Vapor and liquid optical monitoring with sculptured Bragg microcavities

1 MS JNP 17102P-R revised (oct 17th, 2017) Vapor and Liquid Optical Monitoring with Sculptured Bragg Microcavities Manuel Oliva-Ramirez*, Jorge Gil-Rostra, Maria C. López-Santos, Agustín R. González- Elipe, Francisco Yubero* Instituto de Ciencia de Materiales de Sevilla (CSIC, University of Seville) , CICCartuja, Avda. Américo Vespucio 49, E-41092, Spain Abstract. Sculptured porous Bragg Microcavities (BMs) formed by the successive stacking of columnar SiO2 and TiO2 thin films with a zig-zag columnar microstructure are prepared by glancing angle deposition. These BMs act as wavelength dependent optical retarders. This optical behavior is attributed to a self-structuration of the stacked layers involving the lateral association of nanocolumns in the direction perpendicular to the main flux of particles during the multilayer film growth, as observed by Focused Ion Beam Scanning Electron Microscopy. The retardance of these optically active BMs can be modulated by dynamic infiltration of their open porosity with vapors, liquids or solutions with different refractive indices. The tunable birefringence of these nanostructured photonic systems have been successfully simulated with a simple model that assumes that each layer within the BMs stack has uniaxial birefringence. The sculptured BM have been incorporated as microfluidic chips for optical transduction for label free vapor and liquid 2 sensing. Several examples of the detection performance of these chips, working either in reflection or transmission configuration, for the optical monitoring of vapor and liquids of different refractive index and aqueous solutions of glucose flowing through the microfluidic chips are described. Keywords: Sculptured thin films, porous films, Bragg Microcavities, optofluidics, vapor sensing, liquid sensing. *Francisco Yubero, E-mail: y[email protected]s *Manuel Oliva-Ramírez, Email: [email protected] 3 1 Introduction Thin film growth by physical vapor deposition methods such as electron beam, ion-beam evaporation, or magnetron sputtering in glancing angle deposition (GLAD) geometry can be used to fabricate fancy porous microstructured thin films. Several reviews have been written on this topic focusing either on the physics of the deposition process [1,2] or the optical performance [3, 4] and other potential applications of the deposited porous materials [5,6]. One of the most interesting effects of GLAD is the capability to grow strongly birefringent or dichroic coatings because of the possibility to control the anisotropic microstructure of the deposited materials. Thus porous columnar thin films (CTF) or more sophisticated sculptured thin films (STF) can be manufactured to work as retarders [7], mirrors showing circular Bragg phenomenon [8,9], omnidirectional reflection mirrors [10], or tunable circular polarizers [11]. However, a commonly claimed drawback for the implementation of nanostructured GLAD thin films as reliable optical elements is precisely their high porosity and their associated optical instability due to ambient water vapor adsorption and condensation in the pores. The motivation of this work is to convert this weakness in a strength for the development of label free vapor and liquid sensing devices based on a tunable fluidic actuated optics. Vapor or fluid sensors based on porous thin films have been proposed in the literature [12,13,14]. More specifically, fluid sensors have been proposed previously by Lakhtakia et al. [15], high speed optical humidity sensors have been developed by Liu et al. [16] and several other authors have applied optically complex GLAD structures for the monitoring of liquids or vapors using different methods of interrogation [6]. In previous works we have also shown that 1D-photonic crystals and Bragg Microcavities (BM) formed by the successive stacking of porous TiO2 and SiO2 CTFs deposited by electron beam evaporation in a GLAD geometry change their optical behavior when 4 they are infiltrated with liquids [17,18,19]. This response was dependent on the refractive index of the infiltrated liquid and we proposed using this effect for liquid monitoring [18]. An outstanding possibility by the fabrication of these structures is to make them either optically active or inactive with respect to the polarization of light [19], a feature that has not been used previously in a systematic way for liquid or vapor detection. Thus, in the present work we provide a deeper insight into the birefringence behavior of some porous sculptured BM prepared by GLAD and discuss how to benefit from the ability of tuning their optical anisotropy to develop robust, cheap and reliable optofluidic transducers that, incorporated into microfluidic chips, serve to monitor vapors and liquid solutions. 