SPIE Membership Get updates from SPIE Newsroom
  • Newsroom Home
  • Astronomy
  • Biomedical Optics & Medical Imaging
  • Defense & Security
  • Electronic Imaging & Signal Processing
  • Illumination & Displays
  • Lasers & Sources
  • Micro/Nano Lithography
  • Nanotechnology
  • Optical Design & Engineering
  • Optoelectronics & Communications
  • Remote Sensing
  • Sensing & Measurement
  • Solar & Alternative Energy
  • Sign up for Newsroom E-Alerts
  • Information for:

SPIE Photonics West 2019 | Register Today

SPIE Defense + Commercial Sensing 2019 | Call for Papers



Print PageEmail PageView PDF

Micro/Nano Lithography

Forming tiny 3D structures for micro- and nanofluidics

Dielectric breakdown and wet etching provide a method of fabricating 3D nano- and microscale channels inside transparent dielectric materials.
19 March 2007, SPIE Newsroom. DOI: 10.1117/2.1200612.0510

Three-dimensional structures with predesigned optical functions such as waveguides, photonic crystals, and micromechanical parts can be combined into single microdevices with a number of functions. Researchers have made microfluidic devices and sensors that combine optical and microfluidic functions. For practical implementation of these devices, however, the current efficiency of production needs to be increased for 3D laser microfabrication and nanofabrication methods. We expect that a two-step approach—similar to the photographic process of exposure and development—will offer the most effective solution: first, subjecting the sample to a maskless exposure by interference and/or direct laser writing to record the structure, then wet processing to retrieve the written 3D structure. This fabrication route is compliant with batch processing, which is usually required for any practical implementation.1

To localize a light-matter interaction with high sub-wavelength spatial resolution for 3D structuring of materials, we need a laser source that provides a high irradiance per pulse. It became obvious long ago that femtosecond laser pulses are indispensable for this task, due to their ability to create a huge irradiance (of about a terawatt per square centimeter) with a small pulse energy,2 while limiting the thermally affected region around the irradiation spot.

Fluidic microdevices and micro total analysis systems (µTAS) are expected to find an increasing number of applications in biomedical research and diagnostics. 3D architectures allow more integration and compactness, as well as increasedsensitivity due to the small amount of analyte necessary to reliably recognize specific compounds and molecules. The function and versatility of µTAS devices can be augmented by combining optical, electrical, and fluidic systems. Thus, high optical transmissivity—especially to UV light, which can excite fluorescence of an analyte—is required. For electrophoretic applications, the host medium must have high electrical resistivity to avoid polarization, charging, and screening ofthe applied electrical fields. Sapphire, quartz, and high-purity synthetic or fused-silica glasses are good candidates forsuch applications.

We demonstrated 3D patterning of sapphire,5 quartz,6 and glass7–10 by direct laser writing with high spatial resolution. Empty channels with an aspect ratio of more than 1:100 were formed after wet etching in an aqueous solution of hydrofluoric acid (HF) of different compositions. (The etching rate can be controlled by adding HNO3 and NH4HF2.4,5,11) The achievable aspect ratio is mainly determined by the etching contrast between the irradiated and unexposed regions and depends on pulse energy and pulse-to-pulse separation, as well as on the etching solution.4 In optimized conditions, we can form the channels without any apparent taper in silica (see Figure 1). Silica densification is responsible for increased etchability in glass.4,12 In crystalline materials, amorphization created by dielectric breakdown is linked with a higher etching rate and high etching contrast, which allowed us to record channels in sapphire without a taper (see Figure 2). An etching contrast of approximately 1:500 was observed in quartz and 1:103 – 104 in sapphire. (It was difficult to determine the etching contrast in sapphire due to the extremely low etching rateof pristine sapphire.5) There was no widening of channels even after 100h of etching in a 10% aqueous solution of HF: the width determined by optical confocal microscopy remained 520 ± 10nm.3 The exact physical and chemical enhancement of the etching mechanisms in sapphire and quartz are still under discussion.13

Figure 1. A confocal 3D image of photoluminescence from a rhodamine solution within a spiral structure.3 The spiral was written in the synthetic quartz Viosil by 3D direct laser writing at tight focusing conditions, using an objective lens of a numerical aperture of 1.35. Using 800nm irradiation with 180fs pulses, the volume was scanned with a separation of 200nm between pulse locations, at pulse energy equal to the two thresholds of an observable photomodification. Then the modified areas were etched with an aqueous solution of HNO3 and HF.4 The spiral depth spans 0 — 62µm.

