Procedia Engineering 87 ( 2014 ) 963 – 966

Available online at www.sciencedirect.com

1877-7058 © 2014 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/3.0/).Peer-review under responsibility of the scientific committee of Eurosensors 2014doi: 10.1016/j.proeng.2014.11.318

ScienceDirect

EUROSENSORS 2014, the XXVIII edition of the conference series

CNT-Ni-Pd nanocomposite films for optical gas sensor

E.Czerwosz1*, E.Kowalska1, M.Kozłowski1, J.Radomska1, H.Wronka1, M. Angiola3, A. Martucci3, W.Włodarski 2

1Department of Nanotechnology, Tele&Radio Research Institute, Warsaw, Poland 2 School of Electrical and Computer System Eng., RMIT University, Melbourn, Australia

3Department of Industrial Engineering, University of Padova, Padova, Italy

Abstract

We present the optical properties of the nanocomposite nickel - carbonaceous (Ni-C) and carbon nanotubes, palladium-fullerene (CNT- Ni-Pd) films obtained by various methods. The initial film was prepared by physical vapor deposition (PVD) method using two separated sources: fullerene C60 (as carbon source) and nickel acetate (as metal source). Next, the initial film was modified by chemical vapor deposition (CVD) using xylene at temperature ~ 650°C, as a result carbon nanotubes containing inside Ni grains were synthesized. Then, on the carbon nanotubes, palladium-fullerene films were deposited using PVD process, these samples were named by us as (CNT- Ni-Pd) films. All (Ni-C) films and (CNT-Ni-Pd) films were studied by SEM. These nanostructures were analyzed by optical method using measuring the variation of the absorbance of different gases (H2, CO, NO2). © 2014 The Authors. Published by Elsevier Ltd. Peer-review under responsibility of the scientific committee of Eurosensors 2014.

Keywords: PVD; CVD; CNT; Ni; Pd; optical sensor

1. Introduction

Recently, carbon nanotubes (CNTs) due to their unique properties are applied as chemical sensors, biosensors and hydrogen storing materials. Their highly developed surface, many dangling bond and defects make CNTs an ideal

* Corresponding author. Tel.: +48 22 8316746; fax: +48 22 8319231.

E-mail address: elzbieta.czerwosz@itr.org.pl

© 2014 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/3.0/).Peer-review under responsibility of the scientific committee of Eurosensors 2014

964 E. Czerwosz et al. / Procedia Engineering 87 ( 2014 ) 963 – 966

material for different gases adsorption. Sensing properties of materials containing or consisting of nanotubes are associated with an change in CNTs conductivity or optical properties due to gas absorption. To improve sensing properties of such materials and enlarge number of detected gases the functionalization of carbon nanotubes with suitable metal nanoparticles is often proposed. For example, to improve hydrogen adsorption of CNTs doping of transition metals like Pd, Ni, Pt [1-3] or Ti [4] is applied. The most popular is palladium which demonstrate selective absorption of hydrogen gas and ability of formation a chemical species known as a palladium hydride.

Several methods have been developed for the functionalization CNTs by palladium nanoparticles like electroless deposition [5], thermal decomposition [6], vapor deposition [7], chemical reduction in supercritical CO2 solutions [8], impregnation [9] and electrodeposition [10], arc-discharge in solution process [11].

Nanostructural films containing carbon nanotubes covered with palladium are promising because of their surface highly developed and covered with palladium surface. This covering enhances sensing properties toward hydrogen, CO2, CO and ethylene.

In this paper we present results of studies of structure, morphology and topography of CNT-Ni-Pd films as well as the results of optical absorption measurements. Optical gas sensing properties of these nanocomposites films have been tested measuring the variation of the absorbance of different gases (H2, CO, NO2).

2. Results and discussion

2.1. Experimental

CNT-Ni-Pd films were prepared by multi-step method described in our patents [12,13]. These films were deposited on quartz substrates by PVD method. The process was performed under dynamic pressure of 10-5 Pa. Nickel acetate was used as a precursor of metal and fullerite C60 powder was applied as a precursor of carbon. All samples were prepared on fused silica. In PVD deposition method parameters such as current flowing through sources (IC60, INi), duration time (t) and the sources-substrates distances (d) affects the quality of structure and thickness of obtained films. The sample obtained with IC60 =1.9A and INi=1.1 A , t = 1min and d = 60mm was sensitive for CO (sample A).

