Hot Embossing of Parallel v-Groove Microstructures on Glass

Hot Embossing of Parallel v-Groove Microstructures on Glass

Ray-Quen Hsu, Hsing-Bun Wang, and Da-Lan Liang

Department of Mechanical Engineering, National Chiao Tung University, 300 Hsinchu, Taiwan

A series of parallel v-groove microstructures was embossed on glass at an elevated temperature. The effects of glass tempera-ture, applied pressure, holding time, and demold temperature on the conformity between the product and the tool were studied. We found that very precise microstructures can be fabricated on glass at temperatures about 351–551C above the glass transition temperatures. However, careful attention should be paid to the demold temperature and the mold release agent in order to pre-vent the tool from sticking to the glass.

I. Introduction

A

have accelerated the development of microcomponents and systems in almost every conceivable field. In optical appli-cations and communication systems, this trend is no exception. Thus the need to fabricate precise microstructures on glass is clear. Hot embossing, which requires only simple equipment, possesses advantages such as the ability to fabricate delicate structures over a fairly large area and having a very short ma-terial flow distance. As such, it has been adopted as one of the most suitable manufacturing methods for producing microcom-ponents.

Many hot-embossed optical devices use polymers, because they are low weight and their glass transition temperatures (Tg)

are lower than those for glasses. However, compared with poly-mers, glasses have better transparency and chemical stability, making them an obvious choice in many applications. Studies on the hot embossing of polymers have revealed the importance of uniform embossing temperature and pressure. Lin and col-leagues1–3investigated the possibility of hot embossing for pro-ducing microstructures. Scheer and Schulz4 studied polymer behavior during hot embossing. Koro and colleagues5–7 inves-tigated the deformation behavior of the glass, but their devices had geometric features of a much larger scale. Youn and col-leagues8–11used glassy-carbon (GC) micromolds to investigate the surface profile and filling rate of microstructure hot emboss-ing on glass. In order for the structural patterns of the tool to be precisely duplicated on glasses, the temperature of the glasses must be above their respective Tgvalues. If the temperature is

too low, glass flow tends to be retarded, and the detail of the structure will be obscured. On the other hand, if the temperature is too high, then the rapid cooling rate after embossing tends to cause the glass to crack. Thus, the temperature of the glass dur-ing embossdur-ing is critical in the process. Furthermore, the pres-sure applied to the glass must be retained for a certain period of time to reduce the formation of voids in the tool cavity caused by material shrinkage during cooling. Specific embossing con-ditions like temperature, pressure, and holding time with respect to forming materials have not been thoroughly studied. Thus, the purpose of this study was to further investigate the effects of the aforementioned parameters on the embossed products.

II. Experimental Procedure

Glasses selected for the experiments were FCD 1 (Hoya) and SF2 (Ohara), and their physical properties are listed in Table I. Both glasses have a glass transition temperature (Tg) lower than

5001C.

Experimental parameters chosen in this study include glass temperature, holding pressure, holding time, and parting tem-perature of the glass (the demold temtem-perature). After several preliminary tests, the ranges of the parameters were selected, and the experiments were conducted with the conditions shown in Table II.

The microstructure to be fabricated on the glass was an array of parallel v-grooves. The microstructure patterns were obtained by cutting into the surface of a 10.0-mm-thick brass plate with a precision diamond cutter, which in turn functioned as an elec-trode to produce a thin nickel plate. Figure 1 shows the detail dimension of the microstructure.

The nickel plate was used as an embossing die, and was at-tached to a copper heating plate, which in turn was fastened to the ram of the press, and was spaced off with an insulating layer. The glass, on which the structures were to be duplicated, was positioned on the top of another copper heating plate, which in turn was fixed to the bolster of the press with an insulator in between.

The temperatures of each heating plate were automatically adjusted by signals fed back from three thermo couples embed-ded in the plate. Temperature variation over the entire area of the heat plate (250 mm 125 mm) was controlled with a pre-cision of711C. The embossing pressure and the holding time were numerically controlled. The maximum load of the press was 5 tons. The glasses were heated in an oven to the temper-ature specified for the experiment, after which they were moved to the heating plate for 5 min, thereby allowing the heat to be evenly distributed before the pressure was applied.

