Anal. Chem. 1999, 71, 2452-2458
Chemical Imaging of Patterned Inorganic Thin-Film Structures by Lateral Force Microscopy Mikko Utriainen,† Antti Leijala,‡ and Lauri Niinisto 1 *,†
Laboratory of Inorganic and Analytical Chemistry, P.O. Box 6100, and Laboratory of Processing and Heat Treatment of Materials, P.O. Box 6200, Helsinki University of Technology, FIN-02015 Espoo, Finland Raija Matero
Laboratory of Inorganic Chemistry, University of Helsinki, P.O. Box 55, FIN-00014 Helsinki, Finland
Factors influencing the chemical image formation by lateral force microscopy (LFM, or friction force microscopy, FFM) under normal ambient conditions were studied by applying LFM to patterned specimens of inorganic thin films deposited predominantly by atomic layer epitaxy. The patterned steps on SnO2/Si, CaS/Si, CeO2/Si, and Pt/Al2O3 samples were formed by chemical etching or lift-off processing. The results of semiquantitative AFM and LFM studies were compared to the static contact angle studies using capillary force evaluation. The chemical contrast in LFM images of the patterned specimens was the highest in cases where silicon was present. This is in accordance with contact angle data, which showed much higher hydrophilicity on Si than on the other materials studied. Further experiments with a patterned SnO2/Si specimen indicated that chemical contrast can be significantly affected (i) by whether the surface was pretreated with ethanol, (ii) by the loading force (2-50 nN or 1-10 µN) applied, and (iii) by using SnO2-coated AFM probes instead of the conventional Si probes. High-resolution chemical imaging microscopy is a powerful tool in materials research. Lateral force microscopy (LFM, or friction force microscopy, FFM) is a scanning probe method that offers operational ease such as working in ambient air with minimal sample preparation. Chemical identification of hard inorganic materials by LFM1-3 has been demonstrated in the following cases of thin-film heterostructures: MoO3/MoS2,4 mica/ MoS2,5,6 Al2O3/MoS2,6 C60/NaCl(001),7 GaAs/ZnSe,8 GaAs/InGaP,8 * Corresponding author: (phone) +3589 - 451 2600; (fax) +3589 - 462 373; (e-mail)
[email protected]. † Laboratory of Inorganic and Analytical Chemistry. ‡ Laboratory of Processing and Heat Treatment of Materials. (1) Mate, C. M.; McClelland, G. M.; Erlandsson, R.; Chiang, S. Phys. Rev. Lett. 1987, 59, 1942-1945. (2) Meyer, G.; Amer, N. M. Appl. Phys. Lett. 1990, 57, 2089-2091. (3) Overney, R.; Meyer, E. MRS Bull. 1993, (May), 27-34. (4) Kim, Y.; Lieber, C. M. Science 1992, 257, 375-377. (5) Scandella, L.; Schumacher, A.; Kruse, N.; Prins, R.; Meyer, E.; Lu ¨ thi, R.; Howald, L.; Gu ¨ ntherodt, H.-J. Thin Solid Films 1994, 240, 101-104. (6) Schumacher, A.; Kruse, N.; Prins, R.; Meyer, E.; Lu ¨ thi, R.; Howald, L.; Gu ¨ ntherodt, H.-J.; Scandella, L. J. Vac. Sci. Technol. B 1996, 14, 12641267.