2 Fabrication of birefringent thin film optical elements with open porosity In general, glancing angle deposition of a vapor flux at low enough process pressures give rise to porous thin film growth with open porosity and anisotropic microstructure. If the mobility of the deposited moieties of material is weak, open regions are formed by the shadowing of the initial nucleation centers to the arriving flux of material. In the incidence plane (formed by the flux direction and normal to the sample surface) the competition between this shadowing mechanism and adatom mobility leads to the growth of tilted nanocolumns. On the other hand, in the direction perpendicular to the incidence plane, there is no shadowing and a homogeneous supply of vapor takes place. Thus, the columns fan out laterally due to the lack of shadowing perpendicular to the deposition plane inducing an enhanced connectivity of the deposited material in this direction. This anisotropic column aggregation is known as bundling. This phenomenon was thoroughly discussed 5 in ref.[1]. Figure 1a (plane of incidence) and 1b (front view) shows a scheme of the deposition process and an ideal scheme of the developed thin film microstructure. If no movement is applied to the substrate during the GLAD of a thin film (as it is assumed in the description below), tilted columnar growth takes place. In general, this tilted columnar growth is considered a disadvantage for the development of optically active elements that are going to operate at normal incidence of the light. This is due to the fact that standard isolated nanocolumn arrays present limited in-plane birefringence Δn, defined this as the difference in effective refractive index for light incidence normal to the surface for polarizations either parallel np or perpendicular ns (bundling direction) to the plane of incidence (p and s directions in Figure 1b, respectively). In a 3D coordinate system (Figure 1a, ns = n3 and np = 1/sqrt(sin2ψ/n12+cos2ψ/n22) where ψ is the column tilt angle, and n1, n2 and n3 the principal refractive indices of the film (n1 along the column direction, n2 in the direction perpendicular to the column direction in the plane of incidence and n3 in the direction perpendicular to the deposition plane, see figure 1). Taking into account depolarization factors within an effective medium approximation it is expected that n1 < n3 < n2. This situation can be modified varying the standard GLAD protocol. Thus, for example, the socalled serial bideposition [20], the use of substrates with a patterned surface or the use of increased 6 process pressures may alter the lateral interconnection of material and therefore, the columnar growth to enhance in-plane birefringence. Fig 1. Scheme of GLAD deposition process used in this work and developed thin film microstructure. (a) Plane of incidence and tilted thin film columnar growth (main optical axis of CTF are included). (b) Lateral columnar association in the direction s perpendicular to the plane of incidence. 3 Sensing with birefringent porous BM optical elements Although individual single material GLAD thin film architectures can be used to fabricate vapor or liquid sensing elements, we will show that the traditional optical multilayer design combined with CTF porous materials of different refractive index can be effectively used to develop optical devices with enhanced sensitivity and robustness. In particular, BMs are especially suited for the development of sensing elements. A BM consists of two Bragg mirrors separated by a central layer (cavity) whose thickness and refractive index is chosen to open a narrow transmission peak (defect peak) at the reflection band of the Bragg mirrors. In this paper we report on sculptured porous BMs fabricated by the successive stacking of SiO2 and TiO2 CTF prepared by GLAD electron beam evaporation of SiO2 or TiO pellets, respectively. The zenithal angle of evaporation was 70º. The Bragg reflectors consists of seven stacked CTFs 7 of SiO2 and TiO2 (individual thicknesses of 80 nm) sandwiching the thicker (180 nm) SiO2 CTF cavity as