Open Access
Add to BibSonomyAbstract
A microfabricated directional coupler (DC) was used for the detection of DNA conjugated with quantum dots. Output optical signals from DCs of a wide range of device lengths correspond well to theoretical and simulation results. Even 20 µm-long DC devices could detect changes in the output optical intensity by monitoring the near-field pattern using a CCD camera. The signal was enhanced 60 × using a 1500 µm-long DC device. For large cladding refractive-index changes between air and water, the normalized signal changed cyclically several times between 0 and 1. The results suggest that the DC can be the basis for miniaturized two-dimensionally integrated biochemical sensors.
© 2015 Optical Society of America
1. Introduction
Over the last several decades, a variety of micro-devices capable of sensing biochemical species have been developed based on several photonic technologies in the research fields of medical diagnostics, human health, environmental monitoring, and food safety [1, 2]. Since the late 1990s, photonic biochemical sensors based on fiber-optics, opto-fluidics, interferometry, and plasmonics have attracted much attention because of their specific features such as high sensitivity, label-free and real-time monitoring ability, compactness, and capability of easy planar integration. Among these attractive features, the planar integration ability inspired many photonic engineers to develop miniaturized and integrated multichannel waveguides combined with microfluidics. As a result, multi-functionalized smart biochemical sensors that can recognize biochemical species with high speed and high selectivity is particularly important [3]. The development of photonic biosensors began with interferometric waveguides such as the directional coupler (DC) [4], Mach-Zehnder interferometer (MZI) [5, 6], Young’s interferometer [3], bimodal waveguide interferometer [7, 8], grating coupler [9], and straight waveguide-coupled ring resonator [10, 11]. A slot waveguide was reported as a sensing platform, which has made these sensors more highly qualified, for example, in detection sensitivity [12]. More recently, a slot waveguide DC was reported as a biochemical sensor [13].
Currently, there is a demand for single cell analysis systems to complement conventional bulk cell analysis methods. This is based on the concern that conventional bulk cell analysis uses samples with large populations of cells, and as a result, only measurements of mean signals in time as well as in space are recorded. Since living cells have their own unique functions and characteristics, their mean signals make it more difficult to detect or identify independent cell features and abnormalities. This heterogeneity between single cells has been reported to be an important indicator of the onset of cancer and other diseases [14]. In this respect, devices that enable the detection of molecules secreted from a single cell in a small volume and small interaction area are critical for this new type of disease diagnostic. In addition, the measurement of single cell expression from a large population of cells requires a biosensor configuration capable of measuring responses of the individual cells from a large sub-population of the cell culture. A droplet-based miniaturized microfluidic channel (MC) [15, 16] will be one of the promising approaches to challenge such a requirement. In contrast, an interferometric biochemical sensor reported to date was too large (typically 1 to 10 mm in size) to conduct single cell analysis in a small interaction region with the droplet-based MC.
The DC mentioned above has a simple structure consisting of two closely coupled parallel waveguides that can be miniaturized to less than several tens of µm. In the DC, incident light propagating in the waveguide is coupled to the adjacent waveguide through an evanescent field in the gap between the two waveguides. The energy in one waveguide is completely transferred to the other within a distance that is called the complete coupling length Lc. Since the Lc depends on the cladding refractive-index nclad of the biochemical species to be analyzed, an optical signal at the output port changes depending on the biochemical species binding to the surface of the waveguide. The first DC waveguide-based biosensor was reported in 1996 [4]. In [4], Luff et al. adopted an asymmetric DC structure to enhance the sensitivity to a particular material, thus fabricated low index-contrast waveguides by Ag+-Na+ ion exchange in glass. Due to the low index-contrast system, the sensing area is designed to be as large as 10 mm in length, 150 µm in width with a 7 µm separation gap between two parallel waveguides.
In order to realize the DC-based biochemical sensor for the ultimate goal of single cell analysis, several issues have to be overcome. The first is to design sensitive DC devices with micron scale dimensions. The second is to accurately measure small changes of an optical output intensity caused by the adsorption of target biochemical to the sensor. For this purpose, the observation of the near-field spot intensity at the endpoint of the output waveguide using a CCD camera is a useful technique.
