Nano gas sensors paper

A A

A S

a

A R R A A

K A Z S P F

1

m o i t t t d e q s e t b t

i d P t Y

h 0

Sensors and Actuators B 222 (2016) 536–543

Contents lists available at ScienceDirect

Sensors and Actuators B: Chemical

j o u r n a l h o m e p a g e : w w w . e l s e v i e r . c o m / l o c a t e / s n b

novel flexible acetylene gas sensor based on PI/PTFE-supported g-loaded vertical ZnO nanorods array

.S.M. Iftekhar Uddin, Usman Yaqoob, Duy-Thach Phan, Gwiy-Sang Chung ∗

chool of Electrical Engineering, University of Ulsan, 93 Daehak-ro, Nam-gu, Ulsan 680-749, South Korea

r t i c l e i n f o

rticle history: eceived 23 April 2015 eceived in revised form 6 July 2015 ccepted 24 August 2015 vailable online 28 August 2015

a b s t r a c t

The paper presents a novel flexible acetylene gas sensor consisting of Ag-loaded vertical ZnO nanorods (NRs) supported by a polyimide/polytetrafluoroethylene substrate. The fabricated sensor exhibits a high response magnitude of 27.2 (1000 ppm), short response/recovery time of 62/39 s, and broad detection range from 3 to 1000 ppm at a low operating temperature of 200 ◦C. The enhancement of sensing perfor- mance was mainly attributed to the large effective surface area, small grain size, ordered arrays of ZnO

eywords: cetylene gas sensor nO nanorods ilver

NRs, and the successful immobilization of tiny sized Ag nanoparticles. At optimum conditions (8 s Ag- loaded ZnO NRs at 200 ◦C), the sensor showed a negligible response degradation of ∼2.1% at a curvature angle of 90◦, and ∼8% for 5 × 104 bending/relaxing cycles. The outstanding flexibility of the as-fabricated sensor ensured stable sensing performances to extreme bending stress, indicating the possibility of fabricating highly efficient and practical flexible C H gas sensors.

I

lexible

. Introduction

With the expansion of industries and technological advance- ents, the development of solid state gas sensors for the detection

f toxic, flammable, and corrosive gases has become increasingly mportant for environmental monitoring, personal safety protec- ion, and smooth industrial functionalization. Recently, in addition o the traditional solid state gas sensor properties, these applica- ions require a number of additional features such as cost-effective evice fabrication, lighter weight, flexibility, and stretchability to nable smart wearable, handheld and portable gas sensors. Conse- uently, numerous researchers in the field of sensor devices have hown considerable interest in flexible substrates, which have been xplored for a wide range of applications due to their low produc- ion cost, light weight, and remarkable mechanical flexibility, while eing considered as a potential alternative to expensive silicon echnology [1–3].

Among the various reported flexible substrates, polymers ncluding polyimide (PI), polyethylene terephthalate (PET), poly- imethylsiloxane (PDMS), and polyethylene naphthalate (PEN),

I has attracted immense interest because of its extraordinary hermal, mechanical, and chemical properties, including high oung’s modulus (2.5 GPa), wide operating temperature (−269

∗ Corresponding author. E-mail address: [email protected] (G.-S. Chung). URL: http://home2.ulsan.ac.kr/user/gschung (G.-S. Chung).

ttp://dx.doi.org/10.1016/j.snb.2015.08.106 925-4005/© 2015 Elsevier B.V. All rights reserved.

2 2

© 2015 Elsevier B.V. All rights reserved.

to 400 ◦C), high resistivity (1.7 × 1017 �-cm), low coefficient of thermal expansion (30 × 10−6 K−1), and low thermal conductivity (0.16 Wm−1 K−1) [4–6].

Sensing materials on the sensor surface exhibit either inherent or modified structural and chemical characteristics that directly affect the sensing mechanism and play a central role in device performance. One-dimensional (1D) oxide semiconductor nano- structures such as nanowires, nanorods, nanotubes, nanofibers, and nanobelts based solid state and flexible gas sensors have been investigated extensively for decades. 1D ZnO nanostructures have shown promising advancement due to their extraordinary char- acteristics such as large length-to-diameter aspect ratio and large surface-to-volume ratio, making them persistent materials for gas sensing applications [7–10]. In addition, surface-controlled 1D ZnO exhibits more quantum effects than other dimensional nanostruc- tures and has a profound influence on gas sensing performances, in which the grain size, defects, and oxygen-adsorption quan- tities are crucial [10]. Numerous synthesis techniques (namely spray pyrolysis, hydrothermal, solution, plasma enhanced chemical vapor deposition, thermal evaporation, vapor transport, electrode- position, spin-spray, etc.) have been reported in the literature to synthesize these highly crystalline 1D nanostructures with high yield [9–14]. However, similar to other metal oxide nanomaterials, 1D ZnO also has inherent shortcomings including high operat- ing temperature, low response magnitude, and poor selectivity.

The drawbacks are usually overcome by incorporating sensitizers or promoters, using noble metals (Au, Ag, Pt, Pd) [15–18], metal oxides (In2O3, SnO2) [19,20], or carbon materials (graphene) [2].

nd Act

P a o w

m i h r t o a r t a n t a l

P g t e s w c e s

2

2

i w

( p w l s u s a 2 s 4 a ( T t t p s f N t

( R A t

A.S.M.I. Uddin et al. / Sensors a

articularly, the addition of such additives provides additional ctive absorption sites for better surface reactions (chemisorptions, xidation, and reduction) and an additional charge transfer path- ay, while facilitating sensing performance improvements [21].