After embossing, the glass was removed from the heating plate and left to air cool. The dimensions of the v-grooves on the glasses were measured by a Kosaka ET-4000 (Kosaka Labora-tory Ltd., Chiyoda-ku Tokyo, Japan) surface profiler. We col-lected a surface profile for five 0.8 mm sections (Fig. 2). The average peak was then calculated and converted to a percentage of the original tool groove depth (h/H%), as shown in Fig. 3.

Table I. Physical Properties of FCD1 and SF2 Glasses

Glass Hoya FCD1 Ohara SF2 Glass transition temperature (Tg) 4551C 4301C Yield temperature 4851C 4651C

Average index of refraction 1.49700 1.64769 Linear coefficient of thermal

expansion a 301 to 701C: 133 10 7 a 301 to 701C: 88 10 7 a1001–3001C: 155 10 7 a₉₈1001–3001C_{ 10} 7: Specific weight 3.7 3.85 Hardness (Hk) 345 400 S. Bhandarkar—contributing editor w

Author to whom correspondence should be addressed. e-mail: [email protected] Manuscript No. 25267. Received September 23, 2008; approved June 10, 2009.

Journal

J. Am. Ceram. Soc., 92 [11] 2605–2608 (2009) DOI: 10.1111/j.1551-2916.2009.03257.x r2009 The American Ceramic Society

III. Results and Discussion

The purpose of embossing is to duplicate the tool form on the surface of glass. The better the conformity is between the glasses and tools, the better the quality of the product. Figure 4 is a plot of the percentage of the groove depth (h/H%) versus embossing pressure at various glass temperatures. Pressure holding time was set at 120 s. It is clear that for FCD1 at 4901C, the per-centage of the groove depth increased with the applied pressure and reached a maximum of 80% for pressures above 15 MPa. The result indicates that high temperature and high pressure promote better conformity for FCD1 glass.

Figure 5 is the percentage of the groove depth duplicated on a 4.0-mm-thick FCD1 glass at various holding times. The glass temperature was kept at 4901C, and the applied pressure was 15 MPa throughout the embossing process. We found that increas-ing the holdincreas-ing time meant that the percentage of the fill-up of the mold cavities also increased. For holding times longer than 180 s, the groove depth on the glass surface could achieve 96% of the actual profile.

For FCD1 glass that was thicker than 4.0 mm, as depicted in Fig. 6, the percentage of groove depths imprinted was less satisfactory. This might arise from the fact that the thicker the glass is, the more difficult it becomes to direct flow of the glass in the horizontal direction, thus restricting the fill-up of the mold cavity.

The cooling rate of the glass after embossing must be care-fully controlled in order to avoid the formation of a crack. In the above experiments, the embossed glasses were left to cool at room temperature. We found that demold temperatures higher than 1601C promoted the initiation of cracks, as shown in Fig. 7.

Some of the cracks started from the glass surface, while others initiated within the glass. Additionally, sticking sometimes oc-curred between tools and glass surfaces at high temperatures, as shown in Fig. 8. In this case, a suitable boron nitride-based mold release agent can be helpful.

For SF2 (Ohara) glass, which has a lower glass transition temperature but higher softening temperature (5641C) than FCD1 (Hoya), embossing must be carried out at an even higher temperature (5201C) to avoid crack formation. One possible ex-planation is that the difference in the thermal expansion coeffi-cient between the tool and the SF2 glass is larger than it was for FCD1.

IV. Conclusions

An array of parallel v-groove microstructures was fabricated on FCD1 and SF2 glasses at elevated temperatures. For FCD1, a Table II. Experimental Conditions

Glass temperature (1C) 480 485 490 Embossing pressure (MPa) 5 10 15 20

Holding time (s) 60 90 120 150 180 240 Glass thickness (mm) 4 5 6 7 8 Demold temperature (1C) 150 160 170 180 50 µm 50 µm 50 µm 50 µm 50 µm 25 µm 1 mm (a) (b) (c)

Fig. 1. A series of parallel arrays of v-grooves on a nickel plate: (a) dimensions of the v-groove; (b) top view; (c) side view. Fig. 2. Illustration of surface profile sampling sections.

very precise duplication of the templates could be fabricated at a temperature of about 351C above its glass transition tempera-ture, with a medium pressure and a longer holding time. On the other hand, for SF2, which has a lower Tgbut higher softening

temperature than FCD1, embossing needs to be carried out at 551C above Tg in order to obtain a satisfactory duplication

of the template. Finally, a mold release agent was helpful in preventing sticking between glasses and tools at elevated temperatures.