Samples obtained by PVD process with parameters IC60 =1.9 A and INi=1.1A , t = 1 and 1.5min and d = 60mm were modified in a quartz tube reactor by CVD method using xylene C8H10 as an additional source of carbon for the growth of carbon nanotubes on Ni grains. CVD process was performed at atmospheric pressure in the temperature of 650°C applying argon flow (a rate of flow = 40l/h) as the carrier gas for xylene vapor. After CVD modification the carbon nanotubes film was grown. The feed rate of xylene (ν) was 0.1ml/2 min and duration of CVD was t=15 minutes for both samples. The amount of using xylene was 0.5 ml. In such a way the samples A1 and B1 were obtained. These samples were sensitive for CO and NO2. These last A1 and B1 samples were covered with palladium in PVD process performed in the same way as initial PVD process with this difference that in a place of nickel acetate was used palladium acetate. These films were marked as A2 and B2 respectively.

All these films were investigated by scanning electron microscopy (SEM) with JEOL JSM-7600F field emission scanning electron microscope operating at 5keV incident energy. The topography of film’s surface was imaging using secondary electrons (SE) detector. Low angel backscattered electrons (LABE) detector enabled to define samples morphology by observing changes of the contrast composition. The contrast depends on the atomic number of elements of the observed microstructure. Areas with higher atomic number are brighter, while those of lower atomic number are dark.

Optical gas sensing tests were conducted by measuring optical absorbance in the 250-2500 nm wavelength range on films deposited on SiO2 substrates using a Harrick gas flow cell coupled with a Jasco V-650 spectrophotometer. The operating temperature (OT) was varied between room temperature (RT) and 80 °C and different gas concentrations up to 1% volume of CO and H2 and up to 0.01% volume of NO2 in dry air were utilized. The substrate size was approximately 1 cm × 2 cm and the incident spectrophotometer beam was normal to the film surface and covered a 9 mm × 1.5 mm area of the film.

For all the samples it was first measured the transmittance in air and then in gas in order to determine the wavelengths at which the absorbance variation was larger. For such wavelengths, dynamic measurements have been performed in order to compare the response time for the different gases.

965 E. Czerwosz et al. / Procedia Engineering 87 ( 2014 ) 963 – 966

2.2. Results and discussion

SEM image of samples A1, B1, A2 and B2 are shown in Fig.1 a-d.

a) b) c) d)

Fig.1 SEM images of (a) A1, (b) B1, (c) A2, (d) B2 samples

Detail studies (in SE and LABE modes) of A1 sample showed that it is composed of short and broad nanotubes placed in carbonaceous matrix while the sample B1 contains longer and more defected nanotubes. Ni nanograins with diameter of few tens of nm are placed inside of nanotubes what is visible in back scattered electrons SEM image. SEM studies of samples A2 and B2 showed that covering with palladium and carbon A1 and B1 films smoothes a surface and makes CNTs almost invisible what can prevent different reactions.

Fig.3 Dynamic scan for samples: (a) A, (b) A1, (c) B1 at OT=80°C, λ=2325nm

For all the samples at the beginning the transmittance in air was measured next the measurements in gas were performed. The wavelengths at which the absorbance variation was highest was found in this way. For such wavelengths, dynamic measurements have been performed. The measurement of response time were performed for different gases. All the samples were tested at λ=2325nm at the temperature of 80oC.

For the sample A, we registered the response for CO (1%v/v) with a relative absorbance value 1,71*10-3 after 33s and a recovery of 2,02*10-3 after 28s. The recovery value is higher than the one relative to physisorption due to the drift of absorbance signal occurring during this test (Fig.3a). For sample B no response for tested gases was

966 E. Czerwosz et al. / Procedia Engineering 87 ( 2014 ) 963 – 966

observed. For the sample A1 a sharp response for CO gas was registered (Fig.3b). However, the relevant drift of the signal was seen. For this sample we found the response for CO (1%v/v) ~ 9,71*10-4 and 1,08*10-3 in respectively 31s and 41s with recoveries value of 1,08*10-3 and 9,6*10-4 in 41s and 36s respectively. For the sample B1 a answer for CO and NO2 gases was found. In the presence of CO (1%v/v) the changes of the relative absorbance was 9,5*10-4 after 32s and return to air the recovery occurs at the value 1,21*10-3 at 41s. The changes with NOx are unstable and the effect of increasing of relative absorbance of 2,14*10-3 was followed by a continuous decay at every switch to the air was observed (Fig.3c). It seems that the sample with longer CNTs are sensitive both for CO and NOx what could be connected to higher number of defects and dangling bonds available for attachment of various molcules.

