---
title: "Imaging biosensor based on planar optical waveguide"
authors: ["Valery N. Konopsky", "Elena V. Alieva"]
affiliation: "Institute of Spectroscopy, Fizicheskaya, 5, Troitsk, Moscow, 108840, RUSSIA."
journal: "Optics and Laser Technology"
year: 2019
volume: ""
issue: ""
article_number: ""
pages: ""
doi: "10.1016/j.optlastec.2019.02.034"
type: journal-article
site_group: ""
url_abstract: ""
url_pdf: "https://valery.konopsky.com/kvnlocal/KonopskyAlieva_JOLT_2019.pdf"
language: en
source_tex: "Z:\\ValeryData\\Valery_New\\my_articles\\OLT2019\\OLT\\manuscript\\send\\Konopsky2OLT.tex"
source_pdf: "Z:\\ValeryData\\Valery_New\\my_articles\\OLT2019\\OLT\\published\\KonopskyAlieva_JOLT_2019.pdf"
---
Label-free optical biosensors ,Surface wave imaging ,Planar optical waveguide

# Introduction

Optical label-free biosensors are an important tool for the selective recognition of target biomolecules and for biomolecular interaction analysis, providing kinetics data of biological binding events in real time without labelling. The advantages of the label-free concept include elimination of undue detrimental effects from labels that may interfere with fundamental interaction and the absence of time consuming pretreatment [1].

Registration of a bounded optical wave propagating along a bioactive surface is the most popular method used in commercial label-free optical biosensors [2]. In surface plasmon resonance (SPR) biosensors [3], this surface-bound wave is a surface plasmon-polariton (SPP) propagating along a gold or silver surface. In planar optical waveguide (POW) biosensors, this wave is a waveguide mode excited in a high refractive index dielectric layer either via the frustrated total internal reflection from a low refractive index spacer [4, 5] or via a grating coupler, as in the integrated-optical waveguide technique [6, 7, 8].

In recently developed photonic crystal surface mode (PC SM) biosensors, this surface optical wave is the excitation of an optical mode at the external surface of a multilayer structure [9, 10]. In this case, the confinement of the optical field near the interface is due to the photonic gap in the multilayer structure on the internal side of the outer surface and from total internal reflection on the external side, which is in contact with the liquid.

At present, most commercial label-free optical biosensors are based on the SPR method [11], while only one PC SM biosensor is commercially available [12]. Moreover, the SPR biosensors have an imaging modification (SPR*i*), which was first demonstrated in the late 1980s by two groups: Yeatman and Ash [13] and Rothenhäusler and Knoll [14, 15]. These SPR*i* biosensors have two-dimensional (2D) spatial resolution and have attracted much attention recently [16]. Commercial SPR*i* devices are also available, for example from GWC Technologies [17] and HORIBA Scientific [18], and may be used in such fields as high-throughput biosensing and chemical sensor arrays.

Increasing the propagation range of the surface wave permits to get a more sensitive biosensor. As an example, biosensors based on long-range SPPs (with a decreased metal damping) have better sensitivity [19]. However, PC SMs and POWs do not suffer from metal damping at all and consequently have a much longer propagation length than SPPs, which makes it possible to develop more sensitive biosensors based on PC SMs and POWs. Nevertheless, until recently, no imaging modification of optical biosensors based on PC SM and POWs had been proposed. The reason is straightforward: in a standard SPR*i*, the biochip surface is illuminated by a parallel light beam, the angle of incidence of which is tuned on the slope of the resonance dip in the SPR reflectivity curve. A two-dimensional black and white camera then detects the reflected intensity from each point on the investigated surface. Thus, the width of the angular SPR curve determines the dynamic range of the SPR*i*. However, PC SMs and POWs have a longer propagation length and a much smaller resonance angle curve width. Therefore, the dynamic range of the PC SM and POW sensor in a similar scheme would be negligible and impractical.

Recently we presented an imaging biosensor based on PC SM with dynamic range and sensitivity compared or better than ones in the SPR*i* biosensor [20]. PC SM resonance registration by a color camera was used to obtain 2D spatial resolution of the biosensor signal. In this study, a similar scheme is presented, where POW resonance is used for imaging and both parameters of POW and PC  SM biosensors and the simplicity of practical implementation of these biosensors is compared.

