A photonic-crystal waveguide sensor is presented for biosensing. The sensor is applied for refractive index measurements and detection of protein-concentrations. Concentrations around 10 μg/ml (0.15μMolar) are measured with excellent signal to noise ratio, and a broad, dynamic refractive index sensing range extending from air to high viscous fluids is presented.
©2007 Optical Society of America
During the last three decades, optical sensor elements for refractive index (RI) measurements have been subject to research interest, and still today new technologies  are suggested. The application of RI sensors is interesting for both gaseous and aqueous samples and includes measurement of parameters like temperature, humidity, chemical composition and biosensing.
The application of RI sensors for biosensing has especially gained a high degree of interest within the last two decades. It includes detection of DNA, proteins, antibody-antigen interactions, cells and bacteria. In that connection quite a few commercial sensor systems have also emerged , where especially the SPR (surface plasmon resonance) sensor and the conventional waveguide sensor are well established and well known technologies.
The trend for biosensing systems, when considering the commercial market, aims at specific and label-free detection in various complex media to ease the operation of the sensor. In addition, the sensor elements should preferably be cheap in fabrication, the system should be operated without any moving parts, and the sensitivity should be in the picomolar regime .
The SPR sensor was already applied for biosensing back in 1982  and the continuous development has resulted in reported detection limits down to 1 pg/mm2 using angular interrogation , and 5 pg/mm2 for wavelength interrogation [1,5,6]. Also the conventional optical waveguide sensor has been applied for protein sensing with detection limits down to 5 pg/mm2 [7,8].
Recently, a wide range of photonic crystal (PC) sensing devices has been presented in the literature. Photonic crystal fiber (PCF) is one class of PC devices that has been demonstrated for RI measurements and biosensing [9–11]. However these are difficult to implement in a compact, automated system and the fabrication of PCF sensors is rather tedious. So far, the biosensing presented by use of PCF’s includes detection of fluorescently labelled objects like DNA  and specific antibody detection .
Sensors based on planar bulk PCs have also been presented for sensing and biosensing in particular [15–17]. Here, white light is incident perpendicular to the surface of the crystal and the reflected light is measured. This method has been presented for immobilized protein detection  and detection of antibody/antigen multi-layers with a reported detection limit of 500 ng/mm2 .
Components based on planar PCs, also referred to as planar PBG components, where the light is guided along defects such as missing rows of holes can be designed to obtain a very high and spatially selective sensitivity to changes in the RI of the surroundings, superior to the planar bulk devices. The sensing properties of PC-WGs have already been exploited for different simple parameters. Recent applications of PC-WGs include nanofluidic tuning, RI measurements, and optical characterization of molecule orientation [18–20], however, they have not yet been used for biosensing. Thus, one interesting field of application for PBG components is for biosensing purposes. In this paper, we demonstrate a planar PC-WG for biosensing fabricated on a silicon-on-isolator (SOI) wafer. We present both RI-measurements and protein detection performed with the PC-WG. PC-WGs can be realized on SOI wafers and have the advantages compared to both conventional evanescent biosensors [12,15,17], PCFs and planar, bulk PC biosensors that they can be made ultra-compact, and integrated with both additional optical and electronic components onto one single chip.
The PC-WG has a large sensing range making it applicable within a broad dynamic range of RI measurements extending from air to high viscous fluids like oil and it is operated without any rotation of the included parts.
The ultimate sensitive and compact PC based sensor is a cavity sensor where a point defect is introduced in the periodic hole structure [12,13]. The PC cavity sensor has successfully been presented for biosensing, however, the practical application and fabrication of this sensor is not straightforward due to the imperfection of present fabrication techniques and the high sensitivity of the cavities to many different effects. The current fabrication techniques make the PC cavity devices very difficult to fabricate as single devices with the exactly intended properties like resonance wavelength, Q-factor and sensitivity . In addition, mass-fabrication of the PC cavity sensors with exactly the same properties of the individual sensors is a great challenge, where variations in the individual cavities result in different resonance wavelengths and Q-factors of the individual fabricated devices and thus, a manual characterization (combined with tuning or selection) of the single devices is needed. Also the integration of a reference PC cavity sensor can be a very challenging task due to the dependence of very precise fabrication. In these respects the PC waveguide is superior, since it has much more robust properties.
