Available online at www.sciencedirect.com

008) 2939–2942www.elsevier.com/locate/tsf

Thin Solid Films 516 (2

Search for a wetting layer in thin film growth ofpara-hexaphenyl on KCl(001)

Paul Frank a, Gerardo Hernandez-Sosa b, Helmut Sitter b, Adolf Winkler a,⁎

a Institute of Solid State Physics, Graz University of Technology, A-8010 Graz, Austriab Institute of Semiconductor and Solid State Physics, University of Linz, A-4040, Austria

Received 13 April 2007; received in revised form 27 August 2007; accepted 31 August 2007Available online 16 September 2007

Abstract

The non-existence of a strongly bound wetting layer for the system para-hexaphenyl (6P)-KCl(001) was verified by thermal desorptionspectroscopy (TDS) and X-ray photoelectron spectroscopy (XPS). The 6P films were grown by physical vapour deposition under ultrahighvacuum conditions at 400 K. TDS showed just a single desorption peak even down to a coverage of 0.1 nm mean film thickness. The heat ofevaporation for 6P was determined to 2.3 eV. From the change of the XPS K2p substrate signal as a function of the mean film thickness one canconclude that a strong de-wetting exists at the very beginning of the 6P layer growth at 400 K. Additional investigations with atomic forcemicroscopy reveal that at this initial stage the film grows in a form of needle like islands with a high aspect ratio and subsequently terracedmounds of 6P are formed. No dissociation of 6P on KCl was observed.© 2007 Elsevier B.V. All rights reserved.

Keywords: Hexaphenyl; Potassium chloride; KCl(001); Film growth; Wetting layer

1. Introduction

Thin films of para-hexaphenyl (6P) have attained consider-able interest in the recent past, due to their excellentluminescence properties in the blue visible region. Successfulapplications of this material for light emitting diodes [1,2], fieldeffect transistors [3], photovoltaic elements [4] and nonlinearoptics [5] have been reported. In this context the morphology ofthe 6P films and the particular properties of the interfacebetween the substrate and the film are of utmost importance. Itis well known that 6P films can either grow in the form ofneedle like islands, where the molecules are oriented with theirlong axes parallel to the substrate, or in form of terrace likeislands where the molecules are oriented upright [6]. Thedevelopment of these film structures depends on the type of thesubstrate, the substrate structure and the chemical compositionof the substrate. In addition, very often the first layer of the film

⁎ Corresponding author. Institute of Solid State Physics, Graz University ofTechnology, Petersgasse, 16, A-9010 Graz, Austria. Tel.: +43 316 873 8463;fax: +43 316 873 8466.

E-mail address: a.winkler@tugraz.at (A. Winkler).

0040-6090/$ - see front matter © 2007 Elsevier B.V. All rights reserved.doi:10.1016/j.tsf.2007.08.142

is more strongly bound to the substrate than the subsequentlayers. This special first layer is called the wetting layer.Such a wetting layer plays a dominant role for the thin filmgrowth and may also largely determine the electronic propertiesof electronic devices. Since this monomolecular layer is noteasily detectable with techniques usually applied in thin filmtechnology [atomic force microscopy (AFM), X-ray diffraction(XRD), transmission electron microscopy (TEM)], the exis-tence of such a layer has often been the subject of controversialdiscussions.

Potassium chloride (KCl) is frequently used as a substrate tostudy the thin film morphology of 6P. Films have been grownby physical vapour deposition [1,7–12] and by hot wall epitaxy[13–15]. It has been observed by AFM that the grown filmsconsist of needles and terraced mounds. In some recent works itwas proposed from phase imaging AFM [15] and from XRDinvestigations [13] that both needles and terraced mounds growdirectly on the substrate surface, without a wetting layer. How-ever, no direct proof could be given. In this letter we address thequestion of a wetting layer by combined X-ray photoelectronspectroscopy (XPS) and thermal desorption spectroscopy (TDS)studies under well defined ultrahigh vacuum (UHV) conditions,

Fig. 1. Series of thermal desorption spectra for 6P from KCl(001). Adsorptiontemperature: 400 K, heating rate: 1 K/s. Parameter is the mean film thicknessprior to desorption. The arrow points to a peak in curve a which originates fromdesorption from the Ta clamps. The inset shows the linear dependence betweenadsorbed and desorbed 6P for two different adsorption temperatures, 100 K and400 K, respectively.

