Femtosecond laser processes were optimized for nonlinear interactions with various optical materials to develop a novel biophotonic lab-on-a-chip device that integrates laser-formed waveguides (WGs), microfluidic channels and photonic crystals (PCs). Such integration seeks the unique demonstration of dual PC functionalities: (1) efficient chromatographic separation and filtration of analytes through a porous PC embedded inside a microfluidic channel and (2) optofluidic spectroscopy through embedded WGs that probe PC stopband shifts as varying analyte concentrations flow and separate. The building blocks together with their integration were demonstrated, providing embedded porous PCs through which electrochromatography drove an accelerated mobile phase of analyte and an optical stopband was probed via integrated buried WGs. Together, these laboratory results underpin the promise of simultaneous chromatographic and spectroscopic capabilities in a single PC optofluidic device.
©2013 Optical Society of America
Optofluidic devices build on miniaturized-total-analysis-system concepts that combine microfluidic channels with optical components to harness the novel sensing opportunities of micro- and nano-scale photonics [1, 2]. Integrated optical components can bypass fluorescence-based detection methods that require molecular tagging of analytes, thereby enabling rapid, in situ optical characterization without chemical modification . A currently promising direction in optical sensing is aiming to characterize a varying fluidic medium that dynamically modifies the stopbands of integrated one-dimensional (1D) [3, 4], two-dimensional (2D) [5, 6] and three-dimensional (3D) [7–9] photonic crystal (PC) devices. Wu and Gu  review state-of-the-art microfluidic optical sensing devices and highlight PC-based sensors [5, 11–14] with potential opportunity for high detection sensitivity and localized analyte confinement inherent to PC-integrated devices . Significant opportunity further arises when such PCs are porous to fluidic media to facilitate the separation of different proteins, cellular content and other analytes as available in high pressure liquid chromatography (HPLC). While HPLC typically requires a separate detection module such as optical spectroscopy, a porous and well-ordered PC may simultaneously separate and optically detect analytes to offer the significant advantage of a single process step .
Femtosecond laser micro-machining  is a novel direct-laser-write (DLW) and rapid prototyping technique that provides great potential for optofluidic device fabrication . Direct-laser-writing uses ultrashort laser pulses to confine strong nonlinear optical interactions [7, 8, 18] that may induce (1) positive refractive index changes in bulk transparent materials for creating optical waveguides (WGs) [1, 16, 19–25], (2) self-ordered nanogratings  to facilitate differential chemical etching [27, 28] of microfluidic channels [28–31], and (3) multiphoton polymerization of epoxies for PC fabrication [8, 32–34]. While WG-channel integration has recently been explored [17, 31], PC integration with WGs and microfluidics has yet to be fully demonstrated with simultaneous all-on-chip refractive index sensing and capillary electrochromatography (CEC).
In this paper, we enhance lab-on-a-chip (LOAC) multifunctionality, integration and miniaturization by employing various advanced 3D femtosecond DLW processes to create a novel optofluidic platform towards simultaneous CEC and optical sensing of analytes. Porous 3D DLW PCs were integrated within DLW microfluidic channels and probed with DLW WGs for two major LOAC benefits: (1) CEC driven through a 3D PC stationary phase of a chromatographic column, and (2) real-time optical sensing of the PC's stopband as it shifts with the separation of the mobile phase. The building blocks for such integration were demonstrated through an accelerated mobile phase of analyte that was recorded through an embedded porous PC and the optical characterization of a PC's Γ–Z stopband via integrated probing WGs. Together, these laboratory results give promise for simultaneous chromatographic and spectroscopic capabilities in a single PC optofluidic device that would be of interest in analysis of cancerous or exogenous diseases, biohazardous mixtures and other biochemical compounds for clinical, forensic, pharmaceutical or security applications [17, 35, 36].
2. Photonic crystal sensor design
Figure 1(a) shows a schematic of a typical capillary electrophoresis (CE) chip defined by two mutually perpendicular channels (separation and injection) that form microfluidic “pinches” and “plugs” under appropriate voltage biasing of the reservoirs located at the channels ends. Electrophoretic separation begins with analyte injection from reservoir A to C, followed by rapid voltage switching for plug formation and flow from the cross to the separation channel towards reservoir B. The effectiveness of separating the analytes is substantially enhanced by CEC when a PC is embedded inside the separation channel by serving as the stationary phase of a chromatographic column .
Figure 1(b) shows a schematic of the proposed optofluidic device to be fabricated by DLW processing, chemical etching and assembly. The open microfluidic channel in fused silica is intercepted by laser-formed WGs with the objective to optically probe the channel or an embedded porous woodpile PC. The channels and embedded PCs are sealed with polydimethylsiloxane (PDMS). The two examples of porous PC designs (Fig. 1(d) and 1(f)) both serve as the stationary phase of a chromatographic column where the stopbands have been tailored to simultaneously respond to refractive index sensing of analytes via the probing WGs (Fig. 1(c)) or free-space optics (Fig. 1(e)) as they separate during flow through the porous PCs. Three-dimensional periodic PCs possess significantly larger surface area over their 1D or 2D counterparts and offer increased analyte-PC interaction for improved CEC efficiency. However, 3D PCs, as typically generated in photoresists or epoxies, have relatively low refractive index (nSU-8 = 1.6) and consequently provide low refractive index contrast (~0.27) to aqueous solutions (nw = 1.33 for water) that inhibit the opening of a complete photonic band gap. For the present purpose, a woodpile PC written in SU-8 epoxy is subsequently designed to exhibit a stopband targeted near the 1.55 μm telecommunication wavelength along Γ–Z axes oriented to be parallel with the optical probing directions (Fig. 1(b)).