Acetylene (C2H2) has many applications as a fuel or as a raw aterial in various mechanical and chemical industries. Because of

ts high flammability and restrained toxicity, considerable interest as been focused on the development of high-performance, accu- ate, low-power consuming, practical, and wearable C2H2 sensors o avoid unexpected explosions where possible ignition sources are ften plentiful. A number of solid state based C2H2 gas sensors have lready been reported with remarkable improvements such as high esponse magnitude, broad detection range, fast response/recovery ime, good selectivity, and long-term stability [22–25]. Our group lso reported a number of research results on C2H2 sensing with otable enhancement (for details see Table-S1 in the suppor- ing information) [26–29]. However, to the best of our knowledge,

flexible type C2H2 gas sensor has not yet been reported in the iterature.

In this work, an Ag-loaded vertical ZnO nanorods array on a I/PTFE substrate based flexible C2H2 sensor was reported. The rown structure was synthesized through a simple, rapid and low- emperature hydrothermal-RF magnetron sputtering method. The ffect of various bending angles (curvature) and various conditions uch as the times of bending/relaxing on the sensor performance as investigated systematically. The outcome of the device indi-

ates the potential means to fabricate mechanically stable, highly fficient, and reliable, flexible C2H2 sensors that can be easily up- caled with minimum financial requirements.

. Experimental

.1. Material synthesis and sensor fabrication

All the chemicals used in the synthesis process were of analyt- cal grade purchased from Sigma–Aldrich Co. Inc., and were used

ithout further purification. To synthesize the vertical ZnO nanorods, commercial polyimide

PI) film (length: 30 mm, width: 10 mm, thickness: 10 �m) was first asted onto a polytetrafluoroethylene (PTFE) sheet (length: 40 mm, idth: 13 mm, thickness: 1 mm). A template ZnO thin film (seed

ayer) with a thickness of ∼120 nm was deposited on a PI/PTFE upport layer through a metal mask (mask size: 15 mm × 8 mm) sing RF magnetron sputtering (50 Watt, 7 mTorr working pres- ure) in an Ar environment at a deposition rate of 500 nm/h from

ZnO target (99.99%). The ZnO/PI/PTFE film was then annealed at 50 ◦C for 1 h. ZnO NRs were grown on the ZnO seed-coated PI/PTFE ubstrate using the hydrothermal method. In a typical process,

mM zinc nitrate hexahydrate (Zn(NO3)2·6H2O) and 4 mM hex- methylenetetramine (C6H12N4) were added to 150 mL de-ionized DI) water using vigorous stirring at room temperature for 30 min. he mixer solution was poured into a Teflon-lined autoclave and he ZnO seed-coated PI/PTFE template sample was then placed ver- ically inside the autoclave. The Teflon autoclave was then carefully laced in the laboratory autoclave oven and maintained at a con- tant temperature of 80 ◦C for 4 h. The sample was then removed rom the autoclave and rinsed several times with DI water. The ZnO Rs/PI/PTFE was placed on a hotplate at 100 ◦C for 30 min to remove

he residual water. To synthesize the Ag-loaded vertical ZnO NRs, Ag nanoparticles

NPs) were coated on the as-grown ZnO NRs/PI/PTFE sample using F magnetron sputtering (125 W, 7 mTorr working pressure) in an r environment from the Ag target (99.99%), with various loading

imes of 6, 8, and 10 s.

uators B 222 (2016) 536–543 537

Finally, to fabricate a simple resistivity-type flexible sensor, two gold (Au) electrodes with a thickness of 200 nm at 9 mm apart were deposited on top of the sensor layer using a metal mask and RF magnetron sputtering (150 W, 7 mTorr working pressure). An opti- cal photograph of the fabricated flexible sensor and measurement environment is shown in Fig. 1.

2.2. Characterization and sensor test

The structural properties of the sensing materials were inves- tigated using an X-ray diffractometer (XRD) (Rigaku Ultima IV) with Cu K� (� = 0.154 nm) radiation with a 2� scanning range of 10–70◦. The surface morphology was examined using field emis- sion scanning electron microscopy (FESEM; JEOL JSM-7600F) and transmission electron microscopy (TEM; JEOL JEM-2010F) with an accelerating voltage of 10 kV. An energy dispersive spectrometer (EDS, JEOL JEM-2010F) was used for compositional analysis.

For gas sensing measurement, a ceramic heater was carefully glued on the rear-center of the fabricated sensor, and placed in the center of a U-shaped support-stand in an open air environment. The right sidewall of the support-stand was moved left/right to vary and adjust the bending angle of the sensor. A narrow tube was placed on the top of the sensor device to ensure flow of the target gas. A computerized mass flow controller (ATO-VAC, GMC 1200) system was used to vary the concentration of C2H2 in synthetic air. The gas mixture (synthetic air and C2H2) was delivered on top of the sensor device at a constant flow rate of 50 sccm (standard cubic centimeters per minute) with different C2H2 concentrations. The gas concentration was controlled and measured using the following equation:

GAScon. (ppm) = Flow rateair + Flow rategas

Total flow rate .

The gas flow was stopped between each C2H2 pulse to allow the surface of the sensor to return to an atmospheric condition. A Keithley probe station (SCS-4200) with a bias voltage of 1 V was used for all measurements and data acquisition. The device was tested at temperatures ranging between 25 ◦C and 250 ◦C for var- ious C2H2 concentrations. The sensor response (S) was calculated using S = Ra/Rg, where Ra and Rg are the resistances in the presence of air and carrier gases, respectively. The response time and recov- ery time of the sensor were defined as the time taken to reach 90% of the total resistance change.