The samples A2 and B2 coated with Pd did not give a stable variation of the absorbance for different gases, showing irreversible variation and/or very large drift, especially in the case of H2. It could be due to the effect of absorption of hydrogen by palladium and also by nanotubes or carbonaceous matrix.

3. Conclusions

We have shown that our CNT films are sensitive for CO as well as for NOx. The films with longer CNTs can detect both these gases while the film with shorter CNTs detect only CO. The same detection of CO was observed for C-Ni nanocomposite film.

To understand unstable effect of hydrogen absorption by CNT film covered with palladium and carbon we plan to perform more experiments.

Acknowledgements

This work was partly supported by the European Regional Development Fund within the Innovative Economy Operational Programme 2007-2013 (title of the project “Development of technology for a new generation of the hydrogen and hydrogen compounds sensor for applications in above normative conditions”) No UDA-POIG.01.03.01-14-071/08-10

References

[1] J. Kong, M. G. Chapline, and H. Dai, Functionalized carbon nanotubes for molecular hydrogen sensors, Adv. Mater., 2001, 13, 18, 1384-1386. [2] N. Karousis, G.-E. Tsotsou, F. Evangelista, P. Rudolf, N. Ragoussis, and N. Tagmatarchis, Carbon nanotubes decorated with palladium

nanoparticles: synthesis, characterization, and catalytic activity, J. Phys. Chem. C, 2008, 112, 13463–13469 [3] S. Mubeen, T. Zhang, B.Yoo, M. A. Deshusses, and N. V. Myung, Palladium nanoparticles decorated single-walled carbon nanotube

hydrogen sensor, J. Phys. Chem. C, 2007, 111 (17), 6321-6327. [4] Francisco P. Rouxinol, Rogerio V. Gelamo and Stanislav A. Moshkalev, Gas Sensors Based on Decorated Carbon Nanotubes, Carbon

Nanotubes, Jose Mauricio Marulanda (Ed.), (2010) ISBN: 978-953-307-054-4, InTech, Available from: http://www.intechopen.com/books/carbon-nanotubes/gas-sensors-based-ondecorated-carbon-nanotubes

[5] H. C. Choi, M. Shim , S Bangsaruntip and H. Dai, Spontaneous reduction of metal ions on the side- walls of carbon nanotubes, J. Am. Chem. Soc. 2002, 124, 9058.

[6] B. Xue, P Chen, Q. Hong, J. Lin, K. Lee and K. Tan, rowth of Pd, Pt, Ag and Au nanoparticles on carbon nanotubes, J. Mater. Chem. 2001, 11, 2378.

[7] J. Kong, M. G. Chapline and H. Dai, Functionalized carbon nanotubes for molecular hydrogen sensors, Adv. Mater. 2001, 13, 1384. [8] Ye, X. R.; Lin, Y.; Wai, C. M., Decorating catalytic palladium nanoparticles on carbon nanotubes in supercritical carbon dioxide, Chem.

Commun. 2003, 9, 642 [9] Tessonnier, J.-P.; Pesant, L.; Ehret, G.; Ledoux, M. J.; Pham- Huu, C., . Pd nanoparticles introduced inside multi-walled carbon nanotubes for

select ive hydrogenation of cinnamaldehyde into hydrocinnamaldehyde. Appl. Catal., A 2005, 288, 203. [10] Guo, D. J.; Li, H. L, Electrochemical synthesis of Pd nanoparticles on functional MWNT surfaces,. Electrochem. Commun. 2004, 6, 999 [11] Bera, D.; Kuiry, S. C.; McCutchen, M.; Kruize, A.; Heinrich, H.; Meyyappan, M.; Seal, S., In-situ synthesis of palladium nanoparticles-

filled carbon nanotubes using arc discharge in solution, Chem. Phys. Lett. 2004, 386, 364. [12] Kowalska E., Radomska J., Czerwosz E., Wronka H., Preparation of carbon-metal nanoobjects, patent appl. P.405802, 25.09.2013 [13] Czerwosz E., Wronka H., Preparation of films containing Ni and Pd nanograins, patent appl. P. 401249, 16.10.2012