While many groups [7–9] have demonstrated free-space optical sensing of the Γ–Z stopband in conventional woodpiles (Fig. 1(f)), practical functionality for refractive index sensing in the LOAC requires compact optical components such as integrated buried WGs that directly probe the PC-embedded channels. To facilitate such integration, a new PC design is introduced that rotates the Γ–Z axis of the conventional woodpile (vertical in Fig. 1(e)) to align horizontally with the probing WG's optical axis (Fig. 1(c)), defining the “rotated” woodpile PC design as shown by the unit cell in Fig. 1(d). Integrated probing WGs uniquely remove the need for free-space alignment of external sources and collecting lenses by locking the WG's optical axis to the PC's Γ–Z. The conventional and rotated woodpile PCs were explored separately for accelerating a mobile phase of analyte and for stopband sensing, respectively.
Multiple design constraints were considered to permit dual chromatographic and spectroscopic PC functionality: (1) a sufficiently small Γ–Z periodicity (c of Fig. 1(d) and 1(f)) to position a stopband near the targeted 1.55 μm wavelength, (2) a sufficiently porous PC to permit electroosmotic flow (EOF), and (3) sufficiently connected and rigid woodpile logs that do not collapse during fabrication or CEC flow. For the conventional PC structure, transverse laser scanning yielded the “horizontal rods” with approximately elliptical cross-section of major radius RH and minor radius RV as seen in Fig. 1(f). For a sufficiently connected and porous structure, the log dimensions RH and RV must satisfy:Fig. 1(f)). For the rotated PC, longitudinal scanning relative to the laser's optical axis yielded cylindrically symmetric “vertical pillars” with cross-sectional radius rV while transverse scanning yielded approximately “horizontal rods” with elliptical cross-section of major radius rHb and minor radius rHc as seen in Fig. 1(d). Connectivity and porosity for this structure requires:Fig. 1(d)). The asymmetry in the rHb / rHc = wo / zR ratio, where wo is the beam waist (1/e2 Gaussian intensity radius) and zR is the longitudinal length (1/e2 intensity radius) of the focal volume, was inferred from the laser focusing condition which, for the case of 522-nm laser writing (Section 3) with 0.55 numerical aperture (NA) focusing into SU-8 photoresist (refractive index nSU-8 = 1.6), yielded calculated values of wo = 285 nm, zR = 1.96 μm and RV / RH = 6.9. The smaller ratio of RV / RH = 5.5 that was observed experimentally was attributed to a 20% SU-8 shrinkage during chemical development that was commensurate with the rates reported by other groups [37–41]. For this 0.55 NA lens focusing, structurally stable PCs could be fabricated but with overly large log sizes and large c periodicity that prevented the tuning of stopbands to the targeted 1.55 μm wavelength. Tighter focusing with a 0.9 NA lens overcame this limitation with calculated values of wo = 150 nm, zR = 565 nm and rHb / rHc = 3.7 while a smaller rHb / rHc = 3.0 ratio was observed experimentally that was equal to the previously observed 20% shrinkage rate. The FDTD simulations were therefore directed to tailor the rotated woodpile design for the 0.9 NA focusing case and tune the Γ–Z stopband to 1.55 μm wavelength.
Laser exposure conditions were considered that varied the a, b and c periodicities and tuned the horizontal rod and vertical pillar dimensions of the rotated woodpile PC (Fig. 1(d)), while simultaneously satisfying the connectivity and porosity conditions represented in Eqs. (3) and (4). Figure 2(a) shows an example of the 3D FDTD simulated transmission spectrum for Ea and Eb polarized light propagation through a rotated woodpile PC of 10c thickness having dimensions a = b = 1.90 μm, c = 1.00 μm, rV = 0.4 μm, and rHb = 3rHc = 0.45 μm and refractive index of nSU-8 = 1.60 filled and surrounded by water (nw = 1.33). Perfectly matched layers (PMLs) bound the PC along the probing optical axis while periodic boundary conditions surrounded a single PC unit cell transverse to the probing axis. FDTD modeling of transmission spectra through much thicker (50c) PCs were limited by a long simulation time of ~10 days per device. A cursory study for a 50 μm thick water-filled PC yielded an identical center wavelength, slightly narrower width, and stronger stopband reflectance of R = 18 dB in contrast with the 10 μm thick PC. However, imperfections in the fabricated 50 μm thick PC described in Section 4.1 led to much weaker stopbands that were better matched with the simulations with 10 μm length.
Figure 2(a) predicted 1st and 2nd order stopbands at 2.4 μm and 1.5 μm wavelengths, respectively. Since further downscaling of the rod and pillar dimensions was not possible with the 0.9 NA lens, the 2nd order stopband was targeted for optical sensing at 1.55 μm wavelength. With high precision peak finding spectral analysis, the center wavelength of the stopband may be identified to 1 nm resolution.
The 3D FDTD simulation was further applied through a 10c wide rotated PC with dimensions varied in the ranges of rHb = 3rHc = 0.15 to 0.75 μm and rV = 0.2 to 0.5 μm, which satisfied the connectivity and porosity conditions expressed in Eq. (3) and (4). The calculated 2nd order stopband position (λo), full-width-half-maximum (FWHM) bandwidth (Δλ) and peak reflection (R) expected over these dimensions are shown in Fig. 2(b) for both probing polarizations and demonstrate the design's flexibility in targeting λo in the range of 1.4 to 1.6 μm. To generate the strongest 2nd order stopband as shown in Fig. 2(a), laser exposures must be directed to fabricate a woodpile geometry with dimensions of rHb = 3rHc = 0.45 μm ± 0.15 μm and rV = 0.4 μm ± 0.1 μm according to Fig. 2(b) and thus provide R ≈7 dB and Δλ ≈110 nm at λo ≈1.50 μm (Fig. 2(a)). The small ~7 nm dichroic offset between the two probing polarizations, Eb and Ea, is small relative to the corresponding stopband widths and become insignificant for the present purpose of optical sensing in the CEC chip if probing with linear polarization.