3. Results and discussions

3.1. Structural and morphological studies

Phase analysis of the unloaded and the 6, 8, and 10 s Ag-loaded ZnO NRs was carried out, and the observed characteristic pat- terns are shown in Fig. 2. The observed characteristic diffraction reflections at different 2� values indicate the existence of (1 0 0), (0 0 2), (1 0 1), (1 0 2), (1 1 0), and (1 0 3) planes of wurtzite hexag- onal ZnO (JCPDS card No. 36-1451). The characteristic diffraction peaks appeared at 2� = 36.9◦ and 44.32◦, corresponding to the cubic phase Ag (1 1 1) and (2 0 0) planes, respectively (JCPDS card No. 04- 0783). A sharp, strong, and dominant peak of ZnO (0 0 2) centered at 34.2◦ was observed for all characterized samples, which is com- paratively higher than any other peak, indicating the preferential growth of single crystalline ZnO NRs along the 〈0 0 1〉 direction. The strong and narrow (0 0 2) reflection peak demonstrates the orien- tation of the ZnO NRs normal to the substrate surface and is related

to the crystal plane of the wurtzite ZnO, while 〈0 0 1〉 indicates the growth direction [10]. However, no significant shift in peak position indicates that the Ag NPs were attached to the surface of wall of the ZnO NRs without being incorporated into the lattice of

538 A.S.M.I. Uddin et al. / Sensors and Actuators B 222 (2016) 536–543

Fig. 1. Optical images of (a) fabricated sensor and (b)

F

Z l r m h p s p o

o i N h s a F o h w a r t i t t [

m d

◦

ig. 2. XRD patterns of (a) unloaded, (b) 6 s, (c) 8 s, and (d) 10 s Ag-loaded ZnO NRs.

nO. Besides, a slight drop in peak intensity was observed in 6 s Ag- oaded sample compared to the unloaded one. The peak intensity educed more with increasing Ag loading times (8 and 10 s), which ight be attributed to the surface coverage of the ZnO NRs wall by

uge number of Ag NPs. In addition, an additional flat and broad eak appeared at 2� = 18.4◦, which might have originated from the upporting PI substrate. Moreover, the absence of the characteristic eaks of the intermediates in the patterns indicates the formation f high purity sensing materials.

Representative cross-section and in-plane FESEM micrographs f the 8 s Ag-loaded ZnO NRs on the PI/PTFE substrate are depicted n Figs. 3(a) and (b), respectively. Small sized Ag NPs-coated ZnO Rs were observed growing uniformly in a vertical direction with igh yield over the entire substrate. A magnified view of the grown tructure is shown in the inset of Fig. 3(a), which confirms the ttachment of the Ag NPs to the surface of the ZnO NRs. The inset of ig. 3(b) confirms that the as-grown ZnO NRs have a perfect hexag- nal surface throughout their lengths and a preferential growth of exagonal ZnO NRs along the c-axis direction, which is consistent ith the XRD result. The typical average diameter and length of the

s-grown nanorods were measured to be 30 ± 5 and 600 ± 50 nm, espectively. It is believed that seed diameter in the template, reac- ion time, concentration of the precursors, and temperature play mportant roles in the utilization the hydrothermal conditions and he neutralization of chemical complexation in the basic medium o control the dimension of the ZnO NRs and its growth process

11,18].

The smaller ZnO seed particle sizes (see Fig. S1) in the template ight be responsible for the formation of ZnO nuclei at a smaller

iameter of the ZnO NRs, whereas the reaction time controlled

experimental setup at certain bending angles.

the length of the nanorods in the hydrothermal process. It was calculated that the nanorods were grown in length at a growth rate of ∼2.5 nm/min. Moreover, proper hydrothermal conditions substantially controlled the interfacial tension and caused the pref- erential anisotropic growth along c-axis direction. The 1:1 molar ratio of the zinc-nitrate-hexahydrate/hexa-methylene-tetramine and the synthesis temperature in the hydrothermal process might be responsible for the growth of ZnO NRs along 〈0 0 1〉 direction, whereby the side surfaces of the ZnO NRs were enclosed by (2̄ 1̄ 1 0) [11,30].

The sensor was bent (∼40–60◦ bending angle) several times, and no breaking of the grown structure was observed less than 105 times bending/relaxing. This phenomenon indicates excellent stability of the sensor and excellent adhesion of the ZnO NRs to the PI film. Fig. S2 shows the condition of ZnO NRs arrays structure after 102, 103, 104, and 105 times bending and relaxing processes. Neither any significant fracture nor any variation of the structure was observed up to 104 times of deformation. However, after 105

bending cycles, the ZnO NRs showed breakage and less adhesion to the polyimide film, as shown in Fig. S2(d) and Fig. 3(c), both of which affect sensor performance.

The elemental analysis of the as-synthesized Ag-loaded ZnO NRs was carried out using an EDS, and the result is shown in Fig. 3(d). The presence of various well-defined peaks confirmed the exist- ence of Zn, oxygen (O), and Ag in the final product. Absence of any unwanted peaks from intermediates confirmed the formation of a high purity final product, and matched the XRD results.

TEM and high-resolution TEM (HRTEM) analysis of the 8 s Ag-loaded ZnO NRs was carried out for detailed morphological investigations and the results are shown in Fig. 4. It is observed that the measured dimensions of ZnO NRs from FESEM analysis are perfectly matched with the TEM observations. Fig. 4(a) and the inset of Fig. 4(a) clearly show that Ag NPs with an average diam- eter of 15 ± 5 nm were closely attached to the surface of the ZnO NRs. The lattice fringes with an interplanar spacing of 0.52 nm and 0.28 nm correspond to the (0 0 2) and (1 0 0) planes, respectively, of the hexagonal ZnO (Fig. 4(b)), and the spacing of 0.235 nm corre- sponds to the (1 1 1) plane of face centered cubic (fcc) Ag (Fig. 4(c)).

