plasmonic crystal gas sensor incorporating graphene oxide for ...
PLASMONIC CRYSTAL GAS SENSOR INCORPORATING GRAPHENE OXIDE FOR DETECTION OF VOLATILE ORGANIC COMPOUNDS Shawana Tabassum, Qiugu Wang, Wentai Wang, Seval Oren, Md. Azahar Ali, Ratnesh Kumar, and Liang Dong Department of Electrical and Computer Engineering, Iowa State University, Ames, Iowa, USA ABSTRACT This paper reports a gas sensor consisting of a thin layer of graphene oxide (GO) assembled on the top surface of mushroom plasmonic nanostructures. The plasmonic crystal structure is formed by an array of polymeric nanoposts with gold disks at the top and perforated nanoholes in a gold film at the bottom. The optical response of the plasmonic nanostructure is altered due to absorbing different concentrations of gas in presence of GO coating. The refractive index change of the GO layer with different thicknesses allows for selective detection of different gas species with the help of the principal component analysis-based pattern recognition algorithm. The present plasmonic nanostructure offers a promising approach to detect various volatile organic compounds.
INTRODUCTION Gas phase chemical detection has been extensively researched based on many sensing mechanisms, such as change in electrical conductance and mechanical resonance frequency shift of microscale structure, due to absorbing analyte gases by different sensing materials [1, 2]. Many organic semiconductors change their conductivity by several orders of magnitude when adsorbing a gas on their surface. But, there are some downsides associated with this type of sensors, such as phase shifts due to the segregation of additives doped with sensing materials, poisoning triggered by chemical reactions, and noise and resistance introduced by the contacts. Plasmonic nanostructures have been extensively used in liquid phase biological and chemical detection [3]. These structures provide an efficient way of controlling and manipulating light in the vicinity of metal surface below the diffraction limit through the excitations of surface plasmons. Their optical resonance wavelength can shift due to specific binding between receptor and target molecules. However, there are few reports on using plasmonic nanostructures for sensing of gas phase volatile organic compounds (VOCs). This may be attributed to lack of suitable sensing materials able to provide large changes in refractive index when responding to subtle concentration variations of a target gas [4]. Graphene and its derivatives such as graphene oxide (GO) has attracted much attention interests for many sensor applications, due to their exceptional optical, electrical, mechanical and chemical properties, and high surface-area-to-volume ratio [5]. This paper reports on the development of a GO coated plasmonic crystal as a gas sensor and elaborates the identification of different gas species using an array of such structures (Figure 2) [6, 7]. The present plasmonic structure is formed by an array of
polymer nanoposts with gold disks at the top and perforated nanoholes in a gold thin film at the bottom. By coating the surface of multiple identical plasmonic crystals with different thicknesses of GO layer, the effective refractive index of the GO layer on each plasmonic crystal will be differently modulated when responding to a specific gas. This allows identifying various gas species using the principal component analysis based pattern recognition algorithm. It is observed that the introduction of a gas to the device surface resulted in shifting plasmonic resonance mode excited at the Au disks. We studied optical responses of the plasmonic structures with four different thicknesses of GO to four different target analytes, including acetone, ethanol, chloroform and isopropyl alcohol. The sensitivities for acetone, ethanol, chloroform and isopropyl alcohol (IPA) are found as 0.012, 0.017, 0.02, and 0.011 nm/ppm, respectively. We also evaluated the ability of the plasmonic crystal array to discriminate between different VOCs.
