In the last decade, two-photon polymerization (2PP) has gained increasing interest for the production of individually shaped 3D structures. For the successful implementation of 2PP within a production chain for the industrial manufacturing of optical microstructures, advanced material properties of the photo polymer are required. The usable laser dynamic range, shape accuracy as well as surface roughness play a crucial role. In this publication, we present the results of an iterative optimization process aiming at a material applicable for 3D structures written by 2PP. The special focus here lies on the influence of the photo initiator and stabilizer on the usable laser dynamic range and shape fidelity. The application of the optimized photo polymer for the production of a complex prism array, which could be used as master mold, is also shown.
© 2015 Optical Society of America
In many areas of research individual microstructures are gaining in importance. Significant application fields are tissue engineering [1–3] and multifunctional hybrid optics for beam shaping and illumination. New design software allows for complex features like high aspect ratios, steep aspheric surfaces, asymmetric geometries, arbitrary freeform surfaces and varying lattice constants. Structuring features like these is very challenging and cannot always be realized by conventional patterning techniques such as cutting technologies or traditional lithographic methods. Two-photon polymerization (2PP) enables the fabrication of very complex 3D-structures due to its unprecedented high degree of freedom and excellent resolution. During the last decade, the photochemical effect of two- (and multi) photon absorption was used to develop a commercial structuring technology for fabricating prototypes of cell scaffolds [2–4], photonic materials [5, 6] and microoptical elements  mainly in the field of research. There is still a need to implement this beneficial technique within industrial production chains. This demands special material properties like a high two-photon absorption ability, a superior resolution and shape accuracy after development as well as strictly reproducible results.
With regard to molding applications, inorganic-organic hybrid polymers obtained via solgel processes have already proven high suitability for realizing masterstamps of microoptical elements . Due to their inorganic parts, they provide excellent mechanical, chemical as well as thermal stability while their organic backbone renders them flexible, less brittle and easy to process. Thus, it is of great interest to establish these materials for 2PP. In this paper, the authors present the results of iterative studies regarding the influence of different photoinitiators (PI) and stabilizer on the 2PP behavior of a commercial hybrid polymer. Based on standard OrmoComp®, type and concentration of photoinitiator (PI), and the concentration of a stabilizing agent were varied. The different modifications were tested concerning their usability in 2PP on glass substrates. The structures were fabricated with the Dip-in Laser Lithography (DiLL), developed by Nanoscribe GmbH, because this special variation of two-photon lithography enables writing on opaque surfaces like metal or silicon and allows heights of the structures up to 300 μm and even more, if the z-drive of the microscope is used (standard configuration enables max. 170 μm heights).
Standard OrmoComp® was stirred with a magnetic stirring bar for 18 h to incorporate the respective PI and, if used, the stabilizing agent 3,5-Di-tert-butyl-4-hydroxytoluene (BHT) (see Table 1). The clear solutions were filtrated through 1.0 μm PTFE filter to remove all remaining particles. Irgacure 819, Irgacure OXE01 and Darocur 4265 were purchased from BASF Schweiz AG, Diphenyl (2,4,6-trimethylbenzoyl)phosphine oxide (TPO) was purchased from Lehmann Voss Co. KG and BHT from Sigma-Aldrich.
For structures written with DiLL technology special glass substrates with dimensions of 25 mm x 25 mm x 0.6 mm were used. To improve the adherence of the structures, the substrate was cleaned with isopropanol and a lint-free cloth. After fixing the substrate in the sample holder, a drop of the corresponding photosensitive polymer was deposited onto the surface. In the following the sample holder was inserted into the two-photon lithography device “Photonic Professional” upside-down to enable fabrication of the structures with DiLL technology. In DiLL technology, infrared laser light with a wavelength of 780 nm is guided through an inverted microscope objective with a numerical aperture of 1.3 and focused on the bottom surface of the substrate. Only in the focus point, the laser intensity is high enough to enable two-photon absorption (providing energy equal to one photon of 390 nm), which occurs within a voxel of about 150 nm lateral size and 400 nm axial size. The photosensitive polymer serving also as immersion liquid is placed between the microscope’s objective and the substrate. By moving the sample relative to the focus in three dimensions, arbitrary geometries can be realized. The high intensity laser light was provided by a pulsed Femtosecond-Fibre-Laser with a maximum output power of 80 mW, a pulse duration of 120 fs and a repetition rate of 80 MHz. After writing the structures with the “Photonic Professional” device, a development step with the developer solution OrmoDev is necessary to free the desired structures from the unexposed polymer. The sample is placed into the developer bath for three minutes and then dried with a gentle flow of air.