3.2. Gas sensor studies

The operating temperature has a significant influence on the fundamental sensing mechanism of metal oxide based gas sen- sors. In order to determine the optimum working temperature, the fabricated sensors were investigated at a temperature range

of 25–250 C. Fig. 5 depicts the relationship between the response magnitudes of the sensors and the operating temperature mea- sured at 100 ppm C2H2 gas concentration. All the tested samples showed a negligible response at a lower temperature. This can

A.S.M.I. Uddin et al. / Sensors and Actuators B 222 (2016) 536–543 539

F NRs, A loade

p c w p g d A i m f

o o d p t v

ig. 3. SEM images: (a) cross-section and (b) in-plane view of 8 s Ag-loaded ZnO g-loaded ZnO NRs. Inset: (a) side view of 8 s Ag-loaded ZnO NRs, (b) top view of un

robably be accounted for the potential barrier formed by the hemisorbed oxygen species (O2−, O−, O2−) on the sensing layer, hich prevented the C2H2 molecules from reacting at low tem- erature. With the increasing temperature, the response values radually increased, reached the maximum at 200 ◦C, and then ecreased with further increase in temperature. Compared to the g-loaded ZnO NRs, the unloaded ZnO NRs showed a continuously

ncreasing trend within the temperature range. This phenomenon ight be attributed to the high working temperature requirement

or the pure ZnO based sensors. The gas sensing property is strongly influenced by the capability

f the adsorption-desorption of the chemisorbed oxygen species n the sensing surface, which is closely related to the structural

efects of the sensing material. It was reported that the 〈0 0 1〉 polar lane of ZnO NRs is considered to be an exposed facet that con- ains more oxygen vacancies than other facets [31]. These oxygen acancies can deal with a higher number of oxygen atoms to form

Fig. 4. (a) TEM and (b, c) HRTEM imag

(c) unloaded ZnO NRs after 105 time bending/relaxing. (d) EDS spectra of the 8 s d ZnO NRs.

chemisorbed oxygen anions to react with C2H2 molecules and are able to return more trapped electrons to the ZnO surface, resulting in an abrupt change in the conductivity of the sensor. Addition- ally, directional long nanorods can provide a large surface area and pores on the sensor surface, and can help to trap more electrons on the ZnO surface. The size (diameter) of the ZnO NRs also plays an important role in the sensing mechanism. An improved surface- to-volume ratio and higher aspect ratio can be obtained due to the smaller diameter of the ZnO NRs, which leads to insufficient sur- face atomic coordination and high surface energy, resulting in more oxygen adsorption and thus better sensor response [32].

Moreover, due to the catalytic reactions, the interface between Ag NPs and ZnO NRs can generate additional charge carriers or

oxygen vacancies on the sensing surface. It is well known that metal-metal oxide surface conductivity is directly connected with the redox states of the metal additives [33]. In the Ag-loaded ZnO NRs, the ZnO surface might be changed from an electron depletive

es of the 8 s Ag-loaded ZnO NRs.

540 A.S.M.I. Uddin et al. / Sensors and Actuators B 222 (2016) 536–543

F A

s t t p t i m T v t c e o h h p m t r t r

r s C r a t C h a N A t

C r 1 c c t b i c G c

Fig. 6. Response versus C2 H2 concentration for 8 s Ag-loaded ZnO NRs at a working temperature of 200 ◦ C.

ig. 5. Temperature versus response relationship of unloaded and 6, 8, and 10 s g-loaded ZnO NRs to 100 ppm C2 H2 gas concentration.

tate to a nearly flat band state with a redox change of Ag. In addi- ion, Schottky barrier height between the grains of ZnO and Ag in he Ag-loaded ZnO NRs array is also responsible for better sensing erformances toward C2H2 at an elevated temperature [34,35]. Due o the addition of Ag nanoparticles the barrier height might be ncreased up to a certain level. The barrier height can be deter-

ined by work function difference of Ag cluster and ZnO grain. he work function of ZnO is 4.65 eV and the work function of sil- er nano-particles depends on their oxidation state (usually higher han ZnO (∼4.7–5.2 eV) [36]. In the Ag-loaded ZnO NRs array Ag lusters were in the form Ag/AgO in air atmosphere where AgO cov- red the surface of the clusters, which increased the work function f the Ag-loaded ZnO NRs system. At this stage the barrier height is igh and showed increased sensor resistance. The work function is ighly influenced by operating temperature. At an optimum tem- erature of 200 ◦C, in the presence of C2H2 the adsorbed oxygen olecules are removed from the cluster surfaces by reacting with

he gas. This decreases the work function. Thus the Schottky bar- ier height between Ag clusters and ZnO grains decreases leading to he reduction of sensor resistance and hence increases the sensor esponse [36].

The enhancement of the Ag-loaded ZnO NRs could also be egulated by adjusting the amount of Ag concentration for the sen- itive change of the oxidation state of Ag during the exposure of 2H2 gas. Fig. 5 reveals that 8 s Ag-loaded ZnO NRs has a higher esponse (13.8–100 ppm) than all other tested samples (3.21, 9.33, nd 9.57 for the unloaded, 6 s, and 10 s Ag-loaded samples, respec- ively, to 100 ppm C2H2). With the increasing Ag loading time, the 2H2 molecules’ adsorption-desorption kinetic was facilitated, and ence showed an enhanced response value. On the other hand, with

further Ag loading time, a thick layer of Ag might form on the ZnO Rs surface, resulting in a less-effective ZnO open surface porosity. s a consequence, the high density electron cloud might obscure

he ability to detect gas, thereby lowering the response magnitude. The responses of 8 s Ag-loaded ZnO NRs in terms of varied

2H2 concentrations at 200 ◦C are presented in Fig. 6. A maximum esponse of 27.2 (1000 ppm) and broad detection range from 3 to 000 ppm were obtained. The linearity behavior of the response urve (from 100 to 1000 ppm) indicates that the fabricated sensor an be more suitable for the detection of C2H2 at high concentra- ions. The sensitivity of the sensor can be represented empirically y the formula Sg = Ag(Pg)k, where Sg denotes the gas sensitivity, Pg

s the target gas partial pressure (which is proportional to the gas oncentration), Ag is a prefactor, and k is the exponent on Pg [37]. enerally, the exponent k has an ideal value of around 1 for single harged and 0.5 for double charged oxygen adsorbed species. This is

Fig. 7. Response variation of Ag-loaded ZnO NRs for various C2 H2 concentrations at 200 ◦ C at different bending angles.

derived from the surface interaction between chemisorbed oxygen anions and reducing gas. In this case, the value of k was found to be about 0.512 ± 0.004. The deviation from the ideal value is probably due to the agglomeration of nanostructures or less sensitive areas in some specific sensing surface in the tested sample.