DEVICE FABRICATION We utilized soft lithography based replica molding process to form a nanoposts array made of polydimethylsiloxane (PDMS) elastomer. The first step of fabrication involves forming a PDMS mold. A silicon master mold was silanized with (tridecafluoro-1, 1, 2, 2tetrahydrooctyl)-1-trichlorosilane (T2492-KG, United Chemical Technologies) in a desiccator under active vacuum for 20 min. Then, an h-PDMS precursor solution was prepared by mixing poly (7-8% vinylmethylsiloxane)- (dimethylsiloxane) (Gelest # VDT-731), (1, 3, 5, 7-tetravinyl-1, 3, 5, 7-tetramethylcyclotetrasiloxane) (Gelest # SIT7900.0), platinum catalyst Xylene (Gelest # SIP6831.2) and poly (25-30% methylhydro-siloxane)(dimethylsiloxane) (Gelest # HMS-301) at the weight ratio of 3.4: 0.1: 0.05: 1. Air bubbles were removed from the mixture in a degassing chamber for 10 min, followed by spin-coating of the mixture onto the silicon mold at 1000 rpm for 40 s and cured at 70oC for 10 min. Subsequently, an s-PDMS precursor solution was prepared by mixing Sylgard 184 (Dow Corning) and curing agent at the weight ratio of 10: 1 and degassed in a vacuum desiccator for 20 min. The s-PDMS mixture was then poured onto the top surface of h-PDMS and cured on a hotplate at 65oC for 2 hrs. After that, the PDMS slab containing a square array of nanoholes was peeled from the silicon mold. The next step was to use a UV curable polymer (ZIPCONE™ UA or ZPUA) to imprint the nanoposts from the PDMS mold to a silicon wafer. The ZPUA was dropped on the wafer at desired location and the soft PDMS was placed upside down on that location. Then the
wafer was exposed to ultraviolet light for 5 min at an intensity of 3.3 mW/cm2. The nanoposts had the period of 500 nm, the post diameter of 250 nm, and the post height of 210 nm. The entire surface of the device was then coated by a 5 nm thick titanium and a 50 nm thick gold layer. To interpret the specificity of this structure to different gas species, an array of four plasmonic sensors was fabricated on the same substrate. The four sensors were coated with different thickness of GO using a dropcoating method. A brief process flow is shown in Fig. 1. To prepare well-dispersed GO suspension solution, GO nanosheets were mixed in deionized water and sonicated for 90 min. This solution was then diluted subsequently to make four different concentrations. The surface of the device was made hydrophilic by using oxygen plasma treatment. Finally, the device surface was drop coated with GO solution and the device was left for an hour so that water was evaporated and GO coatings were formed on the device surface. Four thicknesses of GO coatings were obtained, including 16.3 nm, 18.0 nm, 18.6 nm, and 28.0 nm. The scanning electron microscopy (SEM) images for a fabricated sensor are shown in Fig. 2. The optical measurement setup includes a bifurcated fiber to illuminate the GO coated sample from a white light source (150 watt quartz halogen lamp) through a collimator, as well as to collect the reflected light into a spectrometer (USB-4000, Ocean Optics).
RESULTS AND DISCUSSION For the developed plasmonic nanostructure with a square lattice, the free space incident wavelength to excite a plasmon resonance mode is given by: SPP =
a i12 j12
d m d m
(1)
where ɛd and ɛm are the dielectric constants of the medium and Au, respectively, a is the lattice constant, and (i1, j1) corresponds to the order of SPPs. Figure 3 shows the plasmonic resonance shift as a function of surrounding refractive index. The higher the refractive index, the larger the resonance shift is observed. For acetone, ethanol, IPA, and chloroform, the sensor exhibits an index sensitivity of 435.3, 434.5, 429.9, and 411.8 nm/RI, respectively.
Figure 3: Refractive index sensitivity measurements
Figure 1: Fabrication flow for an array of four plasmonic crystal structures.
Figure 2: Scanning electron microscopy images for an array of plasmonic nanoposts with a GO coating.
The resonance spectra for all the four gases are illustrated in Fig. 4. At the original position, the resonances occur at an identical wavelength for both xand y-polarized modes due to the symmetry of the plasmonic device. It is evident that the shift in the resonance frequency is almost saturated at higher concentrations (317.6 ppm for chloroform, 630.9 ppm for isopropyl alcohol (IPA) and 823.1 ppm for ethanol) when the GO coating is 16.3 nm thick, as illustrated in Fig. 5. For acetone, the resonance is not saturated at this GO thickness which may be attributed to the structural difference among the gas molecules and hence the bonding with GO nanosheets. As the thickness of GO increases, more quantity of gas is absorbed and therefore, no saturation effect was observed with this thickness. The response time of the sensor to acetone, ethanol, IPA and chloroform was approximately ~120 s, ~150 s, ~165 s and ~115 s, respectively. The fast response may be attributed to the adsorption of molecules at the lowenergy binding sites, such as sp2 carbon domains, while the slow response was mainly caused by interactions between gas molecules with high-energy binding sites, such as vacancies, defects, and oxygen-containing functionalities. Once the sensor is exposed to a new analyte, the analyte molecules may be retained at some of these high-energy binding sites after the first cycle, eliminating the contribution of these sites to the sensor response in the following cycles. When GO is drop casted
Figure 4: Reflectance spectra of four plasmonic crystal gas sensors with different thickness of GO when exposed to gaseous (a) acetone, (b) chloroform, (c) IPA and (d) ethanol of different concentrations. on top of the plasmonic structure, the wavelength at which it resonates is found to be higher than the wavelength at which a bare structure resonates [8]. This is shown for acetone in Fig. 5a. Also as the surrounding medium of the sensor changes from air to a gas, a reflection dip emerges at a longer wavelength which agrees with equation (1).
