Nanometer-level high-accuracy molding using a photo-curable silicone elastomer by suppressing thermal shrinkage

Abstract

Although the so-called “labs-on-a-chip” or micro total analysis systems (micro TAS) fields hold high promise for applications in many fields, conventional fabrication processes based on the semiconductor industry such as photolithography have limitations in terms of productivity. Silicone elastomers are widely used for micromodeling and offer biocompatibility and chemical stability, but they are generally thermosetting and undergo unacceptable levels of shape deformation during curing. In this study, a photocurable silicone elastomer that has recently become commercially available was examined, and its basic optical, mechanical, and other related characteristics, along with its shape transfer capabilities, particularly its nanostructure replication characteristics, were measured in comparison with those of a representative existing thermosetting silicone elastomer. As a result, the photo-cured elastomer was shown to be superior to existing heat-cured silicone elastomers, having mechanical strength approximately three times greater, and was shown to have the same optical transmittance, extending from the near-IR to the near-UV regions. In addition, it was shown that the elastomer is sensitive to light in a wide range of wavelengths, from 254 to 600 nm, with no large difference in its curing characteristics, indicating that curing can be performed under a variety of common forms of illumination. Most importantly, the photocured elastomer provided extremely high replication accuracy due to its thermal shrinkage of less than 0.02%, compared to 2.91% in the heat-cured elastomer.

---

1. Introduction

The microchips referred to as labs-on-a-chip or micro total analysis systems (μTAS) have become a focus of intense research. These chips are fabricated by simultaneously downsizing and integrating all the micro/nanocomponents necessary for a given function of a reaction or analysis including both fluid-control elements, such as pumps and valves and functional elements, such as heaters and separation structures. The chips hold high promise for applications in many fields including analytical chemistry, life science, and medicine.^1–5 Their first major application was in the form of bioanalytical chips for DNA sequencing and protein analysis,^6–9 but the trend has more recently shifted to other areas such as medicine-engineering collaboration fields.¹⁰ This trend is rapidly growing in point-of-care, personalized medicine, and medical implant devices, and in an expanding range of component materials.^11–13

In this context, seed research for microprocessing technology and materials that will enable effective fabrication of the related devices and systems has become a fundamental requirement. The microchips have been fabricated by photolithography and other microprocesses adopted from the semiconductor industry; however, these fabrication processes involve certain limitations: (1) they require large-scale equipment based on vacuum processing, (2) they involve long takt times, and (3) their batch sizes are constrained by the size of the vacuum equipment, and productivity is therefore too low, particularly for the many medical applications requiring disposability.^10,14 Fabrication processes with higher productivity at lower cost have become a key goal for industrialization. In fabrication materials, the goals that have emerged include lower cost, together with low (if possible, no) biotoxicity and thus biocompatibility, and functional properties such as low adsorption of biomaterials.

In the research to achieve these goals, attention has turned to technologies, such as embossing, molding, and other processes, to facilitate mass production and reduce costs.^15,16 These technologies include nanoprinting and injection molding, which has already been in use for some time.¹⁷ In laboratory prototyping, silicone elastomers are widely used for micromodeling because of their good moldability and relatively simple, low-cost equipment required,^16,18,19 and also because of their superiority to other polymers in biocompatibility and chemical stability, despite their characteristic disadvantages as soft elastomers.^20,21 Silicone elastomers are generally thermosetting, and therefore require heating for effective curing.¹⁸ Their characteristic elasticity and deformability as elastomers facilitates the transfer of fine structures over large surface areas without mold damage. Moreover, as thermosetting resins, they undergo significant shape deformation under thermal residual stress imposed by heat curing, which substantially reduces their replication accuracy.^22,23 For micro- and nano-devices requiring accuracies on the micrometer (10⁻⁶ m) and nanometer (10⁻⁹ m) scale, even 1% shrinkage usually exceeds permissible tolerances. Research has generally focused on incorporating allowances for the dimensional changes induced by this thermal residual stress into the device design, but this has proved to be difficult and finding a solution to this problem has become a bottleneck for the practical development of micro- and nano-devices, particularly nanostructure fabrication from silicone elastomers.