2940 P. Frank et al. / Thin Solid Films 516 (2008) 2939–2942

and additional ex-situ AFM measurements, where we unambig-uously prove that indeed no wetting layer exists for 6P on KCl(001).

2. Experimental details

The hexaphenyl (6P) thin films on KCl(001) were preparedunder ultrahigh vacuum conditions (base pressure≈1×10−10

mbar) by physical vapour deposition. The apparatus wasequipped with an Auger electron spectrometer and an X-rayphotoelectron spectrometer (Leybold Heraues, EA10/100), alow energy electron diffraction (LEED) optics with a multichannel plate (Omicron) and an in-line quadrupole mass spec-trometer (Balzers QMA 400) [16]. A KCl crystal was cleavedinto a substrate of 10×10×1 mm3 which was attached bytantalum wires to a steel plate (10×10×1 mm3) for good ther-mal contact. Further tantalum wires were spot welded to theback side of the steel plate which allowed resistive heating ofthe plate. The temperature was measured at the backside of thesteel plate by a Ni/NiCr thermocouple. A home-made quartzglass Knudsen cell was used to evaporate the organic material.The film thickness was monitored by a quartz microbalance(Inficon XTM/2) which was positioned next to the sample.Typically, an evaporation rate of 0.3 nm/min was used. Thermaldesorption spectroscopy (TDS) was performed by desorbing theadsorbed 6P molecules into the mass spectrometer, which wastuned to the mass of 6P with m=458 amu. The heating rate forTDS was 1 K/s. XPS was used to determine the cleanliness ofthe substrate and LEED to check the surface structure. Atomicforce microscopy (Easy Scan 2, Nanosurf) on the 6P coveredKCl(001) was performed ex-situ under ambient conditions.

3. Results and discussions

After the installation of the KCl(001) substrate into the UHVchamber, small amounts of carbon and oxygen were detected onthe surface by XPS. However, after annealing of the sample for1/2 h at 420 K, no contaminants could be detected on thesurface. A series of thermal desorption spectra obtained afterevaporation of different amounts of 6P on KCl(001) at 400 K,ranging from 0.1 nm to 7 nm, are shown in Fig. 1. This ad-sorption temperature was chosen because it is frequently used inthe literature. The asymmetric peak shape, the common leadingedge of the spectra and the shift of the peak maxima to highertemperature with increasing coverage are characteristic of a zeroorder desorption, as expected for multilayer desorption. Similardesorption spectra were also obtained for multilayer desorptionof 6P from Au(111) [17] and mica(001) [18]. In those cases,however, additional desorption peaks at higher temperaturewere observed at small coverage, which indicated the existenceof the more strongly bound wetting layer. In Fig. 1 no evidencefor a strongly bound layer can be seen. Even for the smallestcoverage of 0.1 nm the peak is located at the common leadingedge. The small peak (marked by an arrow) which appears forthe 7 nm TD spectrum is an experimental artefact: It stems fromdesorption of a 7 nm thick 6P film from the tantalum clampswhich are used to hold the KCl sample to the heating plate. It

serves as a calibration point for the temperature scale since itcan be assumed that the temperature of the metal wires corre-sponds quite well to that of the steel plate where the temperatureis measured. The temperature maximum for the 6P multilayerdesorption at the individual coverages is known from literature[17].

From the common leading edge (after proper correction of thetemperature scale) one can easily determine the desorption energyEdes (heat of evaporation) and also the pre-exponential factor forzero order desorption ν0 (because we know the desorption ratequantitatively) by applying the Polanyi–Wigner equation:

Rdes ¼ m0 �Hmax � exp �Edes=RTð Þ: ð1Þ

We find that Edes=2.3 eV and ν0=3×1024 s−1, by assuming

a surface density of 5.4×1013 6P-molecules/cm2. This is ingood agreement with the recently obtained values for 6P on Au(111) [17]. The relatively high value of the pre-exponentialfactor is typical for the desorption of large molecules [19,20].Most importantly, however, is the fact that down to the smallestcoverage of 0.1 nm mean film thickness the 6P molecules showa multilayer desorption behaviour, i.e. the molecules desorbfrom small islands and not from a more strongly bound wettinglayer. Quite similar desorption spectra were obtained after 6Padsorption at 100 K. This demonstrates that no wetting layerexists, independent of the adsorption temperature.

