Hierarchy of adhesion forces in patterns of photoreactive surface layers
G. Hlawacek, Q. Shen, C. Teichert, A. Lex, G. Trimmel, W. Kern
HHierarchy of adhesion forces in patterns of photoreactive surface layers
Gregor Hlawacek, Quan Shen, and Christian Teichert ∗ Institute of Physics, University of Leoben, 8700 Leoben, Austria
Alexandra Lex † and Gregor Trimmel Institute for Chemistry and Technology of Materials,Graz University of Technology, 8010 Graz, Austria
Wolfgang Kern
Institute of Chemistry of Polymeric Materials, University of Leoben, 8700 Leoben, Austria (Dated: November 29, 2018)Precise control of surface properties including electrical characteristics, wettability, and friction isa prerequisite for manufacturing modern organic electronic devices. The successful combination ofbottom up approaches for aligning and orienting the molecules and top down techniques to structurethe substrate on the nano and micrometer scale allows the cost efficient fabrication and integrationof future organic light emitting diodes and organic thin film transistors. One possibility for thetop down patterning of a surface is to utilize different surface free energies or wetting propertiesof a functional group. Here, we used friction force microscopy (FFM) to reveal chemical patternsinscribed by a photolithographic process into a photosensitive surface layer. FFM allowed thesimultaneous visualization of at least three different chemical surface terminations. The underlyingmechanism is related to changes in the chemical interaction between probe and film surface.
Keywords: Friction force microscopy, photoreactive thin layers, photo-lithography, isomerization
I. INTRODUCTION
Modern low cost devices are increasingly based on or-ganic semiconductors. This important class of materialsallows to achieve well-priced thin film transistors and op-tical components such as light emitting diodes. An im-portant intermediate step in this technology is the pos-sibility to control the growth behaviour of active organicmaterials in terms of orientation and structure on shortlength scales. To do so patterned thin surface layers orself-assembled monolayers can be used. The patterning ofsuch films can be achieved in various ways including butnot limited to soft lithography, scanning probe tech-niques (dip pen lithography, nanografting, etc.), en-ergetic beams (UV-light, electrons, etc.) and manymore. Here, we present a friction force microscopy (FFM)study of thin surface layers of a photosensitivethiocyanate-functionalized trialkoxysilane on silicon ox-ide (SiO x ). These films can be reliably prepared on thistechnological important surface with a high degree of con-trol over the final film thickness. However, more impor-tant is the fact that these films can easily be modified byUV light and subsequently functionalized. As different end groups of the molecule will have dif-ferent interaction with the probe of the atomic force mi-croscope, FFM allows to differentiate between them.Using this approach, four different terminations could bedistinguished and hierarchically ordered on a sample thathas been subsequently irradiated twice using line maskswith different feature spacings. As a result the differ-ent terminations could be hierarchically ordered by theirinteraction strength with the AFM probe. II. EXPERIMENTAL
For the preparative work of the organic thin sur-face layers, hazardous chemicals and solvents are used(ammonium thiocyanate, methanol, propylamine, 2,2,2-trifluoroethylamine, and piranha solution). In addition,piranha solution is explosive, and its preparation is highlyexothermic (up to 120 ◦ C). Therefore, reactions must becarried out in a fume hood, and protective clothes andgoggles must be used! UV irradiation causes severe eyeand skin burns. Precautions (UV protective goggles andgloves) must be taken!The photoreactive surface layers were prepared byimmersion of pretreated (Piranha solution) borondoped silicon wafers into a solution of trimethoxy[4-(thiocyanatomethyl)phenyl]silane (Si-SCN) in toluene.X-ray reflectivity measurements revealed a film thicknessof 6 nm for these films. It has to