomposition of S-H bonds during interaction of the molecules with the d-band metal surfaces.39,40 The benzenethiolate SAM was grown on the Cu(100) surface with a constant exposure rate ranging from 0.2 to 0.45 L/min at 300 K, at 5 × 10-9 mbar (with 1 L ) 1.33 × 10-6
288 J. Phys. Chem. C, Vol. 111, No. 1, 2007
Kanjilal et al.
Figure 2. AES peak intensity evolution as a function of exposure to benzenethiol (left panel) and subsequent pentacene coverage (right panel). Relevant Auger peaks: Cu-MVV at 60 eV (open squares), S-LMM at 152 eV (open circles), and C-KLL at 272 eV (open triangles).
mbar × s). The saturation coverage was reached for 10 L exposure of benzenethiol and 33 L exposure of DPDS; in both cases, identical UPS spectra and LEED patterns were observed. Pentacene was kept in resistively heated quartz crucibles, mounted in UHV evaporators. Pentacene was evaporated in situ on the benzenethiolate-Cu(100) system at 300 K (see the schematic sketch in Figure 1b), keeping the crucible temperature at about 460 K for achieving constant evaporation rate and reproducibility. The evaporation rate (1.4 nm/min) was monitored by a quartz microbalance, keeping the pressure in the UHV chamber in the 10-10 mbar range. The surface morphology of the present heterostructure was investigated ex situ at 300 K, by means of a Digital Dimension D5000 AFM equipped with Nanoscope IV controller at L’Aquila. The system was operated in tapping mode using commercial silicon nitride cantilevers (resonance frequency range ) 200300 kHz) with a scan rate of 1.5 Hz. Results and Discussion A. Growth Morphology. The growth morphology of the pentacene/Bt-SAM heterostructure on Cu(100) has been investigated following the evolution of the Auger line intensity and the variation of the work function at the interface. Upon increasing exposure of benzenethiol on the clean Cu(100) surface at room temperature (RT), a distinct variation in the intensity of the AES transitions for the Cu-MVV (60 eV), S-LMM (152 eV), and C-KLL (272 eV) peaks is observed, as reported in the left part of Figure 2. The S-LMM and C-KVV Auger intensities increase up to saturation (about 10 L), with concomitant decrease of the Cu-MVV intensity. The observed changes in the AES peak intensities imply the formation of a complete benzenethiolate layer on the Cu(100) surface at about 10 L, in agreement with previous photoemission results.32 After pentacene deposition on the Bt buffer layer (Figure 2, right panel), the S and Cu Auger peak intensities are found descending monotonically, while the C peak intensity grows, up to 20 Å of pentacene. At that thickness, a single layer of pentacene covers the SAM surface. The formation of the pentacene/Bt-SAM heterostructure adsorbed on Cu(100) at RT is also followed by analyzing the work function change as a function of thiol and subsequent pentacene deposition (Figure 3). The work function of the BtSAM at saturation coverage on Cu(100) lowers by 1.02 ( 0.02 eV, signifying a large dipolar change. Basically, the observed decrease in WF can be attributed to the competing effects of
Figure 3. UV photoemission spectra of the high-binding-energy cutoff for the clean Cu(100) surface, the benzenethiolate at saturation coverage, and as a function of pentacene coverage (from 6 to 50 Å). The cutoffenergy variation measures the work function change.