This paper describes experimental and simulation results of the fundamental relationship between a wide range of the DC lengths and their optical responses to refractive index changes, as well as experimental results of DNA conjugated with quantum dots (QD) bound to the sensing surface. Particularly important are the experimental results that include optical responses from seventeen different DC samples with lengths ranging from 28 µm to 1540 µ m, the estimation of the average refractive index of the QD-DNA layer from the output response, and the observation of the cyclically time-dependent optical response of the refractive-index difference of a moving droplet.
2. Experimental
2.1 Fabrication process
Figure 1(a) shows the schematic diagram of the DC-based biochemical sensor formed on a 2 µm-thick silicon dioxide (SiO2) layer deposited on a silicon substrate. Biochemical sensing is performed on the surface of the DC structure. In this study, sample solutions were directly made in contact with the DC. However, as shown in this figure, a sensing window and a MC (microfluidic channel) will be formed on the DC waveguide to delineate the contact area in our next step. Figure 1(b) shows the cross section of the DC waveguide formed with a negative electron-beam (EB) polymer resist SU-8 (MicroChem, Newton, MA, USA), patterned with EB lithography on the SiO2 layer. A 700 nm-thick core SU-8 layer (nc = 1.59 at 633 nm wavelength) is spin-coated on the cladding SiO2 layer (nclad = 1.46 at 633 nm), followed by post-baking at 95 °C for 60 sec after the EB lithography. The waveguide width and gap distance are designed to have dimensions of 600 nm and 300 nm, respectively. The top view of the DC and additional input and output waveguides is schematically shown in Fig. 1(c). Here, input light is incident into one of the parallel waveguides separated by 100 µm and output light intensity is monitored at both ends of the output parallel waveguides using a CCD camera. For experiments to check the response to different liquid droplets, DC waveguides of different lengths in the range of 20-1500 µm are used to examine a detailed relationship between the output optical intensity and waveguide length of the DC structure with nclad as a parameter.
Fig. 1 (a) Schematic image of the DC waveguide biosensor; (b) and (c) Cross section and top view of the device, respectively.
Figures 2(a) and 2(b) show scanning electron beam microscope (SEM) images of a cleaved cross section and the top view of the DC structure, respectively. The SU-8 DC waveguide is fabricated with precision within 5%. For the characterization of the device, 10 mm × 10 mm chips with DC waveguides of 20 - 30 different geometries and dimensions were cut out from a wafer using a dicing machine and mounted on an optical micro-manipulation stage.
2.2 Optical measurement
Figure 3(a) shows the set-up used for the optical measurement of the DC sensor. A semiconductor laser light (λ = 635 nm, 5 mW) is injected into one of the parallel waveguides through a long focus objective (50 × , 4 mm focal length). After propagating along the coupled parallel waveguides, optical beams are observed at two output ports using a CCD camera through another objective (20 × , 10 mm focal length). Figure 3(b) shows two near-field spots separated by 100 µm taken at the output cleaved edge. For minimizing irregular stray light, the output optical intensities s1 and s2 were recorded only within the 10 µm-diameter dashed circles shown in Fig. 3(b). In some of the biosensing experiments using interferometric waveguides [3, 7], output optical beams were detected and analyzed using far-field patterns. Compared with this, collection of light within a certain spot size of the near-field pattern realizes highly accurate measurements of the output signal. In the figures shown later, the following relative optical intensity Ii (i = 1, 2) is used instead of the measured signals s1 and s2 for minimizing the effect of data fluctuation in the measurement.
Fig. 3 (a) Schematic images of measurement instruments; (b) Near-field image of the output signals s1 and s2.
As a first step to characterize the DC sensor, a liquid droplet is placed on the sensing area using a micro-pipet. Figure 4(a) shows a position of the liquid droplet that covers several DC waveguides at a time, while Fig. 4(b) shows a cross section of the droplet that wets the entire SU-8 parallel waveguide surface. The droplet (dotted outline) shwon in Fig. 4(a) covers three to six sets of DC waveguides depending on the sample area.
