Chemical Sensors

Chemical Sensors

Chemical Sensors 2015, 5: 9

Gas sensing characteristics of nanostructured ZnO thin film influence of manganese doping Azhagia Nambi Gopala Krishnan#, Ganesh Kumar Mani#, Prabakaran Shankar#, Brahmaiah Vutukuri, John Bosco Balaguru Rayappan* Centre for Nanotechnology & Advanced Biomaterials (CeNTAB) and School of Electrical & Electronics Engineering (SEEE) SASTRA University, Thanjavur - 613 401, India # The authors have contributed equally to this work *

Author for correspondence: John Bosco Balaguru Rayappan, email: [email protected] Received 20 Jun 2014; Accepted 15 Aug 2014; Available Online 15 Aug 2014

Abstract Undoped and manganese (Mn)-doped nanostructured ZnO thin films were deposited on glass substrates using spray pyrolysis technique at the substrate temperature of 523 K and subsequently annealed at 673 K for 3 h. The deposited films were found to be polycrystalline in nature with hexagonal wurtzite structure. The FE-SEM images showed the influence of Mn doping on the surface morphology of the uniformly deposited ZnO nanograins. The transmittance of the film was found to be diminished as the concentration of Mn was increased. The room temperature sensing characteristics of the doped and undoped ZnO films were studied towards acetaldehyde, hexanol, ammonia, acetone, ethanol and monoethanolamine vapours. The results revealed that the sensing element was highly selective towards ammonia vapour. High sensing response of 1250 was achieved for 0.008 M Mn-doped ZnO sample with the response and recovery time of 60 and 68 s respectively. Keywords: Thin films; Spray pyrolysis; Nanostructured ZnO; Doping; Gas sensor

1. Introduction The global production of ammonia during 2012 was estimated to be 198 million tons. Ammonification, nitrification, combustion from chemical plants and motor vehicles are the main sources of ammonia. It is extensively used in the production of nitrogenous fertilizer, industrial coolant and in explosive industries. In this perspective high performance, low cost and low power ammonia sensors can find many applications in the environmental quality analysis, explosive detection, mining, automotive industries, petrochemical industries, etc. [1]. The detection of ammonia in industries is necessary to ensure safety of the employees. Several agencies have explored the ammonia exposure limits in industries categorized such as permissible exposure limit (PEL – 50 ppm for 8 h workday), threshold limit value (TLV – 25 ppm for 8 h workday), recommended exposure value (REL - 25 ppm for 8 h workday), short term exposure limit (STEL – 35 ppm for 15 min during workday), immediately dangerous to life or health concentration (IDLH – 300 ppm) [2]. Many types of ammonia sensors like electrochemical, metal oxide, surface acoustic wave (SAW), surface plasmon resonance (SPR), catalytic, conducting polymer, spectrometric sensors are available. But the major attention has been given to metal oxides due to their excellent sensitivity, selectivity, easy fabrication and compatibility with electronics circuits [3–5]. Over the past few decades, ZnO has become an important metal oxide material due its wide tunability of structural, electrical and optical properties [6–11]. Besides being cheap and abundant in nature, its multitude of advantages such as large band gap, thermal/chemical stability, high catalytic nature and biocompatibility, ZnO has found itself in wide applications ranging from transparent Cognizure www.cognizure.com/pubs