2452 Analytical Chemistry, Vol. 71, No. 13, July 1, 1999
GaAs/InGaAs,9 InSb/InP(001),10 C60/GeS,11 Si/In,12 SiO2/Si(Hterminated),13 glass/PbS,14 and Si/SnO2.15 In addition, several papers have shown chemical imaging of organic LangmuirBlodgett-type layers containing patterned or randomly separated domains of surfactants with different functional end groups such as CH3, COOH, or CF3.16-19 The chemical recognition function of LFM in all these cases is based on the different friction coefficients of the materials according to the Amontons’s fundamental friction law
µ ) FF/FL
(1)
where µ is friction coefficient, FF friction force, and FL loading force. This law relies on the fact that true contact area at low loads is less than 1% of the apparent contact area. As the loading force increases, the contact area increases due to broadening of existing contacts and increasing amount of multiasperity contacts. In the case of LFM, the tip-sample contact is, however, rather a singleasperity type and thus FF exhibits a nonlinear behavior as a function of FL. It has been found that the Hertzian-type relation (7) Lu ¨ thi, R.; Haefke, H.; Meyer, E.; Howald, L.; Lang, H.-P.; Gerth, G.; Gu ¨ ntherodt, H.-J. Z. Phys. B 1994, 95, 1-3. (8) Bratina, G.; Vanzetti, L.; Franciosi, A. Phys. Rev. B 1995, 52, R8625-R8628. (9) Tamayo, J.; Gonzalez, L.; Gonzalez, Y.; Garcia, R. Appl. Phys. Lett. 1996, 68, 2297-2299. (10) Tamayo, J.; Garcia, R.; Utzmeier, T.; Briones, F. Phys. Rev. B 1997, 55, R13436-R13439. (11) Allers, W.; Hahn, C.; Lo¨hndorf, M.; Lukas, S.; Pan, S.; Schwarz, U. D.; Wiesendanger, R. Nanotechnology 1996, 7, 346-350. (12) Marti, O. Phys. Scr. 1993, T49, 599-604. (13) Scandella, L.; Meyer, E.; Howald, L.; Lu ¨ thi, R.; Guggusberg, M.; Gobrecht, J.; Gu ¨ ntherodt, H.-J. J. Vac. Sci. Technol. B 1996, 14, 1255-1258. (14) Resch, R.; Friedbacher, G.; Grasserbauer, M.; Kanniainen, T.; Lindroos, S.; Leskela¨, M.; Niinisto ¨, L. Appl. Surf. Sci. 1997, 120, 51-57. (15) Utriainen, M.; Lehto, S.; Niinisto¨, L.; Du ¨ csoˆ, Cs.; Khanh, N. Q.; Horva´th, Z. E.; Ba´rsony, I.; Pe´cz, B. Thin Solid Films 1997, 297, 39-42. (16) Overney, R.; Meyer, E.; Frommer, J.; Brodbeck, D.; Lu ¨ thi, R.; Howald, L.; Gu ¨ ntherodt, H.-J.; Fujihara, M.; Takano, H.; Gotoh, Y. Nature 1992, 359, 133-135. (17) Frisbie, C. D.; Rozsnyai, L. F.; Noy, A.; Wrighton, M. S.; Lieber, C. M.; Science 1994, 265, 2071-2074. (18) Wilbur, J. L.; Biebuyck, H. A.; MacDonald, J. C.; Whitesides, G. M. Langmuir 1995, 11, 825-831. (19) Green, J.-B. D.; McDermott, M. T.; Porter, M. D.; Siperko, L. M. J. Phys. Chem. 1995, 99, 10960-10965. 10.1021/ac980325m CCC: $18.00
© 1999 American Chemical Society Published on Web 05/11/1999
of FF ∝ FL2/3 is valid in this case20 and the friction force can be expressed by21
FF ) sA ) sπ(RFL/K)2/3
(2)
where s is the shear stress of the contact points, A is the contact area, R is the tip radius, and K is the effective elastic modulus of contact given by
K ) 4/3((1 - ν12)/E1 + (1 - ν22)/E2)-1
(3)
where νi is Poisson’s ratio and Ei is the Young’s (elastic) modulus of the tip (i ) 1) and the sample (i ) 2). Equations 2 and 3 express the chemical recognition function through material-dependent elastic properties which, on the other hand, can vary considerably due to variations in microstructure. Besides, between two adhering surfaces a finite friction can also exist without any externally applied load, leading to the following modification of eq 1:
FF ) µ(F0 + FL)
(4)