In the present study, we propose and investigate an alternative solution to this problem. To facilitate high-accuracy nanoscale replication, we propose the use of a photocurable silicone elastomer recently found in the market, which eliminates the need for heat curing. Its basic optical, mechanical, and other related characteristics were investigated, and its shape transfer capabilities were quantitatively assessed, particularly its nanostructure replication characteristics in comparison with those of a thermosetting silicone elastomer.

2. Experimental

In this study, X-34-4184 (Shin-Etsu Silicone) was used as the photo-initiative silicone elastomer and Silpot 184 (Dow Corning Toray Co.) as a representative of existing thermosetting silicone elastomers. Exposing light activates the photo-catalyst, and then the polymerization starts, which is controlled by the catalytic activity, in the case of X-34-4184. This means that the polymerization itself does not progress by photon energy; however, the mechanism of polymerization should be almost as same as that of Silpot 184. The differences lie in the photo-initiative properties and the very high activity of the catalyst. Silpot 184 is only partly cured at room temperature, as a result it is very soft, whereas X-34-4184 is almost fully cured and it also gets cured much faster (hundreds of times or more). This means that X-34-4184 requires considerably less thermal energy compared with Silpot 184. Therefore, we called X-34-4184 as photo-curable PDMS in this study.

As the light source for the photosetting, we used a low-pressure mercury lamp (Multilight: ML-251, Ushio Inc.) equipped with a 365 nm wavelength bandpass filter. For the measurement of optical transmittance, we used a spectrophotometer (V-660, JAS Co), and for the Fourier transform infrared spectrophotometry (FTIR), we used an FTIR measurement system (FT/IR-4100ST, Jasco Corp.). We used a durometer (Hardmatic HH-332, Mitutoyo) to measure the elastomer hardness, as represented by Shore A hardness. To evaluate the mechanical strength, we formed dumbbell-shaped samples in accordance with JIS K6251-7 and measured their tensile strength using a force gauge (Z2-200N, Imada Co., Ltd.). The surface morphology and shrinkage was directly evaluated by the visualized image with a 3-dimensional confocal laser microscope (VK-X200, Keyence Co.).

3. Results

3.1 Photocuring characteristics

We determined the light exposure dose needed to cure the X-34-4184 silicone elastomer by measuring the Shore A hardness evaluated under exposure to 365 nm-wavelength UV light. The results are shown in Fig. 1. In this figure, the blank region in the first few minutes represents the time when the elastomer was still in liquid form, and its hardness therefore could not be measured. Therefore, the figure shows the data obtained from the time it became possible to measure the Shore A hardness. Fig. 2 shows typical results of our measurement of the change in viscosity of the elastomer while still in liquid form, corresponding to the blank region in Fig. 1.

Fig. 1 X-34-4184 curing characteristics. The time course of the post-exposure curing was obtained by the measurement of Shore A hardness.

Fig. 2 Changes in viscosity and temperature of the initial post-exposure X-34-4184 in liquid form.

As shown in Fig. 1, the rate of the hardening of the elastomer initially varied with the light-exposure dose; however, after 30 min of exposure, a Shore A hardness near 4.5 was attained for all doses that remained constant thereafter. This result shows that the X-34-4184 silicone elastomer is extremely sensitive and shows reciprocity law failure in its response to light exposure. As shown in Fig. 2, immediately following the exposure dose of 1.87 J cm⁻², also shown in Fig. 1, the increase in viscosity was accompanied by a temperature rise of several degrees Celsius, apparently caused by heat evolution while curing in the liquid form. This increase is presumably too small to impose thermal stress on the cured silicone elastomer. It may also be noted that this period of several minutes in liquid form following exposure would be highly advantageous in mass production because it would allow sufficient time for the mold casting of the exposed elastomer.

In the tested range, the results also showed that the elastomer is sensitive to light in a wide range of wavelengths, extending from 254 to 600 nm, with no large difference in curing characteristics even under exposure to visible light. This indicates that curing of the elastomer can be performed under fluorescent lighting, sunlight, or other common forms of illumination, which would allow a substantial reduction in equipment cost. This also indicates, however, that care in its handling will be needed to ensure protecting from light, even in a yellow room.

