Conductive SU8 Photoresist for Microfabrication

FULL PAPER Conductive SU8 Photoresist for Microfabrication** By SØbastien Jiguet,* Arnaud Bertsch, Heinrich Hofmann, and Philippe Renaud A conductive composite photoresist has been developed for the direct photopatterning of electrodes. It is based on a dispersion of silver nanoparticles in SU8, a non-conductive, negative-tone photoresist. Manufactured structures have an electrical conductivity at a low silver content of around 6 vol.-%. 1. Introduction Table 2. Examples of photosensitive ECPCs. A dispersion of a conductive filler (a metallic or inorganic powder or conductive polymer) in a non-conductive polymer matrix has the particular effect of producing a composite material with an enhanced electrical conductivity compared to the conductivity of the matrix. Electrically conductive polymer composites (ECPCs) (see Table 1) have been investigated for use in various fields: in inks, paints, coatings, electronics, etc., and for electrical or electromagnetic applications.[1,2] Matrix Table 1. Examples of thermally polymerized ECPCs. Matrix HDPE PVC Epoxy Epoxy Epoxy Epoxy Epoxy Epoxy Nature Filler characteristics Shape Carbon black Carbon black Carbon black Carbon Al, Cu, Zn Ni Ag Ag ± Sphere Sphere Nanotube Sphere Sphere Sphere Needle Applications Refs. Size ± Electromagnetic 29 nm Electrical conductivity 90±100 nm Percolation studies ± Research 5±15 lm Percolation studies 10±15 lm PTC effect 10 lm Electrical conductivity ± Electrical conductivity [24] [25] [26] [27] [28] [4] [29] [3] ECPCs based on epoxy resins are of particular interest for understanding the properties of binary composites (morphology, electrical and thermal conductivity), which are obtained, in most cases, after a heat treatment to crosslink the ECPC materials. Only a few of them are based on a photopolymerizable matrix (see Table 2), and are essentially a dispersion of silver particles in an acrylate polymer. The main application of these ECPCs is as conductive adhesives in electronicsÐfor connecting and bondingÐbut they have also been used in the manufacture of sensors.[3±5] ± [*] Dr. S. Jiguet, Dr. A. Bertsch, Prof. P. Renaud École Polytechnique FØdØrale de Lausanne EPFL-STI-IMM-LMIS4, CH-1015 Lausanne (Switzerland) E-mail: [email protected] Prof. H. Hofmann École Polytechnique FØdØrale de Lausanne EPFL-STI-IMX-LTP, CH-1015 Lausanne (Switzerland) [**] This work was supported by the Swiss National Science Foundation. Adv. Funct. Mater. 0000, 00, 1±7 Nature Acrylate Acrylate/epoxy Acrylate Acrylate Polyimide Ag Ag, Cu on glass Au Ag Fullerene Filler characteristics Shape Size Refs. Percentage Needle Sphere 1”30 lm 10±50 lm 8±40 vol.-% 15±77 vol.-% [30,31] [32] Sphere Sphere Sphere 0.1±5 lm Nano ± ± 33 wt.-% 0.7 wt.-% [33] [13] [34] Tables 1,2 show that conductive composite materials based on metallic particles are principally composed of particles with a mean size in the micrometer range. Their electrical conductivity becomes significant for filler concentrations higher than 8 vol.-%. Herein, we report a new, conductive, photosensitive, composite material, containing at least 6 vol.-% of silver nanoparticles dispersed in SU8, an epoxy photopolymer essentially used for fabricating high-aspect-ratio structures by UV-LIGA. LIGA (or lithographie galvanoformung abformung) is a process for making high-aspect-ratio structures in metal by combining X-ray lithography for patterning a sensitive polymer and a galvanoplastic method.[6,7] Because X-ray lithography is expensive, a UV radiation source is used in place of X-rays in the UV-LIGA process. SU8 components are well-known and widely used for MEMS (microelectromechanical systems), fluidic and packaging applications, and also as masters for microinjection molding.[8±12] The conductive photoresist presented herein can be used in the microfabrication field for the direct manufacture of electrically conductive microcomponents, without the use of traditional metal micropatterning methods (electroplating, sputtering, etc.). 2. Results and Discussion Photosensitive composite materials loaded up to 40 vol.-% have been produced to investigate the silver filler effect on the photopolymerization phenomenon and on the electrical conductivity of the composite photopatterned structures. 