Figure 2(c) shows the tuning of the 2nd order stopband's center wavelength of a rotated woodpile with increase in the background fluid refractive index as simulated with 3D FDTD. Smaller log sizes of rV = 0.22 μm and rHB = 3rHc = 0.36 μm were used to improve the refractive index sensitivity here, yielding 725 nm / refractive index unit (RIU). The 1 nm resolution in specifying the stopband center wavelength indicates a minimum detectable shift of 1.4 × 10−3 RIU is available for sensing applications. This detection limit is suitable for observing cancerous cell (1.392-1.401 RIU)  and hazardous organic (1.35-1.38 RIU)  solutions among other refractive index detection applications. Further, optical Bragg filters may be introduced into the probing waveguide to calibrate for temperature and other noise factors as considered in  and thereby ensure that the photonic crystal stopbands offer sufficient precision for sensing.
A commercial femtosecond fiber laser (IMRA America μJewel D-400-VR) with 1045 nm center wavelength, 100 kHz or 1 MHz repetition rate and 400 fs or 220 fs (Lorentzian FWHM) pulse duration, respectively, was used for device fabrication in glass substrates and SU-8 photoresist (Microchem 2050). The optical delivery system used (1) an acoustic-optic-modulator (AOM) for on-off laser switching, (2) a lithium triborate nonlinear crystal for second harmonic generation (SHG) to 522 nm to drive stronger nonlinear laser-material interactions , (3) air-bearing linear motion stages (Aerotech ABL1000, ALS130) for sample positioning and translation, and (4) a CCD camera for laser alignment and imaging of the target surface. Shah et al. provide a full description of the optical delivery system . Laser exposures were optimized for (1) minimizing wall roughness of microfluidic channels, (2) minimizing insertion and propagation losses of single mode WGs operating at 1.55 μm, (3) maximizing porosity of the conventional woodpiles for CEC testing and (4) tuning the 2nd order stopband of rotated woodpiles near the 1.55 μm telecommunication wavelength. Microfluidic channel templates and WGs were fabricated with 1 MHz repetition rate while PCs were formed with 100 kHz to avoid damage due to heat accumulation effects. The rotated woodpile PCs were fabricated with a 0.9 NA objective (Nikon 333387 BD Plan 100 × ) to generate the smallest pillars and rods while the conventional woodpile, channels and WGs were exposed through a 0.55 NA aspherical lens (40 × , f = 4.5 mm). A summary of all laser processing conditions is shown in Table 1. The properties of laser modification tracks varied with the scanning direction, thus requiring all scan lines in a single device to be written in only one optimal scan direction to avoid generating non-uniform PC structures by this directional quill effect .
In one LOAC approach, Micralyne chips (unbounded MC-BF4-SC) with semi-circular channels of 50 × 20 μm2 cross-sectional area provided the conventional CE design of Fig. 1(a). In an alternative approach, rectangular microchannels of 50 × 25 μm2 area were formed into polished glass substrates (1” × 2” Corning 7980) by laser writing of nanograting tracks (Microfluidic Channel recipe in Table 1) to form templates for guiding hydrofluoric acid (HF) etching. An 11 × 33 array of parallel modification tracks were fabricated according to Ho et al.  with laser polarization aligned perpendicular to the scan direction and thus facilitate nanograting plane alignment parallel with the channel flow direction for rapid HF etching . High and low exposure laser recipes were applied to define the channel interior and boundary, respectively, for the advantage of forming smooth side walls (12 nm rms roughness) around a rapidly etching channel interior. The laser modification tracks were etched with 5% aqueous HF solution (Aeros Organics) for 1 hour to open into channels.
Optical WGs were fabricated with the laser exposure in Table 1, with polarization aligned parallel with the scan direction. In this way, the laser-formed nanograting planes aligned perpendicular to the optical axis of the WGs and parallel with the microfluidic channels and substrate facets to inhibit HF etching into the WG . Additionally, the WGs were terminated ~5 μm from microfluidic channel walls and chip end-facets to further prevent undesired WG etching.
Waveguide mode profiles and transmission properties were characterized with a 1250-1650 nm broadband source (Agilent 83437A) coupled to the laser-formed WG end-facet with Corning SMF28e fiber and index matching oil (nn = 1.464, LRIA-165 Cargille Laboratories) to reduce Fresnel reflection losses. The light leaving the laser-formed WG was similarly collected to an optical spectrum analyzer (OSA) (Ando AQ6317B). Alternatively, near-field mode profiles of a monochromatic laser (Photonetics Tunics-BT) propagating through the WGs at 1.55 μm were captured by imaging the output facet with a 60 × aspherical lens onto a phosphor-coated CCD camera (Spiricon FireWire SCOR 20-1550). Eaton et al. previously described further details of the characterization arrangement .
