- Email: [email protected]
Surface acoustic wave based H2S gas sensors incorporating sensitive layers of single wall carbon nanotubes decorated with Cu nanoparticles
Accepted Manuscript Title: Surface Acoustic Wave Based H2S Gas Sensors Incorporating Sensitive Layers of Single Wall Carbon Nanotubes Decorated with Cu Nanoparticles Author: Mohsen Asad Mohammad Hossein Sheikhi PII: DOI: Reference:
S0925-4005(14)00281-0 http://dx.doi.org/doi:10.1016/j.snb.2014.03.024 SNB 16664
To appear in:
Sensors and Actuators B
Received date: Revised date: Accepted date:
16-6-2013 5-3-2014 6-3-2014
Please cite this article as: M. Asad, M.H. Sheikhi, Surface Acoustic Wave Based H2S Gas Sensors Incorporating Sensitive Layers of Single Wall Carbon Nanotubes Decorated with Cu Nanoparticles, Sensors and Actuators B: Chemical (2014), http://dx.doi.org/10.1016/j.snb.2014.03.024 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Surface Acoustic Wave Based H2S Gas Sensors Incorporating Sensitive Layers of Single Wall Carbon Nanotubes Decorated with Cu Nanoparticles
1
ip t
Mohsen Asad1,2,*, Mohammad Hossein Sheikhi1,2 School of Electrical and Computer Engineering, Shiraz University, Shiraz, Iran Nanotechnology Research Institute, Shiraz University, Shiraz, Iran
cr
2
us
*Corresponding author Email: [email protected]
Abstract
an
This paper presents a surface acoustic wave (SAW) based H2S gas sensor with excellent selectivity and recovery/response time developed using single wall carbon nanotube decorated
M
with copper nanoparticles (Cu NP-SWCNT). A thin film of Cu NP-SWCNT was deposited onto
d
a lithium niobate (LiNbO3) piezoelectric substrate via a drop-casting technique. Sensing of H2S
te
gas was carried out by measuring acoustoelectric perturbation of SAWs traveling along the LiNbO3 piezoelectric substrate and Cu NP-SWCNT sensing layer. H2S gas of concentrations as
Ac ce p
small as 5 ppm could be readily measured. The effect of temperature on the SAW sensor response was also investigated for a range of temperatures from 70 to 200°C. The optimum operating temperature was 175°C, in which, a relatively rapid response (7 s) and recovery time (9 s) was recorded. The selectivity of the proposed Cu NP-SWCNT gas sensor was examined by assessing the sensor response upon exposure to hydrogen, acetone, ethanol, and H2S gas species in air background and a large selectivity towards H2S gas was observed. Keywords: surface acoustic wave; LiNbO3; copper nanoparticles; carbon nanotubes; gas sensor; H2S 1 Page 1 of 39
1. Introduction Surface acoustic wave (SAW) based sensors operate based on the wave modulation as a result of exposure to changes in physical and chemical phenomena [1]. SAW devices were first
ip t
developed nearly five decades ago [2]. Later in early 80s, the SAW transduction effect was employed for sensing applications [3]. Since then, SAW based sensors have been used for a
cr
variety of sensing applications including humidity [4,5], temperature [6], atmospheric pressure
us
[7], blood pressure [8], flow [9], and strain [10] as well as chemical/biochemical agents in gas and liquid media [11-13].
an
In gas media operation, the stress-free boundary conditions that exist on the SAW crystal surface, together with the right choice of the crystal cut, promote the propagation of surface
M
acoustic mode, with energy mostly confined on the surface of the substrate. One of the most
d
commonly used SAW propagation modes is Rayleigh mode, which has been named after Lord
te
Rayleigh who discovered this near surface traveling mode for the first time [14]. The main characteristic of Rayleigh SAWs are their relatively small penetration into the substrate, due to
Ac ce p
near surface energy concentration, ellipsoid displacement path, and mechanical wave traveling speed which is in the order of 10-5 times smaller than that of light [15]. The high energy density confined near the SAW surface plays an important role in obtaining high sensitivity for SAW based sensors. Some of the commonly used piezoelectric crystal types and cuts that allow for the generation and propagation of Rayleigh SAWs are ST cut quartz, YZ LiNbO3, and 128º YX LiNbO3 [16,17]. By coating a thin film of a material, which is sensitive to a certain type of gas, on the piezoelectric substrate the sensing capability of the SAW device to that specific gas is promoted. The applications of nanomaterials as sensitive layers offer distinct advantages. When the 2 Page 2 of 39
nanomaterial coating is exposed to gas or chemical vapors, the gas molecules can interact with the whole bulk of the layer rather than just its surface. As such, the sensitivity is dramatically enhanced. Promising SAW sensors have already been reported for a variety of gas species
ip t
including N2O [18], H2S [19], CO2 [20], and H2 [21]. The efficiency of such sensors is determined by the rate of interaction between the gas molecules and the sensitive layer.
cr
The incorporation of carbon nanotubes (CNTs) as sensitive materials was first reported by
us
Kong in 2000 [22], and since then single-wall carbon nanotubes (SWCNTs) and multi-wall carbon nanotubes (MWCNTs) have been extensively investigated [23, 24]. In order to enhance
an
the selectivity and sensitivity of CNTs towards a variety of gases, different CNT surface modification strategies such as polymer coatings [25], CNT-based nano-composites [26], CNT
M
functionalization using polymer [27], or certain functional groups [28], as well as physical and
d
chemical modifications using metal nanoclusters [29] have been reported.
te
Amongst the aforementioned methods, CNT functionalization with metals is particularly attractive due to the availability of simple, highly efficient and low cost synthesis processes. Li et
Ac ce p
al. reported a composite of Pd-functionalized MWCNT synthesized by a chemical reduction method, which exhibited an efficient and reversible response towards CH4 [30]. Kumar et al. fabricated MWCNT-based H2 sensors functionalized with Pd and Pt using aqueous solutions of PdCl2 and H2PtCl6 [31, 32]. Penza et al. demonstrated NO2 and NH3 sensors based on MWCNTs coated with Pt and Au nanoclusters [33]. Espinosa et al. reported MWCNTs with Ag and Au nanoclusters via electron beam evaporation [34] Electrodeposition method was reported by Sadek et al. for forming gold nanoparticles on MWCNTS to sense H2 [35]. Star et al. designed a gas sensor by electroplating Pt, Pd, Au, and Rh metals on SWCNTs substrate to detect CO, H2S, H2, CH4, NO2, and NH3 [36]. They also employed thermal and electron-beam evaporation of 3 Page 3 of 39
various metals (Mg, Al, Ti, V, Cr, Mn, Fe, Co, Ni, Zn, Mo, Rh, Pd, Sn, W, Pt, Au, and Pb) onto the CNT surface for sensing CH4, H2, CO, and H2S gases. Lu et al. fabricated a sensor for sensing HCN, HCl, NO2, Cl2, acetone, and benzene gases which were composed of pristine and
ip t
Pd- and Au-decorated SWCNTs [37]. However, there is no report on the incorporation of copper (Cu) nanoparticles (NPs) for the decoration of CNTs for sensing different gas species. This is
cr
despite the fact that Cu as a catalyst has been shown to enhance the sensing properties of
us
semiconducting gas sensors based on SnO2 [38], ZnO [39] WO3 [40].
In the current study, we decorated functionalized SWCNTs with Cu NPs to improve the
an
selectivity of SWCNTs towards H2S gas species. We deposited these Cu NPs decorated SWCNTs (Cu-SWCNT) onto a SAW device made of a LiNbO3 piezoelectric substrate to form
2. Theoretical Design
Ac ce p
2.1. Gas Sensitive Layer
te
d
sensitive layer and the fabricated sensors.
M
the SAW sensor. A set of characterization techniques was used for assessing the properties of the
The SAW propagation in a piezoelectric substrate creates mechanical displacements and accompanying electrical potentials as a result of the coupling between the mechanical and electromagnetic fields. Subsequently, changes in the electrical and mechanical properties of any sensitive layer, which is placed on this active area, causes changes in the velocity and magnitude of the propagating waves. If the sensitive layer mass is assumed to be negligible, the acoustoelectric effect predicts a relative change in the velocity and consequently the change in relative frequency as a function of surface electrical conductivity according to [15]: (∆f/f0)σ ≈ (∆υ/υ0)σ ≈ −(K2/2)[(σs2 / (σs2 + υ02Cs2)]
(1)
4 Page 4 of 39
where, K2 is the electromechanical coupling coefficient, σs=σ×h is the surface conductivity (in which σ is the conductivity of the bulk of the sensitive layer and h is the sensitive layer thickness), Cs=εp+εo is the sum of the dielectric permittivity factor in the substrate (εo) and its
ip t
surrounding (εp), vo and fo are the propagation velocities of the SAWs and the center frequency of the device before perturbation. Obviously any change of the operational frequency of the sensor
cr
depends on the electromechanical coupling factor K2, change in the surface conductivity of the
us
layer σs, and the permittivity of the environment. For obtaining a highly sensitive SAW based gas sensor the parameters of the sensitive layer should be tuned in a way that maximum frequency
an
change is obtained as a result of minimum change in the conductivity or permittivity of the sensitive layer. According to the discussion presented in [18], the maximum sensitivity is
M
obtained when σs = υ0Cs. Overall if 0.01υ02Cs2 < σs2 < 100υ02Cs2 [41] an acceptable sensitivity is
d
achieved. In the present study, we used a SWCNT and Cu composite as the semiconductor
te
material on the active area of the SAW device for the detection of H2S. In order to design the SAW sensor, we used a 64º YX cut LiNbO3 piezoelectric substrate, the
Ac ce p
physical characteristics of which are presented in Table 1. Based on equation (1), the acoustoelectric response range is proportionate to K2. 64º YX LiNbO3 was chosen as its K2 is quite large, resulting in a large change in the operational frequency as a result of any electromechanical perturbation.
Table 1.
5 Page 5 of 39
2.2. Design of SAW Sensor’s Geometry The schematic of such a SAW sensor is shown in Figure 1. An important issue in designing of SAW devices is regarding the geometry of the inter-digital transducers (IDTs), where an
ip t
impedance matching between the applied signal source and the input of the SAW device should be established using the geometry of the IDTs [15]. The maximum SAW generation is at the
cr
wavelength equal to the spatial period of the IDTs.
us
Fig. 1.
IDTs are metal electrodes (made of aluminum - Al) which are built on a piezoelectric
an
substrate and for applying and receiving the electric signals onto and from the piezoelectric substrates. Using COSMOL 4.3 software package at two dimensional planes, the mode of
M
piezoelectric device in order to simulate wave penetration in SAW system was chosen. Stress-
d
free conditions for the upper surface of piezoelectric substrate were applied. For the simulation,
te
it was assumed that periodic boundary condition exists on both sides of the piezoelectric substrate. Periodic boundary condition for the electric potential was used along both vertical
Ac ce p
boundaries of geometry. The displacements boundary condition was also defined similarly. The thickness of Al IDTs that was used in simulation was 200 nm. The equation for the center of the operation frequency (f0) was f0=v0/d, where v0 was the propagation velocity and d was the periodicity of the IDTs. Considering the fact that wave velocity of the 64º YX cut LiNbO3 is known (4742 m/s), for a SAW operating at 100 MHz frequency, we can estimate a period length of 48 μm. By considering these parameters, the width of each electrode was determined to be approximately 12 µm as illustrated in Figure 2. Fig. 2.