2. Photonic-crystal waveguide configuration
The term photonic crystal (PC) describes a subclass of materials where a periodic modulation of the refractive index (RI) exists for a given material. Depending on the exact periodic modulation, PCs may posses a photonic band gap (PBG). Thus, a given bandwidth of light cannot be transmitted through such material. A PC-WG comprises a planar PC with a line defect in the periodic structure . The basic property of a PC-WG is that a given bandwidth of light can be guided in the waveguide as the light is confined laterally by the PC and vertically by total internal reflection (TIR), akin to conventional waveguide structures.
Figures 1(A) and 1(B) illustrate the setup and the PC-WG configuration used in our experiments, respectively. The area “PC-WG” in Fig. 1(A) comprises the structure shown in Fig. 1(B). In the setup single-mode tapered, lensed fibres are used to couple light from a tunable laser source [ANDO, model AQ4321] into the tapered ridge feeder waveguides on the sensor-chip. The tapered waveguides have a maximum width of 4 μm at the edge of the SOI wafer and narrows down to 1 μm at the PC-WG and are realized by etching air grooves into the top silicon layer. The polarization of the incident light is selected using a polarizer and polarization controller so it predominantly excites the fundamental TE-like defect mode. The transmission is measured using an optical spectrum analyzer [ANDO, model AQ6317B].
A commercial SOI-wafer is used as platform for the crystal waveguide and the coupling elements. The PC-WG is realized in the top silicon layer of the wafer. The PC structure consists of circular air holes in a triangular lattice structure with lattice pitch, a = 370 nm, and hole diameter, d = 240 nm. The waveguide is obtained by removing a single row of holes in the Γ-K direction of the PC. The components have been fabricated using electron-beam lithography (JEOL-JBX9300FS) and inductively coupled plasma (ICP) etching to define the PC structure into the 320 nm top silicon layer. PCs using a similar fabrication process have previously been presented  but using conventional parallel-plate reactive ion etching. The PC-WG is 25 μm long and has 16 rows of holes on each side of the line defect. This design of the PC gives rise to a large band gap for transverse-electric (TE) polarized light, with the fundamental PBG mode cutting off around 1500 nm. A very small excitation of TM-like modes may occur in the PC-WG due to the incoupling scheme, however these modes are not guided well near the edge of the TE-like band gap, so the potential crosstalk is small.
Figure 2 shows TE-like mode transmission-spectra calculated using 3D Finite-Difference Time-Domain (FDTD) modelling. Spectra are calculated for four homogeneous cover media, air (nC = 1), water (nC = 1.33) and two immersion oils (nC = 1.48 and nC = 1.518). For all four spectra the entire relevant bandwidth is shown. The spectra in Fig. 2 clearly show that a wide range of wavelengths are transmitted through the structure with only a very limited loss. The single dip appearing approximately in the middle of the guided bandwidth is an artificial anti-crossing effect arising from the FDTD calculations .
From Fig. 2 it is seen that the wavelength positions of both the lower and upper band edge increases with increasing cover medium RI with the largest change in wavelength position for the lower band edge. However, the interesting property is the sudden drop in transmission at the upper band edge around 1500 nm where the transmission of the fundamental PBG mode is no longer possible. This sharp change in transmission, in the following referred to as the mode cutoff, is a clear advantage for the actual sensor operation as it is simple to detect and gives the possibility for precise detection even for spectra with a high degree of noise. The penetration depth of the field into the holes depends on the wavelength and distance of the hole from the waveguide defect. Based on FDTD simulations of this structure and SNOM measurements on a similar structure  the penetration depth can be approximated to 100 nm.
The band gab of the PC-WG is highly influenced by changes in RI at the silicon surface, either on the top of the layered structure or at the sidewalls of the holes due to the electric field distribution. This can be an advantage in biosensing as it typically involves an aqueous cover medium containing biological molecules and the detection of specific biological molecules is primarily done by immobilizing molecules at the surface (creating an adlayer of biological molecules), resulting in a RI change in the close vicinity of the silicon surface.
The TE-like and TM-like polarization can, as an approximation, be considered to have the E-field parallel with the surface of the SOI-wafer and perpendicular to the surface, respectively. From Maxwell’s equations and the boundary conditions for TE- and TM- polarized light [24,26]. The change in the normal components of the D-field (Dnx) across an interface should equal the surface charge density. For the biosensor application we are considering an interface between Si and an aqueous protein solution or two dielectric materials without any significant electrical potential difference, thus the surface charge can be approximated to zero. Furthermore the D-field is given by the dielectric constant, ϵ times the E-field and thus we obtain that ϵ 1 E n1 = ϵ 2 E n2. The dielectric constant of Si, ϵSi ≅ 12 while ϵWater ≅ 2 and thus for this interface the change in the E-field is a factor of 6 across the boundary resulting in a very high E-field in the aqueous solution just across the boundary. The E-field decreases exponentially away from the interface (and the exponential tail is very short due to the large index step) and hence, a very high spatial sensitivity is achieved in the very close vicinity of the interface either in the holes or on the SOI wafer surface depending on the used polarization.
