Rhodamine B and Alexa Fluor 430 fluorophores have been used as doping agents for xerogel waveguides defined over an antiresonant (ARROW) filter. This configuration has a significant level of integration, since it merges the waveguide, the light emitter and the filter in a single photonic element. Different technologies have been combined for their implementation, namely soft lithography, standard silicon-based technology and silicon bulk micromachining. The spectral response of 15-mm long waveguides without fluorophore is first analyzed as a function of the waveguide width. Here, it has been observed how the xerogel used has a high transparency in the visible spectra, having only significant absorption at the wavelength where the ARROW filter is in resonance. In a second step, identical waveguides but doped with two different concentrations of Rhodamine B and Alexa Fluor 430 are studied. In addition to the effect of the filter, fluorophore-doped xerogel waveguides show losses close to −2 dB (equivalent to 2 dB of light emission). In addition, it has been observed how an increase of the fluorophore concentration within the xerogel matrix does not provide with a emission increase, but saturation or even a decrease of this magnitude due to self-absorption. Finally, the total losses of the proposed waveguides are analyzed as a function of their width, obtaining losses close to 5 dB for waveguide widths higher than 50 µm.
©2011 Optical Society of America
Using light as an interrogation mechanism has become the predominant detection method in (bio)chemical analysis, mainly due to the ability to provide high sensitivity in robust configurations. There are different working principles in which photonic systems can be based on, as could be changes in intensity , wavelength , phase , polarization , or coherence. Other advantages of photonics systems are (i) systems based on any of these light properties remain unaffected by electromagnetic interferences (EMI). (ii) They cannot generate short circuits and/or spikes. (iii) The electronics and total weight are reduced to a low-cost light source [e.g., a light-emitting diode (LED)] and a standard silicon photodetector when making measurements in intensity. (iv) Fiber optics interconnection assures having the electrical components locally displaced from the measured zone, hence minimizing the risk when scanning harsh environments. Albeit, one of the major issues that prevents the massive implantation of photonic systems is actually the interconnection of the system with the fiber optics, that is, the in/outcoupling of the signals. Nowadays, pig-tailed coupling is the most common method of choice, either with specular facets  or with tapered couplers . Such approach either requires complex technology or extensive back-end processing. In addition, the facet quality of both the photonic system and the fiber optics has to be kept at a very high standard, since even a minor scratch causes a fast degradation of the system performance, thus unfulfilling the low cost approach. The tackling of these issues have recently been partially achieved with the monolithic implementation of self-alignment systems . Nevertheless, they only solve the issue regarding the photonic system facet; the fiber optic still needs being cleaved to ensure efficient light coupling. Optimal configuration would be to have the light source integrated in the photonic system. A labor intensive research has recently been directed to the hybridization of microoptics with laser diodes, with promising results , but still requiring precise handling, hence being hampered by production costs. In contrast, when the photonic system and the light source are defined in the same technological process, the components are inherently self-aligned. Steps towards this approach have already been given , where a laser dye is injected in a microfluidic channel and the emitted light is directly coupled in a SU-8 waveguide. Liquid light sources are undoubtedly interesting, but they are solvent-sensitive and may experience fast ageing. In addition, laser dyes and, more generally, photoluminescence-based light emitters have the inherent drawback of requiring filters so as to block the excitation wavelength. Some steps have been done with the aim of integrating filters, either absorbance-based [10,11] or interferometric . Absorbance-based filters have the advantage of easy processing, but with a very limited adjustability and smooth stopband-passband transition. Multilayered filters have the advantage of a sharp transition, but are angle-dependant and with strong fabrication requirements. A third type of filters, named antiresonant, have so far only been used for confining light in a lower refractive index material, with the so-called Antiresonant Reflecting Optical Waveguides (ARROW) , either with a silicon oxide  or a polymer core . Antiresonant filtering has as a main advantage the relatively sharp transition while reducing the fabrication requirements. In addition, such waveguides are virtually single-moded, since only the fundamental mode propagates in the structure.
