Photonic crystal slabs (PCS) are one of the major transducers for label-free, optical biosensing applications. In this paper we present oblique-angle layer deposition of the high index slab material as a method to improve the PCS sensitivity. In simulations and experiments we consider PCSs composed of a high index silicon monoxide layer on a nanostructured resist layer on a glass substrate. By mounting the substrate at an oblique angle with respect to the evaporation source, the high index material distribution on the nanostructured surface is modified due to shadowing effects. Finite-difference time-domain (FDTD) simulations were performed to predict bulk and surface sensitivities. In order to verify the simulation results we fabricated PCSs at various deposition angles using nanoimprint lithography to replicate a linear grating nanostructure into the resist layer and thermal evaporation for a 60-nm silicon monoxide deposition. The bulk sensitivities of these structures were measured using water-glycerol dilutions. A sensitivity improvement of 281% was obtained for PCSs fabricated at 45° deposition angle compared to normal incidence deposition.
©2013 Optical Society of America
The transducer, which transforms a biological reaction on its surface into a measurement signal, is a central element in label-free experiments. One of the most relevant transducer technologies is based on photonic crystal slabs (PCSs) [1–3]. PCSs found their way to commercial products and are used for molecular interaction experiments  and cell-based assays  (Fig. 1(a)). Besides these applications, they can also serve as the transducer in biosensors for point-of-care diagnostics  showing the broad applicability and general relevance of these structures.
PCSs are composed of a slab waveguide with a periodic nanostructure and provide quasi-guided modes (QGMs) with an intrinsic coupling to the far field. Hence, these modes can be excited in a transmission or reflection measurement and are the origin of resonances in the spectrum known as guided-mode resonances (GMRs) . The near-field distribution of the QGM is a typical waveguide mode with its center inside the waveguide and an evanescent part extending out of the waveguide. In a label-free experiment this part of the mode interacts with the volume close to the surface and is influenced by refractive index changes caused by mass changes. By tracking the GMR’s spectral position, the biological process on the surface is measureable. Beside the biological system and the signal digitalization, the transducer’s sensitivity – defined as the resonance shift divided by the refractive index change of the analyte – determines the detection limit of the experiment . Hence, the transducer’s sensitivity is a crucial quantity for the performance of label-free experiments.
The field distribution of the employed optical mode in the PCS and the modal overlap with the analyte correlate strongly with the sensitivity [9, 10]. The highest sensitivity values have been demonstrated for free-standing PCSs, where the waveguide is not on a substrate, but is surrounded by the analyte . Here, the overlap of the mode with the analyte is highest and sensitivities up to 510 nm/RIU (Refractive Index Unit) have been realized . However, this geometry has its weak point in reduced mechanical stability and complex fabrication. Therefore, most PCSs are fabricated directly on a substrate. In  the authors present a theoretical work, where a PCS design is shown having a high sensitivity of 327 nm/RIU for an individual resonance. They propose to use a self-organizing monolayer opal structure. More suitable for cost-effective fabrication is the use of nanoimprint lithography . A periodic nanostructure is obtained by replication of a nanostructured master. By subsequent high index layer deposition a PCS is realized. This method guarantees mechanical stability as the structure is supported by a thick substrate. With nanoimprint lithography and high index layer deposition a polymer based PCS with a sensitivity of 140 nm/RIU has been demonstrated . In this paper the authors use a low index porous glass layer between the substrate and the high index layer (TiO2). Starting from this easy to fabricate and mechanically stable design researchers have obtained surface sensitivity enhancement of up to 5.5 fold due to enlargement of the sensor’s surface by introducing nanorods to the surface [15, 16].
In the context of organic distributed feedback lasers oblique-angle deposition of the high index layer has been used successfully to increase the photonic bandgap . In this approach the nanostructured substrate is mounted in the vacuum chamber at an oblique angle with respect to the evaporation source. By performing the deposition under an oblique angle the distribution of the high index layer on the nanostructure can be manipulated using shadowing effects as depicted in Fig. 1(b). Inspired by this work, here, we investigate oblique-angle deposition as a means to increase the sensitivity of PCS transducers. Section 2 presents finite-difference time-domain (FDTD) numerical simulation results for the transducer sensitivity as a function of deposition angle. Both bulk sensitivity and surface sensitivity values are evaluated as a function of high index layer thickness and deposition angle. In section 3 the fabrication of the experimental PCSs is described. Section 4 analyzes the bulk sensitivity as a function of the deposition angle for the fabricated PCSs. In section 5 conclusions are given.