Fig. 3. (a) Illustration of the percentage of the original tool groove depth (h/H%). (b) Example of surface profile data measured by the Kosaka ET-4000.

Fig. 4. Groove depth versus embossing pressure at various glass tem-peratures.

Holding time (sec.)

Groove depth (% h /H) Material:FCD1 Thickness:4.0 mm Embossing pressure:15MPa 100 90 80 70 60 50 60 90 120 150 180 240 Glass temperature:490.°C

Fig. 5. The effects of holding time on groove depth.

Groove depth (% h /H) Glass thickness (mm) Material: FCD1 100 95 90 85 80 4 5 6 7 Glass temperature: 490°C Embossing pressure: 15MPa

8

Fig. 6. Groove depth for various glass thicknesses.

Fig. 7. The appearance of cracking when the demold temperature was 1801C.

Fig. 8. (a) SEM image of a normal product’s surface. (b) SEM image of a glass surface with sticking. (c) Sticking occurred between the tools and the glass surfaces at high temperatures.

References 1

L. Lin, Y. T. Cheng, and C. J. Chiu, ‘‘Comparative Study of Hot Embossed Micro Structures Fabricated by Laboratory and Commercial Environments,’’ Microsys. Technol., 4, 113–6 (1998).

2

H. Becker and U. Heim, ‘‘Silicon as Tool Material for Polymer Hot Emboss-ing’’; pp. 228–31 in 12th IEEE International Conference on Micro Electro Mechanical Systems (MEMS 99), January 17–21, IEEE Robot & Automat Soc, Orlando, FL, 1999.

3

M. Heckele, W. Bacher, and K. D. Muller, ‘‘Hot Embossing – The Molding Technique for Plastic Microstructures,’’ Microsyst. Technol., 4, 122–4 (1998).

4

H. C. Scheer and H. Schulz, ‘‘A Contribution to the Flow Behavior of Thin Polymer Films during Hot Embossing Lithography,’’ Microelectron. Eng., 56 [3–4] 311–32 (2001).

5

S. Koro, S. Masao, and S. Tetsuo, ‘‘Aspect of Glass Softening by Master Mold,’’ Proc. SPIE, 2536, 421–33 (1995).

6

H. Shigeru and M. Yoshiharu, ‘‘High-Speed Glass-Molding Method to Mass Produce Precise Optics,’’ Proc. SPIE, 2576, 115–20 (1995).

7

S. Aoyama and T. Yamashita, ‘‘Planar Microlens Arrays Using Stumping Replication Method,’’ Proc. SPIE, 3010, 11–7 (1997).

8

S. W. Youn, M. Takahashi, H. Goto, and R. Maeda, ‘‘A Study on Focused Ion Beam Milling of Glassy Carbon Molds for the Thermal Imprinting of Quartz and Borosilicate Glasses,’’ J. Micromech. Microeng., 16, 2576–84 (2006).

9

S. W. Youn, M. Takahashi, H. Goto, and R. Maeda, ‘‘The Effect of Heat-Treatment Conditions on Mechanical and Morphological Properties of a FIB-Milled Glassy Carbon Mold with Micro Patterns,’’ J. Micromech. Microeng., 16, 1277–84 (2006).

10

S. W. Youn, M. Takahashi, H. Goto, and R. Maeda, ‘‘Microstructuring of Glassy Carbon Mold for Glass Embossing – Comparison of Focused Ion Beam, Nano/Femtosecond-Pulsed Laser and Mechanical Machining,’’ Microelectron. Eng., 83, 2482–92 (2006).

11

M. Takahashi, Y. Murakoshi, R. Maeda, and K. Hasegawa, ‘‘Large Area Micro Hot Embossing of Pyrex Glass with GC Mold Machined by Dicing,’’

Mi-crosyst. Technol., 13, 379–84 (2006). &