2.1. Formulations The SU8/silver composite photoresists were obtained after homogeneously mixing the silver powder with the photosensi- DOI: 10.1002/adfm.200400575  2005 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim 1 tive SU8 resin. As the particle size is in the nanometer range, adding the powder into the liquid resin results in an important increase in the formulation viscosity. Thus, it is necessary to add a certain quantity of solvent (the solvent of the photoresist) to control both the viscosity and the powder sedimentation phenomenon. It has been observed from rheological characterizations that the rheological aspect of these formulations evolves from a plastic behavior to a shear thinning behavior on increasing the volume fraction of silver. Consequently, SU8/silver formulations cannot be spread on quartz wafers by spincoating, and a scraper was thus used to deposit films of the composite resist on the wafers. 2.2. Photopolymerization To obtain electrically conductive photosensitive composite materials, different kinds of filler can be chosen: oxide, metal, or electrically conductive polymer. Silver particles were used in the present study, because of their good electrical conductivity and their low absorption at the irradiation wavelength (365 nm) in comparison to other metallic fillers or oxides. However, silver particles have a strong influence on photopolymerization. Figure 1 shows the changes in the polymerized thickness versus the filler concentration for various irradiation doses. The polymerized thickness was measured by surface pro- 50 Polymerized thickness (µm) -2 900 mJ.cm 40 35 10 vol.% of silver 30 25 20 15 10 20 wt.% photoinitiator 15 wt.% photoinitiator 5 wt.% photoinitiator 5 -2 600 mJ.cm -2 400 mJ.cm 0 30 0 500 1000 1500 2000 -2 Dose (mJ.cm ) 20 Figure 2. Evolution of the polymerized thickness versus the dose and the photoinitiator concentration, for a 10 vol.-% silver composite. The cured depth increases with the dose (as a logarithmic evolution), and with the photoinitiator weight fraction. 10 0 0 5 10 15 20 25 30 35 Silver volume fraction (%) Figure 1. Change in the polymerized thickness versus the silver powder load in the SU8/silver photosensitive composite material. The incident energy ranged from 400 to 900 mJ cm±2. A drastic decrease in thickness was observed on increasing the filler load: this was due to the absorbance and reflectance of silver particles. Higher doses allowed an increase in the cure depth. filometry on structures polymerized in back-side mode on quartz wafers, after their development in the appropriate solvent. For an irradiation dose of 400 mJ cm±2, the polymerized thickness decreased drastically around the percolation threshold (6 vol.-%), as seen in Figure 1. Layer thicknesses in the region of 500 lm for pure SU8 (thick layers were obtained by successive spin-coating) were obtained by using previously 2 described irradiation conditions,[8] whereas adding 3 vol.-% of silver nanoparticles decreased the thickness from 500 to 35 lm. For highly loaded composite, the layer thickness is only 2 lm. This change is related to the absorbance and reflectance of the silver filler, which governs the global absorption and reflection of the light in the composite resist, when the silver content is high enough, as is the case in the formulated media studied here.[13] A higher polymerized thickness can be obtained by applying a higher irradiation dose. In a composite containing 10 vol.-% of filler, the cured depth increased to 12 lm on increasing the irradiation dose to 800 mJ cm±2. A second way to modify the polymerized thickness is to change the photoinitiator concentration. Investigations were made on 10 vol.-% loaded composites with three photoinitiator weight fractions (5, 10, and 15 %); the SU8 photoresist contains 10 wt.