where F0 is the sum of adhesive forces between two surfaces. A contribution of this adhesive force to the friction force can be significant in nanonewton-range loading forces, which are typically applied in scanning force microscopy (SFM) studies. Although F0 may incorporate various long- and short-range electromagnetic interactions, we assume that F0 ) Fcap is valid, where Fcap is the capillary force due to condensed water in humid air. Although this assumption may be highly simplified, it is supported by following facts: the magnitude of the pull-off force (and friction force) depends on (i) the hydrophilicity/hydrophobicity of the surface as well as (ii) on the relative humidity of the ambient. Furthermore, (iii) the disappearance or at least a significant reduction of the force-distance curve hysteresis is obtained by immersing the system in a dry gas atmosphere or into liquids. 22-26 The evaluation of the capillary force has been presented by Thundat et al. 24 with fairly good correlation to the pull-off force measurements. They proposed the following equation for the capillary force of water:
Fcap ) 4πRγ cos θ/(1 + D/(d(H) - D))
(5)
where R is the tip radius, D is the tip-sample distance, θ is the contact angle, γ is the surface tension of water, H is the relative humidity, and d (nm) ) -1.08 cos θ/ln H. (20) Putman, C. A. J.; Igarashi, M.; Kaneko, R. Appl. Phys. Lett. 1995, 66, 32213223. (21) Schwartz, U. D.; Zwo¨rner, O.; Ko¨ster, P.; Wiesendanger, R. Phys. Rev. B 1997, 56, 6987-6996. (22) Binggeli, M.; Mate, C. M. J. Vac. Sci. Technol. B 1995, 13, 1312-1315. (23) Binggeli, M.; Mate, C. M. Appl. Phys. Lett. 1994, 65, 415-417. (24) Thundat, T.; Zheng, X.-Y.; Chen, G. Y.; Warmack, R. J. Surf. Sci. 1993, 294, L939-L943. (25) Friedbacher, G.; Bouveresse, E.; Fuchs, G.; Grasserbauer, M.; Schwarzbach, D.; Haubner, R.; Lux, B. Appl. Surf. Sci. 1995, 84, 133-143. (26) Burnham, N. A.; Colton, R. J.; Pollock, H. M. Nanotechnology 1993, 4, 6480.
According to eqs 4 and 5, the chemical recognition function of the LFM at zero load resides in the terms of the energy of solid/liquid interface. The experimental approach presented in this work appears not to have been often used. Namely, we apply LFM to reference samples prepared by etching or lift-off processing a step on a thinfilm/substrate structure. This appears to be a convenient way to obtain structures with clearly distinguishable domains of different substances. As long as absolute friction coefficients are not available for chemical identification by LFM, methods based on reference material may be the only reliable ones for analytical purposes. The patterned specimens in this study were atomic layer epitaxy (ALE) processed thin films of SnO2, CeO2, and CaS on Si(100) substrates followed by appropriate chemical etching and a lift-off processed PVD Pt thin film on sapphire (11h02) substrate. We measured the static contact angles of water on these surfaces and shall discuss, according to these data, the contribution of the capillary force to the chemical contrast in the LFM images. In addition, the influences of some other potential factors including loading force, surface treatment by organic solvents, and effect of tip material are discussed as well. The modified tips were obtained by coating the silicon-containing AFM microprobes by ALE SnO2. This enabled us to study the effect of both Si and SnO2 as tip materials on the LFM images of the patterned SnO2/Si specimen. The principal difference between chemical vapor deposition (CVD) and ALE27,28 processes may give rise to fascinating new applications for the present technology compared to those CVD-type thin-film processes that have been adapted to AFM probe modification, so far. 29,30 EXPERIMENTAL SECTION SFM Measurements. The SFM instrument used was