the presence of the dipolar molecules and of the charge transfer between the molecule and the copper substrate.41-43 As the S-C axis of the benzenethiolate layer is found almost vertical to the Cu(100) surface,44 the molecular dipole directed along the benzenethiolate S-C axis (S to C direction) plays the dominant role by compensating the expected increase of the WF associated to the adsorption of thiolate species on Cu surfaces.41,42,45,46 On the other hand, soon after the first deposition of pentacene, the cutoff is slightly shifted toward low-binding energies (Figure 3), which is a sign of an increase of WF: it reaches +0.18 ( 0.04 eV with respect to the Bt-SAM, at about 20 Å of pentacene, and it remains stable at higher thickness. The observed change in WF with increasing pentacene coverage indicates that almost a complete single layer (SL) of pentacene is completed at about 20 Å on the underneath Bt-SAM, in agreement with the previously discussed Auger peak evolution. Since the pentacene molecules do not have a dipole moment,47,48 the small increase of WF upon pentacene deposition on the Bt-SAM is the evidence of an induced dipole located at the pentacene/Bt-SAM interface, oppositely oriented to the Bt-SAM/Cu(100) interface dipole. The surface morphology of the SL pentacene deposited on the benzenethiolate buffer layer is examined ex situ by means of AFM, as displayed in Figure 4a (scanned area 200 nm × 200 nm). In the AFM image, the large flat regions represent the tightly packed pentacene molecules in a SL, with random defects (dark areas, marked by arrows), in agreement with our previous results. To quantitatively evaluate the surface morphology of the pentacene layer, an analysis of the AFM root-meansquare (rms) roughness is carried out: analyzing the profile signal between the points “A” and “B” crossing a defects site (Figure 4b), the average estimated pentacene layer height is found consistent with the expected height of a SL of standingon pentacene molecules on top of the standing-up Bt-SAM,49 as schematically drawn in Figure 1b. The topography obtained from the AFM data is in agreement with previous results of pentacene grown on a Bt-covered Au(111) surface, where the vertical orientation of the molecules has been deduced by scanning tunneling microscopy (STM).12 Hence, the presence of the benzenethiolate buffer layer allows preparation of a rather smooth, single layer of standing-up pentacene molecules where
Pentacene Grown on Self-Assembled Monolayer
J. Phys. Chem. C, Vol. 111, No. 1, 2007 289
Figure 4. (a) AFM image (scan area 200 nm × 200 nm) of a 20 Å thick pentacene layer deposited on the benzenethiolate buffer layer; the dark spots (indicated by white arrows) correspond to defects. (b) AFM height profile along the A-B trace drawn in a.
small height fluctuations on a wide flat area are associated to vertical sliding of the molecules.12 B. Long-Range Ordering. The long-range ordering of the pentacene/Bt-SAM/Cu(100) system has been followed by means of LEED at different steps of the interface formation. The diffraction patterns of clean Cu(100), Bt-SAM/Cu(100), and pentacene/Bt-SAM/Cu(100) are displayed in Figure 5. First, we analyze the highly ordered structure of the benzenethiolate saturation layer: the sharp (1 × 1) spots of the clean Cu(100) surface (Figure 5a) transform into a well-ordered commensurate two-domain c(2 × 6) phase (Figure 5b) after Bt-SAM chemisorption at RT.31 Because of the 4-fold symmetry of this substrate, a two-domain structure is present, with the unit cells rotated by 90° to each other, and the LEED pattern is the result of the superposition of the reciprocal nets associated to the c(2 × 6) domains. The c(2 × 6) reconstruction pattern is evidenced by a spot profile along traces 1 and 2 (Figure 5a, b), as reported in Figure 5d and e, for the clean Cu(100) and Bt-SAM systems,
respectively: in this way, we can enlighten the appearance of the (-(1/6)(1/2)) and ((1/6)(1/2)) extra spots between the ((1/2)(1/2)) and (-(1/2) -(1/2)) peaks in the Bt/Cu(100) interface. The spots in the commensurate c(2 × 6) diffraction pattern appear to be streaky, while those corresponding to the c(2 × 2)-symmetry are brighter, and this may be due to the c(2 × 6) reconstruction interpreted as a buckled c(2 × 2) structure, because of restructuring of the underlying Cu(100) surface, as proposed for methanethiolate adsorption on the same Cu(100) surface.33 The first pentacene deposition (3 Å) on the benzenethiolate surface induces a sharpening of the half-order spots of the c(2 × 6) phase (Figure 5c), maintaining the same reconstruction symmetry of the Bt-SAM. Thus, the pentacene molecules order according to the underlying buffer SAM commensurate with the Cu(100) surface. The highly ordered structure slowly reduces its long-range order, and the diffraction spots become more diffuse upon increasing pentacene coverage, from the completed
Figure 5. LEED patterns of (a) (1 × 1) clean Cu(100) surface (electron beam energy Ep ) 45 eV); (b) two-domain c(2 × 6) benzenethiolate buffer layer (Ep ) 37 eV); and (c) two-domain c(2 × 6) 3 Å pentacene/Bt-SAM/Cu(100) system (Ep ) 37 eV). Traces 1 and 2 in a and b represent the directions along which the spot intensity profiles are investigated and are shown in d and e. Only integer-order spots are visible for the clean Cu(100) surface, namely, the (10) and (01) (d), while new half-order spots appear as a consequence of the reconstruction (e). All interfaces prepared at RT, and LEED patterns recorded at 90 K.