PRINCIPAL COMPONENT ANALYSIS
Significant device-to-device variability of the GObased sensors is used as an advantage to realize detection selectivity. An array of GO devices that have sufficiently different properties can be considered as an analogy to a commercial electronic nose where the analyte adsorption events are converted into an array of electrical signals. The signals from all individual sensors with partially overlapping selectivity are collected, and then, a pattern recognition algorithm is used to generate a pattern that successfully separates all the gases from one another.
Figure 5: Resonance wavelength shifts of four plasmonic crystal gas sensors with different thickness of GO as a function of concentration of gaseous acetone, isopropyl alcohol, chloroform, and ethanol. We utilized a principal component analysis method which simplifies multidimensional datasets without crucial loss of information [9] to identify different gas species. For our experiment the data matrix contains the response of each sensor with different GO thickness to a certain gas and concentration. This matrix is reduced to two principle components (PC1, PC2). The variance of PC1 (95.75%) and PC2 (2.94%) is above 98% and therefore, these components already contain significant information to illustrate the data in two dimensions (Figure 6). The figure illustrates a clear separation between the clusters indicating individual gases without overlap. Thus, this algorithm allows the identification of analytes with a set of four sensors [10].
Figure 6: Pattern analyses based on principal component analysis using four different sensors.
CONCLUSION In summary, we have demonstrated a GO coated plasmonic structure which is sensitive to subtle changes in concentration of gas by providing wavelength shift of its plasmonic resonance mode. Due to the changes in the refractive index in presence of different analytes, a red shift was observed for the sensor. Principal component analysis algorithm has been adopted to distinguish between gas species.
ACKNOWLEDGEMENT This work was supported by US National Science Foundation under grants CCF-1331390 and DBI1331390, US Agriculture Department’s NIFA under grant 2013-68004-20374, Iowa Corn Promotion Board, and
Iowa State University’s Plant Sciences Institute Faculty Scholar Program.
REFERENCES [1] G. Chen, T. M. Paronyan, and A. R. Harutyunyan, “Sub-ppt gas detection with pristine graphene”, Appl. Phys., Vol. 101, pp. 053119, 2012. [2] C. Hagleitner, A. Hierlemann, D. Lange, A. Kummer, N. Kerness, O. Brand, and H. Baltes, “smart single-chip gas sensor microsystem”, Nature, Vol. 404, pp. 293,2001. [3] J. N. Anker, W. P. Hall, O. Lyandres, N. C. Shah, J. Zhao, and R. P. Van Duyne, “Biosensing with plasmonic nanosensors”, Nature Materials, Vol. 7, pp. 442, 2008. [4] J. M. Bingham, J. N. Anker, L. E. Kreno, and R. P. Van Duyne, “Gas Sensing with High ResolutionLocalized Surface Plasmon Resonance Spectroscopy”, J. American Chemical Society, Vol. 132, pp. 17358, 2010. [5] A. Lipatov, A. Varezhnikov, P. Wilson, V. Sysoev, A. Kolmakov and A. Sinitskii, “Highly selective gas sensor arrays based on thermally reduced graphene oxide”, Nanoscale, Vol. 5, pp. 5426, 2013. [6] S. Yoo, X. Li, Y. Wu, W. Liu, X. Wang and W. Yi, “Ammonia Gas Detection by Tannic Acid Functionalized and Reduced Graphene Oxide at Room Temperature”, J. Nanomaterials, Vol. 2014, pp. 497384, 2014. [7] R. Zhang, V. Alecrim, M. Hummelgard, B. Andres, S. Forsberg, M. Andersson and H. Olin, “Thermally reduced kaolin-graphene oxide nanocomposites for gas sensing”, Sci. Reports, Vol. 5, pp. 7676, 2015. [8] R. Verma, B. D. Gupta and R. Jha, “Sensitivity enhancement of a surface plasmon resonance based biomolecules sensor using graphene and silicon layers”, Sensors and actuators B. Chem, Vol. 160, pp. 623-631, 2011. [9] A. Ziyatdinov, A. Chaudry, K. Persaud, P. Caminal and A. Perera, “Common Principal Component Analysis For Drift Compensation Of Gas Sensor Array Data”, AIP Conference proceedings, Vol. 566, pp. 1137, 2009.+ [10] A. Zopfl, M. M. Lemberger, M. Konig, G. Ruhl, F. M. Matysik and T. Hirsch, “Reduced graphene oxide and graphene composite materials for improved gas sensing at low temperature”, Vol. 173, pp. 403, 2014.
