вход по аккаунту


High-Throughput Method for the Impedance Spectroscopic Characterization of Resistive Gas Sensors.

код для вставкиСкачать
Sensor Screening
High-Throughput Method for the Impedance
Spectroscopic Characterization of Resistive Gas
Andreas Frantzen, Jens Scheidtmann, Gerald Frenzer,
Wilhelm F. Maier,* Jrg Jockel, Thomas Brinz,
Daniel Sanders, and Ulrich Simon*
Sensors and sensor arrays are increasingly being used for
process control, environmental protection, in medical applications, as well as in automotive and domestic applications.
The development of rapid and sensitive gas sensors with low
cross-sensitivity is the subject of intensive studies. However,
the majority of the research activity is restricted to simple
modification and improvement of known systems and is not
directed towards the search for alternative sensor materials.
The biggest problems with known materials are high crosssensitivity and insufficient long-term stability.[1] The demand
for more selective and stable sensors for a wide range of
applications is therefore very high. Despite the increasing
demand, the commercially used resistive sensor materials are
confined to a few inorganic semiconductors such as titanium
dioxide or tin dioxide. For sensor applications, the doped
[*] Dr. A. Frantzen, Dipl.-Phys. J. Scheidtmann, Dr. G. Frenzer,
Prof. Dr. W. F. Maier
Universit+t des Saarlandes
Postfach 151 150, 66 041 Saarbr/cken (Germany)
Fax: (+ 49) 681-302-2343
oxide is deposited as a thick film on a ceramic support. This
film is electrically connected to a measuring device and a
separate heating unit. The relatively complex setup aggravates the rapid screening of potential sensor materials.
The application of combinatorial methods in materials
research promises significant acceleration, especially in the
area of material and parameter optimization, as well as in the
discovery of new materials.[2] Several reports of successful
studies of oxide layers for electric or electronic applications
were made. Already in 1999, the automated electrochemical
analysis of library plates of an array of 64 electrodes on a
silicon waver was published.[3] A comparable technology for
the electrochemical deposition and characterization of samples of similarly designed library plates was reported by
Symyx Technologies in a patent application.[4] The preparation of gradient libraries has been used for the discovery of
transparent transistors based on ZnO[5] as well as for the
discovery of transparent ferromagnetic mixed oxides (Codoped TiO2).[6] Procedures for the discovery of new microwave dielectrics have also been reported.[7] It is striking that in
all these applications, the library was prepared exclusively
with the help of PVD (physical vapor deposition) and CVD
(chemical vapor deposition) methods. Such materials libraries
are not suitable for the investigation of resistive-sensor
properties. On the one hand, a variable homogeneous
doping below one atom percent is difficult to achieve. On
the other hand, porosity and microstructure of the desired
sensor materials is dependent on the preparation procedure
and is of fundamental importance for the sensor properties.[8, 9] Herein we report the use of liquid-phase deposition
methods in the search for new sensor materials based on
optimized sol–gel synthesis procedures.
Specially developed thin ceramic aluminum oxide plates
with printed interdigitated structures of platinum were used
as substrate materials instead of the commonly used Si-wafers
(Figure 1).[10] The sensor array is composed of a square
arrangement of 64 interdigitated electrodes (see magnification in Figure 1).
The contact of the electrodes is positioned at the rim of
the library plate. The variably doped sensor layers were
deposited in a reactor, whose design was borrowed from
combinatorial catalysis research (Figure 2).[11] The top of the
reactor consists of a mask with 64 drill holes positioned
D. Sanders, Prof. Dr. U. Simon
RWTH Aachen
Professor-Pirlet-Strasse 1, 52 056 Aachen (Germany)
Fax: (+ 49) 241-80-99003
Dr. J. Jockel, Dr. T. Brinz
Robert Bosch GmbH
Zentralbereich Forschung und Vorausentwicklung
Forschung 1, Angewandte Chemie und Energietechnik (Germany)
[**] This work was supported by the German Federal Ministry for
Education and Research (BMBF, PID 03C0305D) as project
2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Figure 1. Library plate with an 8 8 interdigitated electrode array.[10]
DOI: 10.1002/anie.200352424
Angew. Chem. Int. Ed. 2004, 43, 752 –754
Figure 2. Setup of the reactor for direct materials synthesis on the
library plate.