3. Results and discussion
3.1 Theoretical behaviour of the DC chemical sensor
Prior to the experiment, the relationship between si and the position x [Fig. 4(a)] in the DC waveguide was simulated along with its nclad dependence using a commercially available finite-difference time-domain (FDTD) software package (CST Studio). The SU-8 DC waveguide geometry used for the simulation is the same as shown in Figs. 1(b) and 1(c). In the calculation, the light source was placed at the edge of one of the DC waveguides. Figure 5(a) shows calculated I1 and I2 in air (nclad = 1.0), where I1 and I2 show relative optical intensities in the incident and counter waveguides, respectively. Although a little fluctuation is seen near x = 0, the simulation curve of I2 is close to the optical energy distribution function P(x) derived from coupled mode theory for the DC waveguide, that is,
where Lc is the complete coupling length specific to the DC waveguide, as mentioned in Section 1, which is 150 µm in the case of Fig. 5(a). Here, P(x) corresponds to the relative optical intensity I2.Fig. 5 (a) Simulation results of I1 and I2 versus position x along the DC waveguide; (b) Schematic cross section used for the calculation of I2 with nclad; (c) Calculated I2; (d) Changes in I2 at x = 80 µm with respect to Δ nclad; (e) Normalized sensitivity Δ I2 / Δ nclad as a function of x.
When the cladding layer around the DC waveguide was changed from air (nclad = 1.0) to water (nclad = 1.33), the Lc obtained by simulation decreased from 150 µm to 60 µm as a result of enhanced coupling due to the increase of nclad. Furthermore, nclad was changed from 1.33 to 1.37 RIU (refractive index unit), as shown in Fig. 5(b). The corresponding change of the I2 along the waveguide was simulated, as shown in Fig. 5(c), where the Lc is further reduced from 60 µm to 40 µm. At x = 80 µm, I2 varies from 0.78 to 0.24 in optical intensity unit (hereafter referred to simply as OIU). These values are plotted in Fig. 5(d) as a calibration relation between I2 and nclad, where the dotted line indicates the least-squares linear fit. As a result, I2 decreases almost linearly with nclad in this example. Here, the slope of the I2-nclad curve is the sensitivity of the DC sensor and is estimated to be −13.5 OIU using the sensitivity definition [I2 (nclad = 1.37) - I2 (nclad = 1.33)] / (Δnclad = 0.04) from Ref [13]. Furthermore, the sensitivity of our DC is calculated as a function of x, as shown in Fig. 5(e), where the dotted-line curve in the region of x greater than 105 µm has been extrapolated and will be discussed in the next section. In Fig. 5(e), the arrow indicates the sensitivity of the 80 µm-long DC. Additionally, it should be noted that the absolute maximum sensitivities appearing at the x-points (35 µm, 80 µm and 135 µm) increase with x, that is, the DC length. Thus, as will be shown later the sensitivity can be increased by increasing the length of the DC.
3.2 Characterization of the DC sensor using liquid droplet
As a preliminary experiment, the response of the device to water and ethanol was examined. Prior to the experiment, the influence of random fluctuations of I2 due to unexpected contamination in the droplet was examined by alternately placing the droplet and removing it several times (Fig. 6). A 140 µm-long DC waveguide was used. First, a water droplet (3 µL) was placed on the DC waveguide by a micro-pipet and I2 was measured. After the measurement, the water droplet was removed by blowing with dry nitrogen gas. The procedure was repeated three times. The variation of I2 for the water droplet was found to be small around 0.45 OIU. Next, the droplet was changed to ethanol and the same procedure was repeated twice. The signal I2 changed to 0.98 OIU with an ethanol droplet and the variation between measurements is negligibly small. Consequently, the liquid droplets used in these experiments do not contain a significant amount of contamination, which can degrade the reproducibility of the measurements. It is interesting to note that the sensitivity of the DC is approximately 17.7 OIU when the liquid is changed from water to ethanol, which is close to the calculated value of 17.6 OIU that is indicated by the arrow at x = 140 µm in Fig. 5 (e).