Although the primary purpose of this study was to investigate the possibility of using this elastomer to eliminate the need for high-temperature curing, and thus minimize thermal stress, we also investigated the dependence of its curing characteristics on process temperature in view of the possibility that heating the elastomer might provide other significant advantages such as shortened curing times under heating, in applications where the effects of thermal stress would pose no problem. Fig. 3 shows typical curing properties found with a light exposure dose of 1.87 J cm⁻² at process temperatures of 20 (as shown in Fig. 1), 70, 110, and 150 °C. These results showed that the curing properties are strongly nonlinear with temperature. If the temperature is more than 70 °C, the curing is finished very quickly; thus, we found no large differences in either curing rate or hardness beyond that temperature. In summary, curing the elastomer at 70 °C or higher, substantially increased the curing rate as compared with curing at room temperature, and thus reduced the curing time. The curing rate may therefore be a key parameter for improvements in productivity.

Fig. 3 Temperature dependence of X-34-4184 curing rate.

3.2 Mechanical characteristics

We investigated the physical strength of the cured elastomers, as a representative indication of their tensile strength. The tensile strength of viscoelastic bodies such as elastomers generally varies with the rate of elongation, and we therefore performed the measurement in a quasi-static state, at a tensile rate of 1 mm s⁻¹ or lower. Fig. 4 shows the results after curing with an exposure dose of 1.87 J cm⁻² at (a) 20 °C and at (b) 110 °C for X-34-4182, in comparison with (c) Silpot 184 after thermal curing at 110 °C for 1 hour. As shown, X-34-4101 exhibited a higher tensile strength after curing, and its strength after 100 °C curing was about 2 or less times higher; after room-temperature curing, its strength was 4 or more times higher than that of Silpot 184, which clearly indicates that it is superior to Silpot 184 in mechanical characteristics.

Fig. 4 Dependence of mechanical strength on curing temperature.

3.3 Spectral characteristics

Fig. 5 shows the results of the FTIR measurements of the cured X-34-4184 and Silpot 184. No specific information on their molecular structures and additives is publicly available because they are commercial products, but the FTIR spectrum found for the X-34-4184, with its basic structure being polydimethylsiloxane (PDMS), is almost identical to that found for Silpot 184. The only difference between the two elastomer spectra is the strong absorption from 2280 to 2400 cm⁻¹, which is present in the X-34-4184 spectrum but almost absent in the Silpot 184 spectrum. This absorption is most likely attributable to the X-34-4184 photocrosslinking agent. The transmittance spectrum of X-34-4184, as shown in Fig. 6, exhibits a strong peak around 250 nm, which again is most likely attributable to its photocrosslinking agent. Fig. 6 shows that X-34-4184 exhibits good transmittance in the UV region of 300 to 400 nm, which indicates that it is superior to existing heat-cured elastomers such as Silpot 184 as an optically transparent material.

Fig. 5 Comparison of X-34-4184 and Silpot 184 FTIR spectra. The strong peaks observable between 2280 cm⁻¹ and 2400 cm⁻¹ for X-34-4184 are presumably attributable to its photocrosslinking agent. The remainder of its spectrum is nearly identical to that of Silpot 184; thus, indicating that Silpot 184 and X-34-4184 are nearly identical in basic molecular structure.

Fig. 6 Comparison of the light transparency characteristics of Silpot 184 and X-34-4184. The X-34-4184 absorption peaks around 250 nm are presumably attributable to its photocrosslinking agent. The remainder of its spectrum is nearly identical to that of Silpot 184; thus, indicating that it will exhibit nearly the same physical properties.

In summary, the X-34-4184 spectra exhibit characteristic peaks presumably attributable to its photocrosslinking agent, but are otherwise nearly identical to those of Silpot 184; X-34-4184 may therefore be expected to have physical properties nearly the same as those of Silpot 184. The results also demonstrate that the near-UV transmittance of X-34-4184 is superior to that of existing materials, and taken together, these indicate that it will be upwardly compatible with existing materials in terms of transparency.