6 Page 6 of 39
Based on the Eigenfrequency analysis, approximate operational frequency of 104 MHz was selected and the surface displacement associated with this frequency in two- and threedimensions are shown in Figures 3(a) and 3(b), respectively. As a result of the simulation, it was
ip t
observed that mechanical displacement of the piezoelectric substrate at the highest point effectively penetrates 20 μm from surface downward into the substrate. In addition, the
cr
simulation suggests that the maximum of surface displacement is 1.9633×10-9 m and this
us
displacement is periodically spread over the piezoelectric substrate along the horizontal axis. It’s also clear that the maximum displacements occur at the surface of the piezoelectric substrate and
an
Rayleigh waves are confined near the surface.
Fig 3
M
To calculate the effect of the electric field, we used piezoelectric material model in
d
COMSOL 4.3 and the 64º YX cut LiNbO3 substrate was solved using the coupled equations
te
S=sET + dE and D=dtrT + εTE, where T is the stress tensor, S is the strain tensor, D is the electric
Ac ce p
charge density displacement vector, E is the electric field vector, sE, d, and εT are elastic compliance, piezoelectric module (the ratio of strain to applied field or charge density to applied mechanical stress) and the dielectric matrices, respectively (here tr stands for the transposition of the d matrix). From the coupled equations, E can be calculated and the surface voltage may be assessed as V ≈ E×t (in which t is the penetration depth of the electric field). As σs is proportional to electric field, it is subsequently proportional to the calculated voltage. Figures 4(a) and 4(b) show the variation of surface electrical potential as a function of the applied voltages at the electrodes. Fig. 4. 7 Page 7 of 39
As the generated voltage is relatively large, the simulations reveal that the conductivity change can be the dominant effect in comparison to the displacement effect on the active area of the SAW device. As a result, equation (1) is a valid approximation for assessing the frequency shift
ip t
of the sensor.
cr
3. Experimental
us
3.1. Fabrication of Sensitive Substance
CNTs (outer diameter of ~10 nm, length of ~1.5 μm, BET surface of 250-300 m2/g, purity of
an
90%, KNT-GP-BT) was purchased from Parsis Co. (Tehran, Iran – for the proof on the semiconducting properties of the CNTs refer to the Supporting Information). CuSO4 was
M
purchased from Merck Chemical Corporation (Germany) and used as received without further
te
TKA Co. Germany).
d
purification. Deionized (DI) water was prepared using an ultra-pure water system (Smart-2-Pure,
To form the Cu decorated SWCNT sensitive layer, 100 mg SWCNTs were mixed with 40 ml
Ac ce p
of HNO3 and H2SO4 (1:3) mixture at 110ºC. The resulting solution was stirred for 70 min and then sonicated for 5 h. The resulting solution was dried in vacuum oven for 6 h. Cu NPs were decorated onto the SWCNTs using CuSO4 aqueous solution via a chemical reduction process which has been fully described in [42]. 20 mg of functionalized SWCNTs were mixed with 10 mg CuSO4 in 80 ml DI water. The resulting suspension was slowly and steadily stirred at 90°C. Metal Cu, from the chemical point of view, shows different behavior in reduction process than other metals, and can act as one-valence or two-valence metal which process conditions that directly determines the chemical paths [43]. Cu ions as Cu+1 and Cu+2 are obtained after the reduction process and by substituting hydrogen ions of carboxylic group (8 Page 8 of 39
COOH), which are located on the functionalized SWCNT. Sulfuric acid is expected to be one of the main products of such reaction which is separated from the final solution by washing. Subsequently, the resulting material was separated from the solution and was desiccated at 50°C.
ip t
We characterized the electrical conductivity of the sensitive films. For this purpose, we used IDTs on substrates with SiO2 insulating layers. The pristine SWCNTs and Cu NPs-SWCNTs
cr
were dispersed in DI water and then they were spin coated on the IDTs. The conductivity for the
us
pristine SWCNTs and Cu NPs decorated SWCNTs were 2.11 and 10.41 mS/cm2, respectively, at room temperature. As a result, an increase in the conductivity of sensitive film after the
an
decoration with Cu is obvious. The sensitive layer conductivity, after the Cu decoration process, became closer to the optimum value of σs = υ0Cs considering the LiNbO3 relative permittivity of
te
3.2. Fabrication of SAW Device
d
M
~45.
In the current work, the SAW-based sensor is based on a piezoelectric crystal, two sets of
Ac ce p
IDTs and the Cu-SWCNT sensitive layer. We used a 1×2 cm2 optically polished 64º YX cut LiNbO3 substrate of 0.5 cm thickness as the piezoelectric crystal. 50 finger pairs were used in each of the receiver and transmitter IDTs. Electrode of 12 µm width and spacing were patterned that allowed the propagation of waves with the fundamental wavelength of 48 µm. In the fabrication process, the wafers were first cleaned by immersing them into DI water containing a small amount of ammonia under ultrasonication. Then, the wafer was rinsed in ethylene alcohol and dried in vacuum. Afterward, Al coating process (using Blazer e-beam evaporation system) with 200 nm thickness was carried out. Then processes followed by lift-off to complete the IDTs fabrication. 9 Page 9 of 39
To form the nanostructured SWNTs onto the active area of the SAW device, the acoustic propagation path of the SAW delay line was drop-coated with 1.5 wt% poly diallyl dimethyl ammonium chloride (PDDA) solution, as surfactant, for 10 min followed by rinsing with DI
ip t
water and drying. The PDDA treatment introduced positive charges on the LiNbO3 surface which helped to bind the negatively charged carboxyl (–COO–) ends of the SWNTs to the substrate.
cr
Then, Cu-SWNTs were drop-coated onto the LiNbO3 substrate and the resulting film was
us
allowed to completely dry at 60°C temperature overnight.
an
3.3. Characterization
The morphology of the gas sensitive film was revealed by a field emission scanning electron
M
microscopy (FE-SEM, Hitachi S4160) and the elemental structure was analyzed using Zeiss
d
EVO40 SEM equipped with an energy dispersive X-ray detector (EDX, EDAX Oxford, UK).
te
FTIR spectra of the nanomaterials were recorded using a Nicolet Magna IR 550 spectrometer. The gas sensing evaluation was carried out using a set-up as shown in Figure 5. Experiments
Ac ce p
were carried out at room temperature (25°C) and different H2S gas concentrations of 5, 10, 20, 30, 40, 50, 100, and 200 ppm and the frequency shifts were measured using a HP8751 Network Analyzer. We controlled the concentration of the H2S in the chamber by mass flow controller (MFC) against the background moisture free argon gas. Total gas flow at the chamber was maintained at 500 sccm. For investigation on the effect of temperature on the sensor response, the SAW sensors were mounted on a micro-heater inside the chamber. A DC power supply and a PT100 temperature sensor were employed to control the micro-heater temperature. The input gas was humidified by means of a controlled evaporator mixer to investigate the influence of humid
10 Page 10 of 39
environment on the sensor’s response. The input gas in delivered to chamber after flow to a humidity sensor. Fig 5.
ip t
4. Results and Discussions 4.1. Chemical Structure Analysis by FTIR
cr
Figure 6 represents the FTIR spectrum of Cu decorated SWCNT and, as a reference,
us
carboxylic-functionalized SWCNT (f-SWCNT obtained using sulfuric acid and nitric acid treatment as was described in section 3).
an
Spectrum for f-SWCNT in Figure 6 demonstrates that there are peaks corresponding to (CH3)3 functional group with sp3 hybridization at 1390 cm-1, hybridized C=C of sp2 at 1620 cm-1
M
and the vibrant bonds CH2 at 2855 cm-1 and 2920 cm-1. A distinguished peak, appeared at
d
1724 cm-1, belongs to carboxyl group formed onto the CNT walls. This peak shows that carboxyl
te
groups are linked onto the SWCNT wall which proves the possibility of Cu NPs affinity onto SWCNT. Since the tablets of samples are made with KBr (with no IR absorption), 3450 cm-1
Ac ce p
peak belongs to H2O of KBr which is a water absorber and not related to the SWCNT, similarly the peak at 2360 cm-1 which belongs to CO2. In addition, the spectrum of Cu decorated SWCNT is showed in Figure 6 (KBr was not used). Cu nanoparticles are attached onto the carboxyl groups of SWNTs forming CuO at the boundaries, which are identified by a strong peak at 590 cm-1. High frequency mode at 590 cm-1 has previously been reported to be due to Cu-O stretching mode [44, 45]. Fig. 6. 4.2. Morphological and Chemical Composition Analysis by SEM/EDX
11 Page 11 of 39
The morphological structure of Cu-SWCNT was investigated using SEM imaging. Figures 7 (a) and (b) show the SWCNT bundles before and after the formation of Cu nanoparticles, respectively. Figure 7 (b) shows that Cu NPs are successfully linked onto the SWCNT walls. Cu-
ip t
nanoparticles are seen to have covered the bundles of SWCNTs at large densities. Fig. 7.
cr
EDX analysis of Cu-SWCNT was carried out to assess that Cu has been mainly formed on
us
the SWCNT bundles and CuO is a small amount that forms at the boundary of Cu and SWCNTs. Figure 8 demonstrates the composition ratio of Cu, C, and O peaks. Obviously, the oxygen ratio
an
is much less than the Cu ratio confirming that CuO presence, which is formed at the boundaries where Cu particles are attached to SWCNTs is much less than Cu. This means that the majority
M
of particles covering SWCNTs are Cu NPs, SWCNT with oxidized sites onto the SWCNT
Fig. 8.
Ac ce p
te
synthesizing of SWCNT.
d
surface. The Co peak is related to Co impurities which are used as the nucleation element in
4.3. Evaluation of Response to H2S Gas The SAW sensor, which was fabricated using Cu-SWCNT as the sensitive layer and LiNbO3 as the piezoelectric substrate, was investigated for sensing H2S gas and the results are presented and discussed. Absorption of H2S gas onto the SWCNT decorated by Cu NPs cause a change of conductivity in this layer, which according to the acoustoelectric phenomenon leads to frequency shifts due to the perturbation of acoustic waves. The dynamic response of the sensor, as a function of the frequency variation, was measured using a network analyzer. The total pressure change at different concentrations of the H2S gas is 12 Page 12 of 39
shown in Table 2. The response evaluation was carried out seven times and presented result is median value of all responses. Figure 9 illustrates the sensor response for different H2S gas concentrations at room temperature. We assumed here that response and recovery times
ip t
corresponded to the times that sensor outputs reached 90% of their final value and 10% off the baseline value, respectively. The response evaluation was carried out seven times and the
cr
presented results were the median value of all responses. The sensor was exposed to H2S for 20 s
us
each time and purged with dry air afterward. For almost all concentrations, the sensor recovered in less than 20 s to 50% of its initial baseline and in 50 s returned to >90% of its initial baseline.
an
Table 2. Fig. 9.