In order to examine the mechanical flexibility, bending tests of the fabricated sensors were carried out at various curvature angles from 0◦ to 90◦ using a U-shaped support-stand for different C2H2 concentrations at 200 ◦C. Fig. 7 depicts the response variation of 6, 8, and 10 s Ag-loaded ZnO NRs’ sensors with different C2H2 concen- trations (10, 50, 100, 300, 500, 700, and 1000 ppm) as a function of bending angles (�c). As the bending increases, the strain between the PI substrate and the sensing layer results in a minor change in the surface resistance. The initial response magnitude (unbent or �c = 0◦) was measured as 9.33, 13.8, and 9.57 for 6, 8, and 10 s Ag-loaded sensor, respectively, for 100 ppm C2H2. The response magnitudes for all samples were retained without any apprecia- ble degradation up to a bending angle of �c = 30◦ (bending radius ∼12 mm). With the increase of bending angle, the sensor showed a negligible response slump of 1.3% for �c = 45◦, 1.7% for �c = 60◦, and 2.1% for �c = 90◦. This drop possibly attributed to the low bind- ing energy and low charge transfer between strained sensing layer atoms and target gas molecules at higher bending angles. More-

over, at a higher bending state, the ohmic contact distance might be increased, resulting in a slight increase in resistance, and hence a minute drop in response magnitude.

A.S.M.I. Uddin et al. / Sensors and Actuators B 222 (2016) 536–543 541

F t

a t v c m r i l a 6 t s o f

s ( c a i a m e s a n

F l

free electrons to form oxygen adsorbates (O−). At an elevated tem-

ig. 8. Dynamic response of 8 s Ag-loaded ZnO NRs for 3–30 ppm C2 H2 concentra- ions at 200 ◦ C at different bending angles.

The dynamic responses of the 8 s Ag-loaded ZnO NRs sensor t different bending angles are shown in Fig. 8. The figure reveals hat after imposing bending deformation, the deviation in response alue is less significant for a lower gas concentration (3–30 ppm) ompared to a higher concentration (50–1000 ppm). Conversely, a inute enhancement on the response and recovery time of the fab-

icated sensor was obtained with increasing bending angles. Fig. 9 llustrates the response-recovery time characteristics of the 8 s Ag- oaded ZnO NRs at unbent, 45◦, and 90◦. With an increase in bending ngle from 0◦ to 45◦, the response/recovery time reduced from 2/39 s to 58/37 s. To further increase the bending angle (from 45◦

o 90◦), the response/recovery time was reduced to 57/36 s. This light change might have contributed to the low binding energy f the sensor surface that occurred at the bending state, which acilitated the trapping and leaving of C2H2 molecules.

The reliability and the mechanical robustness of the flexible ensor were evaluated through repetitive bending deformation bending/relaxing) with the bending angle of ∼40–60◦ up to 5 × 105 ycles. Fig. 10 shows the real time results of the fatigue test fter several bending and relaxing processes. No significant drop n response value was observed, even for the repetitive bending nd relaxing up to 5 × 104 cycles, which indicates the excellent echanical durability and robustness of the fabricated sensor. This

nhancement might be attributed to the integration of the Ag-ZnO tructure on the PI substrate that allowed superior physical binding nd adhesion to the support substrate. However, the response mag- itude decreased gradually by about 33.7% as the bending/relaxing

ig. 9. Dynamic response and response/recovery time characteristics of the 8 s Ag- oaded ZnO NRs toward 100 ppm C2 H2 at 200 ◦ C at different bending angles.

Fig. 10. Reliability test of the flexible sensor (8 s Ag-loaded ZnO NRs) at 200 ◦ C after several times bending and relaxing.

increased up to 1 × 105 cycles. Factors of the flexibility test and the sensor’s performance deviation are summarized in Table 1 in detail for better clarity. When the process was continued to 5 × 105 cycles, the response magnitude was significantly reduced by about 62.3%. This phenomenon might be attributed to the reduced adhe- sion between the sensing layer and the support substrate, and/or broken sensing layer structure (as shown in Fig. 3(c)).

The selectivity of the sensor was examined by exposing 1000 ppm H2, CO, CO2, C2H2, C3H8, O2, and NO2 to the 8 s Ag-loaded ZnO NRs at an optimum temperature of 200 ◦C. Fig. 11 shows the selectivity histogram of the unloaded and 8 s Ag-loaded ZnO NRs sensors. As expected, the fabricated sensor showed highly selective C2H2 sensing behavior, which might be attributed to the inherent catalytic property of Ag and high selective absorption with the C2H2 molecules.