conductors, solar cell windows, photodiodes, heat mirrors and light emitting diodes, gas sensors, high radiation resistance and surface acoustic wave devices [12–19]. The material properties such as structural, morphological, optical, electrical and mechanical properties can be effectively tuned by adding appropriate dopants [20–29]. Amongst all the transition metals, manganese (Mn) possess a high equilibrium solubility in ZnO (i.e 13 at.%) [30] and Prashant et al. found that the maximum solubility limit is 10% [31]. Ahmed et al. [32] investigated the various properties of Mn-doped ZnO nanorods prepared by microwave hydrothermal method and reported the influence of Mn-doping in enhancing the aspect ratio of nanorods. He has also reported the increased oxygen sensing performance of Mn-doped ZnO nanorods at room temperature. Penga et al. [33] has investigated the sol-gel synthesized Mn-doped (6 at%) Zinc oxide nanopowders showed better response to relative humidity change at room temperature. Also reported the influence of various Mn-doping concentrations (1 and 3 at%) on the humidity sensing performance. Yuzhen Mao et al. [34] reported the effect of 2 wt.% Mn-doping on the acetone sensing characteristics of ZnO nanofibers. This sensing element showed a good response with faster response and recovery time of 17 s and 4 s at 340oC. Again Sandip et al. [35] and Chatterjee et al. [36] studied the effect of Mn-doping on ZnO and confirmed the role of Mn dopants in enhancing the sensitivity. ZnO thin films can be deposited by several methods like thermal evaporation, sputtering, metal organic chemical vapour deposition, spray pyrolysis, sol gel and pulsed laser deposition. However, spray pyrolysis technique stands out among the various techniques because of its cost effectiveness for large scale deposition of homogeneous films with good 1

© Cognizure. All rights reserved.

Chemical Sensors


Table 1. Optimized deposition parameters. Parameters Precursor

Values Zinc acetate dihydrate (0.1 M) Manganese acetate tetrahydrate (0.002 M to 0.01 M in steps on 0.002 M) 523 K Deionized water (50 ml) 2.5 mbar 2 ml.min-1 15 cm 90º 0.2 mm 723 K

Dopant Substrate temperature Solvent Carrier gas pressure Solution flow rate Distance between nozzle and substrate Spray angle Spray nozzle diameter Post annealing temperature

stoichiometry control and easy incorporation of dopants [37– 39]. In this work, the structural, morphological, optical, electrical and room temperature ammonia sensing properties of spray pyrolysis deposited undoped and various concentrations of Mn-doped ZnO thin films were investigated.

measured from weighing method using digital weigh balance (AX 200, Shimadzu, Japan). Film thickness was calculated using the density relation (Eq. 2),

2. Experimental Details

where, Δw is the difference in weight of the substrate before and after film deposition, ρ is the density of ZnO (5.61 g cm-3), l and b are the bare length and breadth of the substrate respectively. Thickness of the films were found to be 450 nm. Morphological features were observed using Field Emission Scanning Electron Microscope (JOEL, JSM-6701F, Japan). The Energy Dispersive X-Ray (EDX) analysis for elemental composition detection was investigated along with FE-SEM system. Optical studies were carried out using UV-Vis spectrophotometer (Perkin Elmer, Lambda 25, USA). Electrical and room temperature gas sensing properties were investigated using an electrometer (Keithley, 6517A, USA) [40,41].

2.1. Film deposition The undoped and Mn-doped ZnO thin films were deposited on ultrasonically cleaned glass substrates using a fully automated spray pyrolysis system (HOLMARC, HO-THO4, India). The 0.1 M of precursor solution for forming undoped ZnO films was prepared by adding zinc acetate dehydrate in 50 ml of deionized water. The solution was stirred with the help of magnetic stirrer for 1 h to obtain homogenous solution. Two drops of acetic acid were also added to prepare a clear solution. Manganese doping was achieved by mixing various concentrations of manganese acetate tetrahydrate (0.002 M to 0.010 M in steps of 0.002 M) along with the zinc acetate dihydrate (0.1 M) in 50 ml of deionized water. The undoped and doped samples are labeled as Mn0, Mn1, Mn2, Mn3, Mn4 and Mn5 respectively. All the glass substrates were ultrasonically cleaned with acetone and deionized water subsequently dried in hot air oven for 1 h. The cleaned substrates were then positioned on the heater maintained at 523 K to deposit the films. Precursor solution was filled in the solution chamber and then sprayed over the preheated substrate using compressed air. The optimized deposition parameters are given in Table 1. The sprayed precursor mist was decomposed to form thin film when reached the hot substrate. The prepared thin films were annealed at 673 K for 3 h to remove the incompletely decomposed precursor salts and also to increase the crystallinity [40]. 2.2. Characterization techniques Structural characteristics of both the Mn-doped and undoped films were studied using X-Ray Diffractomer (X’Pert Pro, Panalytical, Netherlands) with Cu K α radiation (λ=1.5408 Ǻ). The average crystallite size was calculated using the Scherrer’s formula (Eq. 1), (1) where, k is the shape factor, λ is the wavelength of the X-ray, β is the full width half maximum of the diffraction peak and θ is the angle of the diffraction peak. The film thickness was