a DME Rasterscope 4000. The measurements were done by rectangular silicon probes with normal force constants of (i) 0.2 (Oriel Nanosensors, Pointprobe Contact Mode) and (ii) 40 N/m (DME, type 2043). The cantilever dimensions given by the manufacturers were as follows: for (i) w ) 50 µm, d ) 2 µm, L ) 450 µm, and H ) 12.5 µm and for (ii) w ) 38 µm, d )7 µm, L ) 225 µm, and H ) 12,5 µm, where w is the width, d is the thickness, L is the length, and H is the tip height. The given maximum value for the radius of the tip was 15 nm (Oriel). Loading force in the constant deflection mode was determined by assuming FL ) 0 when the tip is far away from surface and by using force constants provided by the manufacturers. The LFM images presented in this paper were recorded at a scanning speed of 3 µm/s. The influence of instrumental parameters, viz. intensity and position of the laser beam and scanning speed variation within 1-10 µm/s, was studied systematically with the 0.2 N/m cantilever but no significant effects were observed in the parameter range applied. The directional cross-correlation in the LFM image was eliminated by scanning the image area in both the 90 and 270° directions and applying the friction-loop (27) Suntola, T. In Handbook of Crystal Growth; Hurle, D. T. J.; Elsevier: Amsterdam, 1994; Vol. 3, pp 601-663. (28) Niinisto¨, L.; Ritala, M.; Leskela¨, M. Mater. Sci. Eng. B 1996, 41, 23-29. (29) Germann, G. J.; McClelland, G. M.; Mitsuda, Y.; Buck, M.; Seki, H. Rev. Sci. Instrum. 1992, 63, 4053-4055. (30) Niedermann, Ph.; Ha¨nni, W.; Blanc, N.; Christoph, R.; Burger, J. J. Vac. Sci. Technol. A 1996, 14, 1233-1236.
Analytical Chemistry, Vol. 71, No. 13, July 1, 1999
2453
Figure 1. (a) Topographic (AFM) and (b) chemical (LFM) maps of the patterned SnO2/Si specimen. The image is taken by using a 0.2 N/m Si probe with 5-nN loading force and 3 µm/s scanning speed. The peak-to-valley height in the AFM image is 206 nm, and the output voltage peak to valley in the LFM image is 13.8 V.
correlation 12,31 to the friction histograms. The LFM images presented in this paper were obtained in the 270° scanning direction. All SFM measurements were performed in laboratory air in a relative humidity of 27-30% (Humicap, Vaisala) and at a temperature of 21-22 °C. Pull-off forces were detected from force-distance measurements of AFM using a cantilever with a spring constant of 0.211 N/m (Oriel). The reported values are averages of 5-10 measurements. Patterned SnO2/Si Specimen. The applied SnO2 film was deposited on n-type Si(100) by ALE using a process described elsewhere.32 The process was performed in a flow-through-type ALE reactor at 1 mbar and 500 °C from SnCl4 and H2O precursors. One ALE cycle consisted of precursor pulses (0.2 s for SnCl4 and 0.6 s for H2O) separated by 3-s inert gas (N2) purging pulses. The number of cycles applied was 6000, which resulted in a 200-nmthick SnO2 film. A patterned interface was obtained by etching the SnO2 film to which a photoresist mask had been applied using the following wet chemical process: (1) dispersing the sample in Zn/HCl (6%) for 20 min, (2) HCl (6%), (3) H2O, (4) HNO3 (13%), and (5) H2O and then repeating steps 4 and 5 two or three times. After these steps, the photoresist was removed in acetone. The sample was cleaned by ultrasonic bath while immersed in ethanol, rinsed with H2O, and then blown dry with air. SnO2 Tip. The 0.2 N/m (Oriel) silicon microprobes were covered with SnO2 layers using the same ALE deposition process of SnO2 thin film32 as described above. The thicknesses of the SnO2 films on the Si probes were 30, 100, and 470 nm as determined