The open microfluidic channels were filled with photoresist by dispensing SU-8 2050 to a uniform 40 μm thick layer at 3000 rpm to completely fill the channels without bubble formation. The conventional and rotated woodpile structures of Fig. 1(f) and 1(d), respectively, were laser exposed in line patterns according to the conditions of Table 1, while following the baking and development procedure of Chanda et al. . In the case of the rotated woodpile, vertical pillars were written to extend above the horizontal rods for top-view imaging with a scanning electron microscope (SEM). During post-exposure development, the PC structure would release from the channel sidewalls due to SU-8 photoresist shrinkage [37–41] and an extra laser writing procedure was applied to form 4 µm thick “adhesion walls” around the PC perimeter with the exposure recipe of Table 1. This adhesion layer was conformed to the shape of the walls to maximize the surface area connecting the PC. For CEC demonstration, a 0.5 μm silica coating was applied to the conventional woodpile PC by chemical vapor deposition (CVD)  following the procedure of Chanda et al  for the benefits of (1) improving the PC rigidness, (2) strengthening the PC-channel adhesion and (3) converting the hydrophobic PC surface to hydrophilic silica for an accelerated analyte flow rate through the porous PC.
The open microfluidic chips were sealed with a single layer (~2 mm) of degassed and cured PDMS (Sylgard 184 Silicone Elastomer Kit). Holes with ~5 mm diameter were first punched through the PDMS to overlay with the LOAC reservoirs and reduce Laplacian pressure for better control of EOF . A radio-frequency oxygen plasma (Harrick Plasma Cleaner PDC-3XG) was used for 30 seconds to functionalize both glass substrate and PDMS surfaces immediately prior to conformal contact and followed with heating for 1 hour at 130°C to create an irreversible bond .
The sealed microfluidic system was flushed with deionized H2O, 1.55 M aqueous HNO3 (Anachemia) and 1.0 M aqueous NaOH (Fisher) by vacuum (Marathon Electric) and immediately filled with 20 mM aqueous sodium tetraborate (Sigma-Aldrich 229946-SG). A 5 μL drop of 10−3 M fluorescein (Eastman Organic Chemicals) was pipetted into reservoir A and a pinch flow was established with 830 V, 680 V, 0 V and 830 V bias at reservoirs A, B, C, and D, respectively. Fluorescein plugs were ejected with 100 V, 0 V, 100 V and 500 V bias and followed along the channel with an inverted microscope (Zeiss AxioObserver) through a filter set (FITC 480 nm excitation) to observe the 521 nm plug emission.
4. Results and discussion
4.1 Probing of photonic crystal stopband
The optical microscope images of Fig. 3 demonstrate the integration of a 3D PC inside a DLW microfluidic channel within a polished glass substrate while facilitating spectral probing along the Γ–Z axis by a buried optical WG (right WG in Fig. 3(b)) to meet the basic design as proposed in Fig. 1(c). Supporting SU-8 adhesion walls are identified in Fig. 3(b) around which a rigid open 3D rotated woodpile structure was formed as seen in the top view SEM image of Fig. 3(c). The inset of Fig. 3(c) shows good correspondence of the calculated isointensity surfaces based on Fig. 1(d) that register accurately with the laser-formed pillars and logs of the rotated woodpile. The structural dimensions of 0.28 μm horizontal rod diameter (rHb = 3rHc = 0.42 μm) and 0.54 μm vertical pillar diameter (rV = 0.27 μm) together with the observed periodicities of a = b = 1.9 μm and c = 1.0 μm suggest a 40% SU-8 filling fraction.
Figure 4(a) shows the 12.3 μm × 12.8 μm mode profile of 1.55 μm wavelength light exiting from the 2.54 cm long WG. Figure 4(b) shows the measured transmission spectra through (1) a 2.54 cm long straight WG (blue), (2) a similar WG (red) intercepted by a 50 μm wide microfluidic channel (left WG of Fig. 3(b)) that was formed by DLW and (3) a similar WG (purple) intercepting the same channel when embedded with the rotated woodpile PC in air (right WG of Fig. 3(b)). All transmission spectra were normalized to the transmission spectrum for fiber-to-fiber coupling in refractive index matching oil and showed a weak 1.2 dB absorption band (23 nm FWHM) at 1.38 μm wavelength arising from OH groups in fused silica . The low 2.8 dB insertion loss of the single WG at 1.55 μm wavelength increased by 1.8 dB (Fig. 4(b) (red)) to match closely with the 1.75 dB of combined Fresnel and diffraction losses expected on crossing the air-filled channel (left WG in Fig. 3(b)). Fresnel reflection by the channel side walls are responsible  for the Fabry-Perot fringes of 25 nm free spectral range and 0.35 dB modulation depth seen in Fig. 4(b) (red) that matched closely with the calculated values of 24 nm and 0.35 dB, respectively.
Probing through the embedded PC yielded the spectrum in Fig. 4(b) (purple) to unveil a definitive 2nd order stopband of R = 3.5 dB strength, Δλ = 65 nm bandwidth and λo = 1.57 μm center wavelength that closely matched the simulated stopband properties for an air-filled rotated woodpile PC with an 85% filling fraction (Fig. 4(b) (green)). A 3rd order stopband at λo = 1.43 µm was also visible as a shoulder on the OH absorption band that also corresponded with the FDTD simulation. Fabry-Perot resonances (0.01 dB) were no longer apparent in Fig. 4(b) (purple), principally owing to a lower Fresnel reflectance (0.04%) between fused silica (nfs = 1.46) and the air-filled PC for which an effective refractive index of neff = 1.52 was estimated. A ~20 dB increase in insertion loss is noted across the PC in Fig. 4(b) (purple) in contrast with 0.7 dB loss expected solely from Fresnel and diffraction through a channel of uniform neff = 1.52 medium. The stopband was further distorted by inhomogeneous SU-8 shrinkage and weakened by optical scattering from photoresist surface roughness. Hence, upon immersing the PC with water, a much lower refractive index contrast led to considerable weakening of the stop band to preclude a definitive measure of the index change. In future work, one may consider laser writing of chirped Bragg gratings into the probing waveguides to create sharper Fabry Perot resonances that improve signal strength and sensing resolution.