Photonics has been paved with the continuous development of smart materials, as could be chalcogenide glasses , photostructurable polymers , elastomers [18,19] or organic-inorganic silicate-based glasses fabricated by sol-gel technology . The main advantage of the latter is that they provide the appropriate solid matrix into which different functional components, even from biological origin, can be easily incorporated. Here, a large number of hybrid xerogel matrices doped with either inorganic or organic components have been reported. Some interesting reports are based on the use of zirconia for defining waveguides  and distributed feedback lasers (DFB) , an erbium/ytterbium/aluminum compound in amplifiers , hydrogen peroxidase (HRP) for full field sensors  or indicators for monitoring local adsorption . Patterning of hybrid organic-inorganic materials has mostly been done by UV-lithography , two-photon absorption  and soft lithography . The latter has been shown to provide with several outstanding properties, such as high fidelity in transferring the patterns from the mould to the substrate or the flexibility of generating arbitrarily-shaped structures in the same chip. In this context, structures higher than 70 µm and with aspect ratios higher than 1:10 (width:height) have been reported .
Despite the good performance of doped xerogel waveguides from one side, and ARROW structures from the other side, up to date there has been no reported approaches based on the mergence of both technologies. In this work, we present a monolithic integration of an ARROW filter with a fluorophore-doped xerogel waveguide. The layers of the antiresonant structure are implemented so as to have antiresonant condition for the emission wavelength assuring the confinement of that wavelength to the waveguide, whereas excitation is fastly transferred to the substrate where it is absorbed. Hence, overall performance of the photonic system is enhanced while keeping the technological requirements at a reasonable level.
2.1 ARROW filters
ARROW consist of a multilayer structure where the light is coupled to the core by means of total internal reflection in the external interfaces, while first and second cladding layers conform a Fabry-Perot mirror that reflects the light in the internal interface. ARROW filters have to be designed to maximize the reflection of the emitted wavelengths while minimizing the reflection for the excitation wavelengths. The Antiresonant condition depends on the working wavelength, the layer thicknesses, the refractive indices and the mode order. Then, if the Fabry-Perot is properly designed, ARROW structures are virtually single-moded, since modes higher than the fundamental have high attenuation.
Figure 1 shows a schematic view of the ARROW multilayer structure with all its elements. Thickness and refractive index of the 1st and 2nd claddings must be tuned to reach high propagation losses for excitation light while emitted light must be coupled to the waveguide core. Silicon (Si) has been considered as the substrate for the ARROW filter, while the material for the 2nd cladding has been fixed as silicon oxide (SiO2; email@example.com) and material for 1st cladding has been fixed as silicon nitride (Si3N4, firstname.lastname@example.org). A xerogel (email@example.com) has been considered as the core of the waveguide while air (firstname.lastname@example.org) as the external environment.
In order to appropriately design the ARROW filters, FIMMWAVE (Photon Design, UK) has been used to simulate such structures. Three different wavelengths have been considered: 430 nm, 530 nm and 625 nm. The first one as the excitation light that has to be filtered, while the second and third ones have to be confined in the waveguide. Figure 2(a) shows the variation of propagation losses vs. the thickness of the 1st cladding (d1) for a fixed thickness of the 2nd cladding (d2) of 1 µm. As it can be observed, the propagation losses are maximized for the wavelength of 430 nm at d1=176 nm with a value of 6.3e−2 dB/cm, while the losses for the wavelengths of 530 nm and 625 nm are of 3.4e−3 dB/cm and 9.3e−3 dB/cm, respectively. Figure 2(b) shows the propagation losses for the three wavelengths as a function of d2 with a fixed value of d1=176 nm. Losses decrease with d2 at a different rate for the three wavelengths. Considering a fixed thickness of 1 µm for the 2nd cladding, the propagation losses for 530 nm and 625 nm are one order of magnitude smaller than these corresponding to a wavelength of 430 nm. Such difference can be even increased by decreasing d2. However, this will result as well in an increasing of the propagation losses for the emitted light. For the ARROW filters presented in this paper a 2nd cladding thickness of 1 µm was used.
Table 1 shows the parameters of the designed ARROW filter with the simulated propagation losses for the three wavelengths. The difference of the propagation losses between the excitation and emission wavelengths assures that the first one will be absorbed rapidly while the latter ones will remain confined into the core of the hybrid xerogel-ARROW waveguide.