-% of photoinitiator (triarylsulfonium hexafluoroantimonate salt; Aldrich). Figure 2 shows a logarithmic evolution of the polymerized thickness versus the dose for Polymerized thickness (µm) FULL PAPER S. Jiguet et al./Conductive SU8 Photoresist for Microfabrication  2005 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim these three compositions. In the dose range 0±400 mJ cm±2, there are no data points because the photoresist is not sufficiently polymerized to allow good adhesion between the composite film and the substrate, consequently after the development step all exposed parts are removed. The logarithmic trend can be explained by a traditional photopolymerization model derived from the Beer±Lambert law and based on a threshold system. The polymerized thickness, Ep, can be expressed by Equation 1[14]   D Ep ˆ Pd ln (1) Dc where Pd is the penetration depth, which is dependent on the absorption of each product contained into composite; D is the dose; and Dc is the critical dose necessary to begin the polymerization. www.afm-journal.de Adv. Funct. Mater. 0000, 00, 1±7 S. Jiguet et al./Conductive SU8 Photoresist for Microfabrication   1 D ln aPh CPh aAg CAg tIa (2) where aPh and aAg are the molar extinction coefficients, respectively, of the photoinitiator and the silver powder; CPh and CAg are the molar concentrations, respectively, of the photoinitiator and the silver powder; D is the dose; t is the minimum time to begin the polymerization; and Ia is the absorbed intensity of the photochemical system. This equation has been obtained by combining the Beer±Lambert law of absorptionÐapplied to the silver particles only (and considering there is no contribution of the photoinitiator to the light absorption)Ðand the photopolymerization as a threshold system. Equation 2 shows that the cured depth is directly proportional to the Naperian logarithm of the dose and of the photoinitiator concentration, and conforms to the experimental results presented in Figure 2, and also in Figure 3, which shows the evolution of the polymerized thickness versus the photoinitiator concentration. The logarithmic trend is similar, whatever the silver filler concentration. As aAg = 105aPh (from literature and experiments, at 365 nm), the approximation in which the contribution of the photoinitiator concentration to the absorption is neglected can be considered true. Adv. Funct. Mater. 0000, 00, 1±7 40 15 vol.% silver 10 vol.% silver 7 vol.% silver Polymerized thickness (µm) 35 30 25 20 15 10 5 1 10 100 Photoinitiator weight fraction (%) Figure 3. Evolution of the polymerized thickness versus the photoinitiator concentration for different silver volume fractions. The cured depth evolves as a logarithmic function of the photoinitiator weight fraction, and it decreases by increasing the filler concentration. The irradiation dose was 900 mJ cm±2. 2.3. Electrical Conductivity Increasing the amount of conductive filler in an insulating SU8 epoxy matrix enhances the electrical conductivity of the resulting composite. This is related to the formation of completely interconnected clusters, composed of silver particles in contact with each other, which have the same size as the dimensions of the sample.[17] It means that the composite sample becomes electrically conductive when there are enough silver particles in the insulating medium to obtain percolation between the particles: A conductive path through the sample occurs. By measuring the electrical resistivity of the composite structures obtained by photolithography of the composite resists, a sudden decrease appears at the percolation threshold, around 6 vol.-% (see Figure 4). The dielectric properties of the 10 4 10 2 10 0 st Before 1 thermal treatment Resistivity (Ω.cm) Ep ˆ FULL PAPER In comparison to the dose effect previously presented, increasing the photoinitiator weight fraction results in a larger polymerized layer. The influence of the photoinitiator concentration on the polymerized thickness seen in the case of the composite resist studied here is very different from the evolution of these parameters for more conventional negative-tone polymer resists: In traditional photopolymerization models, the cured depth decreases when the volume fraction of photoinitiator increases, owing to a higher absorption of the light at the surface of the film, which results in a lower light intensity transmitted deep under the surface.