The offset between the calculated (Fig. 2) and observed (Fig. 4) stopband wavelengths may be attributed to a much lower filling fraction of 40% in contrast with the 85% filling fraction that was inferred to match the FDTD spectrum. This underestimated filling fraction is expected from (1) the formation of thicker photoresist structures around the intersecting rods and pillars that were not visible in top-view SEM images and (2) the laser beam quality of M2 = 1.35 that increased the rod eccentricity by ~15%. Together, these motif changes accumulated to alter the filling fraction and unit cell from the ideal PC model (Fig. 1(d)), thereby shifting the observed stopband wavelength from simulated values. More symmetric and uniform PC structures may be anticipated with (1) using an oil immersion lens to eliminate beam distortions from the spherical aberration at the air-photoresist interface [28, 53], (2) compensating the spherical aberration with higher power exposure at deeper writing depth, and (3) adopting photoresists with lower known shrinkage rates [54, 55]. Despite these structural distortions, the novel means to detect an embedded 3D PC’s stopband with a probing laser-formed WG has been demonstrated for the first time to our knowledge.
Since contamination and clogging of the finely patterned 3D PC will require mass production of disposable chips, the slow DLW processing here will need to be scaled up to higher speed fabrication methods such as holographic or phasemask interference where large volume 3D PC templates that have been reported with strong 30 dB stopbands .
4.2 Acceleration of a mobile phase of analyte
When embedded into microfluidic channels, woodpile structures with the small a, b and c periodicities of 1.0 to 1.9 μm as presented in Section 4.1 were found to impede EOF flow through 3D PC template columns with lengths as short as 400 μm. Hence, it was necessary to increase the periodicity of the 3D PCs to enable CEC observation, which shifted the expected stopbands far outside the targeted 1.55 μm sensing spectrum.
Figures 5(a) and 5(b) show SEM images of a silica-coated conventional woodpile (a = b = 10 μm, c = 19.5 μm, RH = 0.9 μm, RV = 4.6 μm) embedded inside a Micralyne channel before PDMS sealing. After sealing, a fluorescein plug was formed by the voltage sequences of Section 3 to flow at 140 μm/s as seen in the optical images before (Fig. 5(c)) and after (Fig. 5(d)) interaction with the porous silica-coated PC. The fluorescence signal was dramatically weaker from the PC volume than the other channel regions due to a smaller fluid volume and a strong optical scattering loss from the PC structure. Hence, a small region (25 × 25 μm2) of the 3D PC column labelled as the “pocket” in Fig. 5(c) and 5(d) was intentionally opened without a PC structure to serve as an optical probe volume to (1) verify analyte flow through the porous PC, (2) confirm PDMS sealing above and around the PC and (3) verify plug flow speed. The fluorescein plug was found to accelerate through the PC region, doubling the flow speed to 285 μm/s in contrast with an anticipated 190 μm/s flow rate expected from solely considering the 1.33 × volumetric ratios of the open channel and the PC-embedded channel. Here, a PC filling fraction of 25 ± 5% filling fraction was calculated according to detailed SEM analysis of PC top views and cleaved cross-sections. Thus, the accelerated 285 μm/s flow rate is attributed to hydrophilic analyte-PC interactions that underlie adsorption chromatography.
The ~3.5 μm PC pore sizes tested here were found to be sufficiently large to permit plug flow through the silica coated PC surface and facilitate acceleration of a mobile phase of analyte via CEC principles. The flexibility of DLW to vary microfluidic channel widths and PC properties (a, b, c, RH, RV, pore sizes) were essential here to tune the PC structure and enable CEC observations. This flexibility is attractive in further tuning of the PC template structure to optimize electrophoretic mobilities and flow rates for demonstration of multi-analyte separation and eventually for extending CEC flow to PCs with smaller periods to exploit simultaneous refractive index sensing near the 1.43 to 1.55 μm stopbands presented here (Fig. 4), and further into the visible spectrum.
5. Significance and future work
Three dimensional PC templates were designed and embedded inside microfluidic channels for the objective of creating a spectroscopic-chromatographic PC column with simultaneous PC functionality. The flexibility of DLW offered the attractive option to rotate the Γ–Z axis of 3D PCs and enabled optical probing through integrated on-chip WGs, overcoming the need for external optical probing methods as otherwise required with colloidal or phase mask fabrication techniques that create PC templates without complete bandgaps. The results in Fig. 4(b) definitively demonstrate the unique approach for WG probing of the stopbands that have been presented by 3D PCs embedded inside a microfluidic channel. The flexibility of laser patterning in PC design (Fig. 2(b)) permits a wide range of templates to be considered for refractive index sensing and offers a sensitivity as high as 725 nm / RIU (Fig. 2(c)) that exceeds values of (1) 460 nm / RIU reported for a 1D PC-embedded channel , (2) 81 nm / RIU for a 1D evanescently-coupled Bragg-grating-WG , and (3) 125 nm / RIU for a 3D evanescently-coupled PC . However, to fully exploit this sensitivity, a spectral narrowing technique such as inducing Fabry-Perot resonator effects as shown in Fig. 4(b) (red) would be useful to sharpen the otherwise broad 65 nm stopband as observed in Fig. 4(b) (purple).