2. Numerical simulations
First we performed numerical electro-magnetic simulations to compare PCS sensitivity at varied deposition angles. These simulations were carried out with the finite-difference time-domain (FDTD) method using the commercially available software FDTD Solutions from Lumerical. The PCS considered in this work has a linear geometry, which means that the nanostructure is a grating. The periodicity of this grating is 400 nm with a groove depth of 140 nm and a duty cycle of 0.5. The high index layer has a refractive index of n = 2.0, which corresponds to the refractive index of silicon monoxide (SiO) at 600 nm. As the high index layer thickness has a strong influence on the sensitivity [9, 18], it was varied in the numerical simulations from 60 nm to 160 nm.
To describe the entire PCS a 2D simulation domain is used, where the unit cell of the PCS is terminated with periodic boundary conditions. The top and bottom of the simulation domain are terminated with perfectly matched layers (PMLs) to describe the free space around the PCS. A plane wave with normal incidence onto the PCS is used as the excitation source. A power monitor records the transmitted electro-magnetic radiation through the nanostructure. The transmission spectrum is a superposition of thin-film interferences caused by the high index layer and TE-like and TM-like Fano resonances caused by the QGM in the nanostructured high index layer. During this work we focused on TE-like resonances, as they have a lower quality-factor and are hence more dominant in experiments compared to TM-like resonances. In  we introduced a method to convert the resonance shift into an intensity decrease. For this method the dominant TE-like resonance has more impact on the results. However, we expect that the results obtained in this paper for the TE-like resonance may also be applied to TM-like resonances.
In Fig. 2 samples of calculated transmission spectra for two deposition angles are shown. Here, we model the shape of the high index coverage including shadowing effects to simulate material deposition under 0° and 40°. We consider ideal conditions for the deposition resulting in the structure depicted in Fig. 1(b). Characteristic for the structures at higher angles is the rearrangement of the high index material from the groove bottom to the groove sidewall and a gap in the high index material at the bottom of the groove. The volume of high index material on the surface, however, is kept constant for all angles. Figure 2(a) plots the transmission spectra for a PCS with 0° deposition angle. To calculate the bulk sensitivity we perform two simulations and tune the bulk refractive index of the analyte above the PCS from 1.33 to 1.38, as these refractive indices are typical operating points for biological assays. The resulting resonance shift is divided by the refractive index change and we obtain a bulk sensitivity of 36 nm/RIU. The same procedure is performed to calculate the bulk sensitivity for a PCS with 40° deposition angle. As shown in Fig. 2(b) the resonance shift is more than twice and a bulk sensitivity of 74 nm/RIU is obtained.
In Fig. 3 we investigate the influence of the deposition angle on the bulk sensitivity for two different high index layer thicknesses of 60 nm and 100 nm. Transmission spectra as a function of deposition angle are shown in Fig. 3(a). We observe that the resonance in both cases shows a red shift for higher deposition angles as also observed in Fig. 2. For the case of a 60-nm high index layer thickness we observe additionally a discontinuous resonance shift for deposition angles around 15°. The origin of this discontinuity is the existence of two QGMs. While the first mode is dominant for deposition angles from 0° to 20°, the second mode is chosen from 15° upwards. In Fig. 3(b) the calculated sensitivities are plotted as a function of the deposition angle using the algorithm described for Fig. 2. For the PCS with 60-nm high index layer thickness the sensitivity of the first mode improves with higher deposition angles from 36 nm/RIU (Refractive Index Unit) to 61 nm/RIU. The second mode also shows a sensitivity of 36 nm/RIU at 15° rising to 74 nm/RIU at 40°, which is an improvement of 105%. The sensitivity of the PCS with a 100-nm high index layer, however, outperforms the PCS with a 60-nm high index layer at every angle and shows a different characteristic. It starts at a sensitivity of 87 nm/RIU and has its maximum at 12° deposition angle with 94 nm/RIU, which is an improvement of about 8%. Overall, the sensitivity values observed here are lower than the values obtained in , which may be explained by the lower refractive index of SiO compared to TiO2.