4. Refractive index measurements
We tested the sensor by successively applying 3 different solutions on top of the PC-WG: MilliQ water (MQw) and 2 immersion oils with RIs of 1.33, 1.48 and 1.518, respectively. All the RIs used here are for λ = 633 nm and thus, dispersion corrections result in a small off-set (~1%) in the curves presented. The measurements have been performed by applying 6 μl of the solution on top of the PC-WG. Prior to applying each of the solutions, the chip was cleaned by washing in ethanol and measurements on air were performed both before applying the solution and after the cleaning procedure. These measurements (not shown here) showed that the air measurements returned to the same level after the cleaning procedure. Figure 3(A) shows the measured spectra for the three solutions and for air. The curves in Fig. 3(B) show the change in cutoff wavelength, ΔλCutoff vs. change in cover RI, ΔnC , for the calculated and the measured spectra, where the reference wavelength is the cutoff wavelength for air. In Fig. 3(A) the spectra show a loss in transmission at cutoff of approximately -20 dB. A small crosstalk from TM is seen at the transmission minimum after the cutoff wavelength. Measurements performed on a reference waveguide comprising a normal ridge waveguide on the same chip show a transmission of -14 dB over the same wavelength range of which -12 dB is due to the fiber-to-waveguide couplings and -2 dB due to the fiber setup. For the curve in Fig. 3(B) the change in cutoff wavelength has been read out at a transmission of -30 dB and for the calculated curves the change in cutoff wavelength has been read out at a normalized transmission of 0.4. Figure 3(B) shows good agreement between the measured and calculated ΔnC /ΔλCutoff where the measured ΔnC/ΔλCutoff = 0.0157 nm-1 for the fabricated PC-WG. The small deviation seen in Fig. 3(B) can originate from a difference in the hole-diameter of the fabricated compared to the simulated structure or from the existence of a thin (> 10 nm) oxide layer on the SOI-wafer. However, the method for applying the different solutions, the detection, and the reading of the signals give very reliable results.
The experiments are evaluated as a change in wavelength versus change in cover medium RI, which previously has been common practice for evaluation of biosensors. Recently, the change in reflectance vs. change in cover RI ∂R/∂nC is more commonly used to ease comparison between many different devices. For example the SPR sensor has a very large shift in wavelength but also a very broad resonance dip or a low Q-factor. From Fig. 3(A) it is seen that a sharp transmission drop for the PC-WG exists of 30 dB over a wavelength shift of only 2 nm. The entire transmission drop corresponds to a change of only 0.0314 in the cover medium RI, thus using ∂T/∂nC (change in transmission vs. change in cover RI) instead of the shift in wavelength would most likely improve the resolution of the setup.
The protein applied to the PC-WG is bovine serum albumin (BSA) [Sigma]. Two different concentrations of BSA in MQw were applied to the sensor: 10 μg/ml (0.15μMolar) and 100 μg/ml (1.5μMolar). For each protein solution time-measurements have been conducted showing the change in λCutoff vs. time as the solution on the PC-WG is changed from MQw to BSA/MQw and reapplying MQw. The results are shown in Fig. 4. The measurements were performed in the same manner as the RI-measurements: 6 μl of MQw was applied on top of the PC-WG where the fluid covers an area of approximately 10 mm2. After a given time, the larger part of the water was removed, and 6 μl of the BSA solution applied. Later 6 μl of MQw was applied again after removing the larger part of the BSA solution. Between the measurements of different BSA solutions the sensor was cleaned in 5% sodium dodecylsulphate (SDS)/MQw solution, 15 min. of soaking, and 2 min. of ultrasound. Afterwards the sensor is cleaned in ethanol, 5 min. of soaking, and 2 min. of ultrasound. As the detection experiments are performed by letting proteins settle on the surface in still water (without a flow-cell system) and with one specific analyte, we have chosen not to functionalize the surface. However, functionalization schemes for immobilizing proteins on an oxidized Si-surface have already been presented in the literature, which are applicable for this sensor element .