Atomic force microscope (AFM) measurements were taken with a Park Scientific Instruments Autoprobe CP-NC AFM equipped with NCH-tips (thickness 4 μm, length 125 μm) from NanoWorld. Data analysis was performed with ProScan Data Acquisition & Image Processing software.
3. Results and discussion
In this work, we investigated the influence of type and concentration of the PI and the concentration of added stabilizer on the resolution, usability and laser dynamics in a 2PP process. The aim was to find an initiating system, which would allow for a high laser dynamic range enabling writing of lines with different thicknesses and heights. Thicker lines are necessary to speed up volumes filling behavior, thin lines are important for a high resolution influencing shape accuracy or radii of curvature and roughness.
3.1. Chemical modification
The PIs were chosen with respect to different aspects: High absorption at 390 nm, high initiating efficiency and commercial availability were key parameters. For example, on the one hand TPO exhibits like most commercially available PIs, only a very low two-photon absorption coefficient at the required 2PP wavelength set by the laser. This could lead to insufficient initiation . On the other hand, TPO is very efficient in generating free radicals for the initiation of the polymerization . Considering these aspects, also Irgacure 819, Darocure 4265 and Irgacure OXE01 were selected to compare the initiation behavior. Although TPO and Irgacure 819 have the lowest absorption coefficient (below 4 GM), their absorption spectrum matches well and the formation of initiator radicals is very efficient . In comparison to other commercial PIs, Darocur 4265 (mixture of TPO and Darocur 1173, 20 GM) and Irgacure OXE01 (31 GM) exhibit high two-photon absorption coefficients but at less suitable wavelengths than the other two PIs. After establishing an optimal amout of initiator, the influence of increasing stabilizer concentration was studied. All tested variations are listed in Table 1. Usually, the addition of stabilizers prevents premature polymerization during storage. In this study, we found that increasing the amount of stabilizer improves also shape accuracy (see section 3.2).
3.2. Dynamics and usability
In the following, we present the results of the experiments with the different variations of OrmoComp®. By writing prisms (edge length of 10 μm) and semispheres (radius of 5 μm) with increasing laser intensities on glass substrates, we draw conclusions concerning the process window and shape accuracy of each sample. Undoubtedly, the intensity threshold for the initiation of the polymerization decreases with increasing two-photon absorption coefficients being lowest for Irgacure OXE01 (4 mW, Fig. 1, sample 3). The threshold increases in order Irgacure OXE01 < TPO ≈ Irgacure 819 < Darocur 4265. Despite its excellent initiation properties, Irgacure OXE01 leads to a reduced resolution and shape fidelity becoming apparent in Fig. 2(b). The test structures run into each other and the initially intended shape is not recognizable any more. It seems as if the circumference of the polymerization is hard to control due to the high amount of produced initiator radicals. This reduces the effective dynamic range drastically from 22 mW to 10 mW for Irgacure OXE01. A peculiar observation is that all initiators except for TPO, produce a kind of platform between the actual structure and the substrate (Fig. 2). Overall, the best results were obtained with TPO which was used for the following investigations. Next, the influence of PI concentration was investigated by varying the amount of PI between 0.75 wt% and 1.25 wt%. Naturally, increasing the TPO concentration leads to lower laser power thresholds and also to broader process windows (Fig. 1, samples 4–6). The SEM-images show better shape accuracy and a more defined contour between the structures and the substrate surface at higher PI concentration (Fig. 3). This could be due to increased formation of initiating radicals per unit volume, leading to a denser polymerization network. After development the structures remain more accurate.
With TPO concentration fixed at 1.0 wt%, the effect of BHT was investigated. The necessary laser intensity to induce the polymerization increases because the BHT has to be consumed before the polymerization can start (Fig. 1, samples 7–9). We observed an improved shape accuracy with higher amounts of BHT as can be seen in the SEM-pictures (Fig. 4). This effect becomes very clear in the radii of curvature at the edges of the prisms: They decrease from 1.5 μm and 1.1 μm down to 0.9 μm with a higher concentration of BHT. In conclusion, the best results were obtained with 1.0 wt % TPO and 0.5 wt% BHT (sample 9) as this mixture provides the best compromise between usable laser intensity range and achievable shape accuracy.
With respect to industrial molding applications (e.g. hot embossing), tests were carried out to apply the results for writing on polished steel substrates which are more heat and pressure resistant than glass. The results show a much narrower dynamic range, e.g. the usable laser power for OrmoComp® 6 ranges from 8 mW to 30 mW on glass and from 6 mW to 12 mW on polished steel. Thus, the dynamic range of 22 mW on glass decreases by 10 mW for the same photosensitive polymer going from glass to steel substrates. Furthermore, the results on metal are less reproducible, which shows a higher sensitivity of the technology on metal.