be emphasized here,that obviously this is not a monolayer but an oligolayerwith a thickness corresponding to five or six individuallayers (assuming upright standing molecules). The for-mation of oligolayers is attributed to cross linking of thetrimethoxy-silane groups in the presence of water. Ina subsequent step, the samples were illuminated by UVlight under inert gas to avoid photo-oxidation (254 nm,80 mJ / cm ). The illumination leads to an isomeriza-tion of the benzyl thiocyanate (Si-SCN) group to thecorresponding benzyl isothiocyanate (Si-NCS). This il-lumination step was done by utilizing a contact maskwith equidistant lines and spaces to create a patternon the surface which consists of alternating stripes ofSi-SCN and Si-NCS. For selected samples an additionalpost-isomerization modification was performed by expos- a r X i v : . [ c ond - m a t . m t r l - s c i ] O c t FIG. 1: Reaction pathway and products investigated by fric-tion force microscopy. ing the surface to vapours of propylamine. In this pro-cess the isothiocyanate group reacts to the correspond-ing thiourea group (Si-PA). The sequence of the reactionsteps together with the respective molecular structure isshown in Fig. 1. A detailed description of the film prepa-ration has already been given elsewhere. The AFM results were obtained with a Digital Instru-ments Multimode IIIa atomic force microscope. To re-duce damage to the film, the topographic images wererecorded in intermittant contact mode, eliminating effec-tively lateral forces between the tip and the sample sur-face. For intermittant mode normal Si probes with a typi-cal resonance frequency of 300 kHz were used. For rough-ness characterization, the root mean square roughness σ ,the lateral correlation length ξ , and the Hurst parame-ter α were calculated from the images using the height-height correlation function (HHCF). All roughness pa-rameters have been obtained by analyzing at least threeindependent 5 µ m images. FFM (also called lateral forcemicroscopy (LFM) or chemical force microscopy) isa special type of contact mode atomic force microscopy.Lateral forces acting on the tip will twist the cantilever,when scanned perpendicular to its long axis, leading to adeflection of the laser on the four quadrant photodiode inlateral direction. The twist of the cantilever depends onthe friction between the tip and the sample surface. Ascantilevers, specially designed FFM silicon rectangularbeam type cantilevers are used. The cantilevers have anominal length of 225 µ m and a force constant of typical0.2 N/m.The lateral force acting on the tip is influenced bythe friction coefficient between tip and sample surface.This coefficient depends on the interaction between thetip and the terminating group of the molecules form-ing the thin film. For a clearer contrast, FFM im-ages are calculated from trace and retrace images ob-tained simultaneously with the topographic image. This effectively reduces false FFM contrast originat-ing from the surface morphology. The presented im-ages are therefore always calculated from ( trace − retrace ) /
2. For the presented FFM images no scaleis given as no force calibration was performed priorto the measurement. Thus information that can beobtained is purely qualitative, however, with sufficient accuracy allowing to establish a hierarchy of adhesionforces.
III. RESULTS
Figure 2 presents AFM topography images demon-strating the effect of film preparation on surface rough-ness. Homogeneous films of Si-SCN (b), and propylaminemodified Si-NCS (Si-PA) (c) films were prepared andcompared to the surface of the bare substrate (a).The surface of the substrate (Fig. 2(a)) shows a uni-form featureless topography as expected for a silicondioxide surface. The root mean square (rms) roughnessof σ =0.2 nm, the lateral correlation length ξ =30 nm, andthe Hurst parameter α =0.5 confirm the qualitative ob-servation. Other investigations report much larger cor-relation length for SiO measured by X-ray and opticaltechniques. The smaller values given here are relatedto the shorter length scales that can be evaluated withAFM. Deposition of