Fig. 6 Changes in I2 when water and ethanol droplets were placed on the 140 µm-long DC and dried alternately.
Based on these preliminary experiments, Ii was measured in air, water, and ethanol using fourteen different DC waveguides with lengths ranging from 20 µm to 300 µm. The geometry of the waveguide was the same as that shown in Fig. 1(b). Figure 7 shows the relation between the measured I2 and the DC length obtained with air (nclad = 1.0) (black circles), water (nclad = 1.33) (blue squares), and ethanol (nclad = 1.36) (red triangles) droplets, respectively. In each figure, the sinusoidal curve is obtained by least-square fitting using the distribution function P(x) shown in Eq. (2). As shown in the graphs, most of the experimental values fit well to the theoretical curves, although some small deviations exist possibly due to unexpected small variations in the DC structure sizes. In these figures, arrows indicate Lc. The values of Lc obtained from the fitted curves were 131 µm, 58 µm and 48 µm, respectively, These values are close to those obtained by FDTD simulation, which are 140 µm, 56 µm and 51 µm, respectively.
Fig. 7 Dependence of I2 on DC length in (a) Air; (b) Water droplet; and (c) Ethanol droplet. The curves were drawn by least square fitting assuming sinusoidal curves based on Eq. (2).
3.3 Detection of the hybridization of DNA using the DC sensor
Single-stranded DNA (base sequence: 5′-GTGTGTAGTCACCCCAACCT-3′) conjugated with QDs (Qdot® 655 Streptavidin Conjugate, Life Technologies, Carlsbad, CA, USA) is used as a target. Hereafter, it is referred to as QD-DNA complex. Figures 8(a)-8(c) schematically show the hybridization of the QD-DNA complex to complementary DNA probes on the DC surface. Here, QD is used to increase the mass of target DNA and increase nclad [17]. The initial surface of the SU-8 DC waveguide [Fig. 8(a)] was covered with a 3 µL drop volume of HCl aqueous solution (100 mM) and single-stranded DNA (1 µM) with amine-termination. The amino group of the single-stranded DNA and the epoxy group of the SU-8 surface react by a proton catalyst in the HCl solution. An hour later, the drop was washed away with deionized water for 30 sec and the surface was dried using dry nitrogen gas [Fig. 8(b)]. Then, a drop of 100 mM NaCl aqueous solution (3 µL) with target DNA (10 nM) was placed on the DC surface and the target DNA was hybridized with the complementary single-strand DNA in the DC surface [Fig. 8(c)]. An hour later, I2 was measured similarly to the case of Fig. 7. For this purpose, seventeen DC samples of different lengths were prepared; eight samples with 28 to 224 µm lengths, five samples with 560 to 672 µm lengths, and four samples with 1456 to 1540 µm lengths. Lc in water was 56 μm as in the case of the experiment in Fig. 7.
Fig. 8 Scheme of DC surface modification for QD-DNA complex detection. (a) Unmodified DC surface; (b) DC surface modified with complementary single-stranded DNA probes; (c) DC surface following hybridization with QD-DNA complex.
A noticable change occurred when the target DNA was bound to the surface [Fig. 9(b)]. Figure 9(c) shows I2 recorded with a wide range DC lengths. In the figure, sinusoidal curves in blue and red are fitted curves before and after hybridization of the target DNA, respectively. The curves were generated by least-square fitting. Here, the values of I2 obtained with DC lengths ranging from 0 to 250 µm, from 500 to 700 µm, and from 1400 to 1600 µm are represented in Figs. 9(c), 9(d) and 9(e), respectively, for clear visualization.
Fig. 9 Detection of target DNA with the DC sensor in NaCl aqueous solution. (a) Schematic image of the DC surface modified with single-stranded DNA probes; (b) DC surface modified with hybridized target DNA; (c) Change in I2 when the DC length was changed; (d), (e) and (f) Enlarged parts of (c) ranging from 0 to 250 µm, from 500 to 700 µm, and from 1400 to 1600 µm, respectively, extracted from Fig. 9(c).
