Room temperature gas sensing with potassium titanate nanowires Burmistrov I. N., Varezhnikov A. S., Musatov V. Yu., Lashkov A. V., Gorokhovsky A. V.,Yudinceva T. I., Sysoev V. V. Yuri Gagarin State Technical University of Saratov Polytechnicheskaya 77, Saratov 410054, Russia E-mail:
[email protected] Abstract— We have studied the influence of organic vapors, acetone and ethanol, on the resistance of potassium titanate nanowire nonwoven net (PTN) at room temperature. We find that the PTN resistance is reproducibly reduced under appearance of the vapor molecules that might be explained by the influence of the adsorbed species on the electronic/ionic transport. The selectivity of the sensor could be greatly enhanced in order to recognize the kind of the vapors by processing a response of the PTN segment array formed at the same substrate by multiple co-planar electrodes. Keywords— potassium titanate, gas sensor, multisensor array, electronic nose, organic vapors
I.
nanowire,
INTRODUCTION
The gas sensors based on oxide nanomaterials are widely employed to detect the gaseous environment for many applications including the security-checking purposes [1]. In particular, the so-called quasi 1-D nanowires present the advanced interest because their large surface-to-volume ratio and a Debye length comparable to the nanowire radius allow one to observe a strong influence of surface processes on the nanowires’ electronic properties and so to yield often a superior sensitivity when compared to one of thin film counterparts [2]. The sensing mechanism of metal oxides is mainly governed by the fact that the oxygen vacancies on the surfaces are electrically and chemically active. Therefore, the conductivity (or resistivity) of oxide nanomaterials is strongly affected by the adsorbed molecules [3].
properties have not been so far explored in details for gas sensing applications. This material consists of negatively charged layers of titanium octahedron (TiO6) and interlayer potassium and hydronium cations and could be obtained at the nanobelt structure, favorite for gas sensing. The conductivity is primarily managed by hydronium and, partially, by potassium ions. The adsorption of gas molecules at the material surface shifts the electronic configuration and modifies the ionic transport that might be used for gas detection. II.
EXPERIMENTAL
The potassium titanate (K2Ti6O13) was synthesized by molten salt method from titanium dioxide and potassium hydroxide with further annealing at 1050 oC for 1 h. The received agglomerated powder was mechanically grinded by mechanical agate mortar for 3 min in order to make 10 mass. % dispersion in acidic (HCl) aqueous solution at pH=5. This suspension was ultrasonicated (Hielscher UIP1000hd) and washed several times with distilled water to remove the residual KOH and KCl. The separation of washing solution and the potassium titanate powder was carried out by centrifugation. Following the last washing the PTN powder was dried at 90 oC for 10-12 h. The elemental composition of the received PTN powder was carried out with fluorescent microprobe microscope ɊȺɆ30-ȝ. The material surface area was tested by Brunauer– Emmet–Teller method in Quantachrome Nova2200 analyzer.
The most significant number of applications of oxides as gas-sensing materials is known for tin dioxide [4, 5]. This material is chemically stable under the sensor applications and exhibits a high chemiresistive response. However its operation temperature is rather high, around 300-350 oC, that limits some applications. Therefore, recently there is a great interest in funding materials capable to sense gases at room temperature [6] that offers various advantages including a reducing the sensor power consumption. The application of graphene and other 2D materials sounds promising to develop gas sensors operated at temperatures close to room one [7] but their gas response is still rather low one in atmosphere pressure conditions when compared, for instance, to that of metal oxides. Here we consider the potassium titanate which is rather new material formed from the titanium oxide [8] whose
978-1-4799-8203-5/15/$31.00 ©2015 IEEE
Fig. 1. Schematic image of the multielectroded gas-analytical chip fabrication based on the PTN layer.
In order to fabricate the gas-analytical chip, the resulted PTN suspension was diluted and dropped over the Si/SiO2 substrate equipped with multiple Pt electrodes, 100 mkm width and 80 mkm of inter-space (Fig. 1), up to 39 in number, to form the PTN layer (Fig. 2). The morphology of this layer was studied with scanning electron microscopy (SEM) in TescanVega 3SB at 7 kV of accelerating voltage. The Fig. 3 shows the crystal structure of the PTN layer.