2.2 Facet ends
As it was discussed in the previous section, one of the most challenging issues in photonic systems is the definition of high quality facets. Roughness with dimensions similar to the working wavelength causes a dramatic decrease of their performance. Generally, this issue is addressed by either cleaving or polishing of the facet ends. This additional step causes a significant increase of the fabrication costs. Lateral facets are also of key importance, since it has previously been shown that the confinement strongly depends on the verticality and quality of the lateral walls . Taking these issues into account, and also considering that the sol solutions would be injected in liquid phase, the waveguides have been designed considering a microfluidic approach, as shown in Fig. 3 , having a total length of 15 mm (Fig. 3 has been vertically cut for visualization easiness). Fluidic inlet and outlet were directly connected to a serpentine-shaped microchannel, with varying widths of 200 µm, 150 µm, 100 µm, 60 µm and 20 µm. The height of the microchannel was fixed to 30 µm. Connection between two consecutive waveguides was done by means of an auxiliary channel defined at 90°. When the sol reached this region, two unstable meniscuses could appear, which corresponded to the two outer vertices of the facet ends. There, uncontrolled air entrapment might take place, thereby hampering the facet end quality. To control this effect, dead end channels with a width of 20 µm were defined at these corners, as shown at the inset of Fig. 3. Air entrapment were thus displaced inside this channel and hence did not affect the facet ends.
The implementation of fluorophore-doped ARROW structures is based on the combination of soft-lithography and bulk silicon micromachining (for defining the waveguide and the filter, respectively). The substrate used was in both cases a N type, (100) oriented silicon wafer with a diameter of 100 mm. The masters were fabricated using SU-8 negative photoresist. It is known that the adhesion between silicon and SU-8 is not optimal, which causes the SU-8 masters to detach. Adhesion promoters can also be used, but at the expenses of increasing the fabrication steps and cost. To this effect, the SU-8 seed layer approach, already presented in  was used in this work. After cleaning and dehydrating the substrate, a 4-µm thick SU-8 layer (SU-8 2005, MicroChem Corporation, Newton, MA, USA) is spun. Hence, the subsequent polymer layers adhered to this initial SU-8 film, resulting in a better mechanical stability. Then, the wafer was dried at 95°C for 1 hour. When the room temperature was reached again, it was flood-exposed. The Post Exposure Bake (PEB) finished the definition of this SU-8 layer for adhesion enhancement purposes.
An issue that needs to be addressed and solved is the alignment between the fluorophore-doped xerogel and the filter structure. Similar to the situation presented in , a two-level master was fabricated. As a first step, SU-8 2025 was used to obtain layers with thicknesses of 30 µm and exposed to UV light with the mask presented in Fig. 4(a) , which allowed self-alignment with the ARROW filters, as it is explained below. When the PEB of this layer was finished, a second identical SU-8 layer was spun, in this case the mask used corresponded to the microchannels used for defining the waveguides (Fig. 4(b)). After the PEB of this second layer, the wafers were developed in Propylen Glycol Methyl Ether Acetate (PGMEA, MicroChem Corporation, Newton, MA, USA). This step finished the master fabrication (Fig. 4(c)). PDMS pre-polymer solution was made by thoroughly mixing 5 ml of the silicon elastomer and 0.5 ml of the curing agent (Sylgard 184 elastomer kit, Dow Corning, Midland, MI, USA). The mixture was placed under vacuum to remove air bubbles. Then, it was carefully poured onto the SU-8 master to cover its entire area, and finally cured at 80°C for 30 minutes. The fabricated PDMS mold was then peeled off the master with the aid of tweezers and stored in a closed container under clean room conditions (Fig. 4(d)).