290 J. Phys. Chem. C, Vol. 111, No. 1, 2007 SL to the thin film. As the pentacene molecules adopt the standing-up alignment on the benzenethiolate surface from the early stage of pentacene deposition, and since the AFM image shows formation of flat and smooth grains, the gradual disappearance of the well-ordered LEED pattern can be associated to the formation of azimuthally randomly oriented domains within each wide area grain observed in the AFM image. C. Adsorption Energetic and Electronic Properties. The thermal stability of the pentacene/benzenethiolate-SAM/Cu(100) system has been studied by thermal desorption spectroscopy. The TD spectra were recorded at a constant heating rate (β) of 0.05 K/s on the three different systems, namely, a 20 Å thick pentacene deposited at RT on the SAM (pentacene/Bt-SAM/ Cu(100), middle panel of Figure 6), the benzenethiolate-SAM (Bt-SAM/Cu(100), top panel of Figure 6), and a 20 Å thick pentacene layer on bare Cu(100) (pentacene/Cu(100), lower panel of Figure 6). We follow the main desorption fragment (benzene at m/z ) 78), and the whole benzenethiol molecule at m/z ) 110, for analyzing the benzenethiol desorption. However, the main desorption component of the RT-grown Bt-SAM/Cu(100) and pentacene/Bt-SAM/Cu(100) systems is found to be benzene, while the benzenethiol desorption as a whole molecule is negligible (thus not reported in the figure), confirming the C-S bond breaking observed previously32 and reported also for other Bt-SAMs on metallic substrates.39,50 The maximum in the benzene desorption spectrum is centered at 472 ( 15 K (R1), as shown in the top panel of Figure 6. Two components, benzene (m/z ) 78) and the most intense fragment of pentacene (m/z ) 139), have been considered during the desorption of the pentacene heterostructures. In the TD spectra collected for m/z ) 78 from the pentacene/Bt-SAM/ Cu(100) system (Figure 6, middle panel), a broad benzene desorption peak centered at 478 ( 15 K is discernible, which is broader than for the pure Bt-SAM/Cu(100) system. Apart
Kanjilal et al. from this broadening, the shape of the TD spectra for the benzene fragment and the maximum peak position for pentacene/Bt-SAM/Cu(100) and Bt-SAM/Cu(100) are very similar, suggesting that the Bt/Cu system is not significantly influenced by the top pentacene layer. The desorption of pentacene is followed by analyzing the molecular fragment m/z ) 139, as the other pentacene-derived fragments produce very low signals. Two demarcating peaks are observed in the TD spectra for both the pentacene/Cu(100) and the pentacene/Bt-SAM/Cu(100) systems, as shown in the lower and middle panels of Figure 6, respectively. Considering the pure pentacene/Cu(100) interface (lower panel of Figure 6), the lower temperature desorption peak (β1) at 402 ( 15 K can be assigned to desorption of the physisorbed molecules, while the higher temperature desorption peak (β2) at 623 ( 15 K can be ascribed to the desorption of flat-lying pentacene molecules adsorbed on the Cu(100) surface (lower panel of Figure 6). In fact, according to a recent STM experiment,20 the pentacene molecules adsorb in a flat-lying geometry on the Cu(100) surface because of the molecule interaction with the metal substrate, and the whole surface can be paved by depositing a single layer of pentacene, corresponding to about 4 × 1013 molecules/cm2.20 On the basis of these findings, it is expected that a first planar layer is formed, and weakly bonded molecules grow in the 20 Å thick pentacene multilayer adsorbed on the bare Cu(100) surface. The pentacene/Bt/Cu(100) heterostructure (middle panel of Figure 6) gives rise to two broader and shifted thermal desorption peaks, with respect to the pentacene/Cu(100) system, the first peak at 383 ( 15 K, and the second desorption peak at 566 ( 15 K. The first desorption peak, as well, can be attributed to physisorbed pentacene stacked with other pentacene molecules and weakly bonded to the benzene rings of the underneath benzenethiolate layer, while the second desorption peak may be related with the interaction of pentacene molecules with the bare metal-sulfide surface, after the decomposition of the benzenethiolate SAM. Assuming a first-order process, the desorption temperatures measured in these TDS experiments can provide the activation energies, which have been numerically derived with an iterative procedure, from the following equation:51,52
Ea ) kB‚Td2
Figure 6. Thermal desorption (TD) spectra from the benzenethiolate (Bt) layer at saturation coverage on Cu(100) surface (top panel), for the 20 Å pentacene/Bt/Cu(100) heterostructure (middle panel), and for the 20 Å pentacene/Cu(100) system (lower panel). Relevant molecular fragments revealed benzene (m/z ) 78), pentacene fragment (m/z ) 139). All TD spectra are recorded at a constant heating rate of 0.05 K/s.