a thin layer of Si-SCN and subse-quent illumination with 254 nm UV light for 20 min (re-sulting in a Si-NCS film shown in Fig. 2(b)) does notlead to a strong change in the roughness parameters: σ =0.3 nm, ξ =30 nm and α =0.5. As mentioned above,X-ray reflectivity measurements revealed a film thicknessof 6 nm for this layer. Modifying the surface with va-pors of propylamine (Fig. 2(c)) leads to a further slightincrease in surface roughness ( σ = 0 . ξ = 20 nm,and α = 0 . µ m lines and 10 µ m spaces to thesimultaneously recorded FFM image (b). Whereas inthe topography image no stripe pattern is visible, a clearstripe pattern with a 10 µ m pitch appears in the FFMimage. In all FFM images presented, bright areas meanhigher friction and dark ones correspond to lower friction.It will be demonstrated below that the high friction areascorrespond to the illuminated Si-NCS stripes and the lowfriction areas are the pristine Si-SCN stripes covered bythe mask during illumination.Figure 4 shows AFM images obtained after the pat-terned sample described above has been exposed tovapours of propylamine. While the topographic imagepresented in Fig. 4(a) shows a weak stripe pattern withthe expected spacing, the corresponding FFM image inFig. 4(b) allows a clear identification of the line patterncreated by contact lithography and of errors in the mask-ing process (lower right corner). From the combined crosssection in Fig. 4(c) the height difference of 0.5 nm be-tween the modified Si-PA and the pristine Si-SCN stripesis clearly discernible. The addition of an alkyl group tothe molecule will result in an increase of the film thick- FIG. 2: AFM topography images recorded in intermittant mode of (a) the SiO substrate, (b) the film after Si-SCN depositionand subsequent flood illumination with 254 nm UV light, and (c) modification with propylamine. (z-scale in all images is 5 nm.)FIG. 3: Topographic (z-scale: 10 nm) (a) and correspondingFriction Force image (b) from a Si-SCN film patterned with254 nm UV light through a 10 µm mask. In the FFM image(b) bright areas correspond to Si-NCS terminated areas show-ing high friction. The dark stripes are the non-illuminatedSi-SCN areas. ness. An increased height is therefore only expected forthe modified parts of the surface, which are the Si-PAareas. These areas show a lower friction signal. Areas oflower height correspond to non-illuminated (and unmod-ified) zones of the layer containing Si-SCN structures,which give a higher friction than the Si-PA structure.The increase in layer thickness after reaction with propy-lamine was 0.6 nm which equals 10% of the initial layerthickness. This corresponds to results obtained with pho-toreactive polymers containing benzyl thiocyanate sidegroups. In a further experiment, the surface has been exposedtwice to illumination using crossed masks with an inter-mediate propylamine reaction step. The whole samplepreparation process is sketched in Fig. 5(a). In a firststep, a mask with 5 µ m lines and spaces was employedduring UV illumination. After the Si-SCN film has beenexposed to vapors of propylamine, a surface layer madeup of alternating 5 µ m stripes of Si-SCN and Si-PA iscreated. The resulting pattern is similar to the one pre-sented in Fig. 4. This modified film was now illuminatedfor a second time through a mask with 10 µ m lines andspaces oriented perpendicular to the first mask pattern.During this step both stripes, Si-PA as well as Si-SCN,are illuminated partly. It can be expected that the Si-PAsurface will not change significantly during this process. However, the remaining Si-SCN stripes will be convertedinto alternating 10 µ m patches of Si-SCN and Si-NCS.The resulting surface morphology and the friction imageare presented in Fig. 5(b,c). As before, the main fea-tures in the topographic image (Fig. 5(b)) are not relatedto the mask process but result from contamination andsmall long range undulations in the SiO x surface of thewafer. However, the FFM image shows a clear pattern ofregular 5 by 10 µ m patches of four different shadings, i.e.friction levels. We can identify the four areas using theinformation