M
When the sensitive layer is exposed to H2S gas, its resistance changed in comparison to the
d
initial value for the Cu-SWCNT composite in air. As discussed in the theory section, changes in
te
the conductivity of sensitive layer cause perturbation in the propagation of acoustic waves near the surface of the sensor resulting in a change in the operating frequency of the SAW device
Ac ce p
according to Equation (1).
It has been theoretically shown that H2S does not efficiently interact with SWCNTs and the reaction is mostly irreversible, which is not suitable for many gas sensing applications [46]. Decoration of SWCNT bundles using Cu NPs can compensate for this deficiency. It is suggested H2S gas splits onto catalytic copper, forming sulfur (S) element via a stepwise H−S bond cleavage mechanism, resulting in a sulfur adatom and the gas phase H2 [47]: H2S (g) → S + H2 (g)
(2)
13 Page 13 of 39
It is well-known that H2 gas has a high affinity to SWCNTs and by breaking down into H+ ions lends its electrons to the body of the tubes as [48]: (3)
ip t
H2 (g) → 2H+ + 2e−
These two consecutive chemical reactions give rise to the conductivity increase in the Cu NP
us
presence of oxygen rich environment they produce SO or SO2 [49].
cr
SWCNT bundles after the exposure to H2S. As for the S adatoms, it is suggested that in the
an
4.4. Influence of Operating Temperature on the Sensor Performance
Figure 10 illustrates the response of the Cu NP-SWCNT SAW sensor exposed to 300 ppm
M
H2S gas at different operating temperatures ranging from 25 to 200°C with argon background gas. At such temperatures, the Cu NPs don’t oxidize and maintain their catalytic activities. It is
d
seen from the Figure 10, by increasing the temperature, the response of the device increases and
te
the largest frequency shift is observed at 175°C. This is as a result of the catalytic activity of Cu
Ac ce p
NPs that reaches an optimum value at this temperature. Beside the enhancement in frequency shift, the sensor response and recovery times are also improved, which are also associated to the enhancement of catalytic activity of Cu NPs. Fig. 10.
4.5. Evaluation of Sensor Selectivity In order to explore selectivity of the prepared SAW sensor with Cu-SWCNT sensitive layer, the response of the device to several different gas species is evaluated. In this regard, the sensor is exposed to H2S, H2, ethanol and acetone gas species with concentration of 100 ppm for 10 min and the frequency shifts are measured. Figure 11 exhibits the frequency changes for each tested 14 Page 14 of 39
gas. It is clear that the response of the SAW sensor to gases other than H2S is relatively small and there is a large response to H2S (frequency shift of ~260 kHz for H2S vs 20 kHz for H2, 30 kHz for ethanol and 15 kHz for acetone), which means the developed SAW sensor is promisingly
Fig. 11.
us
4.5. Evaluation of moisture influence on the sensors response
cr
ip t
selective to H2S.
In fact, influence of moisture on the sensor response is important in humid environment. The
an
response of the SAW sensors to H2S exposure is shown in Figure 12, while being exposed to 5 ppm H2S, 13000 ppm H2O (or 40% relative humidity at room temperature) at operating
M
temperature from 25°C to 300°C. We have shown that there is little response to H2S gas under
d
humid condition below 100°C temperature. According to presented chemical reactions (2) and
te
(3), we believe that the sensing mechanism of our SAW sensor is based on splitting sulfur from H2S and donating electron to the body of the CNT. Under lower temperature, there is some water
Ac ce p
on the surface of the SAW sensor. When the temperature rise to 100°C, we think the water is completely evaporated from the surface and the response moves towards previous response at non-humid condition with argon background gas. Water has two hydrogen molecules like hydrogen sulfide, but oxygen in H2O is more electronegative than sulfur in H2S and hydrogen atoms are more tightly bonded to the oxygen atom. So, by introducing H2S to the SAW sensor in humid environment, this gas donate some electron to the CNT, where water not. Fig. 12.
5. Conclusion 15 Page 15 of 39
In this study, a SAW-based sensor coated with a sensitive layer made of SWCNT decorated with Cu NPs was designed, fabricated and was successfully evaluated for sensing low concentrations of H2S gas. The optimum operating temperature of 175°C was obtained that
ip t
resulted in the largest frequency shift after exposure to H2S in air ambient. The device was able to sense H2S gas concentrations as small as 5 ppm. Additionally, at this temperature the response
cr
and recovery times of the sensor were 7 and 9 s, respectively, at 300 ppm H2S gas. In addition,
us
the Cu-SWCNT coated SAW sensor selectivity towards H2S was benchmarked against H2, acetone, and ethanol gas species. It was shown that the sensor exhibited a high selectivity to H2S
an
gas. The SAW sensor detect the H2S gas in humid environment at temperatures more than 100°C after evaporation of absorbed water from the surface of the sensor.
M
A potential application for this SAW sensor is sensing and measuring very small amounts of
d
H2S toxic gas wirelessly, which has the capability of being incorporated into sensor networks. It
te
can also be hypothesized that by tuning the SWCNT bundles using other suitable metals, highly
Ac ce p
efficient and selective sensors for sensing various types of gas species can be developed.
Acknowledgement:
We appreciate Prof. Kourosh Kalantar-zadeh from Royal Melbourne Institute of Technology, School of Electrical Engineering because of helping us to prepare more worthy discussion section and for
English
language
assistance
in
preparing
the
manuscript.
16 Page 16 of 39
References [1] M. Penza, M. A. Tagliente, P. Aversa, G. Cassano, L. Capodieci, Single-walled carbon
ip t
nanotubes nanocomposite microacoustic organic vapor sensors, Mater. Sci. Eng. C 26 (2006) 1165–1170.
cr
[2] A. R. Hutson, D. L. White, Elastic wave propagation in piezoelectric semiconductors, J.
us
Appl. Phys. (1962) 40-47.
[3] H. Wohltjen, Mechanism of operation and design consideration for surface acoustic-wave
an
device vapor sensors, Sens. Actuators B 5 (1984) 307–325.
[4] M. Penza, G. Cassano, Relative humidity sensing by PVA-coated dual resonator SAW
M
oscillator, Sens. Actuators B 68 (2000) 300–306.
d
[5] S. J. Ippolito, A. Ponzoni, K. Kalantar-Zadeh, W. Wlodarski, E. Comini, G. Faglia, G.
te
Sberveglieri, Layered WO3/ZnO/36º LiTaO3 SAW gas sensor sensitive towards ethanol vapor and humidity, Sens. Actuators B 117 (2006) 442–450.
Ac ce p
[6] A. Stelzer, S. Scheiblhofer, S. Schuster, R. Teichmann, Wireless sensor marking and temperature measurement with SAW-identification tags, Meas. 41 (2008) 579–588. [7] G. Rodríguez-Madrid, G. F. Iriarte, O. A. Williams, F. Calle, High precision pressure sensors based on SAW devices in the GHz range, Sens. Actuators B 189 (2013) 364– 369. [8] X. Ye, L. Fang, B. Liang, Q. Wang, X. Wang, L. Heb, W. Bei, H. Ko, Studies of a highsensitive surface acoustic wave sensor for passive wireless blood pressure measurement, Sens. Actuators A 169 (2011) 74–82. [9] S. G. Joshi, Flow sensors based on surface acoustic waves, Sens. Actuators A 44 (1994) 191197. 17 Page 17 of 39
[10] H. Oh, K. Lee, K. S. Choa, S. Yang, Improving the insertion loss and sensitivity over existing SAW strain sensor, Procedia Eng. 25 (2001) 567–570. [11] T. Alizadeh, S. Zeynali, Electronic nose based on the polymer coated SAW sensors array
ip t
for the warfare agent simulants classification, Sens. Actuators B 129 (2008) 412–423.
[12] F. Josse, F. Bender, R. W. Cernosek, Guided shear horizontal surface acoustic wave sensors
cr
for chemical and biochemical detection in liquids, Anal. Chem. 73 (2001) 5937–5944.
us
[13] E. Howe, G. Harding, A comparison of protocols for the optimisation of detection of bacteria using a surface acoustic wave (SAW) biosensor, Biosens. Bioelectron. 15 (2000) 641–
an
649.
[14] L. Rayleigh, On waves propagated along the plane surface of an elastic solid, Proc. London.
M
Math. Soc. 1-17 (1) (1885) 4–11.
te
Applications, Academic Press, 1996.
d
[15] D. S. Ballantine, R. M. White, Acoustic wave sensors: theory, design, & physico-Chemical
Ac ce p
[16] H. Wohltjen, Mechanism of operation and design considerations for surface acoustic wave device vapor sensors, Sens. Actuators B 5 (1984) 307-325. [17] M. J. Fernández, J. Fontecha, J. P. Santos, I. Gràciab, C. Cané, M. C. Horrillo, Love-wave sensor array to detect, discriminate and classify chemical warfare agent simulants, Sens. Actuators B 175 (2012) 173–178.
[18] S. J. Ippolito, S. Kandasamy, K. Kalantar-zadeh, W. Wlodarski, K. Galatsis, G. Kiriakidis, N. Katsarakis, M. Suchea, Highly sensitive layered ZnO/LiNbO3 SAW device with InOx selective layer for NO2 and H2 gas sensing, Sens. Actuators B 111–112 (2005) 207–212.
18 Page 18 of 39
[19] X. Wang, W. Wang, H. Li, C. Fu, Y. Ke, S. He, Development of a SnO2/CuO-coated surface acoustic wave-based H2S sensor with switch-like response and recovery, Sens. Actuators
ip t
B 169 (2012) 10–16. [20] C. Lim, W. Wang, S. Yang, K. Lee, Development of SAW-based multi-gas sensor for
cr
simultaneous detection of CO2 and NO2, Sens. Actuators B 154 (2011) 9–16.
us
[21] M. Z. Atashbar, A. Z. Sadek, W. Wlodarski, S. Sriram, M. Bhaskaran, C. J. Cheng, R. B. Kaner, K. Kalantar-zadeh, Layered SAW gas sensor based on CSA synthesized
an
polyanilinenanofiber on AlN on 64° YX LiNbO3 for H2 sensing, Sens. Actuators B 138 (2009) 85–89.