In the direction of demonstrating CEC within PC embedded channels, the periodicity of the 3D photoresist template was opened up and silica-coated to facilitate EOF and accelerate a mobile phase of analyte through adsorption chromatography principles, the first such demonstration in laser-patterned photoresist to our best knowledge. The flexibility of DLW to vary microfluidic channel widths and PC properties were essential to tune the PC structure for enabling such analyte acceleration and will be attractive for further optimization of electrophoretic mobility and flow rate for demonstrating multi-analyte separation. While such separation was previously demonstrated with colloidal crystals [15, 56], the novel flexibility inherent to DLW for tuning the PC structure will offer significant advantages for further advances in 3D PC-based CEC development. The DLW approach promises to create LOACs with multiple and more compact functionalities on a single chip that are flexible in (1) tuning a PC's Γ–Z periodicity, (2) positioning the WGs for probing PC stopbands and microfluidic channels, (3) dynamic tuning of CEC performance, and (4) carving out microfluidic channels and reservoirs into compact 3D networks. In this way, the 3D PC template may be designed and developed towards simultaneous demonstration of CEC and refractive index sensing in the IR and visible spectrum.
The development of simultaneous chromatographic and spectroscopic analysis in PC-embedded microchannels can be extended beyond colloidal columns [15, 56] and laser patterned photoresist to harness the high resolution of DLW and chemical etching directly in fused silica to form all-silica templates of 3D-PC structure  together with probing optical WGs. Such devices may support the higher pressures required for HPLC or be further extended to integrating laser-formed WGs with microfluidic channels that may replace photonic crystal fibers used previously for optical chromatography demonstrations .
Femtosecond laser processes were harnessed to develop a novel LOAC that integrated WGs, microfluidic channels and porous 3D PCs for the objective of simultaneously enabling (1) efficient separation and filtration of analytes through a porous PC embedded inside a microfluidic channel and (2) optical spectral characterization through embedded WGs that probe PC stopband shifts as varying analyte concentrations flow and separate. This paper showed that DLW was flexible in embedding a rotated woodpile PC with IR stopbands aligned for probing with buried laser-formed WGs, thereby removing the need for external optical probing of conventional PCs. The periodicity of conventional PCs were further opened up and silica-coated to facilitate, for the first time to our knowledge, acceleration of a mobile phase of analyte through a photoresist-based PC stationary phase by adsorption chromatography principles. Together, these laboratory results underpin the promise of simultaneous chromatographic and spectroscopic capabilities from a single PC optofluidic device where DLW of multi-functional integrated devices can serve in detection of cancerous or exogenous diseases, biohazardous mixtures and other biochemical compounds for clinical, forensic, pharmaceutical or security application.
The authors are grateful for helpful discussions with Tariq Rafique, Ladan Abolghasemi and Jianzhao Li during early parts of this project. This work was supported by the Natural Sciences and Engineering Research Council of Canada (NSERC), the Canadian Institute for Photonic Innovations (CIPI) and the Edward S. Rogers Sr. Graduate Scholarship.
References and links
2. C. Monat, P. Domachuk, and B. J. Eggleton, “Integrated optofluidics: A new river of light,” Nat. Photonics 1(2), 106–114 (2007). [CrossRef]
4. V. Maselli, J. R. Grenier, S. Ho, and P. R. Herman, “Femtosecond laser written optofluidic sensor: Bragg grating waveguide evanescent probing of microfluidic channel,” Opt. Express 17(14), 11719–11729 (2009). [CrossRef] [PubMed]
6. C. L. Smith, U. Bog, S. Tomljenovic-Hanic, M. W. Lee, D. K. Wu, L. O’Faolain, C. Monat, C. Grillet, T. F. Krauss, C. Karnutsch, R. C. McPhedran, and B. J. Eggleton, “Reconfigurable microfluidic photonic crystal slab cavities,” Opt. Express 16(20), 15887–15896 (2008). [CrossRef] [PubMed]
7. M. Deubel, G. von Freymann, M. Wegener, S. Pereira, K. Busch, and C. M. Soukoulis, “Direct laser writing of three-dimensional photonic-crystal templates for telecommunications,” Nat. Mater. 3(7), 444–447 (2004). [CrossRef] [PubMed]
8. S. Juodkazis, V. Mizeikis, and H. Misawa, “Three-dimensional structuring of resists and resins by direct laser writing and holographic recording,” Adv. Polym. Sci. 213, 157–206 (2008).