indicated by the figure 2a. The thickness of the individual layers was controlled by a calibrated quartz microbalance. Azimuthally turning the substrate by 180º for each deposited layer rendered a zig-zag columnar microstructure and a homogeneous thickness over relatively large areas. The base pressure of the deposition equipment was 1.0x10-6 mbar and the process pressure during film growth was 2x10-4 mbar by intake of a O2 leak to the chamber to warrant that the samples grow transparent (i.e., fully oxidized). The crucible-sample distance was about 50 cm, the deposition rate about 6 nm per minute, and the sample temperature during growth about 100ºC. The resulting zig-zag microstructure is clearly devised in the micrograph of Figure 2a-left corresponding to a FIB-SEM cross section taken normal to the surface along the p direction (i.e., perpendicular to the bundling direction, see inset). Meanwhile, the FIB-SEM taken along the s direction (Figure 2a-right) shows that the nanocolumns are laterally associated (note the higher connectivity of the deposited material in this direction with respect to that in Figure 2a-right) and that this association covers the whole multilayer structure, i.e., it passes from one layer to the next even if the deposited materials are different. We have named this lateral association of nanocolumns comprising the whole thickness of the BM and affecting different material layers as fence-bundling [19]. Figure 2b shows the transmittance at normal incidence of polarized light through the sculptured zig-zag BM. Note that the transmitted defect peak (in this case at 590-620 nm) varies its position and intensity depending on the orientation of the light polarization vector with respect to the optical axes of the sample. Thus, if the polarization of the light is aligned to the p axis of the sample, the transmitted peak is shifted by 23 nm to lower wavelength than for the alignment along the s axis. 8 For intermediate orientations, a double peak is transmitted, confirming the birefringent behavior of the BM. Note that this birefringence has structural origin. Fig 2. a) Cross section FIB-SEM micrographs (left p-axis, right s-axis) of the sculptured BM. b) Experimental transmittance T of polarized light at normal incidence of the BM around its defect peak c) Experimental measurements (left) and simulations (right) of transmittance after aligned Tuu (top) or crossed Tux (middle) polarizers of the BM for incident light normal to the surface and polarization parallel to s, p or φ=45º azimuthal sample orientations. The corresponding intensity ratios between these signals are also shown (bottom). To quantify the optical response of the BM we have developed an optical model to simulate the optical behaviour of this porous sculptured BM. For simplicity, we took each layer within the BM as a porous uniaxial birefringent media. A good consistency with experimental transmittance 9 measurements was achieved considering a multilayer optical model with in-plane birefringence of Δn=0.150 for the TiO2 layers and 0.043 for the SiO2 layers of the BM. Note that these values are about the same than those reported for the same materials deposited as single layer in a serial bideposition process [20]. The consistency of the optical model is demonstrated by the equivalence of the experimental and simulated transmitted spectra reported in Figure 2c. In this experiment, normal incident polarized light impinges onto the sculptured BM and the transmitted light is analyzed with a second polarizer aligned either parallel (Tuu) or perpendicular (Tux) to the polarization vector of the incident light for three azimuthal sample orientations: parallel to s, p or 45º off these main axes (i.e., azimuthal angle φ = 45º, see inset). The spectra reported in this figure can be interpreted in the following terms. For aligned polarizers and azimuthal sample orientation φ=45º, two resonant peaks are generated (red dashed curves in the top panel of Figure 2c) and their separation can be taken as a measure of the optical activity of the system. By contrast, for the polarization vector of the incident light aligned parallel to the s or p azimuthal sample axis, a single defect peak located at different wavelengths was found. One of these peaks appears when the polarization vector of the impinging light aligns with the s sample direction and light travels through the slow optical axis (i.e., parallel to the bundling and presenting