3.4 Characteristics of nanostructure replication

We expected that the replication accuracy of X-34-4184 would be higher than that of Silpot 184, caused by an extremely small shrinkage due to its non-thermal curing effect. Fig. 7 shows a schematic of the general replication procedure. Broadly speaking, the procedure comprises three main steps: (a) casting, (b) curing, and (c) mold release. The procedure for the photocurable elastomer differs from that for the heat-curing elastomer only in the use of light exposure rather than heating in step (b). The photocuring conditions used here may be regarded as typical, and consist of exposing the sample at the room temperature of 20 °C, to 365 nm wavelength, and a dose of 1.87 J cm⁻², and then allowing it to cure for several hours to ensure hardening. The hardened sample was then tested for replication accuracy. For comparison with heat-cured elastomers, we cured Silpot 184 at 100 °C, and submitted it to replication testing. The results shown in Fig. 8 are a typical example, visualized by a 3-dimensional confocal laser microscope. The mold used in the replication was shaped as shown in Fig. 8(a), and consisted of a Si wafer with photoresist structures (SU-3035, Microchem. Inc.).²⁴ Fig. 8(b) shows the result of the shape transfer to the photocurable silicone elastomer X-34-4184 by the procedure shown in Fig. 7. As may be seen from the shape hues representing the height of the structures (with warmer hues indicating greater height), the replication results in a 180° rotation between the master mold and X-34-4184. The image provides qualitative confirmation that the various complex micrometer-order shapes have been accurately transferred.

Fig. 7 Photocurable silicone elastomer replication procedure: (a) pouring the non-polymerized elastomer onto the mold, (b) exposure, and (c) peeling off.

Fig. 8 Typical microstructure replication: (A) Si mold with engraved microstructure pattern, and (B) photocured silicone elastomer microstructure obtained by transfer from the master. As shown, the mold pattern is reversed in the replication process.

Fig. 9 shows the results of a nanoscale replication performed to evaluate the transfer accuracy. The image in Fig. 9(a) shows the nanometer-sized pattern formed on the quartz-glass mold master by electron-beam patterning and plasma etching. Fig. 9(b) and (c) show the results of the pattern replication with X-34-4184 and Silpot 184, respectively. The pattern depth is 200 nm, the line widths in the four groups from left to right are 200, 300, 500, and 800 nm, and the width to space ratio in all four groups is 1 : 3.

Fig. 9 Results of the transfer of nanoscale line-and-space fine structure. The line groups each contain 15 lines, in line widths of 200, 300, 500, and 800 nm, all with a line : space ratio of 1 : 3. (a) Mold master, (b) X-34-4184, and (c) Silpot 184.

The differences between the shapes on the mold master and the replicates were evaluated by 3-D confocal laser microscopy. The results showed shape shrinkages of 2.91% with Silpot 184 and 1.05% with X-34-4184 at 110 °C, whereas the shrinkage was 0.02% or less with X-34-4184 at 20 °C. Silpot 184 was not fully cured and very soft at a curing temperature of 20 °C. Therefore, it was clearly demonstrated that curing X-34-4184 at 20 °C can reduce shrinkage by a factor of 100 or more, indicating that it is an extremely effective material, particularly for nanostructure transfer. Temperature increase should be the major factor for thermal shrinkage of the material in varying degrees. These results indicate that curing without applying any heat is key for low thermal shrinkage, which was realized by X-34-4184 in our work.

In summary, these results confirm that the use of a photocurable silicone elastomer in a non-heated process effectively prevents the shape deformation that poses a major problem with the existing heat-cured silicone elastomers.

4. Conclusions

In this study, we investigated the use of a photocurable silicone elastomer in a non-heated process to obtain increased replication accuracy in the fabrication of micro/nanodevices, and thus eliminating the long-standing problem of accuracy reduction by thermal shrinkage with heat-cured silicone elastomers.

We first investigated the optimum light-exposure dose for photocuring and the time response of the curing. The results showed that curing was effectively completed in 30 min or less after exposure in a wide range of doses, and also that the elastomer remained in a liquid form for several minutes after exposure, which is highly advantageous in applications as a material for mass production by casting and other processes.

We next investigated the optical and mechanical characteristics of the cured elastomer; the results demonstrated that the cured elastomer is superior to existing heat-cured silicone elastomers in optical transmittance extending from the near-IR to the UV regions, and its tensile strength is approximately three times greater. The FTIR spectrum of the cured elastomer showed that it is essentially the same as existing heat-cured elastomers in structure, and the elucidation of the reasons for its higher strength remains for further study.

The accuracy of replication was evaluated, and it was found that photocured X-34-4184 provided extremely high replication accuracy, which is attributable in large part to its thermal shrinkage of just 0.02%, in marked contrast to the value of 2.91% found for the heat-cured Silpot 184 silicone elastomer.