() ( ) Ea ν exp β kBTd
(1)
where kB is the Boltzmann constant, β is the heating rate (0.05 K/s), and ν is a pre-exponential frequency factor (ν ) 1013 s-1).19,53 The desorption temperatures and estimated activation energies are reported in Table 1. The decomposition process of the Bt-SAM/Cu(100) system gives rise to a breaking of the radical (R)-S bond of the thiolateSAM on Cu(100), and the activation energy of the bonding fragmentation approximately corresponds to the R-S bond enthalpy. The temperature of the maximum signal in the desorption spectrum of the benzenethiolate-Cu(100) system at saturation coverage is 472 K, and the corresponding activation energy is 1.44 ( 0.05 eV (139 ( 5 kJ/mol). The activation energy of Bt/Cu(100) system does not change much when a single layer of pentacene is deposited on the Bt-SAM, and the broader benzene desorption peak (roughly centered at 478 K) corresponds to an activation energy of 1.46 ( 0.05 eV (141 ( 5 kJ/mol) for the pentacene/Bt/Cu(100) heterostructure. The desorption energy of pentacene molecules physisorbed on the bare Cu(100) surface, derived from the 402 K desorption
Pentacene Grown on Self-Assembled Monolayer
J. Phys. Chem. C, Vol. 111, No. 1, 2007 291
TABLE 1: Measured Maximum Desorption Temperature (Td) and Calculated Activation Energy (Ea) for the Benzenethiolate/ Cu(100), Pentacene/Benzenethiolate/Cu(100), and Pentacene/Cu(100) Systemsa benzene (m/z ) 78) interface
pentacene (m/z ) 139)
Td
Ea
C6H5S/Cu(100)
472 (15 (K)
pentacene/C6H5S/Cu(100)
478 ( 15 (K)
1.44 ( 0.05 (eV) 139 ( 5 (kJ/mol) 1.46 ( 0.05 (eV) 141 ( 5 (kJ/mol)
pentacene/Cu(100)
Td
Ea
383 ( 15 (K)
1.16 ( 0.05 (eV) 112 ( 5 (kJ/mol) 1.74 ( 0.05 (eV) 168 ( 5 (kJ/mol) 1.22 ( 0.05 (eV) 118 ( 5 (kJ/mol) 1.92 ( 0.05 (eV) 185 ( 5 (kJ/mol)
566 ( 15 (K) 402 ( 15 (K) 623 ( 15 (K)
a
The molecular fragments m/z ) 78 and m/z ) 139 are basically associated to the Bt and pentacene molecular desorption, respectively (see text).
temperature, is 1.22 ( 0.05 eV (118 ( 5 kJ/mol), while the second peak at about 623 K gives an activation energy of 1.92 ( 0.05 eV (185 ( 5 kJ/mol), because of pentacene molecules directly bonded to Cu(100), after desorption of the weakly bonded physisorbed layer. These values are in fairly good agreement with the values found for pentacene adsorption on the Cu(119) surface.54 The (119) surface is a vicinal direction to the (001) surface, and it is constituted of 1-nm-wide terraces of (001) orientation. On the vicinal surface, we recently obtained activation energies (1.27 eV for the physisorbed layer, and 2.09 eV for the first planar adsorbed molecules) about 5% higher than on the completely flat Cu(100) substrate, and this is compatible with the natural surface corrugation at the steps, slightly increasing the potential energy felt by the adsorbed molecules. The two thermal desorption peaks observed for the pentacene/ Bt-SAM/Cu(100) system at 383 and 566 K for m/z ) 139 correspond to activation energies of 1.16 ( 0.05 eV (112 ( 5 kJ/mol) and 1.74 ( 0.05 eV (168 ( 5 kJ/mol), respectively. The first activation energy value can be ascribed to a physisorbed pentacene single layer weakly bonded to the benzene rings of the underneath benzenethiolate layer, and it is comparable to the desorption temperature of the pentacene film physisorbed on the bare Cu(100) surface. The appearance of
Figure 7. (left panel) HR-UPS valence band spectra (HeIR photon energy, 21.218 eV) for the clean Cu(100) surface (bottom spectrum), for the Bt-SAM (bottom but one), and for the pentacene/Bt/Cu(100) system as a function of pentacene coverage. (right panel) Same HRUPS data, zoomed in the low-binding-energy region, also compared to pentacene gas-phase data (top spectrum), aligned to the interface data by shifting the ionization energy scale by -4.93 eV.