a higher refractive index), while the second occurs when the light travels through the fast optical axis with the polarization aligning the p direction (i.e., perpendicular to the bundling). The separation between these two single peaks is 23 nm, a displacement stemming from the combined effect of the optical transmittance of the sculptured BM and the different polarizability of the individual layers for the two optical axis. In particular, we link the observed higher polarizability in the single layers (and therefore a refractive index larger than in the orthogonal direction) to the enhanced connectivity of the material in this direction as observed in the FIB-SEM micrographs 16 refractometer. This corresponds to a sensitivity defined as intensity modification per refractive index unit of 1660/RIU. The absence of any hysteresis as in the case of vapor sensing described above, further support the reliability of the liquid sensing process based on the optically active BM. Fig 4. a) Experimental measurement of the reflectance after cross polarizers Rux and (b) intensity ratio of the reflected defect peaks I(Rux)/I(Ruu) in the porous BM transducer as the concentration of glucose in the solution increases To probe the capability of the previous measurement procedure for real microfluidic applications we have developed an integrated microfluidic system. The scheme of the optical part of the apparatus is drawn in figure 5a and a picture of the actual prototype is presented in figure 5b (a patent claim protects this device and measurement system [23]. It is based on the evaluation of the change of the optical activity of the resonant peak of a porous sculptured BM when the liquid solution circulates through the sample pores. The optical part of the system, schematically presented in Figure 5a, consists of a beam of polarized light that impinges normal to the surface of a fluidic device that contains the BM. The light is supplied by a LED with a wavelength selected 17 to match the resonant peak of the optically active BM. In the scheme of the figure, the system works in transmission, and the polarization state of the transmitted light is analyzed with a polarizing beamsplitter cube that separates polarized light in their two components. An alternative design also exists for a reflection geometry. A virtue of this optical arrangement is to minimize/avoid the use of complex and expensive optical elements (light sources, wavelength analyzing tools, monocromators, or ellipsometers). The simple device is capable of following the evolution of the fermentation process without directly measuring the refractive index of the solution and without requiring an accurate description of the optical parameters of the BM. The measurement relies on determining the variation of the intensity ratio of the BM resonant peak Iux/Iuu asa function of the characteristics of the liquid. This allows an easy implementation together with common electronic appliances. Fig 5. a) Scheme of optical design of microfluidic system for label free liquid sensing: 1 microfluidic chip, 2 porous sculptured BM transductor, 3 polarizing beamsplitter cube, 4 polarized light, 5 and 6 the two components of the polarized light after interaction with the transductor, 7 photodiodes. b) Manufactured prototype 18 4 Summary Porous sculptured Bragg microcavities prepared by e-beam evaporation GLAD by the stacking of successive layers of SiO2 and TiO2 with a zig-zag structural configuration present tunable birefringence (and therefore optical activity) when infiltrated with condensed vapors or liquids so they can be used as transducers for label free vapor and liquid sensing. The outstanding microstructural characteristics of their individual layers, their porosity and the lateral aggregation of their nanocolumns (defined as fence-bundling) confers them a tunable optical activity and the capacity to act as transducers for label free vapor and liquid sensing. We show that the change in birefringence upon condensed vapor infiltration can be used to follow hysteresis loops usually observed in adsorption isotherms, or to monitor the evolution of solutions of varying refractive indices. Acknowledgments We thank Spanish Ministry of Economy, Industry and Competitiveness (AEI-FEDER project MAT2016-79866-R, CSIC 201560E055) for financial support. References 1. H. van Kranenburg, C. Lodder “Tailoring growth and local composition by oblique-incidence deposition: a review