The chemical sensitization mechanism is mainly responsible for the high selectivity of Ag-loaded ZnO NRs array toward C2H2 gas. Ag has good catalytic activity and serves as specific absorp- tion site to dissociate molecular oxygen (O2) and separation of H2 molecules due to spillover effect. When Ag-loaded ZnO NRs array is exposed to air, it absorbs more oxygen on the Ag-loaded ZnO NRs sensing surface than on pure ZnO surface, and captures more

perature of 200 ◦C, the catalytic activity of the Ag-loaded ZnO NRs toward C2H2 decomposition might be strongly facilitated. At this

Fig. 11. Selectivity histogram of the unloaded and 8 s Ag-loaded ZnO NRs toward 1000 ppm test gases at 200 ◦ C.

542 A.S.M.I. Uddin et al. / Sensors and Actuators B 222 (2016) 536–543

Table 1 Details of flexibility test analysis.

Response magnitude Response/recovery time

Measured value (Ra /Rg ) Deviation (%) Measured value (second) Change

Bending angle, �c (◦ ) 0 (flat) 13.8 – 62/39 – 30 13.8 ⊗ 62/39 ⊗ 45 13.62 1.3 ↓ 58/37 ↓ 60 13.56 1.7 ↓ 58/37 ⊗ 90 13.51 2.1 ↓ 57/36 ↓

Bending/relaxing cycle (∼40–60◦ bending angle) 0 13.8 - 62/39 - 1 × 102 13.8 ⊗ 62/39 ⊗ 1 × 103 13.6 1.44 ↓ 62/39 ⊗ 1 × 104 13.1 5 ↓ 63/41 ↑ 5 × 104 12.7 8 ↓ 64/42 ↑ 1 × 105 9.14 33.7 ↓ 69/46 ↑

↓ A aded

s b p f o l t C c s c t i a A n b ( a e d Z

4

fl m l 2 1 t f l a T t f w r

A

t

[

[

[

[

[

[

[

5 × 105 5.20 62.3 ll the measurements were carried out toward 100 ppm C2 H2 at 200 ◦ C for 8 s Ag-lo

tage, the reaction (Ag/ZnO to C2H2 at certain temperature) might e an exothermic and spontaneous process, which preferentially romoted the Ag-loaded ZnO NRs to achieve the sufficient sur- ace energy to deal with the bond energy of C2H2 (bond energy f C2H2 is ∼490 kJ/mol). The first C-H bond energy in C2H2 is quite arge, because the carbon hybridization energy is shared by both he C C and C H bonds [38–40]. The required energy to break the 2H2 bond in the surface reaction between Ag-loaded ZnO and C2H2 ould not be supplied at temperatures below 200 ◦C, and hence the urface reaction of C2H2 with adsorbed oxygen species has been onsidered as the dominant sensing mechanism at 200 ◦C. In addi- ion, due to the higher bond energies compared to C2H2 resulted n less sensitivity toward CO (∼1076.5 kJ/mol), CO2 (∼1498 kJ/mol), nd C3H8 (∼3312 kJ/mol) [39–41]. However, the reactions between g-loaded ZnO NRs and other test gases are endothermic and can- ot be spontaneous, resulting less sensor response. Though the ond energies of H2 (∼436 kJ/mol), O2 (∼498 kJ/mol), and NO2 ∼312.7 kJ/mol) [39–41] are less than or close to the C2H2, inter- ction strength between the sensing layer and target gas is an ffective factor on the sensor response [42]. However, at 200 ◦C, ue to excessive surface energy the work function of the Ag-loaded nO mismatched, which resulted low response toward these gases.

. Conclusions

In summary, Ag-loaded vertical ZnO NRs on a PI/PTFE based exible C2H2 sensor have been demonstrated with excellent echanical bendability, durability, and robustness as well as excel-

ent C2H2 sensing performance at a low operating temperature of 00 ◦C. The C2H2 sensing properties of the unloaded and 6, 8, and 0 s Ag-loaded ZnO NRs were investigated extensively and sys- ematically. Mechanical flexibility and reliability test results of the abricated sensor showed a stable response magnitude with a neg- igible drift of ∼2.1% for a maximum bending angle of 90◦ and

response drop of 8% after 5 × 104 bending/relaxing processes. he negligible variations in the sensing performance characteris- ics after imposing mechanical deformation demonstrate that the abricated sensor can be used in various engineering applications ith conformal wrapping on curved surfaces and with mechanical

obustness.

cknowledgement

This work was supported by the Basic Science Research Program hrough the National Research Foundation of Korea (NRF) funded

[

71/50 ↑ ZnO NRs sample. ↑ = increase; ↓ = decrease; ⊗ = no change.

in 2014 by the Ministry of Science, ICT and Future Planning (NRF- 2014R1A2A2A01002668).

Appendix A. Supplementary data

Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.snb.2015.08.106.

References

[1] M.C. McAlpine, H. Ahmad, D. Wang, J.R. Heath, Highly ordered nanowire arrays on plastic substrates for ultrasensitive flexible chemical sensors, Nat. Mater. 6 (2007) 379–384.

[2] J. Yi, J.M.W. Lee, I.I. Park, Vertically aligned ZnO nanorods and graphene hybrid architectures for high-sensitive flexible gas sensors, Sens. Actuators B 155 (2011) 264–269.

[3] S. Claramunt, O. Monereo, M. Boix, R. Leghrib, J.D. Prades, A. Cornet, P. Merino, C. Merino, A. Cirer, Flexible gas sensor array with an embedded heater based on metal decorated carbon nanofibres, Sens. Actuators B 187 (2013) 401–406.

[4] W.A. MacDonald, Engineered films for display technologies, J. Mater. Chem. 14 (2004) 4–10.

[5] S. Walia, C.M. Shah, P. Gutruf, H. Nili, D.R. Chowdhury, W. Withayachumnankul, M. Bhaskaran, S. Sriram, Flexible metasurfaces and metamaterials: a review of materials and fabrication processes at micro- and nano-scales, Appl. Phys. Rev. 2 (2015) 011303.