# Materials and methods

## Waveguide imaging biosensor

![Schematic of the biosensor. Typical spectra of light as it passes from the LED to the color camera are shown.](media/OLT2019/Fig1.eps)

*Schematic of the biosensor. Typical spectra of light as it passes from the LED to the color camera are shown.*

A new 2D imaging biosensor with spectral registration of POW resonance shift is presented in this paper. Fig. 1 shows a schematic of the biosensor and typical spectra of an optical beam as it passes through the optical elements of the biosensor. A parallel light beam, with polarization $+45^0$ to the plane of incidence, illuminates a waveguide through a prism (Kretschmann-like configuration). After reflection, the light beam passes through the second polarizer, which is $-45^0$ to the plane of incidence, and is then recorded by a color camera (D1312C, Photonfocus, Lachen, Switzerland). Thus, the light beam passes through crossed polarizers and if the waveguide mode does not undergo excitation, there is no signal at the color camera. However, the waveguide presented here is designed so that it supports a $p$ -$\!\!$ polarized mode at $\lambda=500$ nm with an effective refractive index (RI) $\rho=1.5$, corresponding to an in-prism excitation angle $\theta_0=56.4^0$. A fiber-coupled LED (M505F1, Thorlabs, Newton, NJ, USA), with a peak wavelength near $\lambda=500$ nm and a bandwidth (FWHM) of 35 nm, is used for excitation of the waveguide mode. This excitation causes a phase shift of the waveguide wave that re-radiates back to the prism. Therefore, rotation of the resulting polarization of the reflected beam occurs, and, as the result, the light at the excitation wavelength passes through the crossed polarizers. The mean wavelength of the transmitted light peak is determined by balancing of the intensities of the blue and green pixels at each point in the color camera.

## Waveguide structure

The waveguide structure used in these experiments consists of the following: *substrate* $/LH/$ *water*, where substrate is SF11 prism, $H$ is a $TiO_2$ waveguide (thickness $d_2=53.3$ nm), and $L$ is a $SiO_2$ spacer (thickness $d_1=500$ nm). The waveguide structure was created by a SYRUSpro 710 optical vacuum coater (Buhler Leybold Optics, Alzenau, Germany) via electron-beam evaporation and plasma ion-assisted deposition. At $\lambda=500$ nm, the RIs of the SF11 prism, $SiO_2$, $TiO_2$, and water are $n_0=1.8$, $n_1=1.47$, $n_2=2.45$ and $n_e=1.338$, respectively.

## Reagents

The reagents used in the experiment were water, sodium chloride (NaCl), ethanol (EtOH), poly (allylamine hydrochloride) (PAH, 58 kDa), poly (styrene sulfonic acid), sodium salt (PSS, 77 kDa), and poly (allylamine) solution (PAA, 65 kDa, 0.1 mg/mL). All reagents were purchased from Sigma-Aldrich (St. Louis, MO, USA).

## Sample preparation

The waveguide structure deposited on the prism base was washed twice in an ultrasonic bath of pure EtOH solution (each bath was 10 min) and then treated with a home-made UV/ozone cleaner (15 min) immediately before use.

## Data handling

Data acquired by the color camera was processed and presented using custom software. The shift of the waveguide mode wavelength was determined from the next normalized differential value:
``` math
\begin{equation}
   z(x,y)=\frac{B(x,y)-G(x,y)}{B(x,y)+G(x,y)}\, ,
   
\end{equation}
```
where $B(x,y)$ and $G(x,y)$ are the intensities of the blue and green components in each color pixel $(x,y)$ of the camera, respectively.

To convert $z(x,y)$ into $\lambda(x,y)$, the responsivities of the color camera and a procedure described in [20] were used. In short, to obtain $\lambda(z)$, a polynomial fit as follows is used:
``` math
\begin{equation}
   \lambda(z)=\sum_{n=0}^{n=8}p_n \cdot z^n\, ,
   
\end{equation}
```
with the next coefficients: $p_{[0\rightarrow8]}$=[494; -40.5; -48.8; -165.2; 101.9; 716; 94.5; -951; -437.3]. For the particular camera used in these experiments, this fit gives reasonable conversion accuracy. Now, after obtaining the experimental data from the color camera, Eq. (1) was used to determine $z$, followed by Eq. (2) to determine $\lambda$ for each camera pixel to obtain the spatial distribution $\lambda(x,y)$ across the camera that reflects the spatial distribution of the waveguide resonance shifts across the sample.