[14±16] Here, the evolution is opposite, and can be explained by the presence of particles in the matrix: The medium is not a homogenous system, and the silver particles indirectly affect the photopolymerization. The incidence of the initiation on the absorbance of the composite can be considered extremely small compared to the very high absorbance of the silver particles, but the silver particles do not contribute to the initiation of the polymerization phenomenon. The studied system can consequently be considered as a system wherein the absorption of light by the medium and the initiation of polymerization are not linked: The strong absorption of the silver particles governs the absorption of the light in the material. For a given silver content, the way the light is absorbed in the resist layer can be considered the same, regardless of the initiator concentration. The increase in the photoinitiator concentration does not affect the absorption of the light, but increases the probability that polymerization will start deeper in the resist layer. A simple model can be used to describe this behavior. The polymerized thickness Ep can be expressed by Equation 2 10 -2 10 -4 10 -6 st After 1 thermal treatment 0 10 20 30 40 50 Silver volume fraction (%) Figure 4. Evolution of the electrical resistivity of the SU8-silver composite versus silver volume fraction, before and after thermal treatment (isotherm of 90 min at 120 C). A drastic switch in resistivity was observed at the percolation threshold (6 vol.-%), and an electrical conductivity enhancement was obtained after the thermal treatment. www.afm-journal.de  2005 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim 3 composite evolve from insulating to conductive on increasing the filler content. In the case of a regular array on which spherical particles are distributed statistically, simple binary composites show a percolation threshold at volume fractions around 16 vol.-%.[18,19] This difference between theory and measurement is principally related to the shape and size distribution of the silver clusters, as discussed previously.[20±22] The application of a soft thermal treatment (an isotherm at 120 C for 90 min) to the conductive composite structures induced a supplementary decrease in the electrical resistivity of more than one order of magnitude. This is related to the evaporation of a volatile product contained in the composite, which acts as an insulating film around the silver particles.[20] The film thickness does not change further after this thermal treatment. To determine if, after the thermal treatment, the conductive composites have a stable resistivity with temperature, an impedance characterization was made. The evolution of the impedance of a 6 vol.-% loaded composite versus temperature is presented in Figure 5. The first thermal treatment, a ramp from 20 to 120 C at a rate of 5 C min±1, followed by an isotherm of 90 min at 120 C, resulted in a decrease in the impedance with temperature to 42 X, essentially around temperatures of 70± 10 6 10 5 2.4. Structures and Resolution To use this conductive photoresist for microfabrication, resolution tests were performed. Structures in SU8/silver composite have been produced by the UV-LIGA process in back-side and front-side modes. The back-side mode has been preferred because of the better adhesion between the composite microstructures and the substrate. However, because of diffraction, the lateral resolution of structures obtained without direct contact between the mask and the resist was low, around 30 to 50 lm, depending principally on the dose and not on the filler load, as shown in Figure 6. To avoid this problem, a mask in amorphous silicon was patterned at the surface of the quartz wafer, in order 60 50 Lateral resolution (µm) FULL PAPER S. Jiguet et al./Conductive SU8 Photoresist for Microfabrication 40 30 D = 370 mJ.cm-2 20 D = 550 mJ.cm-2 -2 D = 740 mJ.cm D = 830 mJ.cm-2 10 nd Impedance (ohm) 2 thermal treatment 10 0 4 0 5 10 15 20 25 Silver volume fraction (%) 10 3 10 2 10 1 10 0 Figure 6. Variation of the lateral resolution of the test structures with the filler load and the dose, and comparison with the structures photopatterned using an integrated mask (IM). 0 50 100 150 200 250 300 Temperature (ºC) Figure 5. Evolution of the impedance versus temperature during two successive thermal treatments: ramp of 5 C min±1 from 20 to 120 C and isotherm of 90 min at 120 C (first treatment); ramp of 5 C min±1 from 20 to 250 C (second treatment). 