the second peak at higher desorption temperature may be related with the interaction of pentacene molecules with the S-covered metal surface. We suppose that during the decomposition of the benzenethiolate-SAM, a fraction of the pentacene molecules interact with the remaining copper-sulfide layer, giving rise to pentacene bonded to Cu(100). Further information on the electronic properties and on the growth morphology of the pentacene/Bt-SAM/Cu(100) system can be achieved by studying the electronic states by HR-UPS. A set of valence band spectra recorded for the clean Cu(100) surface, Bt-SAM/Cu(100) system at saturation coverage of BtSAM, and for the pentacene/Bt-SAM/Cu(100) heterostructure with increasing pentacene thickness is shown in Figure 7. The Bt-SAM presents several molecular levels and interaction states, discussed in detail elsewhere.32 In particular, we observe BtCu(100) bonding and antibonding states laying across the Cu d-band energy region (2-4 eV BE), a Cu-S bonding state at 2.23 eV BE (I2), and the Bt highest-occupied molecular orbital (HOMO) at 1.59 eV BE (peak I1). After pentacene deposition, a significant number of new features progressively evolve, followed by gradual disappearance of the peaks associated with the underlying Bt-SAM/Cu(100) system. The electronic spectral density of states can be basically ascribed to pentacene molecular orbitals (MOs, labeled A-G) that can be assembled into three energy ranges characterized by the orbital symmetry, namely, π (1-5 eV), π + σ (5-7 eV), and σ (7-13 eV), respectively. The molecular states binding energies are in good agreement with the states observed on pentacene thin films grown on various substrates governed by van der Waals interaction28 and with a direct correspondence to the gas-phase levels.55,56 The energy region of the π-symmetry states is enlarged in the right panel of Figure 7, and the heterostructure data are reported together with photoemission data from pentacene in the gas phase, for comparison. The gas-phase ionization energy scale is related to the BE scale by aligning the high-energy σ-orbitals to band F, a purely σ-like structure in the high BE region of the thin film. In the pentacene/Bt-SAM/Cu(100) system, we observe the HOMO peak at 1.47 eV (peak 1, 3b2g level), and we can attribute the following structure to the other π orbitals: 2au (peak 2) at 2.69 eV, 3b1g (peak 3) at 3.19 eV, 2b2g (peak 4) at 3.86 eV, 3b3u (peak 5) at 4.19 eV, and 1au (peak 6) at 4.43 eV, by comparison with the gas-phase spectrum, as reported in Table 2. Only a light energy shift (e0.3 eV) of the 1au, 2au, and 2b2g levels (peaks 6, 2, and 4, respectively) is observed, with respect to the corresponding MOs in the gas-phase spectrum, attributed to the molecular condensation on the substrate. The absence of peak shifting for the other pentacene related structure confirms the weak interaction between the pentacene molecules in the single layer and the aromatic rings
292 J. Phys. Chem. C, Vol. 111, No. 1, 2007
Kanjilal et al.
TABLE 2: Electronic State Binding Energies for the Pentacene/Benzenethiolate/Cu(100) Interface As Compared to Pentacene Gas-Phase Dataa pentacene/Bt/Cu (100) BE (eV)
orbital number, character
pentacene gas phase (eV)
1.47 2.69 3.19 3.86 4.19 4.43 5.79 6.59 7.86 8.71 9.94 11.17 12.36
1, 3b2g (HOMO) 2, 2au 3, 3b1g 4, 2b2g 5, 3b3u 6, 1au A B C D E F G
1.47 2.79 3.21 3.91 4.28 4.73 5.82 6.57 7.23 8.29 9.96 11.17
a The ionization energy scale of the latter data is referred to the solidstate data BE by aligning the higher-lying σ-like F-band (-4.93 eV energy shift).
of the underlying benzenethiolate layer. It is worth noticing a skewed shape of the HOMO band for the pentacene thin film toward higher BE. This double HOMO structure may be attributed to the energy band dispersion depending on the intermolecular π-π interaction in the randomly oriented pentacene crystallites, as recently suggested for layers of standingup pentacene molecules prepared on weakly interacting substrates.57 However, we may also suggest the presence of surface effects, because of the slightly different energy expected for the HOMO state associated to the inner with respect to the outer molecules in the thin film.58 In the low pentacene coverage phase, we observe progressive quenching of the Cu-S I2 bonding state at 2.23 eV,32 along with the disappearance of the electronic spectral density at the Fermi energy (EF), indicating that the pentacene/Bt-SAM/ Cu(100) system is semiconducting. An estimation of the I2 peak height, of the HOMO intensity, of the intensity of the spectral signal at EF, and of the intensity of band E (characteristic σ-band of pentacene), allows us to obtain information on the growth morphology. The evolution of their normalized intensity is reported in Figure 8 as a function of pentacene thickness on the Bt-SAM/Cu(100) substrate: the intensities of the band E and of the HOMO peak progressively increase as a function of thickness, saturating at about 20 Å. Conversely, the intensities of the Cu-S bonding state and of the spectral density at the Fermi edge are monotonically decreasing with increasing
Figure 8. Intensity evolution of photoemission features from the pentacene/Bt/Cu(100) system, as a function of pentacene coverage: electronic spectral density at the Fermi energy (filled circles), Cu-S I2 electronic state (filled squares), pentacene-induced molecular band E (open circles), and HOMO level (open squares).
