Development of a Rapid Prototyping System for Microneedles Using Moving-mask Lithography with Backside Exposure

Takahisa KAI,* Shunta MORI,* Nobuhiro KATO**, #

Abstract　　Moving-mask lithography with backside exposure was utilized to generate master male mold for biodegradable polymer microneedle production. The microneedle shape was calculated from the exposure dose, mask geometry, and moving trajectory using a newly developed computer simulation. Two conditions (90 µm aperture with 80 µm diameter of circular move-ment, and 90 µm aperture with 90 µm diameter of circular movement) were selected to evaluate the moving-mask exposure ef-fectiveness. By changing the moving trajectory, two different sizes of microneedles were obtained from a single-size aperture mask. The fabricated microneedle and calculated microneedle geometry showed good qualitative agreement. The geometrical difference was 2% in basal diameter and 8%–16% in height. Using the master male mold, biodegradable polymer microneedles made of chondroitin sulfate C sodium salt (CSC) were fabricated by casting from a poly-dimethylsiloxane female mold. The shape of the biodegradable CSC microneedles showed good agreement with the master male mold.

Keywords: moving-mask lithography, biodegradable microneedles, backside exposure.

Adv Biomed Eng. 5: pp. 63–67, 2016.

1.　 Introduction

Skin, which is the largest organ in the human body, is very im-portant for protecting the body against unwanted external attacks. The stratum corneum, the outermost layer of the epidermis, is 15–20 µm thick and an indispensable barrier [1]. In conventional medicine, hypodermic needles pierce through the skin to inject drugs into the body. However, this method causes pain and re-quire cold-chain delivery [2]. To overcome these disadvantages while preserving the advantages of hypodermic needles and con-ventional transdermal drug delivery systems, microneedle tech-nology [3] is a promising alternative. Microneedle technology can be divided into several categories [4]: solid microneedles for skin pretreatment to increase skin permeability, microneedles coated with drugs, hollow microneedles for drug infusion into the skin, and drug-encapsulated polymer microneedles that degrade totally or partially in the skin. Among them, biodegradable mi-croneedles that are produced by molding are the most inexpensive solution. Furthermore, they can deliver a high drug payload and leave no sharp waste materials after use [5]. Photolithographic and related methods are commonly used to fabricate the master male molds of biodegradable microneedles. These techniques in-clude inclined lithography [6], backside lithography utilizing dif-fraction from a photomask pattern [7], backside lithography us-ing a microlens [8], and lithography with a subsequent reactive

ion etching process [9]. These methods are often time-consuming and cost-ineffective, and have some limitations in needle geome-try. Moving-mask lithography using thick positive-tone photore-sist is employed to produce 3D micro structures [10]. However, the height of the structure is restricted up to 50 μm due to the re-sist thickness, which is too low for application to microneedles. To overcome these drawbacks, this study proposes moving-mask UV lithography using thick negative-tone photoresist up to 500 μm in thickness, with backside exposure to form the master male mold of a microneedle. In this method, the geometry of the needle is modi�ed by the mask movement. Various shapes of mi-croneedles can be obtained by this simple scheme. Using this sys-tem, we succeeded to generate the master male mold of a mi-croneedle, and obtain the female replicate mold from the master male mold. Finally, we fabricated biodegradable microneedles by casting a chondroitin sulfate C sodium salt (CSC) solution onto the female mold.

2.　 Methods

2.1　 Resist characterizationIn backside lithography, the thickness of the negative-tone photo-resist after development depends on the exposure dose. The rela-tion between exposure dose and resist thickness was examined by the test pattern exposure and measurement of the resist thickness with a scanning laser microscope (LEXT-OLS3000, Olympus Co., Ltd). An average thickness of 10 points for each exposure dose was utilized to characterize the relation between exposure dose and resist thickness. The dose-thickness relation was �tted to a logarithmic function.