80 C. This is related to the evaporation of volatile products used during the development step (developer, rinsing product). After cooling the samples to room temperature, a second thermal treatment (a ramp from room temperature to 250 C at 5 C min±1) was applied. This resulted in a stable impedance of the sample for temperatures lower than 150 C. At higher temperatures, the impedance decreases again, because of the new mobility of the polymer chains (corresponding to the glass transition). Thus, the conductive composite structures have stable dielectric properties from room temperature to their glass-transition temperatures. 4 D = 830 mJ.cm-2 IM 1st thermal treatment  2005 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim to have direct contact between the composite photoresist and the mask. In that configuration, diffraction is lower and the resolution is significantly improved: it becomes better than 5 lm. The aspect ratio of the resulting conductive composite structures is less than 1 for the standard process, and it was improved to 5 by using an integrated mask (see Figure 7). In comparison to the aspect ratio of SU8, which is around 20, conductive photoresists have a lower aspect ratio, and they allow the direct microfabrication of conductive microstructures. 2.5. Applications SU8/silver composites are electrically conductive for loads above the percolation threshold. Their electrical conductivity ranges from 10 to 104 S cm±1. As the electrical conductivity of bulk silver is around 105 S cm±1, the SU8/silver composites have an interesting conductivity range for use as materials in the microfabrication of electrodes. Figure 8a shows an electrode array fabricated by photolithography in back-side mode on quartz substrates using a 20 vol.-% loaded composite photoresist. www.afm-journal.de Adv. Funct. Mater. 0000, 00, 1±7 S. Jiguet et al./Conductive SU8 Photoresist for Microfabrication -2 D = 370 mJ.cm 4 -2 D = 550 mJ.cm Aspect ratio D = 740 mJ.cm-2 D = 830 mJ.cm-2 3 -2 D = 830 mJ.cm IM 2 1 FULL PAPER silver nanoparticles embedded in SU8, a photosensitive negative-tone resist. When sufficient silver powder is added to the insulating matrix, the composite material becomes electrically conductive. This phenomenon appears for silver volume fractions above the percolation threshold, which is around 6 vol.-%. Depending on the silver volume fraction, the composite structures show a wide range of electrical conductivity. A soft thermal treatment can further enhance the conductivity by one order of magnitude, and the resulting structures have stable dielectric properties on a temperature range limited by the bake temperature. The SU8/silver structures made by UVLIGA in back-side mode and with an integrated mask show a resolution better than 5 lm, and a maximum aspect ratio of 5. 5 0 0 5 10 15 20 Silver volume fraction (%) 25 Figure 7. Evolution of the aspect ratio with the filler load and the dose, and comparison with the structures photopatterned using an IM. The electrode thickness is 2 lm for a dose of 740 mJ cm±2. At higher dose the electrodes are thicker, but the resolution is lower. The contact pads situated at the center of the array (Fig. 8b) are linked. After depositing a droplet of silver paint on the central contact pads, the resistance was measured at the opposite electrodes using an ohmmeter. It ranged from 800 X to 10 kX, mainly because of the inhomogeneity of the composite structures, whereas for platinum, gold, and indium tin oxide (ITO) electrodes, the resistances were, respectively, 200 X, 250 X, and 1.8 kX. To decrease the electrode resistance, a composite with a higher load of silver particles can be used. This first test has shown the feasibility of directly photopatterned electrodes. 4. Experimental Composite Photoresist Formulation: The conductive composite photoresist is a blend obtained by homogeneously mixing the photosensitive SU8 resin (from Gersteltec, Lausanne, Switzerland) with the silver nanoparticles (PUAG119 from METALOR, Neßchatel, Switzerland) first by sonication (Elma, 25 kHz) and then with a mixer (Diax 900, Heidolph). The silver powder has a mean particle size (clusters) of 1.5 lm; the primary particle diameter is in the nanometer range. The main characteristics of the silver powder are presented in Table 3. Table 3. Physical characteristics of the silver powder PUAG119 from METALOR. Characteristic Density Specific area Size distribution at 10 % Size distribution at 50 % Size distribution at 90 % Value 2.5 g cm±3 1.84 m2 g±1 0.2 lm 1.5 lm 2.5 lm 3. Conclusion In order to avoid sedimentation as well as inhomogeneous formulations, the volume fraction of the liquid phase in the formulations was A conductive composite photoresist has been used for the diadapted by adding c-butyrolactone (from Fluka) as solvent of the rect microfabrication of electrodes by UV-LIGA. It is based on SU8 polymer. For example, a formulation containing 5 vol.