coverage and are completely quenched at 20 Å. Thus, 20 Å pentacene is the approximate coverage required to pave completely the underneath Bt-SAM, in agreement with the AES and AFM results discussed in the previous sections. When the single pentacene layer is completed, the heterostructure presents a transition to a semiconducting phase, as no electronic spectral density is present at the Fermi energy. In these conditions, the onset of the HOMO state below EF represents the hole injection barrier for the heterostructure that we estimate at 0.95 eV. Comparing the hole injection barrier of a pentacene SL deposited on this molecular buffer layer, with that of pentacene thin films on metal substrates, we notice an energy reduction: in particular, pentacene deposited on Cu(119) gives rise to a hole injection barrier of 1.05 eV,58 and an interface dipole of 0.85 eV, in agreement with the model proposed for aromatic organic molecules-metal interfaces.59 The model is not straightforward to apply for the present system, where pentacene induces a slight increase of the work function and of the hole injection barrier, because the Bt-SAM buffer layer constitutes a highly polarizable medium. Finally, we studied the temperature-induced modification of the electronic structure of a 60 Å thick pentacene film deposited on the benzenethiolate layer, as shown in Figure 9. Starting from RT, a continuous shift (0.4 eV) in the valence band structures with increasing annealing temperature is detected, up to about 355 K, with no significant change in the peaks line shape. The HOMO energy shift indicates a reduction of the hole-injection barrier from 0.95 to 0.54 eV (see inset to Figure 9), probably because of defects within the layer acting as an unintentional
Figure 9. HR-UPS data from a 60 Å thick pentacene film deposited on the benzenethiolate buffer layer on Cu(100) as a function of annealing temperature. The vertical dotted line is only a guide for the eye. The inset shows the HOMO-onset (hole-injection barrier) energy shifting as a function of annealing temperature.
Pentacene Grown on Self-Assembled Monolayer hole doping. The valence band structure starts modifying above 365 K, suggesting beginning of pentacene desorption; this value is in agreement with the onset of the thermal desorption peak of physisorbed molecular pentacene, reported in Figure 6. Further rise in the annealing temperature leads to a complete modification of the valence band states because of the desorption of the entire pentacene layer and breaking of the underlying benzenethiolate buffer layer, so that the valence band after annealing at 425 K reflects the sulfide/Cu interface.32 Conclusions We prepare a molecular heterostructure constituted by standing-up pentacene molecules on top of a highly ordered benzenethiolate-SAM grown on Cu(100) surface. The pentacene molecules have a low adsorption energy (1.16 eV, 112 kJ/mol) on the underlying aromatic SAM, and they adopt a commensurate two-domain c(2 × 6) structure before growing as a thin film with random orientation in azimuth. The electronic band structure of pentacene is characterized by purely molecular states, a sign of weak interaction between pentacene and the aromatic radical of the