As the most common way to determine the resonance position is to use a spectrometer, the resonance’s spectral width and hence the quality-factor (Q) is a highly relevant parameter specifying the detection limit of the sensor system. We fitted a Lorentzian function to the simulated data and obtain the Q-factor of the resonance as a function of the deposition angle. In Fig. 3(c) the Q-factors for both slab thicknesses are shown. While for small deposition angles the Q-factors for both thicknesses are similar, for deposition angles higher than 20° the Q-factor for resonances provided by the 60-nm PCS are significantly higher. Looking at the product of the Q-factor and the bulk sensitivity we obtain an overall maximum for the 60-nm PCS at 33° (Fig. 3(d)), which also outperforms the value for 100-nm slab thickness. Following  the maximized product of Q-factor and the bulk sensitivity corresponds to the best detection limit for the system.
The change in sensitivity with the deposition angle may be linked to a change in the field distribution of the QGM. In the following we consider the PCS with a high index slab thickness of 60 nm. In Fig. 4 the field distribution of the investigated mode is plotted at the resonance wavelength for 0° and 40° deposition angles. As the sensitivity of the PCS is a function of the interaction area of the mode with the analyte, the higher the modal overlap with the analyte the higher is the sensitivity. In Fig. 4 we observe that the mode is pulled up towards the analyte area at 40° compared to 0° deposition angle. This phenomenon is due to the redistribution of the high index material on the nanostructure. Material, which is at 0° at the bottom of the groove, is deposited at the sidewall at 40°. A further effect causing a sensitivity enhancement may also be the gap in the layer, which appears at the groove bottom for 40° deposition angle. In this gap the mode finds more interaction area with the analyte.
Label-free experiments using the PCS as the transducer can be divided into two categories: cellular assays and molecular interaction experiments. In cellular assays a cell colony is seeded on the PCS surface and the GMR responds to changes in cell number or mass redistribution of the cell itself . Here, the cell is typically higher than the evanescent part of the mode and sensitivity determination may be performed by bulk refractive index measurements. In contrary to this, in molecular interaction experiments the binding process happens in a range only few tens of nanometers above the surface. Hence, the PCS needs to be sensitive to changes in the refractive index directly above the PCS surface. In order to calculate this surface sensitivity we adapt our simulation by evaluating the transmission spectra for a refractive index change in a volume extending 25 nm above the surface.
Figure 5 presents the simulated bulk sensitivity and surface sensitivity values as a function of the deposition angle and the slab thickness. For normal incidence deposition (0°) we observe the bulk sensitivity maximum of 93 nm/RIU for a slab thickness of 125 nm and the surface sensitivity maximum of 50 nm/RIU for a slab thickness of 130 nm. For all slab thickness values the sensitivity is improved employing oblique incidence deposition. The maximum bulk sensitivity value of 100 nm/RIU is predicted for a slab thickness of 125 nm and a deposition angle of 10°. This is an improvement of 7% compared to the normal incidence deposition. In the case of the surface sensitivity the maximum value of 58 nm/RIU is obtained for a slab thickness of 135 nm and a deposition angle of 10° corresponding to a 16% enhancement compared to the normal incidence deposition. The sensitivity enhancement at thinner slab thicknesses is even more pronounced. At 60 nm slab thickness the sensitivity increases by 105% and 2502% for bulk sensitivity and surface sensitivity, respectively. Although, the absolute values for the enhanced sensitivities are still lower compared to 125-nm slab thickness, the Q-factor for this slab thickness increases with higher deposition angles, as shown in Fig. 3(c), which is beneficial for the detection limit.