The fabrication of the filter started using an identical (100) oriented silicon wafer. After its cleaning and dehydration, a 50 nm thick wet thermal silicon dioxide (SiO2) layer was grown; this layer would improve the photoresist adherence and would act as an etching mask. Then, a photolithographic step followed by a SiO2 wet etching was performed to define the windows where the regions for the positioning of the fiber optics would be implemented. A Deep Reactive Ion Etching (DRIE) with the appropriate conditions so as to minimize the scalloping was used in this 250 µm etching step. Then, a 1 µm thermal oxide (n=1.46 @ 633 nm) was grown, followed by a 176 nm silicon nitride (n=2.01 @ 633 nm), which was deposited by Low Pressure Chemical Vapor Deposition (LPCVD). These two layers form the Fabry-Perot of the ARROW structure (Fig. 4(e)). In accordance to the refractive indices and thicknesses, they are tuned in antiresonance for the emitted wavelengths whereas excitation wavelength matches the resonant condition. Hence, emitted wavelength will remain confined in the xerogel waveguide, whereas excitation wavelength will be transferred to the substrate.
At this point, the PDMS mould was placed over the filter structures (Fig. 4(f)). The dimensions of the silicon chip into which the filters were defined were identical to these of the first SU-8 layer. Hence, both structures directly assembled, thus ensuring the waveguide alignment.
Hybrid xerogel glasses were prepared by hydrolysis and polycondensation reactions taking place in a solution containing 615 μL of methyltrimethoxysilane (≥98.0%, MTMOS), 75 μL phenyltrimethoxysilane (purum, PhTMOS) and 60 μL of Tetramethoxysilane (≥99.0%, TMOS) and 2750 µL of water, at room conditions. Methanol, generated by the hydrolysis of the different silane monomers during the preparation of the sol solution, and also water were removed from the solution by stirring for up to 20 hours until it lost around 40% of its initial weight. During this process the viscosity of the sol solution increased, this fact being important for the appropriate fabrication of the xerogel waveguides. At this point 100 µL of the resulting sol solution were added to 30 µL of each fluorophore solution prepared in water at a fixed concentration (Rhodamine B and Alexa Fluor 430, Sigma Aldrich, Germany). The mixture was gently stirred for several minutes using a vortex. After that, 4 μL of the sol solution containing the fluorophore were pipetted at the inlet of the PDMS mould and left to fill the waveguide channels by capillary action, using the so-called Micromolding in Capillaries (MIMIC) soft lithographic approach . The process of filling the five different waveguides (having a length of 15 mm each) and the interconnection microchannels took less than a minute. When the sol solution reached the outlet of the PDMS mold, more sol solution was added to both the inlet and outlet, thus ensuring that during the polymerization process, the structures were not partially emptied (Fig. 4(g)). Then, the PDMS/substrate system was left undisturbed in a sealed contained at 4°C for at least three days to allow the polymer to gel, age and dry, forming a xerogel. After that, the PDMS mold was eventually peeled off the substrate, thus finishing the fabrication process (Fig. 4(h)).
Scanning electron microscopy (SEM) pictures were taken after the fabrication process, and an example is shown in Fig. 5(a) . It can be observed how the waveguides are perfectly defined and that even the dead end channels have also been obtained. Indeed, the dimensions of the fabricated waveguides were the same to those ones of the microstructures in the mold. Figure 5(b) shows a close-up view of the facet ends, where it can be seen that the replicated waveguides defined by soft lithography over ARROW filters are crack-free and with a very low defect level. These results verify the reliability and robustness of the MIMIC fabrication process when working with the xerogel composition and experimental conditions mentioned above.
Two slightly different setups have been implemented for the characterization of the fluorophore-doped and the raw (non-doped) xerogel waveguides. Firstly, in order to test the adequate behavior of the integrated filter, their spectral response was obtained by coupling a broadband light source (Ocean Optics HL-2000, Dunedin, FL, USA) into the input multimode fiber optics (with a core diameter of 50 µm), which was aligned with the help of micropositioners to the facet ends of the waveguides. The readout comprised an identical fiber optics, which carried the signal to a spectrometer (Ocean Optics HR4000, Dunedin, Fl, USA) with a spectral resolution of 0.2 nm. Then, the same setup was used to characterize the Rhodamine B-doped waveguides as a function of the width for two different fluorophore concentrations. For a second characterization of the Alexa-doped waveguides the excitation broadband light source was replaced by a LED with nominal working wavelength of 430 nm. Finally, the total losses of the five different xerogel waveguides fabricated in the same chip, as mentioned above (including insertion losses and attenuation) were determined using a wavelength in which none of the fluorophores have strong excitation bands (632 nm).