Several important results can be concluded from Figs. 9(c) to 9(f). First, the coupling length Lc = 58.6 µm derived from the fitted curve before the hybridization of the target DNA, decreased to 55.8 µm after binding due to the increase of the effective nclad. This is also recognized from the result that the fitted curve in red reduced compared to that in blue in the direction of the abscissa. Second, most of the measured data over the entire range of the DC length fit quite well to the blue and red fitted curves. Taking a careful look at Figs. 9(e) and 9(f), this result is confirmed more clearly: Measured plots in I2 at 615, 640, 675, 1455, 1480 and 1515 µm DC-lengths are significantly close to the corresponding fitted curves both before and after binding of the target DNA. At x = 1540 µm, shown in Figs. 9(c) and 9(f), a measured point (shown in blue) is missing because the measurement of I2 before the hybridization of the target DNA was skipped. Nevertheless, it still provides a useful result since the response after the hybridization of the DNA fits well to the theoretical curve. Such a good aggreement in I2 was observed even with the short DC sample, that is, at six different DC lenghs (28, 55, 80 115, 140 and 195 µm) in Fig. 9(c). Third, from measured and simulated data, we can estimate the effective nclad of the NaCl aqueous solution that involves the target DNA. By comparing the measured I2 at Lc = 80 µm in Fig. 9(d) with simulated I2 at the same DC length in Fig. 5(d), we can estimate the effective nclad to be approximately 1.335 RIU. Considering that nclad of the NaCl solution is 1.33, the shift of nclad, Δnclad, with the addition of the target DNA is estimated to be approximately 0.005 RIU. This is a new finding obtained through this experiments. Lastly, the effect of the DC length on the sensitivity of I2 is considered in the same way as shown in Fig. 5 (e). For this purpose, it is assumed that the refractive index of the NaCl solution is nclad = 1.330 RIU, whereas the refractive index of the NaCl solution with target DNA is 1.335 RIU, thus Δnclad ≈0.005 RIU. Accordingly, the sensitivity of the 28 µm-long DC is approximately 6.0 OIU from ΔI2 = 0.03 OIU between 0.71 OIU (blue point) and 0.74 OIU (red point). On the other hand, the sensitivity of the 1512 µm-long DC amounts to 124 OIU from ΔI2 = 0.62 OIU between 0.32 OIU (blue point) and 0.94 OIU (red point). Thus, it is noted that the sensitivity of the DC is improved by a factor of 21 × when increasing the DC length from 28 µm to 1512 µm.
3.4 Time dependent output signal of the DC sensor
Experimental results mentioned so far were the cases for rather small changes, ΔI2 ~0.01 - 0.61 OIU, versus small refractive index chages, Δnclad ~0.01 - 0.03 RIU. This section describes another case, that is, the case for larger signal changes, ΔI2 > 1.0 OIU, caused by larger Δnclad ~0.3 RIU in such different cases as in air and solution.
As shown in Fig. 9(f), the large ΔI2 of 0.62 OIU was obtained after the binding of the target DNA using the 1456 µm-long DC waveguide. For larger values of Δnclad, I2 varies sinusoidally several times between 0 and 1. Accordingly, the two spots in the near field pattern, as shown in Fig. 3(b), alternate bright and dark states several times. The experiment was carried out using the following procedure: As shown in Fig. 10(a), a 100 mM NaCl solution (nclad = 1.3) was dropped at the drop point [see Fig. 10(a)] of the 1456 µm-long DC waveguide using a micro-pipet, where the liquid front moved along the DC in about twelve seconds. During this time, the NaCl solution droplet expanded gradually on the DC waveguide. Figure 10(b) shows changes in I2 at four different times (t1 - t4), where Lc corresponding to the liquid is compressed in the horizontal direction as the droplet spreads.
Fig. 10 Schematic pictures showing (a) Time-dependent coverage of NaCl droplet for measuring time-dependent I2 ; and (b) Principle of the signal processing in the I2 vs time.
