PHOTOLITHOGRAPHY-FREE LASER-PATTERNED HF ACID-RESISTANT CHROMIUM-POLYIMIDE MASK FOR RAPID FABRICATION OF MICROFLUIDIC SYSTEMS IN GLASS - PubMed (original) (raw)

PHOTOLITHOGRAPHY-FREE LASER-PATTERNED HF ACID-RESISTANT CHROMIUM-POLYIMIDE MASK FOR RAPID FABRICATION OF MICROFLUIDIC SYSTEMS IN GLASS

Konstantin O Zamuruyev et al. J Micromech Microeng. 2017 Jan.

Abstract

Excellent chemical and physical properties of glass, over a range of operating conditions, make it a preferred material for chemical detection systems in analytical chemistry, biology, and the environmental sciences. However, it is often compromised with SU8, PDMS, or Parylene materials due to the sophisticated mask preparation requirements for wet etching of glass. Here, we report our efforts toward developing a photolithography-free laser-patterned hydrofluoric acid-resistant chromium-polyimide tape mask for rapid prototyping of microfluidic systems in glass. The patterns are defined in masking layer with a diode-pumped solid-state laser. Minimum feature size is limited to the diameter of the laser beam, 30 μm; minimum spacing between features is limited by the thermal shrinkage and adhesive contact of the polyimide tape to 40 μm. The patterned glass substrates are etched in 49% hydrofluoric acid at ambient temperature with soft agitation (in time increments, up to 60 min duration). In spite of the simplicity, our method demonstrates comparable results to the other current more sophisticated masking methods in terms of the etched depth (up to 300 μm in borosilicate glass), feature under etch ratio in isotropic etch (~1.36), and low mask hole density. The method demonstrates high yield and reliability. To our knowledge, this method is the first proposed technique for rapid prototyping of microfluidic systems in glass with such high performance parameters. The proposed method of fabrication can potentially be implemented in research institutions without access to a standard clean-room facility.

PubMed Disclaimer

Figures

Figure 1

Figure 1

Fabricated microfluidic devices etched in glass with composite (chrome and polyimide tape) mask. Microchannels with initial mask opening width of 40 μm are etched in 49% HF for 15 min. Etched microchannels are 75 μm deep and 225 μm wide. (a, b) Etched glass wafer is anodically bonded to silicon surface. (c) Surface with etched microchannels is bonded to a plain glass surface in thermal fusion. The wavy microchannels filled with red ink show no leakage or closed channel defects.

Figure 2

Figure 2

Mask for etch optimization consists of twelve regions containing straight lines and chess-style squares. Each region contains nineteen lines of the same width but distributed in spacing. The width of the lines increases from 35 μm, in region one, to 200 μm, in region twelve; the spacing width between lines increases from 40 μm to 373 μm in each region.

Figure 3

Figure 3

SEM images of the microstructures etched in 49% HF with composite mask. (a) Wiggly microchannel. Initial mask opening 130 μm, mask spacing 460 μm. Etched depth 90 μm, final microchannel width 374 μm, and spacing width 217 μm. (b) Straight edge wall separating two 80 μm deep microchannels.

Figure 4

Figure 4

Depth and under etch ratio. (a) SEM image of cross sectional profile of 75 μm deep lines etched in 49% HF with chromium-polyimide mask. (b, c) Profilometer scan of the etched feature at different time points. Initial mask opening width is ~100 μm. (b) Comparison of mask under etch ration for the three masking methods. (c) Baked chromium-polyimide mask withstands much longer etch with minimal mask undercut ratio.

Figure 5

Figure 5

Feature density: minimum required mask spacing between the etched features. The required spacing increases with the etched depth for all masking methods. All etched geometries have width of 98 μm, while spacing width decreases from 300 μm to 33 μm. The chromium-polyimide mask demonstrates the highest feature density and mask stability for deepest etch. (a) SEM image of the scanned geometries. (b) Overlaid profilometer scans of the etched features for the three masking methods.

Figure 6

Figure 6

Comparison of masking methods. (a) Etch rate (μm/min) as a function of mask opening (geometry) width. (b) Mask under etch ratio as a function of mask spacing width between features. (c) Maximum etch time, average depth, under etch ratio, and minimum spacing per etched depth ratio.

Figure 7

Figure 7

Comparison of the quality of the etched features for different masking methods. (a, b) Chromium mask; 30 μm deep etch. Initial square size 90 μm and line width 45 μm. (c, d) Polyimide tape mask; 17 μm deep etch. Initial square size 250 μm and line width 57 μm. (e, f) Chromium and polyimide tape mask; 65 μm deep etch. Initial square size 200 μm and line width 57 μm. (g, h) Baked chromium-polyimide mask; 298 μm deep etch. Edge roughness ~ 20 μm, initial line width 48 μm.

Figure 8

Figure 8

Mask pinhole defects. (a, b) Chromium metal mask; after 6 min. in HF acid. (c, d) Chromium-polyimide baked mask; after 60 min. in HF acid. (e) Characterization of the mask quality with respect to pinhole defects for three mask types in terms of the average area and depth of the pinhole defects, and fraction of affected substrate area.

References

    1. Koh Y, et al. Bead affinity chromatography in a temperature-controllable microsystem for biomarker detection. Analytical and Bioanalytical Chemistry. 2012;404(8):2267–2275. -PubMed
    1. Saarela V, et al. Glass microfabricated nebulizer chip for mass spectrometry. Lab on a Chip. 2007;7(5):644–646. -PubMed
    1. Rusu C, et al. Direct integration of micromachined pipettes in a flow channel for single DNA molecule study by optical tweezers. Journal of Microelectromechanical Systems. 2001;10(2):238–246.
    1. Obeid PJ, et al. Microfabricated device for DNA and RNA amplification by continuous-flow polymerase chain reaction and reverse transcription-polymerase chain reaction with cycle number selection. Analytical Chemistry. 2003;75(2):288–295. -PubMed
    1. Zhu HX, et al. Characterization of deep wet etching of fused silica glass for single cell and optical sensor deposition. Journal of Micromechanics and Microengineering. 2009;19(6)

Grants and funding

LinkOut - more resources