In life sciences, which represent one of the major fields of application for silicone elastomer devices, as well as the mechanical and optical characteristics investigated in this study, the chemical characteristics of silicone elastomers are highly important. In this study, the FTIR spectra showed that the molecular structure of X-34-4184 is nearly identical to that of a heat-cured silicone elastomer widely used in this field, and its chemical characteristics may therefore also be expected to be nearly the same.

In summary, the X-34-4184 photocurable silicone elastomer used in this study clearly surpassed the optical and physical characteristics of materials currently in use, increased their replication accuracy by two orders of magnitude, and demonstrated the effectiveness of photosensitive curing without heating.

Acknowledgements

This work was conducted as part of the Four-University Nano-Micro Fabrication Kawasaki Consortium (Fab-4U). This work was supported by Kakenhi Grants-in-Aid (no. 25289096) from the Japan Society for the Promotion of Science (JSPS), and partly by CREST, JST.

References

1. K. Mogi and T. Fujii, J. Micromech. Microeng., 2010, 20, 055015.
2. H. F. Arata, M. Kumemura, N. Sakaki and H. Fujita, Anal. Bioanal. Chem., 2008, 391, 2385–2393.
3. H. Craighead, Nature, 2006, 442, 387–393.
4. M. I. Mohammed and M. P. Desmulliez, Lab Chip, 2011, 11, 569–595.
5. L. K. Chin, J. Q. Yu, Y. Fu, T. Yu, A. Q. Liu and K. Q. Luo, Lab Chip, 2011, 11, 1856–1863.
6. E. T. Lagally, I. Medintz and R. A. Mathies, Anal. Chem., 2001, 73, 565–570.
7. J. Khandurina, T. E. McKnight, S. C. Jacobson, L. C. Waters, R. S. Foote and J. M. Ramsey, Anal. Chem., 2000, 72, 2995–3000.
8. B. Zheng, L. S. Roach and R. F. Ismagilov, J. Am. Chem. Soc., 2003, 125, 11170–11171.
9. M. Urushidani, Y. Sugii, K. Mogi and K. Hishida, The 17th International Symposium on Applications of Laser Techniques to Fluid Mechanics, 2014.
10. F. B. Myers and L. P. Lee, Lab Chip, 2008, 8, 2015–2031.
11. C. F. Carlborg, T. Haraldsson, K. Oberg, M. Malkoch and W. van der Wijngaart, Lab Chip, 2011, 11, 3136–3147.
12. A. Daridon, V. Fascio, J. Lichtenberg, R. Wutrich, H. Langen, E. Verpoorte and N. F. de Rooij, Fresenius' J. Anal. Chem., 2001, 371, 261–269.
13. K. Mogi, Y. Sugii, T. Yamamoto and T. Fujii, Sens. Actuators, B, 2014, 201, 407–412.
14. A. J. Tudos, G. J. Besselink and R. B. Schasfoort, Lab Chip, 2001, 1, 83–95.
15. S. Y. Chou, C. Keimel and J. Gu, Nature, 2002, 417, 835–837.
16. K. Mogi and T. Fujii, Lab Chip, 2013, 13, 1044–1047.
17. H. Ge, W. Wu, Z. Li, G. Y. Jung, D. Olynick, Y. Chen, J. A. Liddle, S. Y. Wang and R. S. Williams, Nano Lett., 2005, 5, 179–182.
18. G. M. Whitesides, Abstracts of Papers of the American Chemical Society, 1996, 212, 31–Inor.
19. T. Fujii, Microelectron. Eng., 2002, 61–2, 907–914.
20. K. Hosokawa and R. Maeda, J. Micromech. Microeng., 2000, 10, 415–420.
21. M. A. Unger, H. P. Chou, T. Thorsen, A. Scherer and S. R. Quake, Science, 2000, 288, 113–116.
22. Q. He, Z. Liu, P. Xiao, N. He and Z. Lu, Mater. Chem. Phys., 2004, 83, 60–65.
23. J. A. Rogers, K. E. Paul and G. M. Whitesids, J. Vac. Sci. Technol., B: Microelectron. Nanometer Struct.--Process., Meas., Phenom., 1998, 16, 88–97.
24. T. Yamamoto, T. Fujii and T. Nojima, Lab Chip, 2002, 2, 197–202.

---