**5. DNA sensing using photonic crystal waveguides**

For label-free DNA sensing, the experimental protocol is similar to the one previously described for antibody sensing, but now the goal of the experiment is to detect DNA hybridization events occurring on the sensor surface, so single DNA strands (ssDNA) need to be immobilized on the PCW. To do so, the chip's surface is still activated with pure ICPTS as described for the antiBSA sensing in the previous section, then an intermediate layer of streptavidin is deposited on the chip (a concentration of 0.1 mg/ml in 0.1x PBS is used), and finally biotinylated ssDNA probes 10 μM in PBS 0.1x are incubated on the sample, which will bind to the streptavidin layer thanks to the high affinity between the streptavidin and the biotin molecules (Toccafondo et al., 2010). Fig. 13 shows the TE transmission spectrum of the PCW with a PBS 0.1x upper cladding, where FP fringes are shown. In this case, the peak marked with dashed red line has been used for sensing. Fig. 14 shows the steps to be carried out in the experiment and Fig. 15 shows the temporal evolution of the peak position for the different solutions flowed. First of all PBS 0.1x is flowed to obtain the baseline of the measurement. Then a solution containing the complementary ssDNA to that immobilized on the chip, with a concentration of 0.5 μM, is flowed. Binding of the complementary DNA is effectively shown, which induces a shift in the peak position of ΔλDNA = 47.1 pm. The noise level in this experiment is estimated to be σ = 1.865 pm, thus giving an estimated detection limit of 19.8nM for ssDNA hybridization detection.

The strand-end of the complementary ssDNA chosen for this experiment is marked with digoxigenin, in order to allow to perform a control step and confirm the hybridization events. Therefore, after the complementary ssDNA is detected, anti-DIG 10ppm, which has a high affinity with DIG, is flowed. Its binding to the DIG marker of the target ssDNA causes a permanent shift in the peak position of 0.246 nm, which confirms the specific binding of the target DIG-marked ssDNA on the chip.

Label-Free Biosensing Using Photonic Crystal Waveguides 251

structure's spectral response, as previously discussed in this chapter. Therefore, these systems require the use of either a tunable laser source or an optical spectrum analyzer (OSA) to perform the readout of the device, making the total cost of the system significant (above 20.000-30.000 Euros or even higher). Moreover, sweeping times of the order of several seconds up to minutes are needed to acquire each spectrum, preventing an

time

λ

instantaneous observation of the interactions of the target analyte with the sensor.

PBG edge

λ

Fig. 16. Output power evolution using the proposed sensing technique. In the initial state, the source spectrum (black solid line) is filtered by the PBG of the sensing structure (red dashed line) and only a certain amount of power is transmitted (shaded area). When a refractive index variation occurs, the PBG shifts, and the amount of input power filtered

An alternative technique for the development of real-time and low-cost integrated photonic sensing devices using photonic bandgap (PBG) structures is schematically described in Fig. 16. Using this technique, the PBG shift is indirectly determined by using a filtered broadband optical source as excitation instead of a tunable laser, and a power meter at the output instead of an eventual OSA, thus significantly simplifying the system and reducing its cost to values around 2.000-3.000 Euros (or even lower), making it competitive with other sensing systems based on other transduction mechanisms (e.g., electro-chemical based sensors). If the PBG edge is located within the source wavelength range (for the case shown in Fig. 16, the upper edge of the PBG is used for the sensing), the PBG filters the optical source and the overlap of both spectra can be directly measured at the output using a simple power meter. When a variation of the refractive index of the surrounding medium occurs, it induces a shift in the spectral response of the photonic sensing structure. This is translated into a shift in the position of the PBG edge, and thus into a change in the optical power measured at the output, as is illustrated in Fig. 16 where an increase in the RI of the surrounding medium occurs. This power variation is directly used to perform the sensing, without the need to obtain the transmission spectrum of the structure using

transmission

refractive index variation

output power

changes (decreases in this case).

transmission

excitation

Fig. 13. Spectrum of the photonic crystal waveguide in the region of the band edge with a PBS 0.1x upper cladding. The transmission peak used for sensing is marked with dashed red line in the inset, where a higher resolution sweep has been made.

Fig. 14. Scheme of the experiment steps for the ssDNA detection and anti-DIG control.

Fig. 15. (a) Wavelength shift vs time for single strand DNA 0.5 μM sensing, and (b) for anti-DIG 10 ppm sensing. Relative wavelength shift from the initial baseline is represented.