M
[22] J. Kong, N. R. Franklin, C. W. Zhou, M.G. Chapline, S. Peng, K.J. Cho, H. J. Dai,
d
Nanotube molecular wires as chemical sensors, Sci. 287 (2000) 622-625.
te
[23] T. Zhang, M. B. Nix, B.Y. Yoo, M.A. Deshusses, N.V. Myung, Electrochemically
Ac ce p
functionalized single-walled carbon nanotube gas sensor, Electroanal. 18 (2006) 1153–1158. [24] N. Peng, Q. Zhang, Ch. Chow, O.K. Tan, N. Marzari, Sensing mechanisms for carbon nanotube based NH3 gas detection, Nano Lett. 9 (2009) 1626–1630. [25] S. Kim, H .R. Lee, Y. J. Yun, Effects of polymer coating on the adsorption of gas molecules on carbon nanotube networks, Appl. Phys. Lett. 91 (2007) 3126-093126. [26] M. Penza, M. A. Tagliente, P. Aversa, M. Re, G. Cassano, The effect of purification of single-walled carbon nanotube bundles on the alcohol sensitivity of nanocomposite LangmuirBlodgett films for SAW sensing applications, Nanotechnol. 18 (2007) 185502. 19 Page 19 of 39
[27] L. Valentini, V. Bavastrello, E. Stura, I. Armentano, C. Nicolini, J.M. Kenny, Sensors for inorganic vapor detection based on carbon nanotubes and poly(o-anisidine) nanocomposite
ip t
material, Chem. Phys. Lett. 383 (2004) 617–622. [28] E. Bekyarova, M. Davis, T. Burch, Chemically functionalized single-walled carbon
cr
nanotubes as ammonia sensors, J. Phys. Chem. B 108 (2004) 51, 19717– 19720.
us
[29] J. Kong, M. G. Chapline, H. Dai, Functionalized carbon nanotubes for molecular hydrogen sensors, Adv. Mater. 13 (2001) 1384–1386.
an
[30] Y. Li, H. Wang, Y. Chen, M. Yang, A multi-walled carbon nanotube/palladium
Sens. Actuators B 132 (2008) 155–158.
M
nanocomposite prepared by a facile method for the detection of methane at room temperature,
d
[31] M. K. Kumar, S. Ramaprabhu, Nanostructured Pt functionalized multiwalled carbon
te
nanotube based hydrogen sensor, J. Phys. Chem. B 110 (2006) 11291–11298.
Ac ce p
[32] M. K. Kumar, S. Ramaprabhu, Palladium dispersed multiwalled carbon nanotube based hydrogen sensor for fuel cell applications, Int. J. Hydrogen Energy 32 (2007) 2518–2526. [33] M. Penza, G. Cassano, R. Rossi, Enhancement of sensitivity in gas chemiresistors based on carbon nanotube surface functionalized with noble metal (Au, Pt) nanoclusters, Appl. Phys. Lett. 90 (2007) 171231–171233.
[34] E. H. Espinosa, R. Ionescu, C. Bittencourt, Metal decorated multi-wall carbon nanotubes for low temperature gas sensing, Thin Solid Films 515 (2007) 8322–8327.
20 Page 20 of 39
[35] A. Z. Sadek, V. Bansal, D. G. McCulloch, P. G. Spizzirri, K. Latham, D.W.M. Lau, Z. Hu, K. Kalantar-zadeh, Facile, size-controlled deposition of highly dispersed gold nanoparticles on
ip t
nitrogen carbon nanotubes for hydrogen sensing, Sens. Actuators B 160 (2011)1034-1042. [36] A. Star, V. Joshi, S. Skarupo, D. Thomas, J. C. P. Gabriel, “Gas sensor array based on
cr
metal-decorated carbon nanotubes, J. Phys. Chem. B, 110 (2006) 21014–21020.
us
[37] Y. Lu, C. Partridge, M. Meyyappan, J. Li, A carbon nanotube sensor array for sensitive gas discrimination using principal component analysis,” J. Electroanal. Chem., 593 (2006) 105–110.
an
[38] A. Galdikas, V. Jasutis, S. Kačiulis, G. Mattogno, A. Mironas, V. Olevano, D. Senulienė, A. Šetkus, Peculiarities of surface doping with Cu in SnO2 thin film gas sensors, Sens. Actuators B,
M
(1997) 140-146.
d
[39] N. Zhang, K. Yu, Q. Li, Z. Q. Zhu, Q. Wan, Room-temperature high-sensitivity H2S gas
Ac ce p
(2008) 104305.
te
sensor based on dendritic ZnO nanostructures with macroscale in appearance, J. Appl. Phys. 103
[40] S. Zhu, X. Liu, Zh. Chen, C. Liu , C. Feng, J. Gu, Q. Liu, D. Zhang, Synthesis of Cu-doped WO3 materials with photonic structures for high performance sensors, J. Mater. Chem. 20 (2010) 9126-9132.
[41] W. P. Jakubik, Surface acoustic wave-based gas sensors, Thin Solid Films 520 (2011) 986– 993. [42] M. Hamadanian, V. Jabbari, M. Shamshiri, M. Asad, I. Mutlay, “Preparation of novel hetero-nanostructures and high efficient visible light-active photocatalyst using incorporation of 21 Page 21 of 39
CNT as an electron-transfer channel into the support TiO2 and PbS”, J. Taiwan Inst. Chem. Eng. 44 (5), (2013), 748-757.
of CuO nanoparticles, Appl. Catal., A 303 (2006) 273–277.
ip t
[43] J. Pike, S. Chan, F. Zhang, X. Wang, J. Hanson, Formation of stable Cu2O from reduction
cr
[44] K. Borgohain, J. B. Singh, M.V. Rama, T. Shirpthai, S. Mahamuni, Quantum size effects in
us
CuO nanoparticles, Phys. Rev. B 61 (2000) 11093–11096.
[45] T. Premkumar, K. E. Geckeler, Nanosized CuO particles via a supramolecular strategy,
an
Small 2 (2006) 616–620.
M
[46] W. Ruo-Xi, Z. Dong-Ju, W. Jian, L. Cheng-Bu, Theoretical study on the sensing properties of the boron and nitrogen doped carbon nanotubes for hydrogen sulfide, Acta Chim. Sinica, 65
d
(2007) 107-110.
te
[47] T, Qian-Lin, Zh, Si-Ru, L, Yan-Ping, Influence of step defects on the H2S splitting on
Ac ce p
copper surfaces from first-principles microkinetic modeling, J. Phys. Chem. C 116 (2012) 20321-20331.
[48] C. Liu, Y. Y. Fan, M. Liu, H. T. Cong, H. M. Cheng, M. S. Dresselhaus, Hydrogen storage in single-walled carbon Nanotubes at room temperature, Sci. 286 (1999) 127-1129. [49] H. Y. Jung, Y. L. Kim, S. Park, A. Datar, H-J. Lee, J. Huang, S. Somu, A. Busnaina, Y. J. Jung, Y. Kyunkown, High-performance H2S detection by redox reactions in semiconducting carbon nanotube-based devices, Analyst 138 (2013) 7206-7211.
22 Page 22 of 39
Biographies: Mohsen Asad is a PhD candidate and research assistant at the Department of Electrical and
ip t
Computer Engineering, Shiraz University, Iran. He received his BSc (2008) and MSc (2012) in Electrical Engineering from Malek Ashtar University of Technology and Islamic Azad
cr
University, respectively. He has published 15 journal and conference papers and hold 4 patents in the field of chemical sensors and Optoelectronics. His research interests include developing
us
and application of semiconductor and functional materials (1D and 2D based systems) in sensors
an
and optoelectronics.
Mohammad Hossein Sheikhi received his BSc (1994) degree in electrical engineering from
M
Shiraz University, Shiraz, Iran, the MSc (1996) degree from Sharif University of Technology, Tehran, Iran and the PhD degree in electrical engineering from Tarbiat Modarres University,
d
Tehran in 2000. Dr. Sheikhi was selected as the distinguished PhD student graduated from
te
Tarbiat Modarres University in 2000.
Ac ce p
He joined Tohokou University, Sendai, Japan, as a Research Scientist in 2000. After joining Shiraz University in 2001, he focused on the optoelectronics, nanosensors, and nanotransistors. He is currently the Associate Professor and Head of Nanotechnology Research Institute, Shiraz University, Shiraz, Iran.
23 Page 23 of 39
Figure Captions:
Ac ce p
te
d
M
an
us
cr
ip t
Figure 1. Schematic of the H2S SAW sensor coated with a Cu-SWCNT film.
24 Page 24 of 39
Ac ce p
te
d
M
an
us
cr
ip t
Figure 2: Defined mesh structure used for theoretical design with COMSOL 4.3
25 Page 25 of 39
Figure 3: Surface deformation after the SAW propagation (a) two-dimensional and (b) three-
Ac ce p
te
d
M
an
us
cr
ip t
dimensional simulations, operating at the frequency of 104 MHz.
26 Page 26 of 39
27
Page 27 of 39
d
te
Ac ce p us
an
M
cr
ip t
Figure 4: Variation of the surface electrical potential as a function of the applied voltage to the electrodes in: (a) two-dimensional and (b) three-dimensional simulations which are obtained
Ac ce p
te
d
M
an
us
cr
ip t
using COMSOL4.3.
28 Page 28 of 39
29
Page 29 of 39
d
te
Ac ce p us
an
M
cr
ip t
Ac ce p
te
d
M
an
us
cr
ip t
Figure 5: SAW sensor test set-up.
30 Page 30 of 39
Figure 6: FTIR spectra of the SWCNT: after being functionalized (black) and after Cu NPs
Ac ce p
te
d
M
an
us
cr
ip t
decoration (red).
31 Page 31 of 39
Figure 7: SEM images of (a) pristine SWCNT bundles after acid treatment and (b) after Cu NPs decoration on bundles of SWCNTs. As can be seen, the diameters of bundles are larger,
Ac ce p
te
d
M
an
us
cr
ip t
revealing Cu NPs are formed.
32 Page 32 of 39
Ac ce p
te
d
M
an
us
cr
ip t
Figure 8: EDS spectrum of the synthesized Cu-SWCNT.
33 Page 33 of 39
Ac ce p
te
d
M
an
us
cr
ip t
Figure 9: SAW sensor response to various concentrations of H2S at room temperature.
34 Page 34 of 39
Figure 10: Response of SAW sensor exposed to 300 ppm H2S gas at different operating
Ac ce p
te
d
M
an
us
cr
ip t
temperatures.
35 Page 35 of 39
Ac ce p
te
d
M
an
us
cr
ip t
Figure 11: Selectivity of SAW sensor to different gas species.
36 Page 36 of 39
Figure 11: Response of the SAW sensor under 40% relative humid environment at various
Ac ce p
te
d
M
an
us
cr
ip t
temperatures.