3. Sample fabrication
We fabricated PCSs at 10 different deposition angles from 0° to 45° in 5° steps. The fabrication was composed of two steps: imprinting the nanostructure onto a substrate and a subsequent high index layer deposition. The grating nanostructure of a master was transferred to the substrate using nanoimprint lithography. First, the glass substrate was spin coated with a 200-nm imprint resist layer (AMONIL). In the next step we pressed the transparent master to the substrate with the resist and cured the resist with UV radiation . The resulting grating had a periodicity of 400 nm with a duty cycle of 0.5 and a groove depth of 140 nm. Using thermal evaporation in a vacuum chamber we deposited 60-nm silicon monoxide (SiOx) onto the samples. The deposition angle was set using a sample holder allowing for a variable tilt. At angles unequal to zero the resulting layer thicknesses would be less than the desired 60 nm as a smaller effective surface of the sample is facing the material source. To compensate this effect, we evaporated more material for higher tilt angles of the sample. This ensures that the same amount of material is deposited on the sample per unit surface area. The cross-section of a PCS with an evaporation angle of 45° is plotted in Fig. 6, which was captured using focused ion beam (FIB) cutting and scanning electron microscopy (SEM). The Ag layer was deposited onto the sample for imaging. The glass substrate, the AMONIL layer, and the SiOx layer are visible. A deposition of the high index layer at the side wall and the gap in the high index layer at the groove bottom are observed. The sharp spike at the groove, however, which is expected theoretically and shown in Fig. 4, is not obtained. Since in this region at the middle of the groove the mode’s field intensity is very low, we expect no major implication on the results.
4. Experimental characterization
The bulk sensitivity was determined experimentally by applying two liquids with refractive indices of 1.33 (100% water) and 1.38 (71% water and 29% glycerol in volume percent) to the surface of the PCS. The transmission spectra through the PCS were measured by a 4x magnifying microscope setup using a halogen lamp as the light source and a spectrometer as the detector . As shown in Fig. 7(a), the refractive index change on the sensor surface results in a resonance shift. The Q-factor for the resonance obtained with a refractive index of 1.38 was 103. These spectra were used to determine the spectral resonance position. Dividing the resonance shift by |1.33 – 1.38| = 0.05 we obtain the bulk sensitivity. We fabricated two PCSs at every deposition angle and made three sensitivity determinations at different positions on each PCS. Thus, we obtained six values for each deposition angle. Due to fabrication fluctuations, e.g. imperfections in the nanostructure or groove depth variations, we obtain a distribution of sensitivities with mean values and standard deviations depicted in Fig. 7(b). These results are in a good agreement with simulated results plotted in Fig. 3(b). The bulk sensitivity, which was on average 16 nm/RIU at 0° deposition angle, was enhanced by 281% to 61 nm/RIU at 45° deposition angle.
Reducing the detection limit in label-free experiments is of high interest for extending the applicability of such experiments. Using PCSs as transducers the detection limit is directly connected to the sensitivity of the PCS, which is the resonance shift divided by the refractive index change. In this paper we show that using a small change in the fabrication process chain the sensitivity of the PCS is enhanced. In particular, we evaluate mounting the substrate at an oblique angle compared to the deposition source during high index layer deposition. We investigated the effect of oblique-angle deposition using numerical simulations (FDTD). Considering high index slab thicknesses between 60 nm and 160 nm and deposition angles between 0° and 45° we predicted a maximum oblique-incidence bulk sensitivity of 100 nm/RIU, which is an improvement of 7% compared to the maximum normal-incidence bulk sensitivity. For the surface sensitivity a maximum value of 58 nm/RIU is obtained, which is 16% higher than the maximum normal-incidence surface sensitivity. For PCSs with 60-nm slab thickness we calculated normal-incidence bulk sensitivity of 36 nm/RIU and an oblique-angle bulk sensitivity of 74 nm/RIU at 40° deposition angle. Considering the Q-factor of the resonance and using the product of the Q-factor and the bulk sensitivity as a figure of merit, we observed an overall outperformance of the PCS with 60-nm slab thickness compared to the 100-nm slab thickness. For experimental verification we fabricated PCSs with a 60 nm layer thickness at 10 different deposition angles. We measured on average a bulk sensitivity of 16 nm/RIU at 0° deposition angle and 61 nm/RIU at 45° corresponding to a sensitivity improvement of 281%. This demonstrates the potential of oblique-angle layer deposition. Oblique-angle layer deposition may be combined with other approaches for increasing the sensitivity, e.g., the low-index porous glass suggested in , in order to achieve even higher sensitivity values.