(2)

3. Results and Discussion 3.1. Structural studies XRD patterns of the undoped and Mn-doped ZnO thin films are shown in Figure 1 (a). The peaks detected are in agreement with the JCPDS card no. 36-1451 and confirmed the formation of nanostructured polycrystalline ZnO thin films with hexagonal wurtzite crystal structure. The absence of secondary phases in the case of Mn-doped ZnO thin films suggests that the dopant concentration was within the solubility limit [30,31]. The intensity (Figure 1 (b)) of the XRD peaks was found to be decreased with increase in doping concentration which might be due to the deterioration of the crystalline quality of ZnO film caused by the substitution of Zn2+ ions by Mn2+ ions. A similar trend was observed by Nirmala et al and few others [42]. XRD data showed the diffraction peaks corresponding to (100), (002), (101), (102), (110), (103) and (200) planes. The (101) plane was determined to be the most preferential oriented plane as it was found to have the highest texture co-efficient computed using the equation proposed by Barret and Massalski [43]. Increase in the FWHM of (101) crystal plane with doping concentration confirmed the reduction in the crystallite size as shown in Figure 1 (c). The d-spacing for all the undoped and doped thin films was found to be less than the standard value (d = 2.6022 Å). The strain values were found to be constantly increasing with Mn-doping (Figure 1 (d)). This might be due to the

2 © Cognizure. All rights reserved.

Cognizure www.cognizure.com/pubs

Chemical Sensors


Figure 1. a) XRD patterns, b) intensity trend with respect to dopant concentration, c) crystallite size and d) strain values of undoped and Mn-doped ZnO thin films.

smaller divalent Mn substitution in the place of larger divalent Zn atoms in the tetrahedral configuration [44].

might be due to the sputter coated gold elements prior to FESEM measurements to eliminate charging effects.

3.2. Morphological studies The morphological features of the undoped and Mndoped ZnO thin films were observed from FE-SEM images which are shown in Figure 2 (a-f). The undoped ZnO samples showed a homogenous, loosely packed spherical shaped nanograins with large amount of pores. But for the doped films, the grain size was found to be continuously decreased, dispersed and agglomerated with increase in Mn-dopant concentration. At higher dopant concentrations, complete loss of the uniformity can be seen due to increased agglomeration. The size of the undoped ZnO particles were found to be around 25 nm. For Mn-doped samples the particles size could not be measured from SEM images because of agglomerated grains. The grain size observed from the FE-SEM data consistent with the XRD data. Figure 3 (a & b) show the EDX spectra of undoped (Mn0) and Mn-doped ZnO (Mn5) samples, which clearly depicts the presence of Zn, Mn and O elements in the undoped and doped thin films. Presence of Au atoms

3.3. Optical studies Optical absorption properties of the undoped and Mndoped ZnO thin films were investigated in the wavelength range of 300 to 800 nm. From the absorbance spectra of the samples shown in Figure 4 (a), one can observe the edge shift towards higher wavelength with the increase in Mn concentration. The decrease in transmittance with Mn-doping concentration might be due to the defects induced by Mndopants and increased grain boundary scattering owing to smaller crystallite size. Figure 4 (b) shows the Taucs’ plot from which the band gap values were estimated. The change in the band gap with Mn concentration is shown in Figure 4 (c) and it can be seen that the band gap was reduced from 3.12 to 2.65 eV. This trend may be due to the exchange interaction between Mn2+ ions and host ZnO, and also it can be theoretically explained as the second order perturbation as a result of spatial confinement as expected in low dimensional materials [22].