![Experimental spectra of the waveguide mode dips (+450; +450) and peaks (+450; -450) for four different solutions. The responsivities of the blue and green pixels of the color camera are also shown in correspondent colors.](media/OLT2019/Fig2.eps)

*Experimental spectra of the waveguide mode dips (+450; +450) and peaks (+450; -450) for four different solutions. The responsivities of the blue and green pixels of the color camera are also shown in correspondent colors.*

![Waveguide resonance shifts resulting from changes in volume RI.](media/OLT2019/Fig3.eps)

*Waveguide resonance shifts resulting from changes in volume RI.*

# Results

## Spectra of the waveguide mode dips and peaks

Experimental spectra of the waveguide dip and peak are shown in Fig. 2. To register these spectra, the color camera in Fig. 1 was temporarily replaced by the fiber input of a fiber-coupled spectrometer (F30-EXR, Filmetrics, San Diego, CA, USA). The spectrum of the waveguide peak was recorded at the crossed polarizers ($+45^0$; $-45^0$), while the spectrum of the waveguide dip was recorded when the polarization direction of the second polarizer coincided with the first one ($+45^0$; $+45^0$). The narrowness of the resonant dip and peak depends on the thickness of the $SiO_2$ spacer between the SF11 prism and the $TiO_2$ waveguide. A thicker spacer gives a narrower resonant peak. To record the spectra for different RIs of external media, water was replaced by 2%, 4% and 8% solutions of ethanol (EtOH). The increase in the RI of the external media leads to a spectral shift of the waveguide dip and peak to longer wavelengths, as expected. The shift of the waveguide peak was a useful signal in this biosensor. The shift is determined through normalized difference of the blue and green pixels in each camera point. Responsivities of the blue and green pixels of the color camera are also presented in Fig. 2.

## Detection of changes in volume RI 

For initial testing of the biosensor, 4%, 8% and 12% EtOH solutions were injected into a flow cell, alternating with pure water injections. Fig. 3 presents the signals from the wavelength shifts ($\Delta\lambda$) during these injections. This figure shows that in the presented waveguide biosensor a wavelength shift of $\Delta\lambda = 0.9$ nm for 4% EtOH injection was obtained, while in the experiments using the PC SM biosensor a shift of $\Delta\lambda = 0.6$ nm for the same injection was recorded [20]. Thus, the RI sensitivity of the waveguide biosensor is 1.5 times better, according to experimental data. This value is in agreement with a ratio of the theoretical RI sensitivities of these biosensors. The theoretical sensitivity, in units of volume RI, of the waveguide biosensor is $\Delta\lambda/\Delta n_e=583$ nm/RIU, which is 1.48 times better than that of the PC SM biosensor (393 nm/RIU [20]).

![The waveguide resonance shifts due to deposition of polyelectrolytes on the surface of the POW biosensor.](media/OLT2019/Fig4.eps)

*The waveguide resonance shifts due to deposition of polyelectrolytes on the surface of the POW biosensor.*

## Deposition of PAH and PSS polyelectrolytes

A polyelectrolyte sandwich was prepared using the positive polyelectrolyte PAH and the negative polyelectrolyte PSS, both at 1 mM concentration (calculated with respect to the monomer) in an aqueous solution of 0.1 M NaCl. The waveguide structure, placed in flow cell, was alternately exposed to the PAH and PSS solutions for 5 min, starting with the positive polyelectrolyte PAH. After each adsorption step, the substrate was rinsed with an aqueous solution of 0.1 M NaCl for 200 s before exposure to the next electrolyte solution. The results are presented in Fig. 4.

From this figure one can see that the average value of the wavelength shift in the two-layer deposition of PAH/PSS is $\Delta\lambda = 2.7$ nm (excluding the first double layer, which is incomplete). The theoretical sensitivity of the presented waveguide biosensor (in units of the adsorbate thickness and assuming that the RI of the adlayer is $n_a = 1.43$) is $\Delta\lambda/\Delta d_a=0.77$ nm/nm. The same sensitivity of the biosensor PC SM is 2.7 times less (0.284 nm/nm [20]). In the experiments using the PC SM biosensor, a shift of $\Delta\lambda = 1.1$ nm for the PAH/PSS double layer deposition was obtained [20], 2.5 times less than in the presented biosensor. Taking into account the many different parameters in both experiments (surface type, temperature, etc.), the agreement between the theoretical relationship and the experimental data is reasonable.