Figure 11(a) shows the anticipated change in Δnclad and Fig. 11(b) shows the corresponding anticipated change in I2 obtained by the simulation. As nclad increases gradually, I2 varies sinusoidally several times between 0 and 1. In the figure, I2 changes rapidly at the beginning then slowly as Δnclad increases gradually. Figure 11(c) shows the integration of I2, which changes continuously between 0 and 1 as time elapses. It increases monotonously from zero to the saturated maximum value, ΔSmax, which corresponds to the maximum Δnclad. In this case, the 3.5-cycle change in I2 between 0 and 1 corresponds to ΔImax of 7.0 OIU.
Fig. 11 (a), (b), and (c): Schematic pictures showing (a) time-dependence of Δnclad; (b) Time-dependence of s2; and (c) Integration of s2; (d), (e) and (f): Corresponding experimental results showing (d) Frame captures from a real-time movie of I1 and I2 visualized as near-field-patterns; (e) Measured time-dependent I2; and (f) Integrated I2. The DC surface coverage is gradually changed from air to buffer solution in about 12 sec and the 7.5-cycle signal change is detected from the movie.
Figures 11(d)-11(f) show an experimental demonstration of the time-variant Δnclad and I2 mentioned above. Figure 11(d) is a frame capture from a real time movie that shows the change in the near field intensity of the I1 and I2 recorded using the CCD camera. The time resolution of the CCD camera was 60 frames per second. It is confirmed that I1 and I2 images turned bright and dark 7.5 times in twelve seconds. Figures 11(e) and 11(f) show the time course of I2 and the integration of I2 corresponding to Figs. 11(b) and 11(c), respectively. In Fig. 11(e), the shape of I2 is significantly deformed from the sinusoidal curve due to the nonuniform movement of the front of the droplet on the DC surface. Nevertheless, it is confirmed that the change of I2 amounts to about 7.5 cycles between 0 and 1, accordingly the resultant ΔImax is 15.0 OIU. In [7], K. E. Zinoviev reported similar real time monitoring by observation of far-field-patterns using a bimodal waveguide interferometer. In contrast, the real time measurement of the intensity change ΔI2 used here is considered to be accurate and rapid as the sensor length is 1/100 - 1/1000 to that of conventional waveguide-based sensors and the surface-to-volume ratio is much larger. This is very advantageous in sensing rapid chemical reactions.
4. Conclusion
Differentiation between water and ethanol droplets and detection of target DNA were performed using a miniaturized DC waveguide and a time-resolvable CCD monitoring technique. The output optical intensity ratio Ii changed accompanying the change of nclad due to the coverage of the liquid droplet on the DC surface. The Ii was strongly dependent on the DC length. Measured Ii versus DC length correspond well to the sinusoidal fitted curves derived from a large amount of data and the simulated curves calculated from the DC structure used in the experiments. Experiments with water and ethanol droplets verified the change of the complete coupling length Lc specific to the DC structure. Another experiment on the detection of a target DNA clearly confirmed the change of Ii (ΔIi = 0.03 OIU) after the binding of the target DNA even using a 28 µm-long DC waveguide. Such a small change ΔIi was enhanced to 0.61 OIU using a 1512 µm-long DC waveguide. As one of the most characteristic experimental results, the effective index difference Δnclad due to the binding of the target DNA on the DC surface was estimated to be 0.005 RIU. Another distinctive result for the case of large Δnclad was that, when the nclad changed by 0.3 RIU, Ii changed cyclically 7.5 times between 0 and 1 and the integration of the Ii signal amounted to as large as 15.0 OIU. All measured Ii data could be obtained with high accuracy from the output near-field pattern monitored using a time-resolvable CCD camera. Hence, the DC waveguide proposed here is promising for a miniaturized biochemical sensor with high accuracy and suitability for multi-channel integration.
Acknowledgments
This work was supported by a Grant-in-Aid for Scientific Research (No. 25286034) of Japan Society for the Promotion of Science. The authors would like to express their sincere thanks to Dr. Yoshimasa Sugimoto and Dr. Naoki Ikeda for their useful suggestion in fabrication process, Prof. Nobuhiko Ozaki for his instructive advice in optical characterization. The devices were fabricated with support by the University of Tsukuba Nanotechnology Platform.