CONTACT *L. Dong, tel: +1-515-294-0388; [email protected]
INTRODUCTION Gas phase chemical detection has been extensively researched based on many sensing mechanisms, such as change in electrical conductance and mechanical resonance frequency shift of microscale structure, due to absorbing analyte gases by different sensing materials [1, 2]. Many organic semiconductors change their conductivity by several orders of magnitude when adsorbing a gas on their surface. But, there are some downsides associated with this type of sensors, such as phase shifts due to the segregation of additives doped with sensing materials, poisoning triggered by chemical reactions, and noise and resistance introduced by the contacts. Plasmonic nanostructures have been extensively used in liquid phase biological and chemical detection [3]. These structures provide an efficient way of controlling and manipulating light in the vicinity of metal surface below the diffraction limit through the excitations of surface plasmons. Their optical resonance wavelength can shift due to specific binding between receptor and target molecules. However, there are few reports on using plasmonic nanostructures for sensing of gas phase volatile organic compounds (VOCs). This may be attributed to lack of suitable sensing materials able to provide large changes in refractive index when responding to subtle concentration variations of a target gas [4]. Graphene and its derivatives such as graphene oxide (GO) has attracted much attention interests for many sensor applications, due to their exceptional optical, electrical, mechanical and chemical properties, and high surface-area-to-volume ratio [5]. This paper reports on the development of a GO coated plasmonic crystal as a gas sensor and elaborates the identification of different gas species using an array of such structures (Figure 2) [6, 7]. The present plasmonic structure is formed by an array of
polymer nanoposts with gold disks at the top and perforated nanoholes in a gold thin film at the bottom. By coating the surface of multiple identical plasmonic crystals with different thicknesses of GO layer, the effective refractive index of the GO layer on each plasmonic crystal will be differently modulated when responding to a specific gas. This allows identifying various gas species using the principal component analysis based pattern recognition algorithm. It is observed that the introduction of a gas to the device surface resulted in shifting plasmonic resonance mode excited at the Au disks. We studied optical responses of the plasmonic structures with four different thicknesses of GO to four different target analytes, including acetone, ethanol, chloroform and isopropyl alcohol. The sensitivities for acetone, ethanol, chloroform and isopropyl alcohol (IPA) are found as 0.012, 0.017, 0.02, and 0.011 nm/ppm, respectively. We also evaluated the ability of the plasmonic crystal array to discriminate between different VOCs.
DEVICE FABRICATION We utilized soft lithography based replica molding process to form a nanoposts array made of polydimethylsiloxane (PDMS) elastomer. The first step of fabrication involves forming a PDMS mold. A silicon master mold was silanized with (tridecafluoro-1, 1, 2, 2tetrahydrooctyl)-1-trichlorosilane (T2492-KG, United Chemical Technologies) in a desiccator under active vacuum for 20 min. Then, an h-PDMS precursor solution was prepared by mixing poly (7-8% vinylmethylsiloxane)- (dimethylsiloxane) (Gelest # VDT-731), (1, 3, 5, 7-tetravinyl-1, 3, 5, 7-tetramethylcyclotetrasiloxane) (Gelest # SIT7900.0), platinum catalyst Xylene (Gelest # SIP6831.2) and poly (25-30% methylhydro-siloxane)(dimethylsiloxane) (Gelest # HMS-301) at the weight ratio of 3.4: 0.1: 0.05: 1. Air bubbles were removed from the mixture in a degassing chamber for 10 min, followed by spin-coating of the mixture onto the silicon mold at 1000 rpm for 40 s and cured at 70oC for 10 min. Subsequently, an s-PDMS precursor solution was prepared by mixing Sylgard 184 (Dow Corning) and curing agent at the weight ratio of 10: 1 and degassed in a vacuum desiccator for 20 min. The s-PDMS mixture was then poured onto the top surface of h-PDMS and cured on a hotplate at 65oC for 2 hrs. After that, the PDMS slab containing a square array of nanoholes was peeled from the silicon mold. The next step was to use a UV curable polymer (ZIPCONE™ UA or ZPUA) to imprint the nanoposts from the PDMS mold to a silicon wafer. The ZPUA was dropped on the wafer at desired location and the soft PDMS was placed upside down on that location. Then the