In the experiments, the solution is changed from water to the BSA-solution after 6 minutes. From Fig. 4 it is seen that for both measurements the cutoff wavelength increases as the solution is changed. For the low concentration a shift of 0.2 nm is observed when the BSA-solution is applied. The first measurement after the BSA-solution is applied is done after 5 minutes and the following measurements show a steady response. MQw is applied again after the experiment has run for 21 minutes showing no change in the cutoff wavelength. For the concentration 100 μg/ml a similar result is seen. The first measurement after applying the BSA-solution is done already after 2 minutes and here a continuous increase in cutoff wavelength is measured until a steady response is obtained approximately 5 minutes after the BSA solution is applied showing a shift in cutoff wavelength of 0.4 nm. Similar to the result for the lower concentration basically no shift in cutoff wavelength is observed as the solution is changed back to MQw after the experiments have run for 17 minutes. The steady response for both BSA-concentrations indicates that the proteins have adsorbed onto the sensor surface approximately 5 min after the proteins have been applied. In case all the proteins settle and adsorb on the surface, the surface coverage will reach a maximum of respectively 60 and 6 ng/mm2 for the two concentrations.
Ellipsometry measurements on a silicon surface for the two concentrations: 10 μg/ml and 100 μg/ml of BSA results in protein layers of thicknesses 2.6 nm and 3.2 nm, respectively. For the measurements we used a HeNe-laser (λ = 633 nm), a chamber containing 1 ml of the solution and a RI of 1.47  for the protein layer. The Ellipsometry measurements showed a typical Langmuir adsorption isotherm behavior for the BSA adsorption on the Si surface.
From the experimental results the sensor response can be analyzed to show that the sensor has spatial sensitivity with a high sensitivity in the cover medium at the close vicinity of the Si/cover medium interface and decreasing sensitivity with increasing distance from the interface to the position where the change in RI occurs. The 1st estimate is based on that the sensor has the same sensitivity throughout the cover medium, hence the position of the proteins is of no importance and the proteins can be conceived as homogeneously distributed in the water. From the RI measurements, a change in cutoff wavelength of 0.2 nm corresponds to a change in RI of the solution, ΔnC = 0.003. Assuming a RI of the proteins of 1.47 this RI change corresponds to a ratio of 2 × 10-2 between material with RI of 1.47 and material with RI of 1.33. If we make the assumption that the density of proteins and water are equal, this results in a protein concentration of 20 mg/ml, 2000 times higher than the actual used concentration of 10 μg/ml.
A different approach is to analyse the results assuming that the proteins adsorb on the silicon surface as a monolayer of proteins. Our Ellipsometry measurements show layer thicknesses in the same range as reported in  and thus for the analyses of the results we assume a homogeneous adlayer of 2.5 nm with RI of 1.47 is adsorbed on the sensor surface. For a penetration depth of the field into the holes (due to TE-like polarization) of 100 nm we obtain that 2.5% of the field in the solution is sensing the adlayer assuming a square profile of the field. To predict the wavelength shift due to the adlayer we made an additional approximation for how much of the total field extends into the holes. If 10% of the field propagating in the structure extends into the holes, 0.25% of the total field guided in the structure senses the protein layer. Assuming a RI of the structure of 3 and a change in RI from 1.33 to 1.47 for the protein layer the relative change in RI can be found. Considering the percentage of the field sensing the protein layer and setting the wavelength to 1500 nm the total wavelength shift due to the formation of the protein layer will be 0.18 nm This result based on spatial sensitivity of the sensor is very close to what is seen from the experiment. In addition, this approach predicts a logarithmic sensor response, both because of the exponentially decaying field in the holes and because of the normally occurring Langmuir effect giving rise to a logarithmic saturation on the surface.
The huge difference in the estimates of the sensor performance based on the two different approaches shows that the PC-WG sensor has a very high spatial sensitivity. Fortunately, biosensing often involves thin layers of immobilized molecules on the sensor surface, coinciding with the position of the strong field.