References and links
1. M. C. Estevez, M. Alvarez, and L. M. Lechuga, “Integrated optical devices for lab-on-a-chip biosensing applications,” Laser Photonics Rev. 6(4), 463–487 (2012). [CrossRef]
2. V. M. N. Passaro, F. Dell’Olio, B. Casamassima, and F. De Leonardis, “Guided-wave optical biosensors,” Sensors (Basel Switzerland) 7(4), 508–536 (2007). [CrossRef]
3. A. Ymeti, J. S. Kanger, R. Wijn, P. V. Lambeck, and J. Greve, “Development of a multichannel integrated interferometer immunosensor,” Sens. Actuators, B. 83(1–3), 1–7 (2002). [CrossRef]
4. B. J. Luff, R. D. Harris, J. S. Wilkinson, R. Wilson, and D. J. Schiffrin, “Integrated-optical directional coupler biosensor,” Opt. Lett. 21(8), 618–620 (1996). [CrossRef] [PubMed]
5. R. G. Heideman and P. V. Lambeck, “Remote opto-chemical sensing with extreme sensitivity: design, fabrication and performance of a pigtailed integrated optical phase-modulated Mach-Zehnder interferometer system,” Sens. Actuators, B. 61(1–3), 100–127 (1999). [CrossRef]
6. F. Prieto, B. Sepúlveda, A. Calle, A. Llobera, C. Domynguez, 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]
7. K. E. Zinoviev, A. B. González-Guerrero, C. Dominguez, and L. M. Lechuga, “Integrated bimodal waveguide interferometric biosensor for label-free analysis,” J. Lightwave Technol. 29(13), 1926–1930 (2011). [CrossRef]
8. D. Duval, A. B. González-Guerrero, S. Dante, J. Osmond, R. Monge, L. J. Fernández, K. E. Zinoviev, C. Domínguez, and L. M. Lechuga, “Nanophotonic lab-on-a-chip platforms including novel bimodal interferometers, microfluidics and grating couplers,” Lab Chip 12(11), 1987–1994 (2012). [CrossRef] [PubMed]
9. V. M. N. Passaro, R. Loiacono, G. D’Amico, and F. De Leonardis, “Design of Bragg grating sensors based on sub-micrometer optical rib waveguide in SOI,” IEEE Sens. J. 8(9), 1603–1611 (2008). [CrossRef]
10. C.-Y. Chao, W. Fung, and L. J. Guo, “Polymer microring resonators for biochemical sensing applications,” IEEE J. Sel. Top. Quantum Electron. 12(1), 134–142 (2006). [CrossRef]
11. V. M. N. Passaro, F. Dell’Olio, and F. De Leonardis, “Ammonia optical sensing by micro-ring resonators,” Sensors (Basel Switzerland) 7(11), 2741–2749 (2007). [CrossRef]
12. V. R. Almeida, Q. Xu, C. A. Barrios, and M. Lipson, “Guiding and confining light in void nanostructure,” Opt. Lett. 29(11), 1209–1211 (2004). [CrossRef] [PubMed]
13. V. M. N. Passaro, F. Dell’olio, C. Ciminelli, and M. N. Armenise, “Efficient chemical sensing by coupled slot SOI waveguides,” Sensors (Basel) 9(2), 1012–1032 (2009). [CrossRef] [PubMed]
14. C. E. Sims and N. L. Allbritton, “Analysis of single mammalian cells on-chip,” Lab Chip 7(4), 423–440 (2007). [CrossRef] [PubMed]
15. F. Sassa, J. Fukuda, and H. Suzuki, “Microprocessing of liquid plugs for bio/chemical analyses,” Anal. Chem. 80(16), 6206–6213 (2008). [CrossRef] [PubMed]
16. F. Sassa, H. Laghzali, J. Fukuda, and H. Suzuki, “Coulometric detection of components in liquid plugs by microfabricated flow channel and electrode structures,” Anal. Chem. 82(20), 8725–8732 (2010). [CrossRef] [PubMed]
17. N. C. Tansil and Z. Gao, “Nanoparticles in bio-molecular detection,” Nano Today 1(1), 28–37 (2006). [CrossRef]