Chemical Sensors


(a)

(b)

(c)

(d)

(e)

(f)

Figure 2. FE-SEM images a) Mn0, b) Mn1, c) Mn2, d) Mn3, e) Mn4 and f) Mn5 thin film samples.

3.4. Electrical studies Electrical conductivity of the samples at room temperature was measured using an electrometer. The resistance (Figure 5) of the undoped ZnO thin film was found to be 1.90 x 1011 Ω and was reduced to 1.0 x 109 Ω as the Mn concentration was increased to 0.010 M. In general, resistance will increase with decrease in grain size due to the grain boundary resistance. But, in the present case, resistance was found to be decreased as the grain size decreased. It might be due to the influence of Mn-doping which led to an increase in the donor concentration in the films. Hence, the doping effect was dominated over the grain boundary effect. 3.5. Ammonia sensing studies The room temperature gas sensing characteristics of undoped and Mn-doped ZnO sensors were tested using the home made gas / vapour testing chamber which was described clearly in our previous reports [40,41,45]. Conducting silver paste and zero resistance copper wires were used to make electrical contacts over the film surface [40]. 3.5.1. Selectivity Selectivity is one of the key issues in the metal oxide based gas / vapour sensor. For appropriate usage of the sensor, the first entity to be determine is the selectivity. The undoped and Mn-doped ZnO thin films were tested with 100 ppm concentration of various vapours namely acetaldehyde, hexanol, ammonia, acetone, ethanol and monoethanolamine and the observed results are shown in Figure 6. All the sensors exhibited an excellent sensitivity towards ammonia vapour. The selectivity nature of the films towards ammonia might be due to the donation of two lone pair electrons of ammonia to the conduction band of the thin film surface [46]. Hence, the sensor showed better selectivity to ammonia than others.

Figure 3. EDX spectrum of a) undoped (Mn1) and b) Mn-doped ZnO (Mn5) thin films.

3.5.2. Transient response and recovery characteristics The sensing characteristics were tested in the range of 5 to 100 ppm concentration range of ammonia vapour and the observed response is shown in Figure 7 (a). The maximum response was achieved for Mn4 sample. As the dopant concentration was increased, the sensing response also was increased but up to 0.008 M of Mn (Mn4) and further increase in the dopant concentration lead to the decrease in response. The transient resistance response characteristics of Mn4 sample towards various concentrations of ammonia is shown in Figure 7 (b). The transient response of Mn4 sample towards 100 ppm of ammonia alone is shown in Figure 8 (a). The response and recovery time of Mn4 sample towards 5 to 100 ppm ammonia is shown in Figure 8 (b). Initially, at lower concentration levels the response time was very high due to less number of ammonia molecules present inside the chamber and it is very difficult to interact rapidly with the solid surface. As the concentration was increased, the response time was decreased and found to be 60 s for 100 ppm of ammonia. In the case of recovery, at lower concentrations, recovery time was very fast and slowed down as the ammonia concentration was increased. The long term stability of the sensing element (Mn4) was studied over a period of 60 days (Figure 9) and found to be excellent. 3.5.3. Sensing mechanism It is well known that ZnO is an n-type semiconductor. When it is exposed to air atmosphere, the atmospheric oxygen molecules are adsorbed on to the ZnO surface by capturing the



Chemical Sensors


Figure 5. Electrical resistance of the undoped and Mn-doped ZnO thin films.

Figure 6. Selectivity of the undoped and Mn-doped ZnO thin films.

(3)

Figure 4. a) Optical absorbance spectra, b) Tauc’s plot and c) band gap trend with reference to dopant concentration for the undoped and Mn-doped ZnO thin films.

conduction band electrons of ZnO resulted in high resistance. The sensing element was kept in air atmosphere to set the base line resistance for sensing measurements. Once the reducing gases like ammonia was injected into the sensing chamber, the interaction between ammonia and ZnO surface resulted in the following mechanism (Eq.3) where ammonia vapour got oxidized into N2 and H2O thus released electrons to the film surface. This in-turn resulted in the decrease in film resistance.