underneath benzenethiolate-SAM, with the formation of a small induced dipole (0.18 eV). The buildup of this semiconducting heterostructure with a hole-injection barrier of 0.95 eV, lower than that observed for pentacene on bare metals, suggests a way for tailoring the band offset, and it provides a useful benchmark for future theoretical and experimental improvement of hybrid devices. Acknowledgment. We gratefully acknowledge the staff of the Gas-Phase beamline at the ELETTRA Synchrotron Radiation Laboratory. We warmly thank the experimental assistance of Dr. Fabio Bussolotti for the photoemission data and of Dr. Federica Leonardi and Antonio Miriametro for the TDS experiments. This work is partially funded by FIRB-Nomade, PRIN-confin2004 of MIUR, and by “Faculty” and “Ateneo” grants of Roma “La Sapienza” University. References and Notes (1) Gundlach, D. J.; Lin, Y. Y.; Jackson, T. N.; Nelson, S. F.; Schlom, D. G. IEEE Electron DeVice Lett. 1997, 18, 87. (2) Puigdollers, J.; Voz, C.; Orpella, A.; Quidant, R.; Martin, I.; Vetter, M.; Alcubilla, R. Org. Electron. 2004, 5, 67. (3) Soehnchen, S.; Lukas, S.; Witte, G. J. Chem. Phys. 2004, 121, 525. (4) Hu, W. S.; Tao, Y. T.; Hsu, Y. J.; Wei, D. H.; Wu, Y. S. Langmuir 2005, 21, 2260. (5) Jurchescu, O. D.; Baas, J.; Palstra, T. T. M., Appl. Phys. Lett. 2004, 84, 3061. (6) Shaw, J. M.; Seidler, P. F. IBM J. Res. DeV. 2001, 45, 3. (7) Watkins, N. J.; Zorba, S.; Gao, Y. J. Appl. Phys. 2004, 96, 425. (8) Halik, M.; Klauk, H.; Zschieschang, U.; Schmid, G.; Dehm, C.; Schuetz, M.; Maisch, S.; Effenberger, F.; Brunnbauer, M.; Stellacci, F. Nature 2004, 431, 963. (9) Watkins, N. J.; Gao, Y. J. Appl. Phys. 2003, 94, 5782. (10) Dimitrakopoulos, C. D.; Brown, A. R.; Pomp, A. J. Appl. Phys. 1996, 80, 2501. (11) Pratontep, S.; Brinkmann, M.; Nuesch, F.; Zuppiroli, L. Phys. ReV. B 2004, 69, 165201. (12) Kang, J. H.; Filho, D. da S.; Bredas, J.-L.; Zhu, X.-Y. Appl. Phys. Lett. 2005, 86, 152115. (13) Karl, N. Synth. Met. 2003, 133, 649. (14) Fukagawa, H.; Yamane, H.; Kera, S.; Okudaira, K. K.; Ueno, N. Phys. ReV. B 2006, 73, 041302. (15) Kera, S.; Yabuuchi, Y.; Yamane, H.; Setoyama, H.; Okudaira, K. K.; Kahn, A.; Ueno, N. Phys. ReV. B 2004, 70, 085304. (16) Samanta, M. P.; Tian, W.; Datta, S.; Henderson, J. I.; Kubiak, C. P. Phys. ReV. B 1996, 53, 7626. (17) Frey, S.; Stadler, V.; Heister, K.; Eck, W.; Zharnikov, M.; Grinze, M.; Zeysing, B.; Terfort, A. Langmuir 2001, 17, 2408. (18) Wong, K.; Kwon, K.; Rao, B. V.; Liu, A.; Bartels, L. J. Am. Chem. Soc. 2004, 126, 7762.
J. Phys. Chem. C, Vol. 111, No. 1, 2007 293 (19) Lukas, S.; Vollmer, S.; Witte, G.; Woell, C. J. Chem. Phys. 2001, 114, 10123. (20) Gavioli, L.; Fanetti, M.; Sancrotti, M.; Betti, M. G. Phys. ReV. B 2005, 72, 035458. (21) Wang, Y. L.; Ji, W.; Shi, D. X.; Du, S. X.; Seidel, C.; Ma, Y. G.; Gao, H.-J.; Chi, L. F.; Fuchs, H. Phys. ReV. B 2004, 69, 075408. (22) Lukas, S.; Witte, G.; Woell, C. Phys. ReV. Lett. 2001, 88, 028301. (23) Baldacchini, C.; Betti, M. G.; Corradini, V.; Mariani, C. Surf. Sci. 2004, 566-568, 613. (24) Corradini, V.; Menozzi, C.; Cavallini, M.; Biscarini, F.; Betti, M. G.; Mariani, C. Surf. Sci. 2004. 