To demonstrate the feasibility of the 2PP approach, a more complex structure in form of a prism array (designed and analyzed beforehand with VirtualLab™ by LightTrans) was produced on glass. The special features of this prism array are the individual inclination- and tilt-angles of each prism (Fig. 5). Very defined edges and overall smooth prism surfaces were obtained. To confirm the visual impression regarding the surface roughness from the SEM, AFM measurements were performed. Based on SEM, representative prism with different surface smoothness were chosen (Fig. S1). The measurements show Rq-values varying from 25 nm for ”smooth” prisms to 320 nm for ”corrugated” prisms supporting the impressions from the SEM pictures (Fig. 5). The surface quality depends on the orientation and inclination of the prism surface.
4. Conclusion and outlook
Hybrid polymers such as commercially available OrmoComp® are highly suitable for the fabrication of arbitrary shaped structures by 2PP. In order to provide the best possible compromise between writing speed and shape accuracy we investigated the influence of PI and stabilizer. By varying the concentration and type of the PI we can conclude that a higher initiator ratio in OrmoComp® leads to improved shape accuracy and a sharper contour between substrate and structure. The PI Irgacure OXE01 showed the highest sensitivity for two-photon absorption but led to a very narrow effective laser dynamic range and the worst shape accuracy. On metal substrates, which are more industrial relevant than glass substrates, results were ambivalent and not reproducible. Future investigations aim on this problem as well as the replication of 2PP structures by molding techniques. This way, 2PP could serve as reliable fabrication tool for master stamps in the mass production of complex micro structures.
We acknowledge financial support by the Federal Ministry of Education and Research (BMBF) and the Projektträger Karlsruhe (PTKA), funding code: 02PK2294.
References and links
1. A. M. Greiner, M. Jäckel, A. C. Scheiwe, D. R. Stamow, T. J. Autenrieth, J. Lahann, C. M. Franz, and M. Bastmeyer, “Multifunctional polymer scaffolds with adjustable pore size and chemoattractant gradients for studying cell matrix invasion,” Biomaterials 35, 611–619 (2014). [CrossRef]
3. F. Klein, B. Richter, T. Striebel, C. M. Franz, G. v. Freymann, M. Wegener, and M. Bastmeyer, “Two-Component Polymer Scaffolds for Controlled Three-Dimensional Cell Culture,” Advanced Materials 23, 1341–1345 (2011). [CrossRef] [PubMed]
4. F. Klein, T. Striebel, J. Fischer, Z. Jiang, C. M. Franz, G. v. Freymann, M. Wegener, and M. Bastmeyer, “Elastic Fully Three-dimensional Microstructure Scaffolds for Cell Force Measurements,” Advanced Materials 22, 868–871 (2010). [CrossRef] [PubMed]
5. T. Bückmann, R. Schittny, M. Thiel, M. Kadic, G. Milton, and M. Wegener, “On three-dimensional dilational elastic metamaterials,” New Journal of Physics16, 033032, [CrossRef] (2014).
6. I. Staude, C. McGuinness, A. Fröhlich, R. L. Byer, E. Colby, and M. Wegener, “Waveguides in three-dimensional photonic bandgap materials for particle-accelerator on a chip architectures,” Optics Express 20, 5607–5612 (2012). [CrossRef] [PubMed]
7. M. Malinauskas, A. Zukauskas, V. Purlys, K. Belazaras, A. Momot, D. Paipulas, R. Gadonas, A. Piskarskas, H. Gilbergs, A. Gaidukeviciute, I. Sakellari, M. Farsari, and S. Juodkazis, “Femtosecond laser polymerization of hybrid/integrated micro-optical elements and their characterization,” IOP Science, Journal of Optics12, 124010, [CrossRef] (2010).
8. R. Houbertz, G. Domann, C. Cronauer, A. Schmitt, H. Martin, J.-U. Park, L. Fröhlich, R. Buestrich, M. Popall, U. Streppel, P. Dannberg, C. Wächter, and A. Bräuer, “Inorganic-organic hybrid material for application in optical devices,” Thin Solid Films 442, 194–200 (2003). [CrossRef]
9. K. J. Schafer, J. M. Hales, M. Balu, K. D. Belfield, E. W. Van Stryland, and D. J. Hagan, “Two-photon absorption cross-sections of common photoinitiators,” Journal of Photochemistry and Photobiology A: Chemistry 162, 497502 (2004). [CrossRef]
10. J.P. Fouassier and J. Lavele, “Photoinitiators for Polymer Synthesis,” John WileySons, ISBN 3527648267 (2013).
11. W.A. Green, “Industrial Photoinitiators: A Technical Guide,” Taylor Francis, ISBN 1439827451 (2010). [CrossRef]