on the friction contrast obtained from theprevious samples. The dark areas in Fig. 5(b) are theSi-PA areas (labeled PA) created in the first illuminationstep. The neighboring brighter patches (PI) in the 5 µ mstripe are the Si-PA areas that were subsequently illu-minated a second time. In the neighboring 5 µ m stripes(shadowed during the first illumination) one patch hasbeen protected by the mask in both illumination steps(DA) while the other one has been exposed to UV-lighta single time during the second illumination (IL). Thelast two areas are comparable to those shown in Fig. 3and therefore allow the identification of the stripes inFig. 3(b). The areas marked PA and DA are comparableto the combination shown in Fig. 4.The observed friction contrast can therefore be orderedin the following way: The highest tip-film interactionand therefore the largest friction is observed for Si-NCS(IL) followed by Si-SCN (DA) and the two propylaminemodified surfaces (PI and PA in Fig. 5(b)).A possible explanation fort his sequence can be givenby the different polarity of the individual endgroups atthe surface (thiocyanate (Si-SCN), isothiocyanate (Si-NCS), and propyl (CH -CH -CH , Si-PA)) and by thestiffness of the molecular endgroups.Comparing data on the dipole moment of ethyl isothio-cyanate (3.67 Debye) and ethyl thiocyanate (3.33 Debye)as well as on the surface tension γ (at 20 ◦ C) of these com-pounds (ethyl isothiocyanate: γ = 36.0 mN/m, and ethylthiocyanate: γ = 34.8 mN/m) it is found that isothio-cyanates are of higher polarity than the correspondingthiocyanates. Assuming that the friction force betweenthe surface and the silicon tip (which is covered with a na-tive oxide layer) increases with the polarity of the surface,for the UV illuminated regions (containing NCS units at
FIG. 4: 50 µm topographic (z-scale: 5 nm) (a) and FFM (b) image from a patterned Si-SCN/Si-PA film. The pattering hasbeen done through a contact mask with a 10 µm pitch. (c) Indicated cross sections reveal the expected 10 µm pitch in bothtopography and friction contrast. The Si-PA stripes are roughly 0.6 nm higher than the Si-SCN stripes.FIG. 5: (a) Scheme of sample preparation: Illumination of Si-SCN through a 5 µ m mask (Mask 1); exposure to propylamine;second illumination through a 10 µ m mask (Mask 2). (b) resulting topographic (z-scale: 10 nm) and (c) corresponding FFMimage from the double patterned Si-SCN/Si-PA/Si-NCS film. The respective materials are indicated in the FFM image (seetext). the surface) a higher friction force will be recorded thanfor the non-illuminated regions bearing SCN units.After reaction of the photogenerated NCS groups withpropylamine (PA), the surface is terminated with non-polar alkyl groups. It is therefore not surprising that thepost-exposure derivatization with propylamine will resultin lower friction force. Also the stiffness of the molecule isreported to influence the resulting friction coefficient. The flexible alkyl group that terminates Si-PA will there-fore also reduce the observed friction by bending underthe applied normal load. In contrast, the shorter andstiffer thiocyanate and iso-thiocyanate groups (Si-NCShas two double bonds between sulfur, carbon, and ni-trogen; Si-SCN has a triple bond between carbon andnitrogen while the sulfur is linked by two single bonds)can not bend under the applied load.The intermediate friction contrast observed for areas that were illuminated and modified with propylamine,and then illuminated for a second time, can also be ex-plained that way. During the second illumination, resid-ual SCN units – which have remained unreacted in thefirst illumination step – are converted into NCS groupsleading to an increased interaction between the surfaceand the AFM tip. NCS groups which have reacted withpropylamine (to yield propyl thiourea units) are expectedto remain unchanged during the second illumination stepsince N,N’-dialkyl substituted thiourea groups are stableunder UV light. Especially from the last sample, the hi-erarchy in the interaction forces (Si-NCS > Si-SCN > Si-PA) between the individual terminating groups becomesevident.