-% of silver powder requires 35 to 40 vol.-% of added solvent to avoid sedimentation, and a formulation loaded at 20 vol.-% in silver requires a) b) 20 to 25 vol.-% of solvent. The formulation viscosity, determined with a rheometer (rheostress 100 Haak, in (120) cone-plan configuration), ranged from 5 to 40 Pa s at 25 C, for a shear rate of 50 s±1. The formulation viscosity decreases with increasing silver powder content in the final mixing and, from rheological characterizations, it was found that the rheological aspect of these formulations evolves from a plastic behavior to a shear thinning behavior. Microfabrication Process: Because of the shear thinning rheological behavior of the composite formulations, they cannot be spread on wafers by spin-coating, so a scraper was used for this step. The layer thickness was adjusted by using a mold of known thickness. Quartz waFigure 8. Electrodes in a silver/SU8 composite loaded at 20 vol.-% of silver powder. a) 2 lm thick fers were employed to expose the photoresist in electrode array, b) magnification of the center array. the back-side mode through the wafer. This Adv. Funct. Mater. 0000, 00, 1±7 www.afm-journal.de  2005 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim 5 FULL PAPER S. Jiguet et al./Conductive SU8 Photoresist for Microfabrication method allowed good adhesion of the polymerized structures on the substrate after the development step. To manufacture the conductive composite structures, the UV-LIGA process, consisting of the following steps, was applied. Quartz wafers were cleaned by oxygen plasma (7 min at 500 W, 400 mL min±1 of oxygen) before spreading the composite photoresist on them. The composite photoresist film was then prebaked at 95 C for at least 30 min to evaporate the solvent. The deposited layer thickness, measured by surface profilometry (Alpha-Step 500, TENCOR), was 200 lm. The wafers were then exposed through a mask with a Karl Suess UV Mask Aligner, in back-side mode. The irradiation wavelength was 365 nm. The exposed photoresist was baked for 15 min on a hotplate at 95 C, in order to crosslink the resist. The photopatterned structures were developed by dissolving the unexposed photoresist in the appropriate developer (propylene glycol methyl ether acetate: PGMEA, from Fluka). The development took a few minutes, depending on the structure geometry and silver concentration of the formulations. Sonication was sometimes necessary to improve the development speed and to obtain cleaned structures. Finally, the composite structures were rinsed with isopropyl alcohol. Photopolymerization Characterization: To investigate the influence of several parameters (irradiation dose, silver or photoinitiator concentration) on the photopolymerization of the composite photoresists, back-side irradiation was used, as the thickness of the structures obtained after development correspond to the cured depth. Multiple expositions were applied on a single wafer to study the influence of the irradiation dose on the polymerized thickness. The resulting structure thicknesses were measured by surface profilometry (Alpha-Step 500, TENCOR). Electrical Characterization: The electrical resistivity of the conductive composite was evaluated by four-point measurements with a resistivity meter (FPP-5000, VEECO), knowing the thickness of the composite films. The electrical behavior of all SU8/silver composites was characterized by impedance spectrometry, also performed in a fourpoint configuration. For this last method, an LCR-meter (HP4284A) at a fixed frequency of 1 kHz was used. Resolution Characterization: The back-side irradiation resulted in good adhesion between the substrate and the conductive composite structures; but the main drawback is the low resolution of the patterned structures: As there is no contact between the mask and the resist, the diffraction phenomenon greatly