by UV/visible spectrophotometry with appropriate modeling.33 A Zeiss DSM 962 scanning electron microscope was used for tip examination. Other Patterned Specimens. The etching process for creating the film/substrate interface was also performed with ALEdeposited CaS and CeO2 thin films (both about 200 nm thick) on Si(100) substrate. The process details are given in ref 34 for CaS and in ref 35 for CeO2. For CeO2, the etching procedure was the following: (1) SnCl2‚2H2O + HCl (6%), (2) H2O, (3) HCl (6%), and (4) H2O. In the case of CaS, etching was made without photoresist by simply dipping the sample in to 6% HCl. The last 2454 Analytical Chemistry, Vol. 71, No. 13, July 1, 1999
sample containing film/substrate step was lift-off patterned Pt film, deposited by sputtering on sapphire (1 1h02). Contact Angle Measurements. The static contact angles were measured using 10-µL droplets of deionized water. In addition to those samples analyzed by LFM in this study, the contact angles of some other relevant ALE-deposited metal oxide thin films (Al2O3, Ga2O3, NiO, CuOx, La2O3, Sb2O5, TiO2, Ta2O5, HfO2, ZrO2, Nb2O5) were measured, too, to gain a general view of contact angles formed on inorganic substances. Thickness of these films varied between 100 and 300 nm, and they were deposited on soda-lime glass or on Si substrates. The preparation processes of these films are described elsewhere.28,36-42 RESULTS AND DISCUSSION Some General Observations Using 5-nN Loading Force. AFM and simultaneously measured LFM images showing the SnO2/Si interface are presented in parts a and b of Figure 1, respectively. Domains are clearly identifiable, although high interface friction occurred in the narrow SnO2/Si interface regime (Figures 1b and 2a,b). Evidence that friction can be higher at high free energy places like steps is also reported in the literature.2 Similar interface effects were also observed in other measurements of patterned specimens. The LFM images shown in Figure 1b represent the case when the fresh SnO2/Si sample was cleaned in ethanol. After storage (31) Schwartz, U. D.; Ko¨ster, P.; Wiesendanger, R. Rev. Sci. Instrum. 1996, 67, 2560-2567. (32) Viirola, H.; Niinisto¨, L. Thin Solid Films 1994, 249, 144-149. (33) Ylilammi, M.; Ranta-aho, T. Thin Solid Films 1993, 232, 56-62. (34) Rautanen, J.; Leskela¨, M.; Niinisto ¨, L.; Nyka¨nen, E.; Soininen, P.; Utriainen, M. Appl. Surf. Sci. 1994, 82/83, 553-558. (35) Mo ¨lsa¨, H.; Niinisto ¨,L. Mater. Res. Soc. Symp. Proc. 1994, 335, 341-349. (36) Kukli, K.; Ihanus, J.; Ritala, M.; Leskela¨, M. J. Electrochem. Soc. 1997, 144, 300-306. (37) Nieminen, M.; Niinisto¨, L.; Rauhala, E. J. Mater. Chem. 1996, 6, 27-31. (38) Utriainen, M.; Kro ¨ger-Laukkanen, M.; Niinisto ¨, L. Mater. Sci. Eng. B 1998, 54, 98-103. (39) Seim, H.; Mo ¨lsa¨, H.; Nieminen, M.; Fjellva˚g, H.; Niinisto ¨, L. J. Mater. Chem. 1997, 7, 449-454. (40) Kukli, K.; Ritala, M.; Leskela¨, M. Nanostruct. Mater. 1997, 8, 785-793. (41) Ritala, M.; Leskela¨, M.; Nyka¨nen, E.; Soininen, P.; Niinisto ¨, L. Thin Solid Films 1993, 225, 288-295. (42) Viirola, H.; Niinisto¨, L. Thin Solid Films 1994, 251, 127-135.
Table 1. Static Contact Angles of Water on the ALE-Deposited Thin Films and Common Substrate Materials Measured Directly after Washing in Ethanol and after Prolonged (More Than 36 h) Aging in Laboratory Air contact angle/deg material substrates Si(100) with native oxide porous Si, p+Si(100) porosity ∼50% Muscovite mica soda-lime glass sapphire (1 1h02) (R-Al2O3) ALE thin films CaSa Sb2O5 NiO SnO2 TiO2 Nb2O5 Ta2O5 La2O3 CeO2 Ga2O3 HfO2 Al2O3 ZrO2 CuOx Pt a
EtOH treated
after aging
22 ( 10 23 ( 10