[6] Z. Wu, D. Wu, W. Yang, R. Jin, Preparation of highly reflective and conductive metalized polyimide films through surface modification: processing, morphology and properties, J. Mater. Chem. 16 (2006) 310–316.

[7] X.F. Chu, D.L. Jiang, B.D. Aleksandra, H.L. Yu, Gas-sensing properties of thick film based on ZnO nano-tetrapods, Chem. Phys. Lett. 401 (2005) 426–429.

[8] J.X. Wang, X.W. Sun, Y. Yang, C.M.L. Wu, N-P transition sensing behaviors of ZnO nanotubes exposed to NO2 gas, Nanotechnology 20 (2009), 465501 (4pp).

[9] S.D. Shinde, G.E. Patil, D.D. Kajale, V.B. Gaikwad, G.H. Jain, Synthesis of ZnO nanorods by spray pyrolysis for H2 S gas sensor, J. Alloys Compd. 528 (2012) 109–114.

10] C. Gu, L. Shanshan, J. Huang, C. Shi, J. Liu, Preferential growth of long ZnO nanowires and its application in gas sensor, Sens. Actuators B 177 (2013) 453–459.

11] D. Polsongkram, P. Chamninok, S. Pukird, L. Chow, O. Lupan, G. Chai, H. Khallaf, S. Park, A. Schulte, Effect of synthesis conditions on the growth of ZnO nanorods via hydrothermal method, Phys. B 403 (2008) 3713–3717.

12] X. Liu, X. Wu, H. Cao, R.P.H. Chang, Growth mechanism and properties of ZnO nanorods synthesized by plasma-enhanced chemical vapor deposition, J. Appl. Phys. 95 (2004) 3141–3147.

13] M.H. Huang, Y. Wu, H. Feick, N. Tran, E. Weber, P. Yang, Catalytic growth of zinc oxide nanowires by vapor transport, Adv. Mater. 13 (2001) 113–116.

14] L.F. Xu, Q.W. Chen, D.S. Xu, Hierarchical ZnO nanostructures obtained by electrodeposition, J. Phys. Chem. C 111 (2007) 11560–11565.

15] J. Guo, J. Zhang, M. Zhu, D. Ju, H. Xu, B. Cao, High-performance gas sensor based on ZnO nanowires functionalized by Au nanoparticles, Sens. Actuators B 199 (2014) 339–345.

16] Q. Xiang, G. Meng, Y. Zhang, J. Xu, P. Xu, Q. Pan, W. Yu, Ag nanoparticle embedded-ZnO nanorods synthesized via a photochemical method and its

gas-sensing properties, Sens. Actuators B 143 (2010) 635–640.

17] L.C. Tien, H.T. Wang, B.S. Kang, F. Ren, P.W. Sadik, D.P. Norton, S.J. Pearton, J. Lin, Room-temperature hydrogen-selective sensing using single Pt-coated ZnO nanowires at microwatt power levels, Electrochem. Solid-State Lett. 8 (2005) 230–232.

nd Act

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

A.S.M.I. Uddin et al. / Sensors a

18] T.-R. Rashid, D.-T. Phan, G.-S. Chung, A flexible hydrogen sensor based on Pd nanoparticles decorated ZnO nanorods grown on polyimide tape, Sens. Actuators B 185 (2013) 777–784.

19] N. Singh, A. Ponzoni, R.K. Gupta, P.S. Lee, E. Comini, Synthesis of In2 O3 -ZnO core-shell nanowires and their application in gas sensing, Sens. Actuators B 160 (2011) 1346–1351.

20] G. Lu, J. Xu, J. Sun, Y. Yu, Y. Zhang, F. Liu, UV-enhanced room temperature NO2 sensor using ZnO nanorods modified with SnO2 nanoparticles, Sens. Actuators B 162 (2012) 82–88.

21] P. Rai, Y.-S. Kim, H.-M. Song, M.-K. Song, Y.-T. Yu, The role of gold catalyst on the sensing behavior of ZnO nanorods for CO and NO2 gases, Sens. Actuators B 165 (2012) 133–142.

22] X. Wang, M. Zhao, F. Liu, J. Jia, X. Li, L. Cao, C2 H2 gas sensor based on Ni-doped ZnO electrospun nanofibers, Ceram. Int. 39 (2013) 2883–2887.

23] N. Tamaekong, C. Liewhiran, A. Wisitsoraat, S. Phanichphant, Acetylene sensor based on Pt/ZnO thick films as prepared by flame spray pyrolysis, Sens. Actuators B 152 (2011) 155–161.

24] E. Austin, A.V. Brakel, M.N. Petrovich, D.J. Richardson, Fiber optical sensor for C2 H2 gas using gas-filled photonic band gap fiber reference cell, Sens. Actuators B 139 (2009) 30–34.

25] L. Zhang, J. Zhao, J. Zheng, L. Li, Z. Zhu, Hydrothermal synthesis of hierarchical nanoparticle-decorated ZnO microdisks and the structure-enhanced acetylene sensing properties at high temperatures, Sens. Actuators B 158 (2011) 144–150.

26] A.S.M.I. Uddin, G.-S. Chung, Synthesis of highly dispersed ZnO nanoparticles on graphene surface and their acetylene sensing properties, Sens. Actuators B 205 (2014) 338–344.

27] A.S.M.I. Uddin, D.-T. Phan, G.-S. Chung, Low temperature acetylene gas sensor based on Ag nanoparticles-loaded ZnO-reduced graphene oxide hybrid, Sens. Actuators B 207 (2015) 362–369.

28] K.-W. Lee, A.S.M.I. Uddin, D.-T. Phan, G.-S. Chung, Fabrication of low-temperature acetylene gas sensor based on Ag nanoparticles-loaded hierarchical ZnO nanostructures, Electron. Lett. 51 (2015) 572–574.