37 Page 37 of 39
Table Caption:
Table 1: Characteristics of the used LiNbO3 substrate.
x-axis
4742
Coupling coefficient - K2 (%)
ip t
64° YX cut
Surface wave velocity (m/s)
cr
Propagation axis
11.3
Ac ce p
te
d
M
an
us
Description
38 Page 38 of 39
Table 2: Changes in the pressure for different H2S gas concentrations. 5
10
20
30
40
50
100
200
Total pressure (mtorr)
1612
1612
1613
1613
1613
1613
1614
1620
Ac ce p
te
d
M
an
us
cr
ip t
H2S gas concentration (ppm)
39 Page 39 of 39
S0925-4005(14)00281-0 http://dx.doi.org/doi:10.1016/j.snb.2014.03.024 SNB 16664
To appear in:
Sensors and Actuators B
Received date: Revised date: Accepted date:
16-6-2013 5-3-2014 6-3-2014
Please cite this article as: M. Asad, M.H. Sheikhi, Surface Acoustic Wave Based H2S Gas Sensors Incorporating Sensitive Layers of Single Wall Carbon Nanotubes Decorated with Cu Nanoparticles, Sensors and Actuators B: Chemical (2014), http://dx.doi.org/10.1016/j.snb.2014.03.024 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Surface Acoustic Wave Based H2S Gas Sensors Incorporating Sensitive Layers of Single Wall Carbon Nanotubes Decorated with Cu Nanoparticles
1
ip t
Mohsen Asad1,2,*, Mohammad Hossein Sheikhi1,2 School of Electrical and Computer Engineering, Shiraz University, Shiraz, Iran Nanotechnology Research Institute, Shiraz University, Shiraz, Iran
cr
2
us
*Corresponding author Email: [email protected]
Abstract
an
This paper presents a surface acoustic wave (SAW) based H2S gas sensor with excellent selectivity and recovery/response time developed using single wall carbon nanotube decorated
M
with copper nanoparticles (Cu NP-SWCNT). A thin film of Cu NP-SWCNT was deposited onto
d
a lithium niobate (LiNbO3) piezoelectric substrate via a drop-casting technique. Sensing of H2S
te
gas was carried out by measuring acoustoelectric perturbation of SAWs traveling along the LiNbO3 piezoelectric substrate and Cu NP-SWCNT sensing layer. H2S gas of concentrations as
Ac ce p
small as 5 ppm could be readily measured. The effect of temperature on the SAW sensor response was also investigated for a range of temperatures from 70 to 200°C. The optimum operating temperature was 175°C, in which, a relatively rapid response (7 s) and recovery time (9 s) was recorded. The selectivity of the proposed Cu NP-SWCNT gas sensor was examined by assessing the sensor response upon exposure to hydrogen, acetone, ethanol, and H2S gas species in air background and a large selectivity towards H2S gas was observed. Keywords: surface acoustic wave; LiNbO3; copper nanoparticles; carbon nanotubes; gas sensor; H2S 1 Page 1 of 39
1. Introduction Surface acoustic wave (SAW) based sensors operate based on the wave modulation as a result of exposure to changes in physical and chemical phenomena [1]. SAW devices were first
ip t
developed nearly five decades ago [2]. Later in early 80s, the SAW transduction effect was employed for sensing applications [3]. Since then, SAW based sensors have been used for a
cr
variety of sensing applications including humidity [4,5], temperature [6], atmospheric pressure
us
[7], blood pressure [8], flow [9], and strain [10] as well as chemical/biochemical agents in gas and liquid media [11-13].
an
In gas media operation, the stress-free boundary conditions that exist on the SAW crystal surface, together with the right choice of the crystal cut, promote the propagation of surface
M
acoustic mode, with energy mostly confined on the surface of the substrate. One of the most
d
commonly used SAW propagation modes is Rayleigh mode, which has been named after Lord
te
Rayleigh who discovered this near surface traveling mode for the first time [14]. The main characteristic of Rayleigh SAWs are their relatively small penetration into the substrate, due to
Ac ce p
near surface energy concentration, ellipsoid displacement path, and mechanical wave traveling speed which is in the order of 10-5 times smaller than that of light [15]. The high energy density confined near the SAW surface plays an important role in obtaining high sensitivity for SAW based sensors. Some of the commonly used piezoelectric crystal types and cuts that allow for the generation and propagation of Rayleigh SAWs are ST cut quartz, YZ LiNbO3, and 128º YX LiNbO3 [16,17]. By coating a thin film of a material, which is sensitive to a certain type of gas, on the piezoelectric substrate the sensing capability of the SAW device to that specific gas is promoted. The applications of nanomaterials as sensitive layers offer distinct advantages. When the 2 Page 2 of 39
nanomaterial coating is exposed to gas or chemical vapors, the gas molecules can interact with the whole bulk of the layer rather than just its surface. As such, the sensitivity is dramatically enhanced. Promising SAW sensors have already been reported for a variety of gas species
ip t
including N2O [18], H2S [19], CO2 [20], and H2 [21]. The efficiency of such sensors is determined by the rate of interaction between the gas molecules and the sensitive layer.
cr
The incorporation of carbon nanotubes (CNTs) as sensitive materials was first reported by
us
Kong in 2000 [22], and since then single-wall carbon nanotubes (SWCNTs) and multi-wall carbon nanotubes (MWCNTs) have been extensively investigated [23, 24]. In order to enhance
an
the selectivity and sensitivity of CNTs towards a variety of gases, different CNT surface modification strategies such as polymer coatings [25], CNT-based nano-composites [26], CNT
M
functionalization using polymer [27], or certain functional groups [28], as well as physical and
d
chemical modifications using metal nanoclusters [29] have been reported.
te
Amongst the aforementioned methods, CNT functionalization with metals is particularly attractive due to the availability of simple, highly efficient and low cost synthesis processes. Li et
Ac ce p
al. reported a composite of Pd-functionalized MWCNT synthesized by a chemical reduction method, which exhibited an efficient and reversible response towards CH4 [30]. Kumar et al. fabricated MWCNT-based H2 sensors functionalized with Pd and Pt using aqueous solutions of PdCl2 and H2PtCl6 [31, 32]. Penza et al. demonstrated NO2 and NH3 sensors based on MWCNTs coated with Pt and Au nanoclusters [33]. Espinosa et al. reported MWCNTs with Ag and Au nanoclusters via electron beam evaporation [34] Electrodeposition method was reported by Sadek et al. for forming gold nanoparticles on MWCNTS to sense H2 [35]. Star et al. designed a gas sensor by electroplating Pt, Pd, Au, and Rh metals on SWCNTs substrate to detect CO, H2S, H2, CH4, NO2, and NH3 [36]. They also employed thermal and electron-beam evaporation of 3 Page 3 of 39
various metals (Mg, Al, Ti, V, Cr, Mn, Fe, Co, Ni, Zn, Mo, Rh, Pd, Sn, W, Pt, Au, and Pb) onto the CNT surface for sensing CH4, H2, CO, and H2S gases. Lu et al. fabricated a sensor for sensing HCN, HCl, NO2, Cl2, acetone, and benzene gases which were composed of pristine and
ip t
Pd- and Au-decorated SWCNTs [37]. However, there is no report on the incorporation of copper (Cu) nanoparticles (NPs) for the decoration of CNTs for sensing different gas species. This is
cr
despite the fact that Cu as a catalyst has been shown to enhance the sensing properties of
us
semiconducting gas sensors based on SnO2 [38], ZnO [39] WO3 [40].
In the current study, we decorated functionalized SWCNTs with Cu NPs to improve the
an
selectivity of SWCNTs towards H2S gas species. We deposited these Cu NPs decorated SWCNTs (Cu-SWCNT) onto a SAW device made of a LiNbO3 piezoelectric substrate to form
2. Theoretical Design
Ac ce p
2.1. Gas Sensitive Layer
te
d
sensitive layer and the fabricated sensors.
M
the SAW sensor. A set of characterization techniques was used for assessing the properties of the
The SAW propagation in a piezoelectric substrate creates mechanical displacements and accompanying electrical potentials as a result of the coupling between the mechanical and electromagnetic fields. Subsequently, changes in the electrical and mechanical properties of any sensitive layer, which is placed on this active area, causes changes in the velocity and magnitude of the propagating waves. If the sensitive layer mass is assumed to be negligible, the acoustoelectric effect predicts a relative change in the velocity and consequently the change in relative frequency as a function of surface electrical conductivity according to [15]: (∆f/f0)σ ≈ (∆υ/υ0)σ ≈ −(K2/2)[(σs2 / (σs2 + υ02Cs2)]
(1)
4 Page 4 of 39
where, K2 is the electromechanical coupling coefficient, σs=σ×h is the surface conductivity (in which σ is the conductivity of the bulk of the sensitive layer and h is the sensitive layer thickness), Cs=εp+εo is the sum of the dielectric permittivity factor in the substrate (εo) and its
ip t
surrounding (εp), vo and fo are the propagation velocities of the SAWs and the center frequency of the device before perturbation. Obviously any change of the operational frequency of the sensor
cr
depends on the electromechanical coupling factor K2, change in the surface conductivity of the
us
layer σs, and the permittivity of the environment. For obtaining a highly sensitive SAW based gas sensor the parameters of the sensitive layer should be tuned in a way that maximum frequency
an
change is obtained as a result of minimum change in the conductivity or permittivity of the sensitive layer. According to the discussion presented in [18], the maximum sensitivity is
M
obtained when σs = υ0Cs. Overall if 0.01υ02Cs2 < σs2 < 100υ02Cs2 [41] an acceptable sensitivity is
d
achieved. In the present study, we used a SWCNT and Cu composite as the semiconductor
te
material on the active area of the SAW device for the detection of H2S. In order to design the SAW sensor, we used a 64º YX cut LiNbO3 piezoelectric substrate, the
Ac ce p
physical characteristics of which are presented in Table 1. Based on equation (1), the acoustoelectric response range is proportionate to K2. 64º YX LiNbO3 was chosen as its K2 is quite large, resulting in a large change in the operational frequency as a result of any electromechanical perturbation.
Table 1.
5 Page 5 of 39
2.2. Design of SAW Sensor’s Geometry The schematic of such a SAW sensor is shown in Figure 1. An important issue in designing of SAW devices is regarding the geometry of the inter-digital transducers (IDTs), where an
ip t
impedance matching between the applied signal source and the input of the SAW device should be established using the geometry of the IDTs [15]. The maximum SAW generation is at the
cr
wavelength equal to the spatial period of the IDTs.
us
Fig. 1.
IDTs are metal electrodes (made of aluminum - Al) which are built on a piezoelectric
an
substrate and for applying and receiving the electric signals onto and from the piezoelectric substrates. Using COSMOL 4.3 software package at two dimensional planes, the mode of
M
piezoelectric device in order to simulate wave penetration in SAW system was chosen. Stress-
d
free conditions for the upper surface of piezoelectric substrate were applied. For the simulation,
te
it was assumed that periodic boundary condition exists on both sides of the piezoelectric substrate. Periodic boundary condition for the electric potential was used along both vertical
Ac ce p
boundaries of geometry. The displacements boundary condition was also defined similarly. The thickness of Al IDTs that was used in simulation was 200 nm. The equation for the center of the operation frequency (f0) was f0=v0/d, where v0 was the propagation velocity and d was the periodicity of the IDTs. Considering the fact that wave velocity of the 64º YX cut LiNbO3 is known (4742 m/s), for a SAW operating at 100 MHz frequency, we can estimate a period length of 48 μm. By considering these parameters, the width of each electrode was determined to be approximately 12 µm as illustrated in Figure 2. Fig. 2.