and new experimental data” Materials Science and Engineering Rll 295-354 (1994) [doi:10.1016/0927-796X(94)90021-3] 2. A. Lakhtakia, R. Messier, Sculptured thin films, nanoengineering morphology and optics, SPIE Press (2005) 19 3. M.M. Hawkeye, M.T. Taschuk, M.J. Brett, Introduction: Glancing Angle Deposition Technology, in Glancing Angle Deposition of Thin Films: Engineering the Nanoscale, John Wiley & Sons, Ltd, Chichester, UK. (2014) [DOI: 10.1002/9781118847510.ch1] 4. M.W. McCall, I.J. Hodgkinson, Q. Wu, Birefringent thin films and polarizing elements, Imperial College Press 82015) 5. M. Suzuki, “Practical applications of thin films nanostructured by shadowing growth” J. of Nanophotonics 073598 (2013) [doi:10.1117/1.JNP.7.073598] 6. A.R. González-Elipe, A. Barranco, A. Borras, A. Palmero, “Perspectives on oblique angle deposition of thin films: From fundamentals to devices” Progress in Materials Science 76, 59-153 (2016) [doi: 10.1016/j.pmatsci.2015.06.003] 7. R.J. Martin-Palma, f. Zhang, A. Lakhtakia, A. Cheng, J. Xu, C.G. Pantano, “Retardance of chalcogenide thin films grown by the oblique-angle-deposition technique”, Thin Solid Films 517, 5553 (2009) [doi:10.1016/j.tsf.2009.03.200] 8. P.D. McAtee, A. Lakhtakia, “Reflection and transmission of obliquely incident light by chiral sculptured thin films fabricated using asymmetric serial-bideposition technique”. J. Nanophotonics 11, 043502 (2017) [doi:10.1117/1.JNP.11.043502] 9. I Hodgkinson, Q. H. Wu, B. Knight, A. Lakhtakia, K. Robbie, Vacuum deposition of chiral sculptured thin films with high optical activity, Appl. Optics 39, 642 (2000) [doi: 10.1364/AO.39.000642] 10. V. Leontyev, M. Hawkeye, A. Kovalenko, M.J. Brett, “Omnidirectional reflection from nanocolumnar TiO2 films”, J. Appl. Phys. 112, 084317 (2012) [doi:10.1063/1.4759138] 11. K.M. Krause, M.J. Brett, Spatially graded nanostructured chiral films as tunable circular polarizers. Adv. Func. Mat. 18 , 3111 (2008) [doi:10.1002/adfm.200800685]. 20 12. M.E. Calvo, S. Colodrero, N. Hidalgo, G. Lozano, C. Lopez-Lopez, O. Sanchez-Sobrado, H. Mıguez, “Porous one dimensional photonic crystals: novel multifunctional materials for environmental and energy applications”, Energy Environ. Sci. 4, 4800 (2011) [doi:10.1039/C1EE02081A] 13. K. Szendrei , P. Ganter , O. Sànchez-Sobrado , R. Eger , A. Kuhn , B.V. Lotsch , “Touchless Optical Finger Motion Tracking Based on 2D Nanosheets with Giant Moisture Responsiveness”, Adv. Mater. 27, 6341–6348 (2015) [doi:10.1002/adma.201503463] 14. P. Ganter, K. Szendrei, B.V. Lotsch, “Towards the Nanosheet-Based Photonic Nose: Vapor Recognition and Trace Water Sensing with Antimony Phosphate Thin Film Devices” Adv. Mater. 28, 7436–7442 (2016) [doi:10.1002/adma.201601992] 15. A. Lakhtakia, M.W. McCall, J.A. Sherwin, Q.H. Wu, I.J. Hodgkinson, Sculptured thin film spectral holes for optical sensing of fluids, Optics Comm. 194 (2001) 33 [doi: 10.1016/S0030- 4018(01)01225-1] 16. Y.J. Liu, J.Shi, F. Zhang, H. Liang, J. Xu, A. Lakhtakia, “High-speed optical humidity sensors based on chiral sculptured thin films”. Sensors and Actuators B 156, 593 (2011) [doi:10.1016/j.snb.2011.02.003]. 17. L. Gonzalez-Garcia, G. Lozano, A. Barranco, H. Mıguez, A.R. Gonzalez-Elipe TiO2–SiO2 “Onedimensional photonic crystals of controlled porosity by glancing angle physical vapour deposition” J. Mater. Chem. 20, 6408–6412 (2010) [doi:10.1039/c0jm00680g] 18. M. Oliva-Ramirez, L. Gonzlez-García, J. Parra-Barranco, F. Yubero, A. Barranco, A.R. Gonzlez- Elipe, “Liquids Analysis with Optofluidic Bragg Microcavities” ACS Applied Materials and interfaces 5, 6743 (2013) [doi: 10.1021/am401685r] 19. M. Oliva-Ramirez, A. Barranco, M. Loffler, F. Yubero, A.R. González-Elipe, Optofluidic Modulation of Self-associated nanostructural units forming planar Bragg microcavities ACS Nano 10, 1256 (2016) [doi:10.1021/acsnano.5b06625] 21 20. I. Hodgkinson, Q.H. Wu Serial bideposition of anisotropic thin films with enhanced linear birefringence Appl. Opt. 38, 3621 (1999) [doi: 10.1364/AO.38.003621] 21. M.M. Hawkeye, M.J. Brett, Optimized Colorimetric Photonic-Crystal Humidity Sensor Fabricated Using Glancing Angle Deposition, Adv. Funct. Mater. 21, 3652 (2011) [doi: 10.1002/adfm.201100893] 22. S.J. Gregg, K.S.W. Sing, “Adsorption, Surface Area and Porosity”; Harcourt Brace and Co.: Orlando, FL, (1997). 23. M. Oliva-Ramírez et al. “Sensor, apparatus, and method for determining a concentration of solute in a solution”. Patent application number PCT/ES2016/070764; publication number: WO2017072388 A1 . 