2.2　 SimulationThe shape of the microneedle after fabrication was predicted by a computer simulation. The overall exposure dose distribution D(x, y) was obtained by

D(x, y) = I0(x, y) ∗ M(x, y). (1)where I0(x, y), M(x, y), and ∗ represent the light intensity obtained

This study was presented at the Symposium on Biomedical Engi-neering 2015, Okayama, September, 2015. Received on August 2, 2015; revised on November 12, 2015 and January 14, 2016; accepted on February 2, 2016.

* Graduate School of Biology-Oriented Science and Technology, Kinki University, Wakayama, Japan.

** Faculty of Biology-Oriented Science and Technology, Kinki Uni-versity, Wakayama, Japan.

# 930 Nishimitani, Kinokawa, Wakayama 649–6493, Japan. E-mail: nkato@waka.kindai.ac.jp

Original PaperAdvanced Biomedical Engineering5: 63–67, 2016.

DOI:10.14326/abe.5.63

through the photomask, the spatial distribution of the staying time of the photomask at an arbitrary position, and the convolution, respectively. The shape of the photoresist (SU-8; MicroChem) was calculated by substituting the overall exposure dose distribu-tion into the relation between exposure dose and resist thickness obtained experimentally:

H(x, y) = f (D(x, y)). (2)where H(x, y) and f( ) represent the resist thickness and the exper-imentally derived function mapping D(x, y) to H(x, y), respective-ly.

The simulation was coded in the macro language of image processing software (ImageJ 1.84q; NIH) [11] based on the cal-culation procedure shown in Fig. 1. As the �rst step in this proce-dure, the following calculation conditions are set: light intensity, exposure time, mask opening diameter, amplitude of mask move-ment, and driving frequency. Second, 16 bit grayscale images are generated. Third, images are drawn. The images are light patterns modulated by the photomask of each position determined by time. Fourth, the summation of all images is calculated. Finally, the overall dose intensity distribution is converted to resist thick-ness using the experimentally derived relation between exposure dose and resist thickness.

2.3　 Fabrication of master male moldMaster male molds of microneedles were fabricated by backside exposure using a previously reported moving-mask lithography system [12]. A custom-made soda lime glass wafer (50 mm in di-ameter, 170 μm in thickness; Matsunami Glass Ind., Ltd.) was spin-coated with a negative-tone photoresist (SU-8 3050; Micro-Chem) using a spin-coater (1H-D7; Mikasa Co., Ltd). The main spin speed and spin time used to produce a 250-μm thick coating were 1000 rpm and 5 s, respectively. This coating process was repeated twice, and �nally a 500-μm thick resist layer was ob-tained. After relaxation for 1 h at room temperature, the sample was heated at 95°C for 40 min on a hotplate to evaporate the sol-vent. The coated glass wafer was set upside down on the substrate stage, which was covered with a poly-dimethylsiloxane (PDMS) cushion. The substrate stage was then moved in a circular motion (80 μm or 90 μm in diameter, 2 Hz). The mask holder was low-ered to achieve tight contact between the photomask and the glass

substrate. The tight contact was con�rmed by the stage trajectory monitor. The mask holder was then raised 10 μm to provide a proximity gap. A UV light source (EXECURE 4000; Hoya Can-deo) was used to irradiate the photomask at the center of the pat-tern for 15 s with an exposure energy of 150 mJ/cm2. A post-ex-posure bake was conducted at 95°C for 40 min, and the sample was allowed to cool down to room temperature for 2 h. The po-lymerized resist was developed (SU-8 developer; Microchem) for 30 min and then rinsed with isopropyl alcohol. Finally, arrays of tapered cone shaped microneedle master male molds were ob-tained.