6 Page 6 of 39
Based on the Eigenfrequency analysis, approximate operational frequency of 104 MHz was selected and the surface displacement associated with this frequency in two- and threedimensions are shown in Figures 3(a) and 3(b), respectively. As a result of the simulation, it was
ip t
observed that mechanical displacement of the piezoelectric substrate at the highest point effectively penetrates 20 μm from surface downward into the substrate. In addition, the
cr
simulation suggests that the maximum of surface displacement is 1.9633×10-9 m and this
us
displacement is periodically spread over the piezoelectric substrate along the horizontal axis. It’s also clear that the maximum displacements occur at the surface of the piezoelectric substrate and
an
Rayleigh waves are confined near the surface.
Fig 3
M
To calculate the effect of the electric field, we used piezoelectric material model in
d
COMSOL 4.3 and the 64º YX cut LiNbO3 substrate was solved using the coupled equations
te
S=sET + dE and D=dtrT + εTE, where T is the stress tensor, S is the strain tensor, D is the electric
Ac ce p
charge density displacement vector, E is the electric field vector, sE, d, and εT are elastic compliance, piezoelectric module (the ratio of strain to applied field or charge density to applied mechanical stress) and the dielectric matrices, respectively (here tr stands for the transposition of the d matrix). From the coupled equations, E can be calculated and the surface voltage may be assessed as V ≈ E×t (in which t is the penetration depth of the electric field). As σs is proportional to electric field, it is subsequently proportional to the calculated voltage. Figures 4(a) and 4(b) show the variation of surface electrical potential as a function of the applied voltages at the electrodes. Fig. 4. 7 Page 7 of 39
As the generated voltage is relatively large, the simulations reveal that the conductivity change can be the dominant effect in comparison to the displacement effect on the active area of the SAW device. As a result, equation (1) is a valid approximation for assessing the frequency shift
ip t
of the sensor.
cr
3. Experimental
us
3.1. Fabrication of Sensitive Substance
CNTs (outer diameter of ~10 nm, length of ~1.5 μm, BET surface of 250-300 m2/g, purity of
an
90%, KNT-GP-BT) was purchased from Parsis Co. (Tehran, Iran – for the proof on the semiconducting properties of the CNTs refer to the Supporting Information). CuSO4 was
M
purchased from Merck Chemical Corporation (Germany) and used as received without further
te
TKA Co. Germany).
d
purification. Deionized (DI) water was prepared using an ultra-pure water system (Smart-2-Pure,
To form the Cu decorated SWCNT sensitive layer, 100 mg SWCNTs were mixed with 40 ml
Ac ce p
of HNO3 and H2SO4 (1:3) mixture at 110ºC. The resulting solution was stirred for 70 min and then sonicated for 5 h. The resulting solution was dried in vacuum oven for 6 h. Cu NPs were decorated onto the SWCNTs using CuSO4 aqueous solution via a chemical reduction process which has been fully described in [42]. 20 mg of functionalized SWCNTs were mixed with 10 mg CuSO4 in 80 ml DI water. The resulting suspension was slowly and steadily stirred at 90°C. Metal Cu, from the chemical point of view, shows different behavior in reduction process than other metals, and can act as one-valence or two-valence metal which process conditions that directly determines the chemical paths [43]. Cu ions as Cu+1 and Cu+2 are obtained after the reduction process and by substituting hydrogen ions of carboxylic group (8 Page 8 of 39
COOH), which are located on the functionalized SWCNT. Sulfuric acid is expected to be one of the main products of such reaction which is separated from the final solution by washing. Subsequently, the resulting material was separated from the solution and was desiccated at 50°C.
ip t
We characterized the electrical conductivity of the sensitive films. For this purpose, we used IDTs on substrates with SiO2 insulating layers. The pristine SWCNTs and Cu NPs-SWCNTs
cr
were dispersed in DI water and then they were spin coated on the IDTs. The conductivity for the
us
pristine SWCNTs and Cu NPs decorated SWCNTs were 2.11 and 10.41 mS/cm2, respectively, at room temperature. As a result, an increase in the conductivity of sensitive film after the
an
decoration with Cu is obvious. The sensitive layer conductivity, after the Cu decoration process, became closer to the optimum value of σs = υ0Cs considering the LiNbO3 relative permittivity of
te
3.2. Fabrication of SAW Device
d
M
~45.
In the current work, the SAW-based sensor is based on a piezoelectric crystal, two sets of
Ac ce p
IDTs and the Cu-SWCNT sensitive layer. We used a 1×2 cm2 optically polished 64º YX cut LiNbO3 substrate of 0.5 cm thickness as the piezoelectric crystal. 50 finger pairs were used in each of the receiver and transmitter IDTs. Electrode of 12 µm width and spacing were patterned that allowed the propagation of waves with the fundamental wavelength of 48 µm. In the fabrication process, the wafers were first cleaned by immersing them into DI water containing a small amount of ammonia under ultrasonication. Then, the wafer was rinsed in ethylene alcohol and dried in vacuum. Afterward, Al coating process (using Blazer e-beam evaporation system) with 200 nm thickness was carried out. Then processes followed by lift-off to complete the IDTs fabrication. 9 Page 9 of 39
To form the nanostructured SWNTs onto the active area of the SAW device, the acoustic propagation path of the SAW delay line was drop-coated with 1.5 wt% poly diallyl dimethyl ammonium chloride (PDDA) solution, as surfactant, for 10 min followed by rinsing with DI
ip t
water and drying. The PDDA treatment introduced positive charges on the LiNbO3 surface which helped to bind the negatively charged carboxyl (–COO–) ends of the SWNTs to the substrate.
cr
Then, Cu-SWNTs were drop-coated onto the LiNbO3 substrate and the resulting film was
us
allowed to completely dry at 60°C temperature overnight.
an
3.3. Characterization
The morphology of the gas sensitive film was revealed by a field emission scanning electron
M
microscopy (FE-SEM, Hitachi S4160) and the elemental structure was analyzed using Zeiss
d
EVO40 SEM equipped with an energy dispersive X-ray detector (EDX, EDAX Oxford, UK).
te
FTIR spectra of the nanomaterials were recorded using a Nicolet Magna IR 550 spectrometer. The gas sensing evaluation was carried out using a set-up as shown in Figure 5. Experiments
Ac ce p
were carried out at room temperature (25°C) and different H2S gas concentrations of 5, 10, 20, 30, 40, 50, 100, and 200 ppm and the frequency shifts were measured using a HP8751 Network Analyzer. We controlled the concentration of the H2S in the chamber by mass flow controller (MFC) against the background moisture free argon gas. Total gas flow at the chamber was maintained at 500 sccm. For investigation on the effect of temperature on the sensor response, the SAW sensors were mounted on a micro-heater inside the chamber. A DC power supply and a PT100 temperature sensor were employed to control the micro-heater temperature. The input gas was humidified by means of a controlled evaporator mixer to investigate the influence of humid
10 Page 10 of 39
environment on the sensor’s response. The input gas in delivered to chamber after flow to a humidity sensor. Fig 5.
ip t
4. Results and Discussions 4.1. Chemical Structure Analysis by FTIR
cr
Figure 6 represents the FTIR spectrum of Cu decorated SWCNT and, as a reference,
us
carboxylic-functionalized SWCNT (f-SWCNT obtained using sulfuric acid and nitric acid treatment as was described in section 3).
an
Spectrum for f-SWCNT in Figure 6 demonstrates that there are peaks corresponding to (CH3)3 functional group with sp3 hybridization at 1390 cm-1, hybridized C=C of sp2 at 1620 cm-1
M
and the vibrant bonds CH2 at 2855 cm-1 and 2920 cm-1. A distinguished peak, appeared at
d
1724 cm-1, belongs to carboxyl group formed onto the CNT walls. This peak shows that carboxyl
te
groups are linked onto the SWCNT wall which proves the possibility of Cu NPs affinity onto SWCNT. Since the tablets of samples are made with KBr (with no IR absorption), 3450 cm-1
Ac ce p
peak belongs to H2O of KBr which is a water absorber and not related to the SWCNT, similarly the peak at 2360 cm-1 which belongs to CO2. In addition, the spectrum of Cu decorated SWCNT is showed in Figure 6 (KBr was not used). Cu nanoparticles are attached onto the carboxyl groups of SWNTs forming CuO at the boundaries, which are identified by a strong peak at 590 cm-1. High frequency mode at 590 cm-1 has previously been reported to be due to Cu-O stretching mode [44, 45]. Fig. 6. 4.2. Morphological and Chemical Composition Analysis by SEM/EDX
11 Page 11 of 39
The morphological structure of Cu-SWCNT was investigated using SEM imaging. Figures 7 (a) and (b) show the SWCNT bundles before and after the formation of Cu nanoparticles, respectively. Figure 7 (b) shows that Cu NPs are successfully linked onto the SWCNT walls. Cu-
ip t
nanoparticles are seen to have covered the bundles of SWCNTs at large densities. Fig. 7.
cr
EDX analysis of Cu-SWCNT was carried out to assess that Cu has been mainly formed on
us
the SWCNT bundles and CuO is a small amount that forms at the boundary of Cu and SWCNTs. Figure 8 demonstrates the composition ratio of Cu, C, and O peaks. Obviously, the oxygen ratio
an
is much less than the Cu ratio confirming that CuO presence, which is formed at the boundaries where Cu particles are attached to SWCNTs is much less than Cu. This means that the majority
M
of particles covering SWCNTs are Cu NPs, SWCNT with oxidized sites onto the SWCNT
Fig. 8.
Ac ce p
te
synthesizing of SWCNT.
d
surface. The Co peak is related to Co impurities which are used as the nucleation element in
4.3. Evaluation of Response to H2S Gas The SAW sensor, which was fabricated using Cu-SWCNT as the sensitive layer and LiNbO3 as the piezoelectric substrate, was investigated for sensing H2S gas and the results are presented and discussed. Absorption of H2S gas onto the SWCNT decorated by Cu NPs cause a change of conductivity in this layer, which according to the acoustoelectric phenomenon leads to frequency shifts due to the perturbation of acoustic waves. The dynamic response of the sensor, as a function of the frequency variation, was measured using a network analyzer. The total pressure change at different concentrations of the H2S gas is 12 Page 12 of 39
shown in Table 2. The response evaluation was carried out seven times and presented result is median value of all responses. Figure 9 illustrates the sensor response for different H2S gas concentrations at room temperature. We assumed here that response and recovery times
ip t
corresponded to the times that sensor outputs reached 90% of their final value and 10% off the baseline value, respectively. The response evaluation was carried out seven times and the
cr
presented results were the median value of all responses. The sensor was exposed to H2S for 20 s
us
each time and purged with dry air afterward. For almost all concentrations, the sensor recovered in less than 20 s to 50% of its initial baseline and in 50 s returned to >90% of its initial baseline.
an
Table 2. Fig. 9.