Images of the raw (non-doped), Rhodamine B and Alexa 430 waveguides are shown in Fig. 6(a) , 6(b) and 6(c), respectively. Excellent waveguide definition can be observed in the three cases (total length of 15 mm). The small amount of scattered light during the coupling in such waveguides is indicative of the low number of defects (playing the role of scattering centers) and straightforwardly, of the good light confinement. Hence, it can be concluded that waveguides doped with both Rhodamine B (10−3 M, Fig. 6(b)) and Alexa Fluor 430 (10−3 M, Fig. 6(c)) structurally present the same quality when compared with the non-doped ones. Therefore, the inclusion of such fluorophores did not appear to affect the xerogel polymerization process and the resulting material quality.
The spectral response of the five xerogel waveguides with different widths and defined over the ARROW filter are shown in Fig. 7 . When comparing with the fiber-to-fiber spectra (also shown in Fig. 7), the effect of the integrated filter can be easily observed at a wavelength of 435 nm (marked with a black arrow in the graphs). The difference between the experimental and simulated filtering wavelength is due to the shift between the expected thickness of the 1st cladding layer (176 nm) and the experimental one (178 nm). This difference is not critical and the filter is valid for the working range. Hence, this characterization validates the proposed technology for fabricating low-cost xerogel waveguides on ARROW filters. It is observed the expected increase of the coupled light intensity for wider waveguides until the dimensions of both the fiber optics and the waveguide are similar. Beyond this point, a decrease of the intensity is observed. This effect is well known, and is related to the fact that not all the light propagating in the waveguide is coupled to the readout fiber optics.
Once the adequate working behavior of the ARROW filters was shown, the next step was to check the effect of the fluorophores on the optical properties of the xerogel waveguides. Two different concentrations of each fluorophore (Rhodamine B and Alexa 430) were used in the fabrication process described above. To just determine the effect of the fluorophore, the results obtained for the raw xerogel waveguide were used as reference. Figure 8 shows the losses due to this fluorophore (fluorophore losses) as a function of the wavelength for Rhodamine-B (Fig. 8(a)) and Alexa Fluor 430 (Fig. 8(b)) waveguides with different widths. In this situation, it could be concluded that the ARROW filter (shown with a black arrow) caused a decrease of the fluorophore losses. This does not have to be considered as a higher light intensity reaching the output fiber optics at the resonant frequency of the ARROW filter. Conversely, it is a consequence of having used the spectral response of the raw waveguides as reference for determining the fluorophore losses of the doped waveguides. Once this effect is understood, it can be observed how fluorophore losses significantly increase at the region where they have their excitation bands. It has to be noted, however, that there exists a small shifting of the excitation and emission bands as compared to these previously reported. This is because the position of such bands is solvent-dependant . Hence, their inclusion in a xerogel matrix appeared to cause a shift towards longer wavelengths. Nevertheless, their behavior as fluorophores remained unaltered. This point can be verified at the wavelengths were photonic re-emission may occur. At this region, graphs for both fluorophores show negative values of their losses, which have to be considered as emission. Interestingly, the increase of the fluorophore concentration does not mean an increase of the emission, but just the opposite. This effect can be understood if it is taken into account that both fluorophores have a large overlap between the excitation and emission bands. Hence, larger amounts of such dopants may induce a higher self-absorption, thus hampering the overall performance of the proposed xerogel based light emitters.