#### **6. Low-cost sensing technique using photonic crystal waveguides**

One of the problems of using PCWs for biosensing, as well as of other photonic structures like ring resonators, is that the detection is based on the measurement of the shift of the

Fig. 13. Spectrum of the photonic crystal waveguide in the region of the band edge with a PBS 0.1x upper cladding. The transmission peak used for sensing is marked with dashed red

> **Y Y**

0 20 40 60 80 100

antiDIG 10ppm

time (minutes)

**Y Y**

> PBS 0.1x

= 0.246nm

line in the inset, where a higher resolution sweep has been made.

**a) b) c)**

**Y**

DNA 0.5M

0 20 40 60 80

time (minutes)


0

20

PBS 0.1x

= 47.1pm

relative peak position (pm)

40

60

a)

Fig. 14. Scheme of the experiment steps for the ssDNA detection and anti-DIG control.

PBS 0.1x

b)

relative peak position (pm)

Fig. 15. (a) Wavelength shift vs time for single strand DNA 0.5 μM sensing, and (b) for anti-DIG 10 ppm sensing. Relative wavelength shift from the initial baseline is represented.

One of the problems of using PCWs for biosensing, as well as of other photonic structures like ring resonators, is that the detection is based on the measurement of the shift of the

**6. Low-cost sensing technique using photonic crystal waveguides** 

structure's spectral response, as previously discussed in this chapter. Therefore, these systems require the use of either a tunable laser source or an optical spectrum analyzer (OSA) to perform the readout of the device, making the total cost of the system significant (above 20.000-30.000 Euros or even higher). Moreover, sweeping times of the order of several seconds up to minutes are needed to acquire each spectrum, preventing an instantaneous observation of the interactions of the target analyte with the sensor.

Fig. 16. Output power evolution using the proposed sensing technique. In the initial state, the source spectrum (black solid line) is filtered by the PBG of the sensing structure (red dashed line) and only a certain amount of power is transmitted (shaded area). When a refractive index variation occurs, the PBG shifts, and the amount of input power filtered changes (decreases in this case).

An alternative technique for the development of real-time and low-cost integrated photonic sensing devices using photonic bandgap (PBG) structures is schematically described in Fig. 16. Using this technique, the PBG shift is indirectly determined by using a filtered broadband optical source as excitation instead of a tunable laser, and a power meter at the output instead of an eventual OSA, thus significantly simplifying the system and reducing its cost to values around 2.000-3.000 Euros (or even lower), making it competitive with other sensing systems based on other transduction mechanisms (e.g., electro-chemical based sensors). If the PBG edge is located within the source wavelength range (for the case shown in Fig. 16, the upper edge of the PBG is used for the sensing), the PBG filters the optical source and the overlap of both spectra can be directly measured at the output using a simple power meter. When a variation of the refractive index of the surrounding medium occurs, it induces a shift in the spectral response of the photonic sensing structure. This is translated into a shift in the position of the PBG edge, and thus into a change in the optical power measured at the output, as is illustrated in Fig. 16 where an increase in the RI of the surrounding medium occurs. This power variation is directly used to perform the sensing, without the need to obtain the transmission spectrum of the structure using

Label-Free Biosensing Using Photonic Crystal Waveguides 253

chapter, we have shown how PCWs can be used for biosensing purposes and how the use of Fabry-Perot fringes appearing at the guided band's edge instead of the edge itself allows performing the sensing with a higher accuracy and leading to a reduced detection limit. Experimental results have been shown for the detection of refractive index variations (with a detection limit of 3.5x10-6 RIU), label-free antibody sensing (with a surface mass density detection limit below 2.1 pg/mm2 and total mass detection limit below 0.2 fg), and label-free

A technique for the indirect tracking of the guided band's edge shift has also been presented. This technique avoids the use of expensive tunable sources or detectors, which are required when carrying out a direct tracking of the spectral response shift, as usually done when using photonic sensing devices such as ring resonators or photonic crystal based

Financial support from the Spanish MICINN under contract TEC2008-06333, the Universidad Politécnica de Valencia through program PAID-06-09, and the Conselleria

Barrios, C. A., Bañuls, M. J., González-Pedro, V., Gylfason, K. B., Sánchez, B., Griol, A.,

Buswell, S. C., Wright, V. A., Buriak, J. M., Van, V., & Evoy, S. (2008). Specific detection of

Carlborg, C. F., Gylfason, K. B., Kaźmierczak, A., Dortu, F., Bañuls Polo, M. J., Maquieira