The detected protein surface coverage shown here is still a way from the detected protein surface coverage reported for the more established techniques like the SPR, conventional waveguide sensors and the QCM (quartz crystal microbalance) (in gaseous environments protein surface coverage of 0.01 ng/mm2 are reported ). However, the presented sensor-system is not yet optimized for biosening and thus, the measured concentrations do not express the lower sensitivity limit but can be optimized in a number of areas. Initially, from Fig. 4 it is seen that changes in the cutoff wavelength that are 10 times smaller should be possible to detect with acceptable signal-to-noise ratio. Secondly, the sensitivity can be increased by improvements on the sensor setup such as temperature stabilization, coupling stabilization, continuous measurements of a reference waveguide, and topology optimization of the device geometry such as the shape and size of the holes, width of the waveguide and thickness of the Si layer , which all should improve the sensitivity to lower the detected concentration. Furthermore, ∂T/∂nC measurements may most likely result in an increased resolution of the setup.
In conclusion we predict that it is possible to improve the sensor setup to increase the sensitivity considerably.
We have presented a photonic-crystal waveguide used for RI measurements and protein detection fabricated on a commercial SOI wafer using conventional CMOS fabrication procedures. Using different concentrations of BSA proteins we have demonstrated not only detection of proteins but also that it is possible to distinguish between different surface coverage densities. The demonstrated sensor showed a logarithmic response to changes in cutoff wavelength vs. protein-concentrations and a surface coverage of 6 ng/mm2 was easily detected.
The research was supported by the Carlsberg Foundation, PIPE and NEDO grants.
References and links
1. B. Sepulveda, A. Calle, L.M. Lechuga, and G. Armelles, “Highly sensitive detection of biomolecules with the magneto-optic surface-plasmon-resonance sensor,” Opt. Lett. 31,1085–1087 (2006). [CrossRef] [PubMed]
2. BIOS-1, ASI AG, Zrich, Switzerland; IAsys, Affinity Sensors, Cambridge, UK; Analight Bio200, Farfield Sensors, Ltd., Manchester, UK; OWLS 110, Micro Vacuum, Budapest, Hungary.
4. C. Nylander, B. Liedberg, and T. Lind, “Gas-detection by means of surface-plasmon resonance”, Sens. Actuators 3,79–88 (1982). [CrossRef]
5. J. Homola, S.S. Yee, and G. Gauglitz, “Surface Plasmon resonance sensors: review,” Sens. Acuators B 54,3–15 (1999). [CrossRef]
6. J.S. Yuk, S. Yi, H.G. Lee, H.J. Lee, Y. Kim, and K. Ha, “Characterization of surface plasmon resonance wavelength by changes of protein concentration on protein chips,” Sens. Actuators B 94,161–164 (2003). [CrossRef]
7. F. Hook, J. Vorors, M. Rodahl, R. Kurrat, P. Boni, J.J. Ramsden, M. Textor, N.D. Spencer, P. Tengvall, J. Gold, and B. Kasemo ”A comparative study of protein adsorption on titanium oxide surfaces using in situ ellipsometry, optical waveguide lightmode spectroscopy, and quartz crystal microbalance/dissipation,” Colloids and surfaces B 24,155–170 (2002). [CrossRef]
8. R. Kurrat, B. Walivaara, A. Marti, M. Textor, P. Tengvall, J.J. Ramsden, and N.D. Spencer, “Plasma protein adsorption on titanium: comparative in situ studies using optical waveguide lightmode spectroscopy and Ellipsometry,” Colloids and Surfaces B 11,187–201 (1998). [CrossRef]
9. J.M. Fini, “Microstructure fibres for optical sensing in gases and liquids,” Meas. Sci. Technol. 15,1120–1128 (2004). [CrossRef]
10. J.B. Jensen, L.H Pedersen, P.E. Hoiby, L.B. Nielsen, J.R. Folkenberg, J. Riishede, D. Noordegraaf, K. Nielsen, A. Carlsen, and A. Bjarklev, “Photonic crystal fiber based evanescent-wave sensor for detection of biomolecules in aqueous solutions,” Opt. Lett. 29,1974–1976 (2004). [CrossRef] [PubMed]
11. J.B. Jensen, P.E. Hoiby, G. Emiliyanov, O. Bang, L.H. Pedersen, and A. Bjarklev, “Selective detection of antibodies in microstructured polymer optical fibers,” Opt. Express 13,5883–5889 (2005). [CrossRef] [PubMed]
12. E. Chow, A. Grot, L.W. Mirkarimi, M. Sigalas, and G. Girolami, “Ultracompact biochemical sensor built with two-dimensional photonic crystal microcavity,” Opt. Lett. 29,1093–1095 (2003). [CrossRef]
13. L.W. Mirkarimi, S. Zlatanovic, M.S. Sigalas, M.A. Bynum, K. Robotti, E. Chow, and A. Grot, “Toward single molecule detection with photonic crystal microcavity biosensor,” Presented at LEOS summer topical meetings, Quebec, Canada (2006).