In the insert of Fig. 1 the correlation between the integral ofthe TDS peak areas (proportional to the desorbed amount) andthe film thickness as obtained from the quartz microbalance isshown. It is a straight line, both for an adsorption temperatureof 400 K, as well as of 100 K. This shows that the stickingcoefficient is independent of the coverage and of the sampletemperature in this range. It can therefore safely be assumed thatthe sticking coefficient is unity at and below 400 K. The secondimportant result from this experiment is that all of the adsorbedmaterial desorbs again, i.e. no partial dissociation of the 6Pmolecules takes place. Such partial dissociation has been

Fig. 3. A: AFM image of a 1.1 nm thick 6P film on KCl(001) grown by PVD at400K substrate temperature. B: Cross section along the line as depicted in panel A.

2941P. Frank et al. / Thin Solid Films 516 (2008) 2939–2942

observed for 6P and 4P on polycrystalline gold and on steppedgold surfaces [21].

One could argue that a wetting layer very well might exist atthe adsorption temperature but that during the temperatureincrease in TDS the molecules in the wetting layer mightbecome mobile enough to form islands, from where they thendesorb in form of the observed zero order reaction kinetics. Inorder to check this possibility we have followed the change ofthe potassium K2p signal as a function of the mean filmthickness d at 400 K and 100 K, respectively (Fig. 2). At 100 Kthe 6P molecules are not very mobile and a “hit and stick”adsorption behaviour can be assumed. In fact, in this case thedecrease of the substrate K2p signal can be approximated quitewell by an exponential function, as one would expect for anamorphous layer-by-layer growth:

I K2pð Þ=I0 K2pð Þ ¼ exp �d=kð Þ: ð2ÞA mean free path λ of 1.7 nm for the K2p electrons

(Ekin=961 eV) in the 6P film was calculated from thisexponential fit. In contrast, if the 6P film was formed at asubstrate temperature of 400 K the K2p signal remained nearlyconstant up to a mean thickness of about 0.5 nm. If a wettinglayer were formed at this temperature the XPS substrate signalshould decrease at least as fast as at 100 K. The little change ofthe substrate signal shows that the molecules are alreadystrongly de-wetting in this low coverage range by forming highislands with small footprints. Interestingly, with a furtherincrease of the coverage the decrease of the substrate signalbecomes similar to that at 100 K, just shifted by some thicknessvalue. This indicates that further film growth does not proceedin form of islands with high aspect ratios (height vs. footprint),but that rather a continuous layer growth exists. From literatureone knows that 6P on KCl(001) forms long and high needleswith lying molecules as well as extended terraced mounds,consisting of standing molecules [13]. From Fig. 2 it is obviousthat the needles with high aspect ratios are formed at the very

Fig. 2. K2p3/2 XPS signal as a function of film thickness, grown at 100 K and400 K substrate temperature, respectively. For the K2p signal at 100 K the datapoints are fitted by an exponential function. In the other case the line is drawn asa guide to the eye.

beginning of the layer growth and subsequently the film growthmainly proceeds in the form of terraced mounds, as also ob-served for thin 6P films grown by hot wall epitaxy (HWE) [15].

In Fig. 3Awe show an AFM image of a 1.1 nm thick film of6P on KCl(001) prepared at 400 K. Even at this small meanthickness one sees already large irregularly shaped islandsbetween the well developed needles. A cross section along aline as indicated in Fig. 3A is shown in Fig. 3B. The height ofthe broad terrace is 2.6 nm, indicating that this island is formedof standing molecules. The needle like island in this case has afootprint width of about 140 nm and a height of about 20 nm.This AFM result agrees quite well with recently obtainedmorphologies for 6P on KCl(001) grown by HWE [13–15].From XRD and TEM investigations it is known that both, theneedle like islands and the terrace like islands, grow epitaxiallyon the KCl(001) surface [11,13]. It has also been shown thatindeed the needles consist of side tilted molecules with theirlong axes being parallel to the surface, whereas the terracedmounds consist of standing molecules.