according to the library design. Viton fittings (O-rings) are
positioned on the bottom of the mask. Both the library plate
(aluminum oxide) and the mask are screwed to a metal base
plate. A viton foil, a metal plate, and a teflon foil between the
library plate and the metal base plate distribute the pressure
from the screws more evenly. It was of interest to sample the
sensor properties of sol–gel materials, because sol–gel procedures have rarely been used in the preparation of solid sensor
materials. Sol–gel procedures are especially well-suited for
automated synthesis through pipetting robots. The development of suitable sol–gel recipes for the preparation of mixedoxide films has been carried out conventionally in laboratories. Before their application to libraries, the sol–gel recipes
were optimized for good adhesion and to avoid the formation
of cracks on the library plates. The procedure was transferred
to a pipetting robot, which introduced the sol mixtures
directly into the openings of the library reactor. Doping was
achieved by the addition of appropriate salt solutions in the
reactor. After gelation and drying, the reactor was opened
and the library plate was removed. After a visual inspection,
the plate was calcined at 700 8C.
The base material of the library shown here is a
tungsten(vi) oxide doped with tantalum (0.5 atom %). The
surface of the potential sensor layers was doped by impregnation with various chloride-free metal salt solutions.
The sensor characterization was carried out by highthroughput impedance spectroscopy (HT-IS). The setup of
the HT-IS consists of a measuring head in which the library
plate is positioned. Each individual electrode on the library
plate is electrically connected. The measuring head is
positioned in an oven, which allows electrical measurements
at temperatures up to 800 8C in a controlled atmosphere,
prepared by proper mixing of analytical gas mixtures and
reference gases with the help of mass flow controllers. The
experimental flow and data accumulation are controlled by
special software.[12]
The measurements discussed herein start with synthetic
air (20 % O2, 80 % N2) as reference, which is also the carrier
gas of the test gases. The experimental sequence of test gases
is 50 ppm H2, synthetic air, 50 ppm CO, synthetic air, 5 ppm
Angew. Chem. Int. Ed. 2004, 43, 752 –754
NO, synthetic air, 5 ppm NO2, synthetic air, 50 ppm propene,
synthetic air, synthetic air. All gas mixtures have a relative
humidity of 45 % at 25 8C. The measurements in this example
were carried out at 350 8C. The measurements were started
15 minutes after a constant gas flow has been reached. The
test gases flow through the reactor (volume 45 mL) at a
flow rate of 100 mL s 1. Accordingly, 2560 measurements are
made per library plate, which results from 64 samples, ten
different gas-flow compositions, and four temperatures.
The materials on the libraries were electrically analyzed
with an impedance analyzer (Agilent 4192A) within a
frequency range from 10 to 107 Hz and a measurement
voltage of 100 mV under variation of temperature and gas
composition. The impedance spectra show typical electric RC
responses, which allow automated data fitting by the associated circuit equivalent. This allows the evaluation of the
total frequency response of the samples to be restricted to a
few parameters, such as resistance R, capacitance C, and the
related relaxation time t. Figure 3 shows the complex
Figure 3. Complex impedance (Argand plot) of a typical measurement on the
library and its automated fit with the circuitry model (parallel RC couple).
impedance plot of a typical measurement (Argand plot) as
well as the associated parallel RC circuit equivalent. In this
case R = 2.49 > 106 W and C = 6.1 > 10 12 F.
For the description of the response properties towards
various test gases, the sensitivity S was defined as S =
(Rmax Tmin)/Rmax, with Rmax as the maximum resistance
and Rmin as the minimum resistance resulting from data fitting
by the parallel RC element. The minus sign ( ) indicates an
increase in resistance and the plus sign (+) indicates a
decrease in resistance upon contact with gas.
The sensitivities towards the various test gases are
presented as a Manhattan plot. This presentation provides
characteristic fingerprints of the individual materials tested
(Figure 4). Upon exposure to H2, CO, and propene, a
decrease in resistance relative to that of the reference
synthetic air atmosphere is observed, while the material
responds to exposure to NO2 with a small increase in
resistance. After each exposure to a test gas, the measurement
in the reference gas atmosphere shows the reversibility of the
observed response during the time of the experiment
(15 min).
2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Figure 4. Fingerprint of the library position 6-6: WO3 ; bulk dopant Ta
(0.5 atom %), surface dopant Au (0.5 atom %).