14. W. Bogaerts, V. Wiaux, D. Taillaert, S. Beckx, and R. Baets, “Large-scale production techniques for photonic nanostructures (invited),” Proceedings of SPIE (2003).
15. B. Cunningham, P. Li, B. Lin, and J. Pepper, “Colorimetric resonant reflection as a direct biochemical assay technique,” Sens. Actuators B 81, .316–328 (2002). [CrossRef]
16. L.L. Chan, B.T. Cunningham, P.Y. Li, and D. Puff, “Self-referenced assay method for photonic crystal biosensors: Application to small molecule analytes,” Sens. Actuators B 120,392–398 (2007). [CrossRef]
17. O. Levi, W. Suh, M.M. Lee, J. Zhang, S.r.J. Brueck, S. Fan, and J.S. Harris, “Guided-Resonance in Photonic crystal slabs for biosensing applications,” presented at CLEO/QELS Technical Conference. Long Beach, USA, 23 May 2006.
18. J. Topolancik, P. Bhattacharya, J. Sabarinathan, and P.C. Yu, “Fluid detection with photonic crystal-based multi channel waveguides,” Appl. Phys. Lett. 82,1143–1145 (2003). [CrossRef]
20. R. Ferrini, J. Martz, L. Zuppiroli, B. Wild, V. Zabelin, L.A. Dunbar, R. Houdre, M. Mulot, and S. Anand, “Planar photonic crystals infiltrated with liquid crystals: optical characterization of molecule orientation,” Opt. Lett. 31,1238–1240 (2006). [CrossRef] [PubMed]
21. S.G. Johnson, P.R. Villeneuve, S. Fan, and J.D. Joannopoulos, “Linear waveguides in photonic-crystal slabs,” Phys. Rev. B 62,8218–8222 (2000). [CrossRef]
22. P.I. Borel, L.H. Frandsen, M. Thorhauge, A. Harpøth, Y.X. Zhuang, M. Kristensen, and H.M.H. Chong, “Efficient propagation of TM polarized light in photonic crystal components exhibiting band gaps for TE polarized light,” Opt. Express 11,1757–1762 (2003). [CrossRef] [PubMed]
23. A. Lavrinenko, P.I. Borel, L.H. Frandsen, M. Thorhauge, A. Harpøth, M. Kristensen, T. Niemi, and H.M.H. Chong, “Comprehensive FDTD modelling of photonic crystal waveguide components,” Opt. Express 12,234–248 (2004). [CrossRef] [PubMed]
24. V.S. Volkov, S.I. Bozhevolnyi, P.I. Borel, and M. Kristensen, “Near.field characterization of low-loss photonic crystal waveguides,” Phys. Rev. B 72,035118 (2005). [CrossRef]
25. Reitz et al, Foundations of electromagnetic theory (Addison-Wesley, 1993), Chap. 16.
26. Klein and Furtak, Optics (John Wiley & Sons, 1986), Chap. 2.
27. R.A. Stile, T.A. Barber, D.G. Castner, and K.E. Healy, “Sequential robust design methodology and X-ray photoelectron spectroscopy to analyze the grafting of hyaluronic acid to glass substrates,” J. Biomed. Mater. Res. 61 ,391–398 (2002). [CrossRef] [PubMed]
28. N. J. Freeman, L.L. Peel, M.J. Swann, G.H. Cross, A. Reeves, S. Brand, and J.R. Lu, “Real time, high resolution studies of protein adsorption and structure at the solid-liquid interface using dual polarization interferometry,” J. .Phys.: Condens. Matter 16,S2493–S2496 (2004). [CrossRef]
29. I. Chorkendorff and J.W. Niemantsverdriet, Concepts of modern catalysis and kinetics (Viley-vch, 2003). [CrossRef]
30. J. Rickert, A. brecht, and W. Göpel, “QCM operation in liquids: Constant sensitivity during formation of extended protein multilayers by affinity,” Anal. Chem. 69,1441–1448 (1997). [CrossRef] [PubMed]
31. P.I. Borel, A. Harpøth, L.H. Frandsen, M. Kristensen, P. Shi, J.S. Jensen, and O. Sigmund, “Topology optimization and fabrication of photonic crystal structures,” Opt. Express 12,1996–2001 (2004). [CrossRef] [PubMed]