Figure 2. A confocal 3D image shows photoluminescence from a rhodamine solution inside channels in sapphire. The channels were recorded by direct laser writing and then etched.3,5

It is worth noting that a wet etching anisotropy observed in quartz can be used to form channels with circular cross-sections, which is usually difficult due to the elliptical shape of the tightly focused laser beam in direct laser writing. For numerical apertures between 1 and 1.4, the axial length exceeds the lateral cross-section of the focal light intensity distribution by 3 — 4 times, even without spherical abberation. By taking advantage of different etching rates along different crystallographic planes of quartz, we can obtain a channel with a circular cross-section.6

For photostructuring of transparent materials, Gauss-Bessel (GB) laser beams and pulses—see Figure 3(a)—can be used to record linear photomodification in the axial direction for the length, which is determined by the beam diameter, D, and the axial cone angle, λ, as shown in the inset of Figure 3(a). The axial extent of the ‘focus’ can be much longer than the lateral cross-section.16 GB beams can be produced by holograms or axicon lenses. In the latter case, the cone angle is defined by the axicon wedge angle, d, and the refractive index of the material from which the lens is made. We used an axicon lens to record photomodification in silica and slide (borosilicate) glasses.14,16,17 The normalized intensity distribution of a GB beamin cylindrical coordinates is given by:18

where is the zero-order Bessel function of the first kind and the wavevector component perpendicular to propagation along the z direction k- = k sin(γ) is determined by the cone angle γ. Equation 1 is visualized in Figure 3(a) for a particular axicon lens used in our experiments.

Figure 3. (a) The intensity distribution of a Gauss-Bessel (GB) beam calculated for γ = 320mrad and scheme for forming a GBbeam by an axicon lens. (The color maps represent intensity.) The first minimum at 0.383λ/sinγ is the radius ofthe central spot. (b) Funnel channels were etched throughout borosilicate glass by exposure to the GB beam at high irradiance14 and wet etching in a 5% aqueous solution of hydrofluoric acid. Scale bar, 0.5mm.15 The inset shows the narrowest part of the channel before and after the connection. Etching took place from both sides. Scale bar, 10µm.

We used wet etching in a 5% aqueous solution of HF to fabricate funnel-shaped channels along the damage traces recorded by GB pulses.15 The narrowest cross-section of the etched channels was close to 1µm at the waist, as shown in Figure 3(b). The initial linear damage tracewas recorded by multipulse exposure at high irradiance, which resulted in traces composed of almost equidistantly separated breakdown sites.14 Irradiation was carried out with axial translation neither of the beam nor of the sample. The entire trace was exposed simultaneously, and we observed both front and back-sideablation of the slide glass during exposure. The different funnel channels shown in Figure 3(b) were recorded at different initial settings of the glass sample inside the non-diffracting zone of the GB beam. For the lower channels, the GB beam was shifted left by several tens of micrometers.

Figure 4 shows a typical sharply tapered 3D channel recorded in silica by pulses that overlapped to within 1% of their diameter and etched. The tapered, funnel-shaped channels can be used for filter, separation, and selection applications. Also, 3D channels can be recorded in polymers without wet etching. With the proper choice of pulse energy and distance between adjacent irradiation spots, 3D empty-core channels with densified walls have been recorded in poly(methyl methacrylate) with channel diameters ranging from 0.5 to 1µm.19

Figure 4. A 3D fluorescence image shows a funnel spiral structure. The funnel was recorded in silica with overlapping laser pulses and etched for 390min in a 5% aqueous solution of HF.

The 3D laser-assisted nano- and microfabrication of microfluidic structures is expected to become practical in the near future and be used to make µTAS, nanoelectromechanical, and microelectromechanical structures. Hybrid sensor applications can also use the technique for making optically recorded waveguides, photonic crystals, molecular ratchets, and channels that can be integrated in 3D and provide new functionalities.

Saulius Juodkazis, Hiroaki Misawa
Photon Process, Hokkaido University
Sapporo, Japan

Saulius Juodkazis is associate professor at the Research Institute for Electronic Science of Hokkaido University. His current interests include space-time-spectrum-resolved characterization of light-matter interactions in the microdomain, nanophotonics, plasmonics, laser tweezers, and applications of ultrafast laser pulses. He has published 104 scientific papers in ISI journals. In addition, he is an active participant of SPIE conferences and has provided 31 papers, contributed several invited talks, and served as a chairman at LPMconferences.

Hiroaki Misawa is a professor at the Research Institute for Electronic Science and director of the Nanotechnology Center at Hokkaido University.His research interests include photochemistry, light-matter interaction, ultrafast processes in materials, photonic crystals, laser trapping, and plasmonics. He has authored more than 200 papers, including more than 25 presentations and invited talks at SPIE conferences.