Figure 5 shows the fingerprints of the complete library
plate. The surface doping elements are labeled below the
plots. Empty plots indicate library positions whose electric
properties were identified as pure capacitance by the
automated data fitting. Such empty diagrams can be
explained by insufficient electrical contact of the materials
on the library.
Figure 5. Sensor characteristics of the materials on the tungsten oxide
library. See text for details. The shaded (gray) fingerprint corresponds
to the position used in Figure 4.
The multiple deposition of individual samples as well as
the spatial distribution of identical doping elements on the
library are used to compensate for gradients in gas concentration, which result from catalytic conversion of test gases,
such as CO. Especially with CO, a concentration decrease
from the center to the outer plate was noted.
Most of the materials in these studies show a distinct
sensitivity for 50 ppm propene in the form of a decrease in
resistance. They also show a low cross-sensitivity against the
other test gases. Exceptions are the materials impregnated
with gold, rhodium, and palladium. Besides the response to
propene, these materials also show a decrease in resistance
2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
upon exposure to H2 (50 ppm) and CO (50 ppm). The
significant sensitivity deviation of identical materials, as
observed with the palladium-doped positions, is associated
with a concentration gradient of the test gases.
These measurements have shown that the presented
combination of high-throughput materials synthesis on multielectrode arrays with an interdigitated capacitor structure and
high-throughput impedance spectroscopy is a useful technique for the rapid screening for new sensor materials for gassensing applications. This is the first method that allows the
investigation of a large number of new potential sensor
materials. The multiple deposition described herein was used
to compensate for concentration gradients and will not be
necessary in future applications, since an improved method
for the introduction of the test gas has been developed.[12] This
technology allows new materials and applications to be tested,
which previously would have required time-consuming tests
for individual sensors. This technology can also be used for
the discovery of new materials in the areas of electrocatalysis,
ferroelectrics, and dielectrics.
Received: July 21, 2003 [Z52424]
Keywords: combinatorial chemistry · high-throughput
screening · impedance spectroscopy · sensors ·
sol–gel processes
[1] N. Barsan, M. Schweizer-Berberich, W. GGpel, Fresenius J. Anal.
Chem. 1999, 365, 287.
[2] B. Jandeleit, D. J. Schaefer, T. S. Powers, H. W. Turner, W. H.
Weinberg, Angew. Chem. 1999, 111, 2648; Angew. Chem. Int. Ed.
1999, 38, 2494.
[3] M. G. Sullivan, H. Utomo, P. J. Fagan, M. D. Ward, Anal. Chem.
1999, 71, 4369.
[4] C. J. Warren, R. C. Haushalter, L. Matsiev, US 6 187 164 B1,
2001 [CAplus AN 2001, 110095].
[5] J. Nishii, F. M. Hossain, T. Aita, Y. Ohmaki, S. Kishimoto, T.
Fukumura, Y. Ohno, H. Ohno, S. Takagi, K. Saikusa, I. Ohkubo,
A. Ohtomo, F. Matsukura, H. Koinuma, M. Kawasaki, Jpn. J.
Appl. Phys. 2003, 42, 347.
[6] Y. Matsumoto, M. Murakami, T. Hasegawa, T. Fukumura, M.
Kawasaki, P. Ahmet, K. Nakajima, T. Chikyow, H. Koinuma,
Appl. Surf. Sci. 2002, 189, 344.
[7] H. Chang, X.-D. Xiang, Integr. Ferroelectr. 2000, 28, 113.
[8] J. F. McAleer, P. T. Moseley, J. O. Norris, D. E. Williams, J.
Chem. Soc. Faraday Trans. 1987, 83, 1323.
[9] J. F. McAleer, P. T. Moseley, J. O. Norris, D. E. Williams, J.
Chem. Soc. Faraday Trans. 1988, 84, 411.
[10] U. Simon, D. Sanders, J. Jockel, C. Heppel, T. Brinz, J. Comb.
Chem. 2002, 4, 511.
[11] J. Scheidtmann, P.-A. Weiß, W. F. Maier, Appl. Catal. A 2001,
222, 79 – 89.
[12] U. Simon, D. Sanders, J. Jockel, unpublished results.
Angew. Chem. Int. Ed. 2004, 43, 752 –754
Без категории
Размер файла
153 Кб
spectroscopy, impedance, throughput, high, method, characterization, sensore, gas, resistive
Пожаловаться на содержимое документа