2.4　 Casting of biodegradable microneedleA double casting process was used to mold the biodegradable mi-croneedle (Fig. 2). The photoresist microneedle obtained above was used as the master male mold to fabricate the female mold made of PDMS (Sylgard 184; Dow Corning). The volume ratio of the PDMS base and the curing agent was 10:1. The PDMS mix-ture was poured onto the master male mold and cured at 60°C for 3 h. The cured female mold was peeled off from the master male mold. Chondroitin sulfate C sodium salt (CSC) (032-08802; Wako) and DI water were mixed in a mass ratio of 2:1 and poured onto the PDMS female mold. To �ll the cavities of the PDMS mold, the CSC solution was placed in vacuum at −100 kPa for 5 min at room temperature. Then, the CSC solution was covered with an acetylcellulose membrane �lter as the backup material of the CSC microneedle and cured at 60°C for 3 h. After solidi�ca-tion, the CSC microneedle was obtained. The PDMS mold can be reused to make additional microneedles.

3.　 Results

3.1　 Resist CharacterizationThe thicknesses of 10 circular resist patterns 250 μm in diameter were measured at 11 exposure doses (Fig. 3). From the plot, the relation between UV exposure dose and resist thickness was �t by a logarithmic function,

h [µm] = 236.09 ln(d) − 641.39. (3)where h [μm] and d [mJ/cm2] represent resist thickness and ex-posure dose, respectively.

3.2　 Simulation resultsThe 3D pro�les of two types of microneedles in the simulation

Fig. 1　Flowchart of the simulation.

Fig. 2　 Processing steps for fabricating biodegradable microneedles: (a) moving-mask exposure, (b) master male mold of micronee-dle, (c) PDMS molding, (d) female mold of microneedle, (e) pouring CSC solution in female mold, (f) CSC microneedle.

Advanced Biomedical Engineering. Vol. 5, 2016.(64)

are shown in Fig. 4. Type A (90 µm mask with a movement diam-eter of 80 µm) showed a symmetrical bullet-like shape with a long dull tip. Type B (90 µm mask with a movement diameter of 90 µm) showed a symmetrical bullet-like shape with a long sharp tip. The geometric features of both microneedles are shown in Table 1.

3.3　 Microneedle master male mold fabricated by mov-ing-mask lithography

The master male molds obtained by moving-mask lithography are shown in Fig. 5. Each microneedle array was composed of 25 ×  25 cone-shaped microneedles. The geometry of the master male molds was examined by scanning electron microscopy (SEM). The geometric features of both types of microneedles are shown in Table 2. Six needles for each type of microneedle were ana-lyzed.

3.4　 Biodegradable microneedlesBiodegradable microneedles were successfully obtained by cast-ing the CSC solution using the PDMS female mold (Fig. 6). As veri�ed in Fig. 7, the shape of the master male mold was trans-ferred to the CSC microneedles with a high level of precision. The geometry of the CSC microneedles is shown in Table 3.

Six needles were analyzed. To mimic a pharmacological agent in these biodegradable polymer microneedles, red food dye was mixed with the CSC solution. The red color was distributed

Fig. 3　 Relation between UV exposure dose and measured height of photoresist (SU-8). Error bars represent standard deviation of the mean (n =  10).

Fig. 4　 Simulated shapes of microneedles calculated with 90-µm mask and (a) 80-µm diameter circular motion (Type A) and (b) 90-µm diameter motion (Type B).

Table 1　 Geometry of simulated microneedles (µm).

type height basal diameter tip radius

A 518.6 159.3

B 421.1 170.0

Fig. 5　 Electromicrographs of resist microneedles fabricated with 90-µm mask and (a) 80-µm diameter circular motion (Type A) and (b) 90-µm diameter motion (Type B).

Table 2　 Geometry of fabricated resist microneedles

(mean ±  SD; µm).

type height basal diameter tip radius

A 446.1 ±  4.9 156.3 ±  1.8 10.4 ±  0.9

B 388.0 ±  2.9 167.5 ±  2.1 13.6 ±  1.9

Fig. 6　Optical image of cross section of the PDMS female mold.