Despite the reasonable resistance of Rhodamine dyes to photodegradation, together with its pH insensitivity , it has been demonstrated that the conjugation of rhodamine-based dyes to proteins cause a considerable fluorescence quenching . Conversely, Alexa fluor fluorescent dyes have similar resistance to degradation and pH insensitivity, but also a brighter emission that is retained after its conjugation. Hence, only Alexa Fluor 430 doped xerogel waveguides were tested with excitation light sources with smaller bandwidth. Here, a blue LED was used. In Fig. 9 , the spectra obtained with the raw xerogel waveguides and the waveguides doped with two concentrations of the Alexa Fluor 430 fluorophore, are shown. In addition, for comparison purposes, the spectra obtained when the input and output fiber optics were faced is also presented. When comparing the spectra of the doped waveguides with the fiber-to-fiber, the effect of the ARROW filter can be clearly observed (black arrow). This high filtering efficiency validates the proposed technology, since the most standard optical excitation source when working with fluorophore molecules does not use broadband source, but a monochromatic light. Hence, the results herein suggest that if a laser source was used, the excitation wavelength would be highly filtered. In addition, the effect of the increasing self-absorption for higher fluorophore concentrations can again be observed, with a dramatic decrease of the emitted intensity when doping the xerogel waveguides with 10−3M of Alexa Fluor 430.
Finally, the total losses of the proposed waveguides were determined by selecting a wavelength which falls outside the fluorophore absorption bands. Here, all the waveguides should have an identical response. To verify this issue, a 2.5 mW laser source (Thorlabs S1FC635) with working wavelength of 635 nm was used considering the in-house equipment available. The results are presented in Fig. 10 . As expected, the five different types of waveguides (raw, doped with two different concentrations of Rhodamine B and doped with two different concentrations of Alexa fluor 430) showed an identical response (within the experimental error). For large cores the total losses (that include insertion losses and attenuation) have a constant value close to 5 dB for any waveguide width. As the width decreases, the higher order modes were filtered out, hence increasing the total losses. Attenuation was estimated to be close to 1.6 dB/cm. These values are very similar to the previously reported in , where doped sol-gel waveguides were obtained. Then, it can be concluded that xerogel waveguides functionalized with fluorophores and defined over ARROW filters are a very promising configuration for merging light sources, waveguides and filters into a single photonic system.
Fluorophore-doped xerogel waveguides defined over an ARROW filter were designed, fabricated and characterized. Fabrication was based on two complementary technologies: soft lithography and silicon technology. The former was used for defining a PDMS mould with a serpentine-shaped microchannel, whereas the latter was used for defining the ARROW filter with high accuracy so as to select the rejection bands. Finally, with the positioning of the PDMS over the ARROW filter and using again a soft lithographic method, the doped xerogel waveguides were implemented. Facet quality was also assured by a proper modification of the unstable meniscuses. Characterization demonstrated the adequate working behavior of the ARROW filter when broadband light was coupled to raw xerogel waveguides. Similarly, when measuring doped xerogel waveguides, together with the ARROW filter, the properties of either Rhodamine B or Alexa Fluor 430 were transferred to the waveguide, obtaining absorption bands (at the excitation wavelengths) and gain due to photonic reemission. In this latter case, emitted light remained confined in the waveguide, hence merging the waveguide and the active material. These results were obtained either with broadband light or with a LED and confirmed the working behavior of the proposed configuration. In addition, two fluorophore concentrations were tested, and it was demonstrated that due to the overlapping between both bands, self-absorption took place for the highest concentration, resulting in an overall decrease of the waveguide as light emitter.
Finally, the total losses (including insertion and propagation) were measured, obtaining values close to 5 dB for waveguides wider than 50 µm. These results confirm the validity of the proposed fluorophore-doped xerogel waveguides as integrated light emitters on photonic systems.
The research leading to these results has received funding from the European Research Council under IST Programme (P. CEZANNE, IST-2-IP-031867), the European Community's Seventh Framework Programme (FP7/2007-2013) / ERC grant agreement n° 209243 and the Spanish Research Council (CSIC, PIF08-018-3).