III.
RESULTS AND DISCUSSION
The elemental analysis of the synthesized PTN prior and after washing shows that the potassium mass content is reduced in two times from approx. 21 mass. % down to approx. 10 mass. %. The as-synthesized PTN is characterized by Ti/K ratio equal to ca. 3.7:1 while the washing procedure results in ca. 9:1 one. As a result, the structure is exfoliated into layered one due to a reduction of between-in electrostatic forces under potassium replacement by hydroxyls. This is supported by BET data of surface area which is enhanced from 2 m2/g (for assynthesized PTN) up to 20 m2/g (for washed PTN). After placing over the substrate the PTN layer had a good mechanical contact with Pt electrodes; the I-V curves of the PTN/electrode interface were appeared to be linear (Fig. 4).
Fig. 2. Electron microscopy image of the surface of the Si/SiO2 substrate covered with the PTN layer.
Fig. 4. The I-V curves for one of PTN segments in the chip measured in air and mixture of ethanol, ca. 61 kppm, with air. Following the exposure to the vapors of ethanol and acetone the resistance of the PTN layer segments reproducibly decreased up to 2 orders of magnitude (Fig. 5). The response time was in minutes. Fig. 3. The transmission electron microscopy image of the PTN layer (left) and animation of the titanate layered structure (right). The substrate equipped with PTN layer segmented by electrodes was wired into ceramic 50-pin card [9] and placed into the custom-made experimental setup consisting of read-out electronics in combination with a gas delivery system based on barbotage of liquids. The resistance of all the PTN segments in the array was read with rate of 1 sec/segment under exposure to acetone and ethanol vapors, 6-60 kppm concentration, in mixture with air as exemplary test gases at room temperature. In order to yield the measuring signal the dc voltage of 1 V magnitude was applied. The total array signal was processed by linear discriminant analysis (LDA) method [10].
In general, this chemiresistive effect is occurred because the developed surface area of the PTN allows one to effectively adsorb organic molecules (R= ɋɇ3–ɋɈ–ɋɇ3, (ɋɇ3)2ɋɇ2Ɉɇ, etc) as
PTN + R → PTN " R . According to earlier studies the PTN layer interacts with oxygen and hydroxyl groups under air atmosphere with formation of superoxide and hydroxide-radicals which catalyze the oxidation of the adsorbed organic molecules [11]:
( ) PTN (O )" R PTN (O )" R
( ) + H O → PTN (OH )" R
PTN e − " R + O2 → PTN O2− " R + − 2
+
− 2
+
2
−
+
+ PTN − O2 H
→ PTN + oxidised _ products
( )
PTN − O2 H + PTN " R → PTN O2− + oxidised _ products
I-V measurement
I-V measurement
Fig. 5. The change of resistance of PTN segment array under exposure to ethanol of various concentrations in mixture with air. As a result, the oxygen and hydroxyl groups localize the free electrons from the conductance band at the surface layer. When the organic molecules are oxidized the captured electrons return back to the conductance band that results in decrease of both the PTN conductivity and potential barriers in the borders between agglomerates and separate wickers. The resistance of each PTN sensor segment is defined by both the bulk whicker conductivity and the whicker-to-whicker and whicker-to-electrode interfaces. The intrinsic variation of density of the PTN layer over the chip array leads to variations in ratios of the contact resistance versus to PTN bulk one and so to variations in number of percolation nets [12]. Therefore, each PTN sensor segment of the array has a specific value of both the resistance and gas response.
Fig. 6. The linear discriminant analysis of the 11 PTN segment array responses to ethanol and acetone, 35 kppm concentration, in mixture with air. The response to pure air is given for reference.
By measuring the chip under exposure to the test gases the array signal is recorded as a vector one. The Fig. 6 shows the results of the LDA processing of the vector responses from the PTN-based sensor array to the test vapors of same concentration. As one can see, the data clusters related to the different vapors are clearly separated; the average distance between the gas clusters is approx. 12.91 rel. un. This makes it possible to selectively recognize the two organic vapors in difference to employing just a single sensor which is not selective.