Fig. 7　Optical image of CSC microneedles.

Table 3　 Geometry of fabricated chondroitin sulfate C sodium

salt (CSC) microneedles (mean ± SD; µm).

type height basal diameter tip radius

A 424.3 ±  4.0 145.6 ±  1.3 9.6 ±  0.5

Takahisa KAI, et al: Microneedle Prototyping System Using Lithography (65)

uniformly in the microneedles. Therefore uniform distribution of medicine can be expected in the future.

4.　 Discussion

4.1　 Difference between simulation and fabrication resultsThe shapes of the microneedles obtained by simulation and those obtained by fabrication showed good qualitative agreement. As shown in Table 1 and Table 2, the basal diameters of the mi-croneedles were almost the same (the difference was within 2%). However, the height of the fabricated microneedles was 6%–18% shorter than that of the simulated microneedles. These differences are partially due to the simulation assumptions, which ignore dif-fraction by the mask pattern and the collimation angle of UV illu-mination. Because the tips of the microneedles were not fully cross-linked and therefore fragile, the fabrication process affected these delicate tips in submicrometer order, and blunted the sharp apexes.

4.2　 Shape modi�cation by altering mask movementAltering the diameter of the circular trajectory of the stage motion produces an arbitrary UV exposure dose distribution, yielding mi-croneedles with various tapered cone shapes from a single photo-mask pattern. Comparing the two types of fabricated micronee-dles, the basal diameter of Type A was larger than that of Type B. On the other hand, the height of Type A was shorter than that of Type B. Other trajectories produced completely different shapes of microneedles (not shown).

4.3　 Biodegradable microneedlesThe shape of the biodegradable CSC microneedles was in good agreement with the master male mold. However, the CSC mi-croneedles shrank 5.1%–7.6% compared with the master male mold. This shrinkage can be explained by two phenomena: PDMS shrinks approximately 5% after curing, and the CSC slurry shrinks during the heat solidi�cation process. The master male mold should be designed to compensate for this inevitable shrink-age. The ability of the microneedles to penetrate human skin should also be considered. Even after shrinkage, the tip radius of the CSC microneedles were less than 30 µm and the needles were sharp enough to insert into skin [13]. However, the ability to pen-etrate skin is not determined only by the tip radius. A mechanical penetration test should be conducted. Drug payload uniformity was con�rmed in Fig. 7. At present, the microneedles and the base of the array are made of the same material. To improve the ef�ciency of drug injection, the drug should be concentrated in the microneedles only. Using an automated dispensing machine would also improve the precision of �lling the female mold. We are currently improving this system to avoid wasting drugs.

5.　 Conclusion

A computer simulation was developed to predict the shape of mi-croneedles produced by moving-mask lithography. A master male mold of microneedle was made using moving-mask lithography and the shape of the needle showed good qualitative agreement with the simulation results. The master male microneedle mold was successfully transferred to a biodegradable microneedle made of CSC by casting in a PDMS female mold.

Acknowledgement

This study was supported in part by Project Research of the Fac-ulty of Biology-Oriented Science and Technology, Kinki Univer-sity No. 12-IV-16.

Con�ict of interest

We have no con�ict of interest relationship with any companies or commercial organizations based on the de�nition of the Japanese Society for Medical and Biological Engineering.

References

1. Ikejiri A, Ito Y, Naito S, Takada K: Two-and Three-Layered Dis-solving Microneedles for Transcutaneous Delivery of Model Vaccine Antigen in Rats. J Biomater Nanobiotechnol. 3, pp. 325–334, 2012.

2. Nochi T, Takagi H, Yuki Y, Yang L, Masumura T, Mejima M, Na-kanishi U, Matsumura A, Hiroi T, Morita S, Tanaka K, Takaiwa F, Kiyono H: Rice-based mucosal vaccine as a global strategy for cold-chain- and needle-free vaccination. PNAS. 104(26), pp. 10986–10991, 2007.