532-535, 249. (25) Lagoute, J.; Kanisawa, K.; Foelsch, S. Phys. ReV. B 2004, 70, 245415. (26) Menozzi, C.; Corradini, V.; Cavallini, M.; Biscarini, F.; Betti, M. G.; Mariani, C. Thin Solid Films 2003, 428, 227. (27) Schroeder, P. G.; France, C. B.; Park, J. B.; Parkinson, B. A. J. Phys. Chem. B 2003, 107, 2253. (28) Ozaki, H. J. Chem. Phys. 2000, 113, 6361. (29) Casalis, L.; Danisman, M. F.; Nickel, B.; Bracco, G.; Toccoli, T.; Iannotta, S.; Scoles, G. Phys. ReV. Lett. 2003, 90, 206101. (30) Meyer zu Heringdorf, F.-J.; Reuter, M. C.; Tromp, R. M. Nature 2001, 412, 517. (31) Kanjilal, A.; Bussolotti, F.; Crispoldi, F.; Beccari, M.; Di Castro, V.; Betti, M. G.; Mariani, C. J. Phys. IV France 2006, 132, 301. (32) Di Castro, V.; Bussolotti, F.; Mariani, C. Surf. Sci. 2005, 598, 218. (33) Driver, S. M.; Woodruff, D. P. Surf. Sci. 2001, 488, 207. (34) Mariani, C.; Allegretti, F.; Corradini, V.; Contini, G.; Di Castro, V.; Baldacchini, C.; Betti, M. G. Phys. ReV. B 2002, 66, 115407. (35) Blyth, R. R.; Delaunay, R.; Zitnik, M.; Krempasky, J.; Krempaska, R.; Slezak, J.; Prince, K. C.; Richter, R.; Vondracek, M.; Camilloni, R.; Avaldi, L.; Coreno, M.; Stefani, G.; Furlani, C.; De Simone, M.; Stranges, S.; Adam, M.-Y. J. Electron Spectrosc. Relat. Phenom. 1999, 101, 959. (36) Beccari, M. M.S. Thesis, University of Rome ‘La Sapienza’, 20042005. (37) Snyder, H. R.; Geller, H. C. J. Am. Chem. Soc. 1952, 74, 4864. (38) Anderson, S. E.; Nyberg, G. L. J. Electron Spectrosc. Relat. Phenom. 1990, 52, 735. (39) Kane, S. M.; Huntley, D. R.; Gland, J. L. J. Phys. Chem. B 1998, 102, 10216. (40) Kondoh, H.; Saito, N.; Matsui, F.; Yokoyama, T.; Ohta, T.; Kuroda, H. J. Phys. Chem. B 2001, 105, 12870. (41) Konopka, M.; Rousseau, R.; Stich, I.; Marx, D. J. Am. Chem. Soc. 2004, 126, 12103. (42) Konopka, M.; Rousseau, R.; Stich, I.; Marx, D. Phys. ReV. Lett. 2005, 95, 096102. (43) De Renzi, V.; Rousseau, R.; Marchetto, D.; Biagi, R.; Scandolo, S.; del Pennino, U. Phys. ReV. Lett. 2005, 95, 046804. (44) Bussolotti, F. Ph.D. Thesis, University of Modena and Reggio Emilia, 2005. (45) Tibbets, G. G.; Burkstrand, J. M.; Tracy, J. T. Phys. ReV. B 1977, 15, 3652. (46) Leschik, G.; Courths, R.; Wern, H. Surf. Sci. 1993, 294, 355. (47) Tsiper, E. V.; Soos, Z. G. Phys. ReV. B 2003, 68, 085301. (48) Voelkel, A. R.; Street, R. A.; Knipp, D. Phys. ReV. B 2002, 66, 195336. (49) Bilic´, A.; Reimers, J. R.; Hush, N. S. J. Chem. Phys. 2005, 122, 094708. (50) Bol, C. W. J.; Friend, C. M.; Xu, X. Langmuir 1996, 12, 6083. (51) Christmann, K. In Topics in Physical Chemistry; Baumga¨rtel, H., Franck, E. U., Gru¨bein, W., Eds.; Springer-Verlag: New York, 1991. (52) Redhead, P. Vacuum 1962, 12, 203. (53) Vollmer, S.; Witte, G.; Woell, C. Langmuir 2001, 17, 7560. (54) Baldacchini, C.; Mariani, C.; Betti, M. G. J. Chem. Phys. 2006, 124, 154702. (55) Clark, P. A.; Brogli, F.; Heilbronner, E. HelV. Chim. Acta 1972, 55, 140. (56) Deleuze, M. S.; Trofimov, A. B.; Cederbaum, L. S. J. Chem. Phys. 2001, 115, 5859. (57) Fukagawa, H.; Yamane, H.; Kataoka, T.; Kera, S.; Nakamura, M.; Kudo, K.; Ueno, N. Phys. ReV. B 2006, 73, 245310. (58) Baldacchini, C.; Mariani, C.; Betti, M. G.; Gavioli, L.; Fanetti, M.; Sancrotti, M. Appl. Phys. Lett. 2006, 89, 152119. (59) Kahn, A.; Koch, R.; Gao, W. J. Polym. Sci., Part B: Polym. Phys. 2003, 41, 2529.