# Discussion

## Comparison of sensor sensitivities and baseline noises for the waveguide and for the PC SM biosensors

Both biosensors reveal volume and surface changes as a shift in resonance, in nanometers (of wavelength). To convert the wavelength shifts to changes in refractive index unit (RIU) or to changes in the adsorbate layer thickness in nanometers (of thickness), these shifts should be multiplied by the corresponding sensitivities. Additionally, after multiplying these sensitivities by a wavelength noise, $\delta\lambda=10^{-3}$ nm in the present device (for 1-s accumulation time and for the minimal spot size $\sim100\times100$ $\mu$m$^2$), the baseline noise of the device in units of volume RI or adlayer thickness may be obtained (during the integration over this minimal spot size). It should be noted that if spatial integration takes place over a large spot size ($>1$ cm$^2$), the wavelength noise is an order of magnitude smaller ($\delta\lambda=10^{-4}$ nm), and the baseline noises of $n_e$ and $d_a$, measured over large spots, are also accordingly smaller.

Therefore, the baseline noise in RIU is $\delta n_e =1.7\times10^{-6}$ during the integration over the minimal spot size and is $\delta n_e =1.7\times10^{-7}$ during the integration over the larger spot sizes. This RIU baseline noise is 1.5 times better than that of the PC SM biosensors, as the RIU sensitivity of the waveguide biosensor (583 nm/RIU) is 1.5 times higher (the wavelength noise is the same in both devices).

The baseline noise in thickness units is $\delta d_a =1.3\times10^{-3}\,\mathrm{nm}=1.3\,\mathrm{pm}$, corresponding to 1.3 pg/mm$^2$ in surface mass density. For the highest spatial resolution it gives 13 fg, which is deposited in the spot $100\times100$ $\mu$m$^2$. This baseline noise is 2.7 times better than that of the PC SM biosensors since the adlayer’s thickness sensitivity (0.77 nm/nm) is 2.7 times higher for the presented device.

## Dynamic range

The increased sensitivities of the presented waveguide biosensor lead to lower baseline noises, but also to smaller dynamic range (DR). The DR of the biosensor is limited by the LED bandwidth, which is 35 nm (FWHM). Thus, the operating range of the device is 480–515 nm. To obtain the DR, the corresponding change in RI or the change in thickness that shifts the waveguide resonance from 480 to 515 nm is used. For this device, the DR is $\phantom{}_{480}^{515}\Delta n_e\simeq 0.03$ RIU or $\phantom{}_{480}^{515}\Delta d_a\simeq 55$ nm. If a light source with a wider spectrum would be used, the DR can be two to three times larger (450–550 nm) and will be limited only by the spectral range between the maxima of the blue and green pixels.

## Comparison of simplicity of practical implementation of the waveguide and the PC SM biosensors

In the presented waveguide-based biosensor the waveguide mode is excited in a high refractive index ($n_2=2.45$) dielectric layer via the frustrated total internal reflection from a low refractive index spacer ($n_1=1.47$). Therefore, the effective refractive index, $\rho$, of the waveguide mode should be more than RI of the spacer ($\rho>n_1$). To obtain the phase matching with the waveguide mode with such high effective RI, one should choose a prism with high $n_0$ to have a reasonable (i.e., not a grazing) excitation angle inside the prism, since $\sin(\theta_0)=\rho/n_0$. An SF11 glass prism ($n_0 = 1.8$) was used in the presented experiments, providing the excitation angle of $\theta_0=56.4^0$ inside the prism. Next, one should either deposit the spacer and waveguide layers directly onto the SF11 prism base, or use some replaceable SF11 plates and immersion oil with the same RI (1.8). This is a limitation of the presented biosensor, which makes it less practical than the PC SM-based biosensor.

The photonic crystal multilayer structure can be deposited on standard BK-7 plates, and it then possible to place these interchangeable plates on a standard BK-7 prism using standard immersion oil (with RI $n_0 = 1.52$). While in the waveguide biosensor, an irreplaceable prism, or non-standard replacement plates and immersion oils must be used. The reason is that in the PC SM biosensor the effective RI of the surface wave can be in the range $n_1>\rho>n_e$ ($1.47>\rho>1.338$, with the materials used in this study), whereas in the POW biosensor it can be in the range $n_2>\rho>n_1$ ($2.45>\rho>1.47$, with the materials used here).