References and links
2. M. B. Christiansen, J. M. Lopacinska, M. H. Jakobsen, N. A. Mortensen, M. Dufva, and A. Kristensen, “Polymer photonic crystal dye lasers as optofluidic cell sensors,” Opt. Express 17(4), 2722–2730 (2009). [CrossRef] [PubMed]
3. F. Prieto, B. Sepúlveda, A. Calle, A. Llobera, C. Domínguez, A. Abad, A. Montoya, and L. M. Lechuga, “An integrated optical interferometric nanodevice based on silicon technology for biosensor applications,” Nanotechnology 14(8), 907–912 (2003). [CrossRef]
4. H. Joensson, C. Zhang, M. Uhlén, and H. Andersson Svahn, “A homogeneous assay for biomolecule interaction analysis in droplets by fluorescence polarization” 14th International Conference on Miniaturized Systems for Chemistry and Life Sciences, Groningen, The Netherlands 1802–1804 (2010)
5. J. Hübner, K. B. Mogensen, A. M. Jorgensen, P. Friis, P. Telleman, and J. P. Kutter, “Integrated optical measurement system for fluorescence spectroscopy in microfluidic channels,” Rev. Sci. Instrum. 72(1), 229–233 (2001). [CrossRef]
6. F. Z. Lin, Y. J. Chiu, S. A. Tsai, and T. H. Wu, “Laterally tapered undercut active waveguide fabricated by simple wet etching method for vertical waveguide directional coupler,” Opt. Express 16(11), 7588–7594 (2008). [CrossRef] [PubMed]
7. A. Llobera, R. Wilke, and S. Büttgenbach, “Poly(dimethyl siloxane) hollow Abbe prism with microlenses for detection based on absorption and refractive index shift,” Lab Chip 4(1), 24–27 (2004). [CrossRef] [PubMed]
8. N. M. Jokerst, M. A. Brooke, S. Cho, M. Thomas, J. Lillie, D. Kim, S. Ralph, K. Dennis, B. Comeau, and C. Henderson, “Integrated planar lightwave bio/chem OEIC sensors on si CMOS circuits,” Proc. SPIE 5730, 226–233 (2005). [CrossRef]
9. S. Balslev, A. M. Jorgensen, B. Bilenberg, K. B. Mogensen, D. Snakenborg, O. Geschke, J. P. Kutter, and A. Kristensen, “Lab-on-a-chip with integrated optical transducers,” Lab Chip 6(2), 213–217 (2006). [CrossRef] [PubMed]
10. O. Hofmann, X. Wang, A. Cornwell, S. Beecher, A. Raja, D. D. Bradley, A. J. Demello, and J. C. Demello, “Monolithically integrated dye-doped PDMS long-pass filters for disposable on-chip fluorescence detection,” Lab Chip 6(8), 981–987 (2006). [CrossRef] [PubMed]
11. A. Llobera, S. Demming, H. N. Joensson, J. Vila-Planas, H. Andersson-Svahn, and S. Büttgenbach, “Monolithic PDMS passband filters for fluorescence detection,” Lab Chip 10(15), 1987–1992 (2010). [CrossRef] [PubMed]
13. T. Baba and Y. Kokubun, “Dispersion and radiation loss characteristics of antiresonant reflecting optical waveguides-numerical results and analytical expressions,” J. Quant. Elect. 28(7), 1689–1700 (1992). [CrossRef]
14. A. Llobera, Í. Salinas, I. Garcés, R. Alonso, and C. Domínguez, “Large-core single-mode waveguides with cross-sectional antiresonant confinement,” IEEE J. Lightwave Tech. 22(6), 1560–1565 (2004). [CrossRef]
15. N. J. Goddard, J. Hulme, C. Malins, K. Singh, and P. R. Fielden, “Asymmetric anti-resonant reflecting optical waveguides (arrow) as chemical sensors,” Analyst (Lond.) 127(3), 378–382 (2002). [CrossRef]
16. S. Wong, M. Deubel, F. Perez-Willard, S. John, G. A. Ozin, M. Wegener, and G. von Freymann, “Direct laser writing of three-dimensional photonic crystals with complete a photonic bandgap in chalcogenide glasses,” Adv. Mater. 18(3), 265–269 (2006). [CrossRef]
17. T. Woggon, T. Kleiner, M. Punke, and U. Lemmer, “Nanostructuring of organic-inorganic hybrid materials for distributed feedback laser resonators by two-photon polymerization,” Opt. Express 17(4), 2500–2507 (2009). [CrossRef] [PubMed]
18. S. J. Lee, M. Goedert, M. T. Matyska, E. M. Ghandehari, M. Vijay, and J. J. Pesek, “Polymethylhydrosiloxane (PMHS) as a functional material for microfluidic chips,” J. Micromech. Microeng. 18, 025026 (9pp) (2008).