The authors acknowledge support by the German Federal Ministry for Education and Research BMBF (Project No. 0316145B).
References and Links
1. D. Threm, Y. Nazirizadeh, and M. Gerken, “Photonic crystal biosensors towards on-chip integration,” J Biophotonics 16, 1–16 (2012). [PubMed]
2. B. Cunningham, P. Li, B. Lin, and J. Pepper, “Colorimetric resonant reflection as a direct biochemical assay technique,” Sens. Actuators B Chem. 81(2-3), 316–328 (2002). [CrossRef]
3. R. Magnusson, D. Wawro, S. Zimmerman, and Y. Ding, “Resonant photonic biosensors with polarization-based multiparametric discrimination in each channel,” Sensors (Basel) 11(12), 1476–1488 (2011). [CrossRef] [PubMed]
4. B. T. Cunningham, P. Li, S. Schulz, B. Lin, C. Baird, J. Gerstenmaier, C. Genick, F. Wang, E. Fine, and L. Laing, “Label-free assays on the BIND system,” J. Biomol. Screen. 9(6), 481–490 (2004). [CrossRef] [PubMed]
6. Y. Nazirizadeh, U. Bog, S. Sekula, T. Mappes, U. Lemmer, and M. Gerken, “Low-cost label-free biosensors using photonic crystals embedded between crossed polarizers,” Opt. Express 18(18), 19120–19128 (2010). [CrossRef] [PubMed]
7. S. Fan and J. Joannopoulos, “Analysis of guided resonances in photonic crystal slabs,” Phys. Rev. B 65(23), 235112 (2002). [CrossRef]
10. N. A. Mortensen, S. Xiao, and J. Pedersen, “Liquid-infiltrated photonic crystals: enhanced light-matter interactions for lab-on-a-chip applications,” Microfluid. Nanofluid. 4(1-2), 117–127 (2008). [CrossRef]
12. L. Shi, P. Pottier, Y. A. Peter, and M. Skorobogatiy, “Guided-mode resonance photonic crystal slab sensors based on bead monolayer geometry,” Opt. Express 16(22), 17962–17971 (2008). [CrossRef] [PubMed]
13. L. J. Guo, “Nanoimprint lithography: Methods and material requirements,” Adv. Mater. 19(4), 495–513 (2007). [CrossRef]
14. I. D. Block, L. L. Chan, and B. T. Cunningham, “Photonic crystal optical biosensor incorporating structured low-index porous dielectric,” Sensor Actuat. B 120(1), 187–193 (2006). [CrossRef]
15. W. Zhang, N. Ganesh, I. D. Block, and B. T. Cunningham, “High sensitivity photonic crystal biosensor incorporating nanorod structures for enhanced surface area,” Sensor Actuat. B 131(1), 279–284 (2008). [CrossRef]
16. W. Zhang, S. Kim, N. Ganesh, I. D. Block, P. C. Mathias, H.-Y. Wu, and B. T. Cunningham, “Deposited nanorod films for photonic crystal biosensor applications,” J. Vac. Sci. Technol. A 28(4), 996–1001 (2010). [CrossRef]
17. M. Stroisch, C. Teiwes-Morin, T. Woggon, M. Gerken, U. Lemmer, K. Forberich, and A. Gombert, “Photonic stopband tuning of organic semiconductor distributed feedback lasers by oblique angle deposition of an intermediate high index layer,” Appl. Phys. Lett. 95(2), 021112 (2009). [CrossRef]
18. Y. Nazirizadeh, F. Oertzen, K. Plewa, N. Barié, P.-J. Jakobs, M. Guttmann, H. Leiste, and M. Gerken, “Sensitivity optimization of injection-molded photonic crystal slabs for biosensing applications,” Opt. Mater. Express 3(5), 556–565 (2013). [CrossRef]
19. M. Hansen, M. Ziegler, H. Kohlstedt, A. Pradana, M. Raedler, and M. Gerken, “UV capillary force lithography for multiscale structures,” J. Vac. Sci. Technol. B 30(3), 031601 (2012). [CrossRef]