ACKNOWLEDGMENT The authors thank the Russ. Ministry of Education & Science for a support by grant no. 8.236.2014/Ʉ and Dr. Sommer’s group (KIT, Germany) for collaboration to prepare multielectroded chip substrates used in the course of studies.
IV. CONCLUSIONS This work shows the ability to design a gas-analytical multisensor chip based on the potassium titanate nonwoven net layer which exhibits a chemiresistive signal under exposure to organic vapors at room temperature. The gas selectivity is achieved by processing the vector signal recorded from the array of sensor segments in the chip by pattern recognition methods like a linear discriminant analysis.
[2]
REFERENCES [1]
[3]
[4]
G. Korotcenkov, V. Sysoev, Conductometric Metal Oxide Gas Sensors: Principles of Operation and Approaches to Fabrication. In: G. Korotcenkov (ed.), Chemical Sensors: Comprehensive Sensor Technologies. Vol. 4. New York: Momentum Press, 2011, pp. 53-186. A. Ponzoni, G. Faglia, and G. Sberveglieri, Quasi One-Dimensional Metal Oxide Nanostructures for Gas Sensors. In T. Zhai, J. Yao (eds.), One-Dimensional Nanostructures: Principles and Applications. Hoboken: John Wiley & Sons, Inc, 2012. 576 pp. M. A. Carpenter, S. Mathur, and A. Kolmakov, Metal Oxide Nanomaterials for Chemical Sensors. New York: Springer, 2013. 548 pp. N. Barsan, M. Schweizer-Berberich, and W. Gopel, “Fundamental and practical aspects in the design of nanoscaled SnO2 gas sensors: a status report”, Fresenius’s Journal of Analytical Chemistry, vol. 365, pp. 287304, 1999.
[5]
[6]
[7]
[8]
E. Comini, G. Faglia, G. Sberveglieri, Z. Pan, and Z. L. Wang, “Stable and highly sensitive gas sensors based on semiconducting oxide nanobelts”, Applied Physics Letters, vol. 81, pp. 1869-1871, 2002. T. Close, G. Tulsyan, C. A. Diaz, S. J. Weinstaein, and C. Richter, “Reversible oxygen scavenging at room temperature using electrochemically reduced titanium oxide nanotubes”, Nature Nanotechnology, vol. 10, pp. 418-422, 2015. F. Schedin, A. K. Geim, S. V. Morozov, E. W. Hill, P. Blake, M. I. Katsnelson, and K. S. Novoselov, “Detection of individual gas molecules adsorbed on graphene”, Nature Materials, vol. 6, pp. 652-655, 2007. T. Sanchez-Monjaras, A. Gorokhovsky, and J. I. Escalante-Garcia, “Molten salt synthesis and characterization of potassium polytitanate ceramic precursors with varied TiO2/K2O molar ratios”, J. Am. Ceram. Soc., vol. 91, pp. 3058–3065, 2008.
[9]
V. V. Sysoev, I. Kiselev, V. Trouillet, and M. Bruns, “Enhancing the gas selectivity of single-crystal SnO2:Pt thin film chemiresistor microarray by SiO2 membrane coating”, Sensors and Actuators B, vol. 185, p. 5969, 2013. [10] A. Hierlemann, R. Gutierrez-Osuna, “Higher-order chemical sensing”, Chemical Reviews, vol. 108, pp. 563-613, 2008. [11] A. V. Gorokhovsky, E. V. Tret’yachenko, M. A. Vikulova, D. S. Kovaleva, and G. Yu. Yurkov, “Effect of chemical composition on the photocatalytic activity of potassium polytitanates intercalated with nickel ions”, Russ. Journal of Applied Chemistry, vol. 86, pp. 343-350, 2013. [12] V. V. Sysoev, J. Goschnick, T. Schneider, E. Strelcov, and A. Kolmakov, “A gradient microarray electronic nose based on percolating SnO2 nanowire sensing elements”, Nano Letters, vol. 7, iss. 10, pp. 3182-3188, 2007.