M
When the sensitive layer is exposed to H2S gas, its resistance changed in comparison to the
d
initial value for the Cu-SWCNT composite in air. As discussed in the theory section, changes in
te
the conductivity of sensitive layer cause perturbation in the propagation of acoustic waves near the surface of the sensor resulting in a change in the operating frequency of the SAW device
Ac ce p
according to Equation (1).
It has been theoretically shown that H2S does not efficiently interact with SWCNTs and the reaction is mostly irreversible, which is not suitable for many gas sensing applications [46]. Decoration of SWCNT bundles using Cu NPs can compensate for this deficiency. It is suggested H2S gas splits onto catalytic copper, forming sulfur (S) element via a stepwise H−S bond cleavage mechanism, resulting in a sulfur adatom and the gas phase H2 [47]: H2S (g) → S + H2 (g)
(2)
13 Page 13 of 39
It is well-known that H2 gas has a high affinity to SWCNTs and by breaking down into H+ ions lends its electrons to the body of the tubes as [48]: (3)
ip t
H2 (g) → 2H+ + 2e−
These two consecutive chemical reactions give rise to the conductivity increase in the Cu NP
us
presence of oxygen rich environment they produce SO or SO2 [49].
cr
SWCNT bundles after the exposure to H2S. As for the S adatoms, it is suggested that in the
an
4.4. Influence of Operating Temperature on the Sensor Performance
Figure 10 illustrates the response of the Cu NP-SWCNT SAW sensor exposed to 300 ppm
M
H2S gas at different operating temperatures ranging from 25 to 200°C with argon background gas. At such temperatures, the Cu NPs don’t oxidize and maintain their catalytic activities. It is
d
seen from the Figure 10, by increasing the temperature, the response of the device increases and
te
the largest frequency shift is observed at 175°C. This is as a result of the catalytic activity of Cu
Ac ce p
NPs that reaches an optimum value at this temperature. Beside the enhancement in frequency shift, the sensor response and recovery times are also improved, which are also associated to the enhancement of catalytic activity of Cu NPs. Fig. 10.
4.5. Evaluation of Sensor Selectivity In order to explore selectivity of the prepared SAW sensor with Cu-SWCNT sensitive layer, the response of the device to several different gas species is evaluated. In this regard, the sensor is exposed to H2S, H2, ethanol and acetone gas species with concentration of 100 ppm for 10 min and the frequency shifts are measured. Figure 11 exhibits the frequency changes for each tested 14 Page 14 of 39
gas. It is clear that the response of the SAW sensor to gases other than H2S is relatively small and there is a large response to H2S (frequency shift of ~260 kHz for H2S vs 20 kHz for H2, 30 kHz for ethanol and 15 kHz for acetone), which means the developed SAW sensor is promisingly
Fig. 11.
us
4.5. Evaluation of moisture influence on the sensors response
cr
ip t
selective to H2S.
In fact, influence of moisture on the sensor response is important in humid environment. The
an
response of the SAW sensors to H2S exposure is shown in Figure 12, while being exposed to 5 ppm H2S, 13000 ppm H2O (or 40% relative humidity at room temperature) at operating
M
temperature from 25°C to 300°C. We have shown that there is little response to H2S gas under
d
humid condition below 100°C temperature. According to presented chemical reactions (2) and
te
(3), we believe that the sensing mechanism of our SAW sensor is based on splitting sulfur from H2S and donating electron to the body of the CNT. Under lower temperature, there is some water
Ac ce p
on the surface of the SAW sensor. When the temperature rise to 100°C, we think the water is completely evaporated from the surface and the response moves towards previous response at non-humid condition with argon background gas. Water has two hydrogen molecules like hydrogen sulfide, but oxygen in H2O is more electronegative than sulfur in H2S and hydrogen atoms are more tightly bonded to the oxygen atom. So, by introducing H2S to the SAW sensor in humid environment, this gas donate some electron to the CNT, where water not. Fig. 12.
5. Conclusion 15 Page 15 of 39
In this study, a SAW-based sensor coated with a sensitive layer made of SWCNT decorated with Cu NPs was designed, fabricated and was successfully evaluated for sensing low concentrations of H2S gas. The optimum operating temperature of 175°C was obtained that
ip t
resulted in the largest frequency shift after exposure to H2S in air ambient. The device was able to sense H2S gas concentrations as small as 5 ppm. Additionally, at this temperature the response
cr
and recovery times of the sensor were 7 and 9 s, respectively, at 300 ppm H2S gas. In addition,
us
the Cu-SWCNT coated SAW sensor selectivity towards H2S was benchmarked against H2, acetone, and ethanol gas species. It was shown that the sensor exhibited a high selectivity to H2S
an
gas. The SAW sensor detect the H2S gas in humid environment at temperatures more than 100°C after evaporation of absorbed water from the surface of the sensor.
M
A potential application for this SAW sensor is sensing and measuring very small amounts of
d
H2S toxic gas wirelessly, which has the capability of being incorporated into sensor networks. It
te
can also be hypothesized that by tuning the SWCNT bundles using other suitable metals, highly
Ac ce p
efficient and selective sensors for sensing various types of gas species can be developed.
Acknowledgement:
We appreciate Prof. Kourosh Kalantar-zadeh from Royal Melbourne Institute of Technology, School of Electrical Engineering because of helping us to prepare more worthy discussion section and for
English
language
assistance
in
preparing
the
manuscript.
16 Page 16 of 39
References [1] M. Penza, M. A. Tagliente, P. Aversa, G. Cassano, L. Capodieci, Single-walled carbon
ip t
nanotubes nanocomposite microacoustic organic vapor sensors, Mater. Sci. Eng. C 26 (2006) 1165–1170.
cr
[2] A. R. Hutson, D. L. White, Elastic wave propagation in piezoelectric semiconductors, J.
us
Appl. Phys. (1962) 40-47.
[3] H. Wohltjen, Mechanism of operation and design consideration for surface acoustic-wave
an
device vapor sensors, Sens. Actuators B 5 (1984) 307–325.
[4] M. Penza, G. Cassano, Relative humidity sensing by PVA-coated dual resonator SAW
M
oscillator, Sens. Actuators B 68 (2000) 300–306.
d
[5] S. J. Ippolito, A. Ponzoni, K. Kalantar-Zadeh, W. Wlodarski, E. Comini, G. Faglia, G.
te
Sberveglieri, Layered WO3/ZnO/36º LiTaO3 SAW gas sensor sensitive towards ethanol vapor and humidity, Sens. Actuators B 117 (2006) 442–450.
Ac ce p
[6] A. Stelzer, S. Scheiblhofer, S. Schuster, R. Teichmann, Wireless sensor marking and temperature measurement with SAW-identification tags, Meas. 41 (2008) 579–588. [7] G. Rodríguez-Madrid, G. F. Iriarte, O. A. Williams, F. Calle, High precision pressure sensors based on SAW devices in the GHz range, Sens. Actuators B 189 (2013) 364– 369. [8] X. Ye, L. Fang, B. Liang, Q. Wang, X. Wang, L. Heb, W. Bei, H. Ko, Studies of a highsensitive surface acoustic wave sensor for passive wireless blood pressure measurement, Sens. Actuators A 169 (2011) 74–82. [9] S. G. Joshi, Flow sensors based on surface acoustic waves, Sens. Actuators A 44 (1994) 191197. 17 Page 17 of 39
[10] H. Oh, K. Lee, K. S. Choa, S. Yang, Improving the insertion loss and sensitivity over existing SAW strain sensor, Procedia Eng. 25 (2001) 567–570. [11] T. Alizadeh, S. Zeynali, Electronic nose based on the polymer coated SAW sensors array
ip t
for the warfare agent simulants classification, Sens. Actuators B 129 (2008) 412–423.
[12] F. Josse, F. Bender, R. W. Cernosek, Guided shear horizontal surface acoustic wave sensors
cr
for chemical and biochemical detection in liquids, Anal. Chem. 73 (2001) 5937–5944.
us
[13] E. Howe, G. Harding, A comparison of protocols for the optimisation of detection of bacteria using a surface acoustic wave (SAW) biosensor, Biosens. Bioelectron. 15 (2000) 641–
an
649.
[14] L. Rayleigh, On waves propagated along the plane surface of an elastic solid, Proc. London.
M
Math. Soc. 1-17 (1) (1885) 4–11.
te
Applications, Academic Press, 1996.
d
[15] D. S. Ballantine, R. M. White, Acoustic wave sensors: theory, design, & physico-Chemical
Ac ce p
[16] H. Wohltjen, Mechanism of operation and design considerations for surface acoustic wave device vapor sensors, Sens. Actuators B 5 (1984) 307-325. [17] M. J. Fernández, J. Fontecha, J. P. Santos, I. Gràciab, C. Cané, M. C. Horrillo, Love-wave sensor array to detect, discriminate and classify chemical warfare agent simulants, Sens. Actuators B 175 (2012) 173–178.
[18] S. J. Ippolito, S. Kandasamy, K. Kalantar-zadeh, W. Wlodarski, K. Galatsis, G. Kiriakidis, N. Katsarakis, M. Suchea, Highly sensitive layered ZnO/LiNbO3 SAW device with InOx selective layer for NO2 and H2 gas sensing, Sens. Actuators B 111–112 (2005) 207–212.
18 Page 18 of 39
[19] X. Wang, W. Wang, H. Li, C. Fu, Y. Ke, S. He, Development of a SnO2/CuO-coated surface acoustic wave-based H2S sensor with switch-like response and recovery, Sens. Actuators
ip t
B 169 (2012) 10–16. [20] C. Lim, W. Wang, S. Yang, K. Lee, Development of SAW-based multi-gas sensor for
cr
simultaneous detection of CO2 and NO2, Sens. Actuators B 154 (2011) 9–16.
us
[21] M. Z. Atashbar, A. Z. Sadek, W. Wlodarski, S. Sriram, M. Bhaskaran, C. J. Cheng, R. B. Kaner, K. Kalantar-zadeh, Layered SAW gas sensor based on CSA synthesized
an
polyanilinenanofiber on AlN on 64° YX LiNbO3 for H2 sensing, Sens. Actuators B 138 (2009) 85–89.