3. McAllister DV, Wang PM, Davis SP, Park JH, Canatella PJ, Allen MG, Prausnitz MR: Microfabricated needles for transdermal de-livery of macromolecules and nanoparticles: Fabrication. PNAS. 100(24), pp. 13755–13760, 2003.

4. Kim YC, Park JH, Prausnitz MR: Microneedles for drug and vac-cine delivery. Adv Drug Deliv Rev. 64(14), pp. 1547–1568, 2012.

5. Sullivan SP, Koutsonanos DG, Martin MdP, Lee JW, Zarnitsyn V, Choi SO, Murthy N, Compans RW, Skountzou I, Prausnitz MR: Dissolving polymer microneedle patches for in�uenza vaccina-tion. Nature Med. 16, pp. 915–920, 2010.

6. Choi SO, Rajaraman S, Yoon YK, Wu X, Allen MG: 3-D metal patterned microstructure using inclined UV exposure and metal transfer micromolding technology. Proc. of Solid State Sensors, Actuators, and Microsystems Workshop, Hilton Head, SC, pp. 348–351, 2006.

7. Ami Y, Tachikawa H, Takano N, Miki N: Formation of polymer microneedle arrays using soft lithography. J Micro/Nanolitho MEMS MOEMS. 10(1), pp. 0115031–0115036, 2011.

8. Park JH, Yoon, YK, Choi SO, Prausnitz MR, Allen MG: tapered conical polymer microneedles fabricated using an integrated lens technique for transdermal drug delivery. IEEE Trans Biomed Eng. 54(5), pp. 903–913, 2007.

9. Park JH, Allen MG, Prausnitz MR: Biodegradable polymer mi-croneedles: fabrication, mechanics and transdermal drug deliv-ery. J Control Release. 104, pp. 51–66, 2005.

10. Hirai Y, Sugano K, Tsuchiya T, Tabata O: A three-dimensional microstructuring technique exploiting the positive photoresist property. J Micromech Microeng. 20(6), pp. 065005–13, 2010.

11. Schneider CA, Rasband WS, Eliceiri KW: NIH image to ImageJ: 25 years of image analysis. Nature Methods. 9(7) pp. 671–675, 2012

12. Kato N, Kai T, Hirano M: Moving-mask lithography for 3D mi-crostructure molding. J Photopol Sci Technol. 27(1), pp. 85–89, 2014.

13. Davis SP, Landis BJ, Adams ZH, Allen MG, Prausnitz MR: In-sertion of microneedles into skin: measurement and prediction of insertion force needle fracture force. J Biomech. 37(8), pp. 1155–1163, 2004.

Advanced Biomedical Engineering. Vol. 5, 2016.(66)

Nobuhiro KATO

Nobuhiro KATO received his BSc, MSc, and PhD

degrees from Osaka Prefecture University Japan in

1992, 1994 and 1997. In 1997, he joined Kinki

University, where he is currently an Associate Pro-

fessor of Faculty of Biology-Oriented Science and

Technology. His research interests include micro

fabrication, micro �uidics and micro scale medical device. He is a

member of JSMBE, JSME, JSPE, JSAP, and CHEMINAS.

Takahisa KAI

Takahisa KAI received his BSc degree from Kinki

University, Japan, in 2014. He is currently a Mas-

ter’s course student of Graduate School of Biolo-

gy-Oriented Science and Technology, Kinki Uni-

versity. His research interest include micro scale

medical device.

Shunta MORI

Shunta MORI received his BSc degree from Kinki

University, Japan, in 2015. He is currently a Mas-

ter’s course student of Graduate School of Biolo-

gy-Oriented Science and Technology, Kinki Uni-

versity. His research interest include development

of microneedles fabrication process and their ap-

plication.

Takahisa KAI, et al: Microneedle Prototyping System Using Lithography (67)