The increased room temperature sensing performance towards ammonia may be due to the enhancement of the surface catalytic effect of nanostructured ZnO thin film with higher surface to volume ratio [3,47]. The increased sensing response of the Mn-doped ZnO thin films might be due to the decreased crystallite size and enhanced electron density [33]. This may also be due to the excess donor concentrations induced by Mn-doping [22,33]. Since Mn2+ has lower ionization energy than that of ZnO, it might have influenced the oxygen adsorption / desorption process [34]. Hence Mndoped ZnO film showed an excellent response than undoped ZnO film. The sensing responses were found to be 240, 260, 305, 476, 1250 and 400 for Mn0, Mn1, Mn2, Mn3, Mn4 and Mn5 samples respectively. The response was found to be increased with an increase in Mn concentration and found to be maximum for Mn4 sample. Mn4 film with smaller spherical grains and unevenly distributed pores might be one of the reasons for the maximum sensing response towards ammonia vapour. Since smaller grains facilitates inner-grain interaction [48] with the target gas, the response has considerably 5

© Cognizure. All rights reserved.

Chemical Sensors


Figure 7. a) Response trend for the undoped and Mn-doped ZnO thin films and b) transient resistance response for Mn4 sample.

increased for the Mn4 sample. Further, the enhanced pore density might have strongly influenced the sensing performance of this sample [49,50]. Normally, sensing elements with smaller grain size will have an improved sensing response due to the enhanced inner grain interaction. But, in the present case, Mn5 sample with lower crystallite size showed a lower response than that of Mn4 sample. This might be due to the reduction in the mobility of electrons as a result of scattering of excessive carriers generated by Mn doping. Further, Katoch and Liu et al. have reported that the degradation of crystallinity might have an adverse effect in sensing performance [51–53]. In addition, the progressive desorption might also be the reason for the reduced response of the Mn5 sample [54,55]. These observations revealed the constructive and destructive influence of Mn doping on the sensing performance of ZnO and hence the role of dopant concentration in engineering the sensor performance. Furthermore, the low response and recovery time might be due to the enhanced catalytic activity of the smaller crystallites of the film [45]. In order to highlight the enhanced figure of merits of the Mn-doped ZnO thin film based room temperature ammonia sensor, the same has been compared with the existing literature (Table 2). This comparison highlights the enhanced performance of the present sensor in terms of response, response and recovery time than that of existing ammonia sensors. 4. Conclusions Nanostructured undoped and Mn-doped ZnO thin films were successfully deposited on glass substrates using

Figure 8. a) Transient resistance response curve towards 100 ppm of ammonia and b) response and recovery times for 5 to 100 ppm of ammonia for Mn4 sample.

Figure 9. Long term stability of Mn4 sample in air and 100 ppm of ammonia.

spray pyrolysis technique. The impact of manganese doping in structural, morphological and optical properties of the films were studied and correlated. The crystallite size was found to be decreased from 28 to 7 nm as the Mn-doping concentration was increased up to 0.010 M. The resistance of the ZnO thin film was decreased from 1.90 x 1011 Ω to 1 x 109 Ω upon increasing the dopant concentration from 0.002 M to 0.010 M. Room temperature sensing performance of all the undoped and Mn-doped ZnO samples were carried out towards 5 to 100 pm



Chemical Sensors


Table 2. Performance comparison of doped and undoped ZnO ammonia sensors. Material (Thin Film) ZnO ZnO ZnO Al-ZnO Mn-ZnO Ni-ZnO Ni-ZnO Pd-ZnO ZnO-Cr2O3 PANI-ZnO ZnO + Camphor + PANI F Graphite -ZnO

Concentration (ppm) 25 50 17 50 50 100 750 30 300 100 100 40-45

Response (S) 233 53.6 400 64.6 100 ~110 5 ~60 13.7 ~4.6 28 ~12

Response time (s) 20