# Conclusions

A new biosensor method based on wavelength interrogation of the resonant peak of the waveguide mode was proposed and experimentally implemented. The presented biosensor has 1.5-2.7 times better sensitivities and lower baseline noises compared to the similar PC SM biosensor. However, the waveguide biosensor has a dynamic range 1.5-2.7 times worse and is more difficult to put into practice due to the high RI of prism involved.

# Acknowledgement

The authors are grateful to Irina Petrova for the supply and preparation of PSS/PAH polyelectrolytes.

# References

## References

1. M. A. Cooper, Label-free screening of bio-molecular interactions , Anal. Bioanal. Chem. 377 (2003) 834--842.

2. X. Fan, I. M. White, S. I. Shopoua, H. Zhu, J. D. Suter, Y. Sun, Sensitive optical biosensors for unlabeled targets: A review, Anal. Chim. Acta 620 (1-2) (2008) 8--26.

3. J. Homola, Surface plasmon resonance sensors for detection of chemical and biological species, Chem. Rev. 108 (2) (2008) 462--493.

4. R. Cush, J. Cronin, W. Stewart, C. Maule, J. Molloy, N. Goddard, The resonant mirror - a novel optical biosensor for direct sensing of biomolecular interactions. I. Principle of operation and associated instrumentation , Biosensors & Bioelectronics 8 (7-8) (1993) 347--353.

5. A. Nabok, A. Al-Rubaye, A. Al-Jawdah, A. Tsargorodska, J.-L. Marty, G. Catanante, A. Szekacs, E. Takacs, Novel optical biosensing technologies for detection of mycotoxins, Optics & Laser Technology 109 (2019) 212--221.

6. K. Tiefenthaler, W. Lukosz, Sensitivity of grating couplers as integrated-optical chemical sensors, J. Opt. Soc. Am. B-Opt. Phys. 6 (2) (1989) 209--220.

7. W. Lukosz, Integrated optical chemical and direct biochemical sensors, Sens. Actuator B-Chem. 29 (1-3) (1995) 37--50.

8. J. G. Wang"uemert-P'erez, A. Hadij-ElHouati, A. S'anchez-Postigo, J. Leuermann, D.-X. Xu, P. Cheben, A. Ortega-Mo nux, R. Halir, 'I. Molina-Fern'andez, Subwavelength structures for silicon photonics biosensing, Optics & Laser Technology 109 (2019) 437--448.

9. V. N. Konopsky, E. V. Alieva, Photonic crystal surface waves for optical biosensors, Anal. Chem. 79 (12) (2007) 4729--4735.

10. V. N. Konopsky, T. Karakouz, E. V. Alieva, C. Vicario, S. K. Sekatskii, G. Dietler, Photonic crystal biosensor based on optical surface waves, Sensors 13 (2) (2013) 2566--2578.

11. A. P. Turner, Biosensors: sense and sensibility, Chemical Society Reviews 42 (8) (2013) 3184--3196.

12. www.pcbiosensors.com (2011).

13. E. Yeatman, E. Ash, Surface plasmon microscopy, Electronics Letters 23 (20) (1987) 1091--1092.

14. B. Rothenh"ausler, C. Duschl, W. Knoll, Plasmon surface polariton fields for the characterization of thin films, Thin Solid Films 159 (1-2) (1988) 323--330.

15. B. Rothenh"ausler, W. Knoll, Surface plasmon microscopy, Nature 332 (6165) (1988) 615--617.

16. P. Singh, SPR biosensors: Historical perspectives and current challenges, Sensors and Actuators B: Chemical 229 (2016) 110--130.

17. www.gwctechnologies.com.

18. www.horiba.com.

19. O. Krupin, H. Asiri, C. Wang, R. N. Tait, P. Berini, Biosensing using straight long-range surface plasmon waveguides, Optics Express 21 (1) (2013) 698--709.

20. V. N. Konopsky, E. V. Alieva, Photonic crystal surface mode imaging biosensor based on wavelength interrogation of resonance peak, Sensors and Actuators B: Chemical 276 (2018) 271--278.

21. Analytical Methods Committee, Recommendations for the definition, estimation and use of the detection limit, Analyst 112 (1987) 199--204. thebibliography.

---
Generated: 2026-08-30 from Z:\ValeryData\Valery_New\my_articles\OLT2019\OLT\manuscript\send\Konopsky2OLT.tex, validated against Z:\ValeryData\Valery_New\my_articles\OLT2019\OLT\published\KonopskyAlieva_JOLT_2019.pdf.