M
[22] J. Kong, N. R. Franklin, C. W. Zhou, M.G. Chapline, S. Peng, K.J. Cho, H. J. Dai,
d
Nanotube molecular wires as chemical sensors, Sci. 287 (2000) 622-625.
te
[23] T. Zhang, M. B. Nix, B.Y. Yoo, M.A. Deshusses, N.V. Myung, Electrochemically
Ac ce p
functionalized single-walled carbon nanotube gas sensor, Electroanal. 18 (2006) 1153–1158. [24] N. Peng, Q. Zhang, Ch. Chow, O.K. Tan, N. Marzari, Sensing mechanisms for carbon nanotube based NH3 gas detection, Nano Lett. 9 (2009) 1626–1630. [25] S. Kim, H .R. Lee, Y. J. Yun, Effects of polymer coating on the adsorption of gas molecules on carbon nanotube networks, Appl. Phys. Lett. 91 (2007) 3126-093126. [26] M. Penza, M. A. Tagliente, P. Aversa, M. Re, G. Cassano, The effect of purification of single-walled carbon nanotube bundles on the alcohol sensitivity of nanocomposite LangmuirBlodgett films for SAW sensing applications, Nanotechnol. 18 (2007) 185502. 19 Page 19 of 39
[27] L. Valentini, V. Bavastrello, E. Stura, I. Armentano, C. Nicolini, J.M. Kenny, Sensors for inorganic vapor detection based on carbon nanotubes and poly(o-anisidine) nanocomposite
ip t
material, Chem. Phys. Lett. 383 (2004) 617–622. [28] E. Bekyarova, M. Davis, T. Burch, Chemically functionalized single-walled carbon
cr
nanotubes as ammonia sensors, J. Phys. Chem. B 108 (2004) 51, 19717– 19720.
us
[29] J. Kong, M. G. Chapline, H. Dai, Functionalized carbon nanotubes for molecular hydrogen sensors, Adv. Mater. 13 (2001) 1384–1386.
an
[30] Y. Li, H. Wang, Y. Chen, M. Yang, A multi-walled carbon nanotube/palladium
Sens. Actuators B 132 (2008) 155–158.
M
nanocomposite prepared by a facile method for the detection of methane at room temperature,
d
[31] M. K. Kumar, S. Ramaprabhu, Nanostructured Pt functionalized multiwalled carbon
te
nanotube based hydrogen sensor, J. Phys. Chem. B 110 (2006) 11291–11298.
Ac ce p
[32] M. K. Kumar, S. Ramaprabhu, Palladium dispersed multiwalled carbon nanotube based hydrogen sensor for fuel cell applications, Int. J. Hydrogen Energy 32 (2007) 2518–2526. [33] M. Penza, G. Cassano, R. Rossi, Enhancement of sensitivity in gas chemiresistors based on carbon nanotube surface functionalized with noble metal (Au, Pt) nanoclusters, Appl. Phys. Lett. 90 (2007) 171231–171233.
[34] E. H. Espinosa, R. Ionescu, C. Bittencourt, Metal decorated multi-wall carbon nanotubes for low temperature gas sensing, Thin Solid Films 515 (2007) 8322–8327.
20 Page 20 of 39
[35] A. Z. Sadek, V. Bansal, D. G. McCulloch, P. G. Spizzirri, K. Latham, D.W.M. Lau, Z. Hu, K. Kalantar-zadeh, Facile, size-controlled deposition of highly dispersed gold nanoparticles on
ip t
nitrogen carbon nanotubes for hydrogen sensing, Sens. Actuators B 160 (2011)1034-1042. [36] A. Star, V. Joshi, S. Skarupo, D. Thomas, J. C. P. Gabriel, “Gas sensor array based on
cr
metal-decorated carbon nanotubes, J. Phys. Chem. B, 110 (2006) 21014–21020.
us
[37] Y. Lu, C. Partridge, M. Meyyappan, J. Li, A carbon nanotube sensor array for sensitive gas discrimination using principal component analysis,” J. Electroanal. Chem., 593 (2006) 105–110.
an
[38] A. Galdikas, V. Jasutis, S. Kačiulis, G. Mattogno, A. Mironas, V. Olevano, D. Senulienė, A. Šetkus, Peculiarities of surface doping with Cu in SnO2 thin film gas sensors, Sens. Actuators B,
M
(1997) 140-146.
d
[39] N. Zhang, K. Yu, Q. Li, Z. Q. Zhu, Q. Wan, Room-temperature high-sensitivity H2S gas
Ac ce p
(2008) 104305.
te
sensor based on dendritic ZnO nanostructures with macroscale in appearance, J. Appl. Phys. 103
[40] S. Zhu, X. Liu, Zh. Chen, C. Liu , C. Feng, J. Gu, Q. Liu, D. Zhang, Synthesis of Cu-doped WO3 materials with photonic structures for high performance sensors, J. Mater. Chem. 20 (2010) 9126-9132.
[41] W. P. Jakubik, Surface acoustic wave-based gas sensors, Thin Solid Films 520 (2011) 986– 993. [42] M. Hamadanian, V. Jabbari, M. Shamshiri, M. Asad, I. Mutlay, “Preparation of novel hetero-nanostructures and high efficient visible light-active photocatalyst using incorporation of 21 Page 21 of 39
CNT as an electron-transfer channel into the support TiO2 and PbS”, J. Taiwan Inst. Chem. Eng. 44 (5), (2013), 748-757.
of CuO nanoparticles, Appl. Catal., A 303 (2006) 273–277.
ip t
[43] J. Pike, S. Chan, F. Zhang, X. Wang, J. Hanson, Formation of stable Cu2O from reduction
cr
[44] K. Borgohain, J. B. Singh, M.V. Rama, T. Shirpthai, S. Mahamuni, Quantum size effects in
us
CuO nanoparticles, Phys. Rev. B 61 (2000) 11093–11096.
[45] T. Premkumar, K. E. Geckeler, Nanosized CuO particles via a supramolecular strategy,
an
Small 2 (2006) 616–620.
M
[46] W. Ruo-Xi, Z. Dong-Ju, W. Jian, L. Cheng-Bu, Theoretical study on the sensing properties of the boron and nitrogen doped carbon nanotubes for hydrogen sulfide, Acta Chim. Sinica, 65
d
(2007) 107-110.
te
[47] T, Qian-Lin, Zh, Si-Ru, L, Yan-Ping, Influence of step defects on the H2S splitting on
Ac ce p
copper surfaces from first-principles microkinetic modeling, J. Phys. Chem. C 116 (2012) 20321-20331.
[48] C. Liu, Y. Y. Fan, M. Liu, H. T. Cong, H. M. Cheng, M. S. Dresselhaus, Hydrogen storage in single-walled carbon Nanotubes at room temperature, Sci. 286 (1999) 127-1129. [49] H. Y. Jung, Y. L. Kim, S. Park, A. Datar, H-J. Lee, J. Huang, S. Somu, A. Busnaina, Y. J. Jung, Y. Kyunkown, High-performance H2S detection by redox reactions in semiconducting carbon nanotube-based devices, Analyst 138 (2013) 7206-7211.
22 Page 22 of 39
Biographies: Mohsen Asad is a PhD candidate and research assistant at the Department of Electrical and
ip t
Computer Engineering, Shiraz University, Iran. He received his BSc (2008) and MSc (2012) in Electrical Engineering from Malek Ashtar University of Technology and Islamic Azad
cr
University, respectively. He has published 15 journal and conference papers and hold 4 patents in the field of chemical sensors and Optoelectronics. His research interests include developing
us
and application of semiconductor and functional materials (1D and 2D based systems) in sensors
an
and optoelectronics.
Mohammad Hossein Sheikhi received his BSc (1994) degree in electrical engineering from
M
Shiraz University, Shiraz, Iran, the MSc (1996) degree from Sharif University of Technology, Tehran, Iran and the PhD degree in electrical engineering from Tarbiat Modarres University,
d
Tehran in 2000. Dr. Sheikhi was selected as the distinguished PhD student graduated from
te
Tarbiat Modarres University in 2000.
Ac ce p
He joined Tohokou University, Sendai, Japan, as a Research Scientist in 2000. After joining Shiraz University in 2001, he focused on the optoelectronics, nanosensors, and nanotransistors. He is currently the Associate Professor and Head of Nanotechnology Research Institute, Shiraz University, Shiraz, Iran.
23 Page 23 of 39
Figure Captions:
Ac ce p
te
d
M
an
us
cr
ip t
Figure 1. Schematic of the H2S SAW sensor coated with a Cu-SWCNT film.
24 Page 24 of 39
Ac ce p
te
d
M
an
us
cr
ip t
Figure 2: Defined mesh structure used for theoretical design with COMSOL 4.3
25 Page 25 of 39
Figure 3: Surface deformation after the SAW propagation (a) two-dimensional and (b) three-
Ac ce p
te
d
M
an
us
cr
ip t
dimensional simulations, operating at the frequency of 104 MHz.
26 Page 26 of 39
27
Page 27 of 39
d
te
Ac ce p us
an
M
cr
ip t
Figure 4: Variation of the surface electrical potential as a function of the applied voltage to the electrodes in: (a) two-dimensional and (b) three-dimensional simulations which are obtained
Ac ce p
te
d
M
an
us
cr
ip t
using COMSOL4.3.
28 Page 28 of 39
29
Page 29 of 39
d
te
Ac ce p us
an
M
cr
ip t
Ac ce p
te
d
M
an
us
cr
ip t
Figure 5: SAW sensor test set-up.
30 Page 30 of 39
Figure 6: FTIR spectra of the SWCNT: after being functionalized (black) and after Cu NPs
Ac ce p
te
d
M
an
us
cr
ip t
decoration (red).
31 Page 31 of 39
Figure 7: SEM images of (a) pristine SWCNT bundles after acid treatment and (b) after Cu NPs decoration on bundles of SWCNTs. As can be seen, the diameters of bundles are larger,
Ac ce p
te
d
M
an
us
cr
ip t
revealing Cu NPs are formed.
32 Page 32 of 39
Ac ce p
te
d
M
an
us
cr
ip t
Figure 8: EDS spectrum of the synthesized Cu-SWCNT.
33 Page 33 of 39
Ac ce p
te
d
M
an
us
cr
ip t
Figure 9: SAW sensor response to various concentrations of H2S at room temperature.
34 Page 34 of 39
Figure 10: Response of SAW sensor exposed to 300 ppm H2S gas at different operating
Ac ce p
te
d
M
an
us
cr
ip t
temperatures.
35 Page 35 of 39
Ac ce p
te
d
M
an
us
cr
ip t
Figure 11: Selectivity of SAW sensor to different gas species.
36 Page 36 of 39
Figure 11: Response of the SAW sensor under 40% relative humid environment at various
Ac ce p
te
d
M
an
us
cr
ip t
temperatures.
37 Page 37 of 39
Table Caption:
Table 1: Characteristics of the used LiNbO3 substrate.
x-axis
4742
Coupling coefficient - K2 (%)
ip t
64° YX cut
Surface wave velocity (m/s)
cr
Propagation axis
11.3
Ac ce p
te
d
M
an
us
Description
38 Page 38 of 39
Table 2: Changes in the pressure for different H2S gas concentrations. 5
10
20
30
40
50
100
200
Total pressure (mtorr)
1612
1612
1613
1613
1613
1613
1614
1620
Ac ce p
te
d
M
an
us
cr
ip t
H2S gas concentration (ppm)
39 Page 39 of 39