reduces the resolution. To have both a good adherence of the patterned structures on the substrate and a good resolution, the mask can be defined directly on the substrate [23]. This was obtained by depositing a 1.5 lm layer of amorphous silicon on a quartz wafer, and patterning it with a standard dry-etching process using a photopatterned positive photoresist layer (4 lm thick) as the mask. Received: December 8, 2004 Final version: April 19, 2005 ± [1] V. E. Gul', Structure and Properties of Conducting Polymer Composites, New Concepts in Polymer Science, Brill Academic Publishers, The Netherlands 1996. [2] S. K. Bhattacharya, Metal-Filled Polymers: Properties and Applications, Marcel Dekker, New York 1986. [3] D. Lu, Q. K. Tong, C. P. Wong, IEEE Trans. Electron. Packag. Manuf. 2000, 22, 228. [4] R. Strumpler, J. Appl. Phys. 1996, 80, 6091. [5] M. Sun, Microelectron. J. 2001, 32, 197. [6] E. Spiller, R. Feder, J. Topalian, E. Castellani, L. Romankiw, M. Heritage, Solid State Technol. 1976, 19, 62. [7] E. W. Becker, W. Ehrfeld, D. Munchmeyer, H. Betz, A. Heuberger, S. Pongratz, W. Glashauser, H. J. Michel, R. Vonsiemens, Naturwissenschaften 1982, 69, 520. [8] H. Lorenz, M. Despont, N. Fahrni, J. Brugger, P. Vettiger, P. Renaud, Sens. Actuators, A 1998, 64, 33. [9] H. Lorenz, M. Despont, N. Fahrni, N. LaBianca, P. Renaud, P. Vettiger, J. Micromech. Microeng. 1997, 7, 121. [10] S. Metz, S. Jiguet, A. Bertsch, P. Renaud, Lab Chip 2004, 4, 114. [11] M. O. Heuschkel, L. J. GuØrin, B. Buisson, D. Bertrand, P. Renaud, Sens. Actuators, B 1998, 48, 356. [12] J. Thaysen, A. D. Yalçinkaya, P. Vettiger, A. Menon, J. Phys. D: Appl. Phys. 2002, 35, 2698. [13] R. C. Krohn, WO Patent 0 151 567, 2001. [14] P. F. Jacobs, Rapid Prototyping and Manufacturing: Fundamentals of Stereolithography, Society of Manufacturing Engineers, Dearborn 1992. [15] S. P. Pappas, UV Curing: Science and Technology, Technology Marketing Corporation, Fargo, North Dakota 1980. [16] P. F. Jacobs, Stereolithography and Other RP and M Technologies, Society of Manufacturing Engineers, New York 1996. [17] R. Biller, J. Phys. 1985, 18, 989. [18] S. Kirkpatrick, Rev. Mod. Phys. 1973, 45, 574. [19] R. Zallen, The Physics of Amorphous Solids, Wiley, New York 1983. [20] S. Jiguet, A. Bertsch, H. Hofmann, P. Renaud, Adv. Eng. Mater. 2004, 6, 719. [21] F. Carmona, P. Prudhon, F. Barreau, Solid State Commun. 1984, 51, 255. [22] A. Y. Dovzhenko, P. V. Zhirkov, Phys. Lett. A 1995, 204, 247. [23] M. C. Peterman, P. Huie, D. M. Bloom, H. Fishman, J. Micromech. Microeng. 2003, 13, 380. [24] N. Dishovsky, M. Grigorova, Mater. Res. Bull. 2000, 35, 403. [25] E. K. Sichel, J. I. Gittleman, P. Sheng, Phys. Rev. B 1978, 18, 5712. [26] K. Benaboud, M. E. Achour, F. Carmona, L. Salome, Ann. Chim. Sci. Mater. 1998, 23, 315. [27] L. Valentini, D. Puglia, E. Frulloni, I. Armentano, J. M. Kenny, S. Santucci, Compos. Sci. Technol. 2004, 64, 23. [28] G. M. Tsangaris, M. C. Kazilas, Mater. Sci. Tech. 2002, 18, 226. [29] H.-G. Busmann, B. Gunther, U. Meyer, Nanostruct. Mater. 1999, 12, 531. [30] L. D. Embury, GB Patent 2 111 072, 1982. [31] L. D. Embury, EP Patent 0 081 323, 1983. [32] A. D. Bolon, G. M. Lucas, R. L. Bartholomew, US Patent 3 968 056, 1976. [33] J. J. Felten, US Patent 3 877 950, 1975. [34] Y. Tajima, Y. Shigemitsu, H. Arai, E. Takeuchi, K. Takeuchi, Synth. Met. 2001, 121, 1167. ______________________ 6  2005 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.afm-journal.de Adv. Funct. Mater. 0000, 00, 1±7 S. Jiguet et al./Conductive SU8 Photoresist for Microfabrication Microelectronics Microelectrode arrays have been fabricated (see Figure) photolithographically, using a conductive composite photoresist containing a dispersion of silver nanoparticles. The composite structures made have a resolution of better than 5 lm and a maximum aspect ratio of 5. They show a wide range of electrical conductivity, which is dependent on the silver volume fraction. A soft thermal treatment can further enhance the conductivity by one order of magnitude. Adv. Funct. Mater. 0000, 00, 1±7 S. Jiguet,* A. Bertsch, H. Hofmann, P. Renaud ..................................... j ± j FULL PAPER FULL PAPERS Conductive SU8 Photoresist for Microfabrication www.afm-journal.de  2005 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim 7