4. Summary and conclusions

The special feature of the present study is that the kinetics of6P layer growth on KCl(001) has been determined quantita-tively. HWE can be used to prepare well defined thin films, butit is usually not possible to check the surface conditions and todetermine the film thickness quantitatively. On the other handwe have shown in this work that the morphology of the filmsprepared by physical vapour deposition (PVD) is comparable tothat obtained from HWE. It is now clear that no wetting layer isformed at the initial stage of the film growth. On the other hand

2942 P. Frank et al. / Thin Solid Films 516 (2008) 2939–2942

the film starts to grow in form of needle like islands. This is aquite unique result. So far the general rule for 6P growth wasthat only in the case of an existing wetting layer a needle likeisland growth can be observed, where the wetting layer servesas a template for the further film growth [17]. Such a behaviouris generally seen if the bonding between the 6P molecules andthe substrate is quite strong and if the substrate is well ordered.In case of decreased bonding, e.g. due to carbon impurities onthe surface [17] or due to an amorphisation of the substrate [18],the film growth proceeds in form of standing molecules formingterraced mounds from the very beginning, without a wettinglayer of lying molecules. The 6P-KCl(001) system is the firstexample where needle like islands with lying molecules areformed without the existence of a wetting layer. From the verysimilar growth behaviour of 6P on other alkali halide substrates,like NaCl and KBr [22], one can speculate that in general nowetting layers may exist on alkali halide substrates.

Acknowledgements

This work was financially supported by the Austrian ScienceFund FWF, Proj. no. S97 and P19197.

References

[1] H. Yanagi, S. Okamoto, Appl. Phys. Lett. 71 (1997) 2563.[2] S. Tasch, C. Brandstätter, F. Meghdadi, G. Leising, G. Froyer, L. Athouel,

Adv. Mater. 9 (1997) 33.

[3] D.J. Gundlach, T.N. Jackson, D.G. Schlom, S.F. Nelson, Appl. Phys. Lett.71 (1997) 3853.

[4] C. Winder, A. Andreev, H. Sitter, G. Matt, N.S. Sariciftci, D. Meissner,Synth. Met. 139 (2003) 573.

[5] F. Balzer, H.G. Rubahn, Adv. Funct. Mater. 15 (2005) 17.[6] R. Resel, Thin Solid Films 433 (2003) 1.[7] T. Mikami, H. Yanagi, Appl. Phys. Lett. 73 (1998) 563.[8] H. Yanagi, T. Morikawa, Appl. Phys. Lett. 75 (1999) 187.[9] Y. Yoshida, H. Takiguchi, T. Hanada, A. Tanigaki, E.M. Han, K. Yase,

J. Cryst. Growth 198/199 (1999) 923.[10] D.M. Smilgies, N. Boudet, H. Yanagi, Appl. Surf. Sci. 189 (2002) 24.[11] E.J. Kintzel Jr., D.M. Smilgies, J.G. Sofronik, S.A. Safron, D.H. Van

Winkle, J. Vac. Sci. Technol., A 22 (2004) 107.[12] F. Balzer, H.G. Rubahn, Surf. Sci. 548 (2004) 170.[13] T. Haber, A. Andreev, A. Thierry, H. Sitter, M. Oehzelt, R. Resel, J. Cryst.

Growth 284 (2005) 209.[14] T. Haber, M. Oehzelt, R. Resel, A. Andreev, A. Thierry, H. Sitter, D.M.

Smilgies, B. Schaffer, W. Grogger, J. Nanosci. Nanotechnol. 6 (2006) 1.[15] A. Andreev, A. Montaigne, H. Sitter, C. Teichert, J. Vac. Sci. Technol., A

24 (2006) 1660.[16] S. Müllegger, O. Stranik, E. Zojer, A. Winkler, Appl. Surf. Sci. 221 (2004)

184.[17] S. Müllegger, A. Winkler, Surf. Sci. 600 (2006) 1290.[18] P. Frank, G. Hlawacek, O. Lengyel, A. Satka, C. Teichert, R. Resel, A.

Winkler, Surf. Sci. 601 (2007) 2152.[19] K.A. Fichthorn, R.A. Miron, Phys. Rev. Lett. 89 (2002) 196103.[20] S. Tait, Z. Dohnalek, C.T. Campbell, B.D. Kay, J. Chem. Phys. 122 (2005)

164708.[21] S. Müllegger, A. Winkler, Surf. Sci. 600 (2006) 3982.[22] N. Yoshimoto, T. Sato, Y. Saito, S. Ogawa, Mol. Cryst. Liq. Cryst. 425

(2004) 279.