9. J. Wu, D. Day, and M. Gu, “A microfluidic refractive index sensor based on an integrated three-dimensional photonic crystal,” Appl. Phys. Lett. 92(7), 071108 (2008). [CrossRef]
11. C. J. Choi and B. T. Cunningham, “A 96-well microplate incorporating a replica molded microfluidic network integrated with photonic crystal biosensors for high throughput kinetic biomolecular interaction analysis,” Lab Chip 7(5), 550–556 (2007). [CrossRef] [PubMed]
12. L. L. Chan, S. L. Gosangari, K. L. Watkin, and B. T. Cunningham, “Label-free imaging of cancer cells using photonic crystal biosensors and application to cytotoxicity screening of a natural compound library,” Sens. Actuators B Chem. 132(2), 418–425 (2008). [CrossRef]
14. H. J. Kim, S. Kim, H. Jeon, J. Ma, S. H. Choi, S. Lee, C. Ko, and W. Park, “Fluorescence amplification using colloidal photonic crystal platform in sensing dye-labeled deoxyribonucleic acids,” Sens. Actuators B Chem. 124(1), 147–152 (2007). [CrossRef]
15. U. Kamp, V. Kitaev, G. von Freymann, G. A. Ozin, and S. A. Mabury, “Colloidal crystal capillary columns - towards optical chromatography,” Adv. Mater. 17(4), 438–443 (2005). [CrossRef]
17. R. Osellame, H. J. W. M. Hoekstra, G. Cerullo, and M. Pollnau, “Femtosecond laser microstructuring: an enabling tool for optofluidic lab-on-chips,” Laser Photon. Rev. 5(3), 442–463 (2011). [CrossRef]
18. J. Bhawalkar, G. He, and P. Prasad, “Nonlinear multiphoton processes in organic and polymeric materials,” Rep. Prog. Phys. 59(9), 1041–1070 (1996). [CrossRef]
19. J. Chan, T. Huser, S. Risbud, J. Hayden, and D. Krol, “Waveguide fabrication in phosphate glasses using femtosecond laser pulses,” Appl. Phys. Lett. 82(15), 2371–2373 (2003). [CrossRef]
20. S. M. Eaton, H. Zhang, P. R. Herman, F. Yoshino, L. Shah, J. Bovatsek, and A. Y. Arai, “Heat accumulation effects in femtosecond laser-written waveguides with variable repetition rate,” Opt. Express 13(12), 4708–4716 (2005). [CrossRef] [PubMed]
21. R. Osellame, N. Chiodo, G. Della Valle, G. Cerullo, R. Ramponi, P. Laporta, A. Killi, U. Morgner, and O. Svelto, “Waveguide lasers in the c-band fabricated by laser inscription with a compact femtosecond oscillator,” IEEE J. Sel. Top. Quantum Electron. 12(2), 277–285 (2006). [CrossRef]
22. H. Zhang, S. M. Eaton, J. Li, A. H. Nejadmalayeri, and P. R. Herman, “Type II high-strength Bragg grating waveguides photowritten with ultrashort laser pulses,” Opt. Express 15(7), 4182–4191 (2007). [CrossRef] [PubMed]
23. N. D. Psaila, R. Thomson, H. Bookey, A. Kar, N. Chiodo, R. Osellame, G. Cerullo, A. Jha, and S. Shen, “Er:Yb-doped oxyfluoride silicate glass waveguide amplifier fabricated using femtosecond laser inscription,” Appl. Phys. Lett. 90(13), 131102 (2007). [CrossRef]
24. H. Zhang, S. Ho, S. M. Eaton, J. Li, and P. R. Herman, “Three-dimensional optical sensing network written in fused silica glass with femtosecond laser,” Opt. Express 16(18), 14015–14023 (2008). [CrossRef] [PubMed]
25. W. J. Chen, S. M. Eaton, H. Zhang, and P. R. Herman, “Broadband directional couplers fabricated in bulk glass with high repetition rate femtosecond laser pulses,” Opt. Express 16(15), 11470–11480 (2008). [CrossRef] [PubMed]
27. Y. Bellouard, A. Said, M. Dugan, and P. Bado, “Fabrication of high-aspect ratio, micro-fluidic channels and tunnels using femtosecond laser pulses and chemical etching,” Opt. Express 12(10), 2120–2129 (2004). [CrossRef] [PubMed]
28. C. Hnatovsky, R. Taylor, E. Simova, P. Rajeev, D. Rayner, V. Bhardwaj, and P. Corkum, “Fabrication of microchannels in glass using focused femtosecond laser radiation and selective chemical etching,” Appl. Phys., A Mater. Sci. Process. 84(1–2), 47–61 (2006). [CrossRef]
29. A. Marcinkevičius, S. Juodkazis, M. Watanabe, M. Miwa, S. Matsuo, H. Misawa, and J. Nishii, “Femtosecond laser-assisted three-dimensional microfabrication in silica,” Opt. Lett. 26(5), 277–279 (2001). [CrossRef] [PubMed]
30. V. Maselli, R. Osellame, G. Cerullo, R. Ramponi, P. Laporta, L. Magagnin, and P. L. Cavallotti, “Fabrication of long microchannels with circular cross section using astigmatically shaped femtosecond laser pulses and chemical etching,” Appl. Phys. Lett. 88(19), 191107 (2006). [CrossRef]
31. S. Ho, P. R. Herman, and J. S. Aitchison, “Single- and multi-scan femtosecond laser writing for selective chemical etching of cross section patternable glass micro-channels,” Appl. Phys., A Mater. Sci. Process. 106(1), 5–13 (2012). [CrossRef]
32. R. A. Borisov, G. N. Dorojkina, N. I. Koroteev, V. M. Kozenkov, S. A. Magnitskii, D. V. Malakhov, A. V. Tarasishin, and A. M. Zheltikov, “Femtosecond two-photon photopolymerization: a method to fabricate optical photonic crystals with controllable parameters,” Laser Phys. 8(5), 1105–1108 (1998).