29] A.S.M.I. Uddin, K.-W. Lee, G.-S. Chung, Acetylene gas sensing properties of an Ag-loaded hierarchical ZnO nanostructure-decorated reduced graphene oxide hybrid, Sens. Actuators B 216 (2015) 33–40.

30] L. Vayssieres, Growth of arrayed nanorods and nanowires of ZnO from aqueous solutions, Adv. Mater. 15 (2003) 464–466.

31] G.R. Li, T. Hu, G.L. Pan, T.Y. Yan, X.P. Gao, H.Y. Zhu, Morphology-function relationship of ZnO: polar planes, oxygen vacancies, and activity, J Phys. Chem. C 112 (2008) 11859–11864.

32] T.-R. Rashid, D.-T. Phan, G.-S. Chung, Effect of Ga-modified layer on flexible hydrogen sensor using ZnO nanorods decorated by Pd catalysts, Sens. Actuators B 193 (2014) 869–876.

33] M. Batzill, U. Diebold, The surface and materials science of tin oxide, Prog. Surf. Sci. 79 (2005) 47–154.

34] S.M. Majhi, P. Rai, Y.-T. Yu, Facile approach to synthesize Au@ZnO core–shell nanoparticles and their application for highly sensitive and selective gas sensors, ACS Appl. Mater. Inter. 7 (2015) 9462–9468.

35] Y.S. Shim, L. Zhang, D.H. Kim, Y.R. Choi, S.H. Nahm, C.Y. Kang, W. Lee, H.W.

Jang, Highly sensitive and selective H2 and NO2 gas sensors based on surface-decorated WO3 nanoigloos, Sens. Actuators B 198 (2014) 294–301.

36] Z.K. Horastani, S.M. Sayedi, M.H. Sheikhi, E. Rahimi, Effect of silver additive on electrical conductivity and methane sensitivity of SnO2 , Mater. Sci. Semicond. Process. 35 (2015) 38–44.

uators B 222 (2016) 536–543 543

37] R.W.J. Scott, S.M. Yang, G. Chabanis, N. Coombs, D.E. Williams, G.A. Ozin, Tin dioxide opals and inverted opals: near-ideal microstructures for gas sensors, Adv. Mater. 13 (2001) 1468–1472.

38] C.W. Bauschlicher Jr., S.R. Langhoff, Theoretical study of the C-H bond dissociation energies of CH4 , C2 H2 , C2 H4 , and H2 C2 O, Chem. Phys. Lett. 177 (1991) 133–138.

39] T.W. Swaddle, Inorganic Chemistry: An Industrial and Environmental Perspective, Academic Press, USA, 1997.

40] T.L. Cottrell, The Strengths of Chemical Bonds, 2nd ed., Academic Press, New York, USA, 1961.

41] S.J. Blanksby, G.B. Ellison, Bond dissociation energies of organic molecules, Acc. Chem. Res. 36 (2003) 255–263.

42] Z.S. Hosseinia, A. Irajizad, A. Mortezaali, Room temperature H2 S gas sensor based on rather aligned ZnO nanorods with flower-like structures, Sens. Actuators B 207 (2015) 865–871.

Biographies

A.S.M. Iftekhar Uddin received his B.Sc. Eng. from the Faculty of Engineering, Inter- national Islamic University Chittagong, Chittagong, Bangladesh, in 2005 and M.E. from the School of Electrical Engineering, University of Ulsan, Ulsan, South Korea, in 2015. He joined as lecturer in Sylhet International University, Sylhet, Bangladesh, in 2006 and promoted as Assistant professor in 2010. He is now working as a Ph.D. candidature in the School of Electrical Engineering, University of Ulsan, Ulsan, South Korea. His research interests include metal/metal oxide and graphene, localized surface plasmon resonance (LSPR) based nanosensors, and flexible nanosensors.

Usman Yaqoob received his B.Sc. from the School of Electronics Engineering, Inter- national Islamic University Islamabad, Islamabad, Pakistan, in 2013. He is now working as a Ph.D. candidature in the School of Electrical Engineering, Univer- sity of Ulsan, Ulsan, South Korea. His research interests include metal/metal oxide, CNTs, graphene composites based flexible nanosensors, and self-powered wearable nanosensors based on triboelectric and vibration-driven nanogenerators.

Duy-Thach Phan received his B.E. from the School of Electrical Engineering, Hochim- inh University of Technology, Ho Chi Minh, Vietnam, in 2008, M.E. and Ph.D. degree from the School of Electrical Engineering, Ulsan University, Ulsan, South Korea, in 2010 and 2015, respectively. He is now working as a post-doc fellowship in the School of Electrical Engineering, University of Ulsan, Ulsan, South Korea. His research interests include SiC, ZnO, AlN based on SAW sensors, graphene-based sensors, MEMS-based sensors, FEM and atomistic scale modeling.

Gwiy-Sang Chung received his B.E. and M.E. Degrees in Electronic Engineering from Yeungman University, Kyongsan, South Korea, in 1983 and 1985, respectively, and his Ph.D. Degree from Toyohashi University of Technology, Toyohashi, Japan, in 1992. He joined the Electronics and Telecommunications Research Institute (ETRI), Daejon, South Korea, in 1992, where he worked on Si-on-insulator materials and devices. Moreover, he also worked as a visiting scholar at UC Berkeley and Stan- ford University, CA, USA, in 2004 and 2009, respectively. He is now a professor in

the School of Electrical Engineering, University of Ulsan, Ulsan, South Korea. His research interests include Si, SiC, ZnO, AlN-M/NEMS, flexible self-powered wireless sensors nodes, energy harvesting, localized surface plasmon resonance (LSPR), and graphene/MoS2 -based composites. He is the author or co-author of more than 130 scientific and technical SCI international journal papers.