IV. CONCLUSION
For lithographically patterned photoreactive surfacelayers we demonstrated that friction force microscopyis not only able to distinguish between different headgroups, but also that this can be done simultaneously forat least three different terminations. The method is ableto detect minute changes in the molecular structure. Thecase of the thiocyanate (Si-SCN) and isothiocyanate (Si-NCS) head groups is especially relevant since these areeducts and products of an isomerization reaction. FFM isable to discern the two terminations, although, only thesequence of the terminating three atoms is altered. How-ever, the resulting change in polarity of the terminatingmolecular groups is large enough to be detected by FFMas different friction levels. In addition, the post isomer-ization modification with propylamine could be clearlyvisualized with FFM. In this case, a further reduction of polarity together with a change in the stiffness of themolecule leads to the lowest friction observed in the in-vestigated system.Currently, contact angle measurements are underwayto obtain an independent confirmation of the observed hi-erarchy. The next step in future work will be to quantifythe adhesive and frictional forces responsible for the qual-itative results presented here. For this task a well knownapproach suggested in literature will be used to calibratethe AFM probes with sufficient accuracy. These stud-ies shall include the use of functionalized tips to fine tunethe probe sample interaction. Acknowledgments
This work was supported by Austrian Science FundFWF projects S9707-N08 and S9702-N08. ∗ Electronic address: [email protected]; URL: † Current address: Institute of Physical Chemistry, Univer-sity of M¨unster, 48149 M¨unster, Germany N. Larsen, H. Biebuyck, E. Delamarche, and B. Michel, J.Am. Chem. Soc. , 3017 (1997). Y. Xia, J. Tien, D. Qin, and G. Whitesides, Langmuir ,4033 (1996). R. Piner, J. Zhu, F. Xu, S. Hong, and C. Mirkin, Science , 661 (1999). S. Xu, S. Miller, P. Laibinis, and G.-Y. Liu, Langmuir ,7244 (1999). G.-Y. Liu, S. Xu, and Y. Qian, Acc. Chem. Res. , 457(2000). H. Sugimura, T. Hanji, O. Takai, T. Masuda, and H. Mis-awa, Electrochim. Acta , 103 (2001). D. W. Carr, M. J. Lercel, C. S. Whelan, H. G. Craighead,K. Seshadri, and D. L. Allara, in
J. Vac. Sci. Technol. A (AVS, 1997), vol. 15, pp. 1446–1450. M. J. Lercel, R. C. Tiberio, P. F. Chapman, H. G. Craig-head, C. W. Sheen, A. N. Parikh, and D. L. Allara, in
J.Vac. Sci. Technol. B (1993), vol. 11, pp. 2823–2828. R. Smith, P. Lewis, and P. Weiss, Prog. Surf. Sci. , 68(2004). A. Lex, P. Pacher, O. Werzer, A. Track, Q. Shen, R. Schen-nach, G. Koller, G. Hlawacek, E. Zojer, R. Resel, et al.,Chem. Mater. , 2009 (2008). G. Meyer and N. M. Amer, Appl. Phys. Lett. , 2089(1990). Y. Wang and M. Liebermann, Langmuir , 1159 (2003). Y.-P. Zhao, H.-N. Yang, G.-C. Wang, and T.-M. Lu, Phys.Rev. B , 1922 (1998). C. M. Mate, G. M. McClelland, R. Erlandsson, and S. Chi-ang, Phys. Rev. Lett. , 1942 (1987). C. Frisbie, L. Rozsnyai, A. Noy, M. Wrighton, andC. Lieber, Science , 2071 (1994). C. Teichert, J. MacKay, D. Savage, M. Lagally, M. Brohl,and P. Wagner, Appl. Phys. Lett. , 2346 (1995). T. Kavc, G. Langer, W. Kern, G. Kranzelbinder, E. Tou-ssaere, G. Turnbull, I. Samuel, K. Iskra, T. Neger, andA. Pogantsch, Chem. Mater. , 4178 (2002). J. A. Dean, ed.,
Lange’s Handbook of Chemistry (McGraw-Hill, 1999), 15th ed. B. Bhushan and H. Liu, Phys. Rev. B , 2454121 (2001). R. Overney, E. Meyer, J. Frommer, D. Brodbeck, R. L¨uthi,L. Howald, H.-J. G¨untherodt, M. Fujihira, H. Takano, andY. Gotoh, Nature , 133 (1992). M. Varenberg, I. Etsion, and G. Halperin, Rev. Sci. In-strum. , 3362 (2003). A. Noy, C. Frisbie, L. Rozsnyai, M. Wrighton, andC. Lieber, J. Am. Chem. Soc. , 7943 (1995).23