wafer was exposed to ultraviolet light for 5 min at an intensity of 3.3 mW/cm2. The nanoposts had the period of 500 nm, the post diameter of 250 nm, and the post height of 210 nm. The entire surface of the device was then coated by a 5 nm thick titanium and a 50 nm thick gold layer. To interpret the specificity of this structure to different gas species, an array of four plasmonic sensors was fabricated on the same substrate. The four sensors were coated with different thickness of GO using a dropcoating method. A brief process flow is shown in Fig. 1. To prepare well-dispersed GO suspension solution, GO nanosheets were mixed in deionized water and sonicated for 90 min. This solution was then diluted subsequently to make four different concentrations. The surface of the device was made hydrophilic by using oxygen plasma treatment. Finally, the device surface was drop coated with GO solution and the device was left for an hour so that water was evaporated and GO coatings were formed on the device surface. Four thicknesses of GO coatings were obtained, including 16.3 nm, 18.0 nm, 18.6 nm, and 28.0 nm. The scanning electron microscopy (SEM) images for a fabricated sensor are shown in Fig. 2. The optical measurement setup includes a bifurcated fiber to illuminate the GO coated sample from a white light source (150 watt quartz halogen lamp) through a collimator, as well as to collect the reflected light into a spectrometer (USB-4000, Ocean Optics).
RESULTS AND DISCUSSION For the developed plasmonic nanostructure with a square lattice, the free space incident wavelength to excite a plasmon resonance mode is given by: SPP =
a i12 j12
d m d m
(1)
where ɛd and ɛm are the dielectric constants of the medium and Au, respectively, a is the lattice constant, and (i1, j1) corresponds to the order of SPPs. Figure 3 shows the plasmonic resonance shift as a function of surrounding refractive index. The higher the refractive index, the larger the resonance shift is observed. For acetone, ethanol, IPA, and chloroform, the sensor exhibits an index sensitivity of 435.3, 434.5, 429.9, and 411.8 nm/RI, respectively.
Figure 3: Refractive index sensitivity measurements
Figure 1: Fabrication flow for an array of four plasmonic crystal structures.
Figure 2: Scanning electron microscopy images for an array of plasmonic nanoposts with a GO coating.
The resonance spectra for all the four gases are illustrated in Fig. 4. At the original position, the resonances occur at an identical wavelength for both xand y-polarized modes due to the symmetry of the plasmonic device. It is evident that the shift in the resonance frequency is almost saturated at higher concentrations (317.6 ppm for chloroform, 630.9 ppm for isopropyl alcohol (IPA) and 823.1 ppm for ethanol) when the GO coating is 16.3 nm thick, as illustrated in Fig. 5. For acetone, the resonance is not saturated at this GO thickness which may be attributed to the structural difference among the gas molecules and hence the bonding with GO nanosheets. As the thickness of GO increases, more quantity of gas is absorbed and therefore, no saturation effect was observed with this thickness. The response time of the sensor to acetone, ethanol, IPA and chloroform was approximately ~120 s, ~150 s, ~165 s and ~115 s, respectively. The fast response may be attributed to the adsorption of molecules at the lowenergy binding sites, such as sp2 carbon domains, while the slow response was mainly caused by interactions between gas molecules with high-energy binding sites, such as vacancies, defects, and oxygen-containing functionalities. Once the sensor is exposed to a new analyte, the analyte molecules may be retained at some of these high-energy binding sites after the first cycle, eliminating the contribution of these sites to the sensor response in the following cycles. When GO is drop casted
Figure 4: Reflectance spectra of four plasmonic crystal gas sensors with different thickness of GO when exposed to gaseous (a) acetone, (b) chloroform, (c) IPA and (d) ethanol of different concentrations. on top of the plasmonic structure, the wavelength at which it resonates is found to be higher than the wavelength at which a bare structure resonates [8]. This is shown for acetone in Fig. 5a. Also as the surrounding medium of the sensor changes from air to a gas, a reflection dip emerges at a longer wavelength which agrees with equation (1).