22 Manuel Oliva-Ramirez received his Ph.D. degree in Material Science from the University of Seville in 2016. His research was related to the design and fabrication of nanostructured porous thin films of different oxides for their implementation in optofluidic and photovoltaic devices. He is currently a postdoctoral researcher at the Leibniz-Institute for New Materials in Saarbruecken working on the development of conductive composites. Jorge Gil-Rostra received his degree in Chemistry from de University of Burgos in 1998 and his Ph. D. degree in Materials Science from the University of Seville in 2013. From 1998 to 2007 he has worked as responsible of Surface Treatments and Quality Manager at several Spanish companies. Since 2007 he works as a researcher at the Institute of Materials Science of Seville. His main research interest focus on magnetron sputtering deposition methods, mixed oxides thin films, luminescent materials, and optofluidic devices. Agustin R. González-Elipe is Research Professor of CSIC at the Institute of Materials Science of Seville. He is expert in thin film technology and plasma and vacuum deposition methods that, among other applications, he has used for the fabrication of photonic sensor films and optofluidic detection systems. He is author or co-author of more than 400 publications and author and editor of several books and reviews on thin film growth and applications. Carmen López-Santos is a postdoctoral researcher at the Institute of Materials Science of Seville (Spain). She received her BS degree in physics and her PhD degree from the University of Seville in 2004 and 2009, respectively. Her current research interests include surface modification by plasma and vacuum technologies for multifunctional applications of photoactive oxides, anisotropic nanostructures, biomaterials and protective coatings with antifreezing properties. Francisco Yubero received his Ph.D. degree in physics at the Universidad Autónoma de Madrid in 1993. Since 1996, he is a Research Scientist from CSIC at the Instituto de Ciencia de Materiales 23 de Sevilla. His research focuses on theoretical aspects related to surface electron spectroscopies and the practical implementation of functional thin films into technological devices. 24 Caption List Fig 1. Scheme of GLAD deposition process used in this work and developed thin film microstructure. (a) Plane of incidence and tilted thin film columnar growth (main optical axis of CTF are included). (b) Lateral columnar association in the direction s perpendicular to the plane of incidence. Fig 2. a) Cross section FIB-SEM micrographs (left p-axis, right s-axis) of the sculptured BM. b) Experimental transmittance T of polarized light at normal incidence of the BM around its defect peak c) Experimental measurements (left) and simulations (right) of transmittance after aligned Tuu (top) or crossed Tux (middle) polarizers of the BM for incident light normal to the surface and polarization parallel to s, p or φ=45º azimuthal sample orientations. The corresponding intensity ratios between these signals are also shown (bottom). Fig 3. Scheme of the set-up used for vapor sensing in reflection mode (a) and picture of the experimental arrangement (b): (1) light source, (2) (2’) polarizers, (3) substrate, (4) BM transducer, (5) transducer liquid level distance. Reflectance measurements through aligned Ruu (c) or crossed Rux (d) polarizers at several distances between the sculptured BM and the acetone level, as it approaches to the liquid level. (e) peak position (left axis), full width at half maximum (right axis) of the resonant peak and (f) intensity of the defect peak after crossed polarizers I(Rux) at several distances BM-acetone level, for the BM approaching (full symbols) or receding (hollow symbols) from the liquid level. 25 Fig 4. a) Experimental measurement of the reflectance after cross polarizers Rux and (b) intensity ratio of the reflected defect peaks I(Rux)/I(Ruu) in the porous BM transductor as the concentration of glucose in the solution is increases Fig 5. a) Scheme of optical design of microfluidic system for label free liquid sensing: 1 microfluidic chip, 2 porous sculptured BM transductor, 3 polarizing beamsplitter cube, 4 polarized light, 5 and 6 the two components of the polarized light after interaction with the transductor, 7 photodiodes. b) Manufactured prototype