19. Y. Xia and G. M. Whitesides, “Soft lithography,” Annu. Rev. Mater. Sci. 28(1), 153–184 (1998). [CrossRef]
20. R. Houbertz, “Laser interaction in sol–gel based materials 3-D lithography for photonic applications,” Appl. Surf. Sci. 247(1-4), 504–512 (2005). [CrossRef]
21. E. Z. Faraggi, Y. Sorek, O. Levi, Y. Avny, D. Davidov, R. Neumann, and R. Reisfeld, “New Conjugated Polymer Sol-Gel Glass Composites: Luminescence and Optical Waveguides,” Adv. Mater. 8(10), 833–837 (1996). [CrossRef]
22. C. Ye, L. Shi, J. Wang, D. Lo, and X.-L. Zhu, “Simultaneous generation of multiple pairs of transverse electric and transverse magnetic output modes from titania zirconia organically modified silicate distributed feedback waveguide lasers,” Appl. Phys. Lett. 83(20), 4101–4104 (2003). [CrossRef]
23. X. Orignac, D. Barbier, X. M. Du, R. M. Almeida, O. McCarthy, and E. Yeatman, “Sol-gel silica/titania-on-silicon Er/Yb-doped waveguides for optical amplification at 1.5 µm,” Opt. Mater. 12(1), 1–18 (1999). [CrossRef]
24. A. Llobera, V. J. Cadarso, M. Darder, C. Domínguez, and C. Fernández-Sánchez, “Full-field photonic biosensors based on tunable bio-doped sol-gel glasses,” Lab Chip 8(7), 1185–1190 (2008). [CrossRef] [PubMed]
26. C. Sanchez, B. Lebeau, F. Chaput, and J. P. Boilot, “Optical properties of functional hybrid organic-inorganic nanocomposites,” Adv. Mater. 12(23), 1969–1993 (2003). [CrossRef]
27. S. Maruo and J. T. Fourkas, “Recent progress in multiphoton fabrication,” Laser Photon. Rev. 2(1-2), 100–111 (2008). [CrossRef]
28. C. Fernández-Sánchez, V. J. Cadarso, M. Darder, C. Domínguez, and A. Llobera, “Patterning of high aspect-ratio sol-gel structures by micro transfer molding,” Chem. Mater. 20(8), 2662–2668 (2008). [CrossRef]
29. A. Llobera, I. Salinas, I. Garcés, A. Merlos, and C. Domínguez, “Effect of wall tilt on the optical properties of integrated directional couplers,” Opt. Lett. 27(8), 601–603 (2002). [CrossRef]
30. A. Llobera, V. Seidemann, J. A. Plaza, V. J. Cadarso, and S. Büttgenbach, “SU-8 optical accelerometers,” J. MicroElectro. Mechan. Syst. 16(1), 111–121 (2007). [CrossRef]
32. E. Kim, Y. Xia, and G. M. Whitesides, “Micromolding in capillaries: applications in materials science,” JACS 118(24), 5722–5731 (1996). [CrossRef]
33. M. Fikry, M.M. Omar, and L.Z. Ismail, “Effect of host medium on the fluorescence emission intensity of rhodamine b in liquid and solid phase,” J. Fluoresc. 19(7), 41–746 (2009). [CrossRef]
34. A. Entwistle and M. Noble, “The use of Lucifer yellow, bodipy, FITC, TRITC, RITC and Texas Red for dual immunolabeling visualized with a confocal scanning laser microscope,” J. Microsc. 168, 219–238 (1992). [CrossRef]
35. O. Valdes-Aguilera and D. C. Neckers, “Aggregation phenomena in xanthene dyes,” Acc. Chem. Res. 22(5), 171–177 (1989). [CrossRef]
36. J. Jasieniak, J. Pacifico, R. Signorini, A. Chiasera, M. Ferrari, A. Martucci, and P. Mulvaney, “Luminescence and amplified stimulated emission in CdSe–ZnS-nanocrystal-doped TiO2 and ZrO2 waveguides,” Adv. Funct. Mater. 17(10), 1654–1662 (2007). [CrossRef]