PRINCIPAL COMPONENT ANALYSIS
Significant device-to-device variability of the GObased sensors is used as an advantage to realize detection selectivity. An array of GO devices that have sufficiently different properties can be considered as an analogy to a commercial electronic nose where the analyte adsorption events are converted into an array of electrical signals. The signals from all individual sensors with partially overlapping selectivity are collected, and then, a pattern recognition algorithm is used to generate a pattern that successfully separates all the gases from one another.
Figure 5: Resonance wavelength shifts of four plasmonic crystal gas sensors with different thickness of GO as a function of concentration of gaseous acetone, isopropyl alcohol, chloroform, and ethanol. We utilized a principal component analysis method which simplifies multidimensional datasets without crucial loss of information [9] to identify different gas species. For our experiment the data matrix contains the response of each sensor with different GO thickness to a certain gas and concentration. This matrix is reduced to two principle components (PC1, PC2). The variance of PC1 (95.75%) and PC2 (2.94%) is above 98% and therefore, these components already contain significant information to illustrate the data in two dimensions (Figure 6). The figure illustrates a clear separation between the clusters indicating individual gases without overlap. Thus, this algorithm allows the identification of analytes with a set of four sensors [10].
Figure 6: Pattern analyses based on principal component analysis using four different sensors.
CONCLUSION In summary, we have demonstrated a GO coated plasmonic structure which is sensitive to subtle changes in concentration of gas by providing wavelength shift of its plasmonic resonance mode. Due to the changes in the refractive index in presence of different analytes, a red shift was observed for the sensor. Principal component analysis algorithm has been adopted to distinguish between gas species.
ACKNOWLEDGEMENT This work was supported by US National Science Foundation under grants CCF-1331390 and DBI1331390, US Agriculture Department’s NIFA under grant 2013-68004-20374, Iowa Corn Promotion Board, and
Iowa State University’s Plant Sciences Institute Faculty Scholar Program.
REFERENCES [1] G. Chen, T. M. Paronyan, and A. R. Harutyunyan, “Sub-ppt gas detection with pristine graphene”, Appl. Phys., Vol. 101, pp. 053119, 2012. [2] C. Hagleitner, A. Hierlemann, D. Lange, A. Kummer, N. Kerness, O. Brand, and H. Baltes, “smart single-chip gas sensor microsystem”, Nature, Vol. 404, pp. 293,2001. [3] J. N. Anker, W. P. Hall, O. Lyandres, N. C. Shah, J. Zhao, and R. P. Van Duyne, “Biosensing with plasmonic nanosensors”, Nature Materials, Vol. 7, pp. 442, 2008. [4] J. M. Bingham, J. N. Anker, L. E. Kreno, and R. P. Van Duyne, “Gas Sensing with High ResolutionLocalized Surface Plasmon Resonance Spectroscopy”, J. American Chemical Society, Vol. 132, pp. 17358, 2010. [5] A. Lipatov, A. Varezhnikov, P. Wilson, V. Sysoev, A. Kolmakov and A. Sinitskii, “Highly selective gas sensor arrays based on thermally reduced graphene oxide”, Nanoscale, Vol. 5, pp. 5426, 2013. [6] S. Yoo, X. Li, Y. Wu, W. Liu, X. Wang and W. Yi, “Ammonia Gas Detection by Tannic Acid Functionalized and Reduced Graphene Oxide at Room Temperature”, J. Nanomaterials, Vol. 2014, pp. 497384, 2014. [7] R. Zhang, V. Alecrim, M. Hummelgard, B. Andres, S. Forsberg, M. Andersson and H. Olin, “Thermally reduced kaolin-graphene oxide nanocomposites for gas sensing”, Sci. Reports, Vol. 5, pp. 7676, 2015. [8] R. Verma, B. D. Gupta and R. Jha, “Sensitivity enhancement of a surface plasmon resonance based biomolecules sensor using graphene and silicon layers”, Sensors and actuators B. Chem, Vol. 160, pp. 623-631, 2011. [9] A. Ziyatdinov, A. Chaudry, K. Persaud, P. Caminal and A. Perera, “Common Principal Component Analysis For Drift Compensation Of Gas Sensor Array Data”, AIP Conference proceedings, Vol. 566, pp. 1137, 2009.+ [10] A. Zopfl, M. M. Lemberger, M. Konig, G. Ruhl, F. M. Matysik and T. Hirsch, “Reduced graphene oxide and graphene composite materials for improved gas sensing at low temperature”, Vol. 173, pp. 403, 2014.
CONTACT *L. Dong, tel: +1-515-294-0388; [email protected]