Castelló, J.G., Toccafondo, V., Pérez-Millán, P., Losilla, N.S., Cruz, J.L., Andrés, M.V., &

Chow, E., Grot, A., Mirkarimi, L. W., Sigalas, M., & Girolami, G. (2004). Ultracompact

De Vos, K., Bartolozzi, I., Schacht, E., Bienstman, P., & Baets, R. (2007). Silicon-on-Insulator

*Letters*, Vol. 29, No. 10, (May 2004), pp. 1093-1095, ISSN 0146-9592

No. 12, (June 2007), pp. 7610-7615, ISSN 1094-4087

Maquieira, A., Sohlström, H., Holgado, M., & Casquel, R. (2008). Label-free optical biosensing with slot-waveguides. *Optics Letters*, Vol. 33, No. 7, (April 2008), pp. 708-

proteins using photonic crystal waveguides. *Optics Express*, Vol. 16, No. 20,

Catala, A., Kresbach, G. M., Sohlström, H., Moh, T., Vivien, L., Popplewell, J., Ronan, G., Barrios, C. A., Stemme, G., & van der Wijngaart, W. (2010). A packaged optical slot-waveguide ring resonator sensor array for multiplex label-free assays in labs-on-chips. *Lab on a Chip*, Vol. 10, No. 3, (February 2010), pp. 281-290, ISSN 1473-

García-Rupérez, J. (2011). Real-time and low-cost sensing technique based on photonic bandgap structures *Optics Letters*, Vol. 36, No. 14, (July 2011), pp. 2707-

biochemical sensor built with two-dimensional photonic crystal microcavity. *Optics* 

microring resonator for sensitive and label-free biosensing. *Optics Express*, Vol. 15,

ssDNA sensing (with a detection limit of 19.8 nM).

d'Educació through program GV-2010-031 is acknowledged.

(September 2008), pp. 15949-15957, ISSN 1094-4087

structures.

**8. Acknowledgment** 

710, ISSN 0146-9592

2709, ISSN 0146-9592

**9. References** 

0197

expensive tunable elements. Moreover, since the output power can be continuously monitored (several power values per second can be measured), a real-time sensing is performed, which allows an instantaneous observation of the interactions taking place in the sensing structure.

The initial spectral alignment between the source and the sensor will determine the sensitivity and the linearity of the device. Eq. 1 describes the relative power variation at the output (in dB) as a function of the initial spectral overlap between source and sensor (*BW*) and the shift of the guided band's edge due to a change in the refractive index (Δ*λ*).

$$10 \times \log\_{10} \left( 1 - \frac{\Delta \mathcal{U}}{BW} \right) \tag{1}$$

Fig. 17 shows the output power variation depending on the initial overlap between the source and the sensor. A high initial overlap leads to a linear response of the sensor, although a lower sensitivity is obtained. On the other hand, as the initial overlap is reduced, the sensitivity increases but a more non-linear behaviour is observed. However, a proper modeling and calibration of the sensor response will allow working in the non-linear regime, with a significant increase in the sensor sensitivity.

Fig. 17. Power variation versus wavelength shift for different initial alignments between the source and the sensor. BW indicates the initial overlapping bandwidth.

Another advantage of this readout technique is the possibility of continuously acquire the output power of the device (rigorously talking, many output power samples are taken each second). This will not only allow us to instantaneously observe any interaction taking place within the sensing device, but it will also allow to perform a temporal averaging of the power values and reduce the noise, thus leading to a significant reduction of the detection limit of the device.

#### **7. Conclusion**

Integrated planar photonic structures are one of the main candidates for the development of label-free biosensing devices, and among them, photonic crystal based structures. In this chapter, we have shown how PCWs can be used for biosensing purposes and how the use of Fabry-Perot fringes appearing at the guided band's edge instead of the edge itself allows performing the sensing with a higher accuracy and leading to a reduced detection limit. Experimental results have been shown for the detection of refractive index variations (with a detection limit of 3.5x10-6 RIU), label-free antibody sensing (with a surface mass density detection limit below 2.1 pg/mm2 and total mass detection limit below 0.2 fg), and label-free ssDNA sensing (with a detection limit of 19.8 nM).

A technique for the indirect tracking of the guided band's edge shift has also been presented. This technique avoids the use of expensive tunable sources or detectors, which are required when carrying out a direct tracking of the spectral response shift, as usually done when using photonic sensing devices such as ring resonators or photonic crystal based structures.