33. J. Serbin, A. Egbert, A. Ostendorf, B. N. Chichkov, R. Houbertz, G. Domann, J. Schulz, C. Cronauer, L. Fröhlich, and M. Popall, “Femtosecond laser-induced two-photon polymerization of inorganic-organic hybrid materials for applications in photonics,” Opt. Lett. 28(5), 301–303 (2003). [CrossRef] [PubMed]
37. D. Chanda, L. Abolghasemi, and P. R. Herman, “One-dimensional diffractive optical element based fabrication and spectral characterization of three-dimensional photonic crystal templates,” Opt. Express 14(19), 8568–8577 (2006). [CrossRef] [PubMed]
38. D. C. Meisel, M. Diem, M. Deubel, F. Perez-Willard, S. Linden, D. Gerthsen, K. Busch, and M. Wegener, “Shrinkage precompensation of holographic three-dimensional photonic-crystal templates,” Adv. Mater. 18(22), 2964–2968 (2006). [CrossRef]
39. J. Jang, C. Ullal, M. Maldovan, T. Gorishnyy, S. Kooi, C. Koh, and E. Thomas, “3D micro- and nanostructures via interference lithography,” Adv. Funct. Mater. 17(16), 3027–3041 (2007). [CrossRef]
40. D. Chanda, N. Zachari, M. Haque, M. Ng, and P. R. Herman, “Flexible fabrication of three-dimensional optical-domain photonic crystals using a combination of single-laser-exposure diffractive-optics lithography and template inversion,” Opt. Lett. 34(24), 3920–3922 (2009). [CrossRef] [PubMed]
41. R. G. Denning, C. F. Blanford, H. Urban, H. Bharaj, D. N. Sharp, and A. J. Turberfield, “The control of shrinkage and thermal instability in SU-8 photoresists for holographic lithography,” Adv. Funct. Mater. 21(9), 1593–1601 (2011). [CrossRef]
42. X. J. Liang, A. Q. Liu, C. S. Lim, T. C. Ayi, and P. H. Yap, “Determining refractive index of single living cell using an integrated microchip,” Sens. Actuators A Phys. 133(2), 349–354 (2006).
43. D. H. Luo, R. A. Levy, Y. F. Hor, J. F. Federici, and R. M. Pafchek, “An integrated photonic sensor for in situ monitoring of hazardous organics,” Sens. Actuators B Chem. 92(1–2), 121–126 (2003). [CrossRef]
44. L. Shah, A. Arai, S. Eaton, and P. R. Herman, “Waveguide writing in fused silica with a femtosecond fiber laser at 522 nm and 1 MHz repetition rate,” Opt. Express 13(6), 1999–2006 (2005). [CrossRef] [PubMed]
45. P. Kazansky, W. Yang, E. Bricchi, J. Bovatsek, A. Arai, Y. Shimotsuma, K. Miura, and K. Hirao, “Quill writing with ultrashort light pulses in transparent materials,” Appl. Phys. Lett. 90(15), 151120 (2007). [CrossRef]
46. S. M. Eaton, H. Zhang, M. L. Ng, J. Li, W. J. Chen, S. Ho, and P. R. Herman, “Transition from thermal diffusion to heat accumulation in high repetition rate femtosecond laser writing of buried optical waveguides,” Opt. Express 16(13), 9443–9458 (2008). [CrossRef] [PubMed]
47. D. Chanda, L. E. Abolghasemi, M. Haque, M. L. Ng, and P. R. Herman, “Multi-level diffractive optics for single laser exposure fabrication of telecom-band diamond-like 3-dimensional photonic crystals,” Opt. Express 16(20), 15402–15414 (2008). [CrossRef] [PubMed]
48. N. Tétreault, G. von Freymann, M. Deubel, M. Hermatschweiler, F. Pérez-Willard, S. John, M. Wegener, and G. Ozin, “New route to three-dimensional photonic bandgap materials: silicon double inversion of polymer templates,” Adv. Mater. 18(4), 457–460 (2006). [CrossRef]
50. A. Plecis and Y. Chen, “Fabrication of microfluidic devices based on glass-PDMS-glass technology,” Microelectron. Eng. 84(5–8), 1265–1269 (2007). [CrossRef]
51. I. H. Malitson, “Interspecimen comparison of the refractive index of fused silica,” J. Opt. Soc. Am. 55(10), 1205–1208 (1965). [CrossRef]
52. H. Abu-Safia, R. Al-Tahtamouni, I. Abu-Aljarayesh, and N. A. Yusuf, “Transmission of a Gaussian beam through a Fabry-Perot interferometer,” Appl. Opt. 33(18), 3805–3811 (1994). [CrossRef] [PubMed]
53. S. Wong, M. Deubel, F. Pérez-Willard, S. John, G. Ozin, M. Wegener, and G. V. Freymann, “Direct laser writing of three-dimensional photonic crystals with a complete photonic bandgap in chalcogenide glasses,” Adv. Mater. 18(3), 265–269 (2006). [CrossRef]
54. A. Hayek, Y. Xu, T. Okada, S. Barlow, X. Zhu, J. H. Moon, S. R. Marder, and S. Yang, “Poly(glycidyl methacrylate)s with controlled molecular weights as low-shrinkage resins for 3D multibeam interference lithography,” J. Mater. Chem. 18(28), 3316–3318 (2008). [CrossRef]
55. A. Ovsianikov, J. Viertl, B. Chichkov, M. Oubaha, B. MacCraith, I. Sakellari, A. Giakoumaki, D. Gray, M. Vamvakaki, M. Farsari, and C. Fotakis, “Ultra-low shrinkage hybrid photosensitive material for two-photon polymerization microfabrication,” ACS Nano 2(11), 2257–2262 (2008). [CrossRef] [PubMed]
57. S. Ho, M. Haque, P. R. Herman, and J. S. Aitchison, “Femtosecond laser-assisted etching of three-dimensional inverted-woodpile structures in fused silica,” Opt. Lett. 37(10), 1682–1684 (2012). [CrossRef] [PubMed]
58. P. C. Ashok, R. F. Marchington, P. Mthunzi, T. F. Krauss, and K. Dholakia, “Optical chromatography using a photonic crystal fiber with on-chip fluorescence excitation,” Opt. Express 18(6), 6396–6407 (2010). [CrossRef] [PubMed]