Sensors and Actuators B 126 (2007) 557–561

Improving the stability of surface acoustic wave (SAW) chemical

sensor coatings using photopolymerization
Dmitry Pestov, Ozge Guney-Altay, Natalia Levit, Gary Tepper ∗
Virginia Commonwealth University, Department of Mechanical Engineering, Richmond, VA 23284, United States
Received 26 January 2007; received in revised form 3 April 2007; accepted 3 April 2007
Available online 12 April 2007

Abstract
Submicron particles of a diethyl ester of p-phenylenediacrylic acid (EPA) were produced by the technique known as rapid expansion of
supercritical solutions (RESS) and were deposited onto surface acoustic wave (SAW) resonators. The EPA particles were polymerized in the
solid-state directly on the surface of the SAW devices in order to stabilize the mechanical properties of the sensor coatings. The sensor affinity
to particular vapors was enhanced by performing the polymerization step in the presence of organic template vapors. The resulting SAW sensors
exhibited high quality factors (Q0 ) and we have demonstrated the sensor stability for periods up to 3.8 years.
© 2007 Elsevier B.V. All rights reserved.

Keywords: Gas sensor; Surface acoustic waves; Photopolymerization; Rapid expansion of supercritical fluids; Coating stability

1. Introduction susceptible to time-dependent mechanical changes caused by

Gas-phase chemical sensors based on surface acoustic wave
(SAW) devices were first reported more than two decades ago
[1]. Current SAW-based chemical sensors employ an array of
sensors to detect and identify gas-phase analytes. Each sensor
element of the array is coated with a different thin polymer film
selected to provide preferential affinity to a specific vapor class
based on its polarity, polarizability or ability to participate in
hydrogen bonding. The combined output pattern of the array is
used to recognize the presence of specific vapors with high reli-
ability and sensitivity [2,3]. The thin polymer film plays the
key role for achieving optimum overall sensor performance.
For optimum performance, a sensor coating must have strong
adhesion to the SAW device surface, should reversibly absorb
the analyte vapor and must maintain mechanical integrity over
a range of operating conditions and temperatures. Finally, a
good coating must be stable over time and resilient to vari-
ous conditions in the storage and measurement media [4–6].
Highly viscous linear (jelly-like or rubbery) polymer coatings
are normally preferred in chemical sensors since they provide
high vapor permeability [6]. However, rubbery coatings are
∗ Corresponding author. Tel.: +1 804 827 4079; fax: +1 804 827 7030.
E-mail address: gctepper@vcu.edu (G. Tepper).
interactions with the surface of the SAW device, resulting in
unwanted changes in sensor performance [4,7]. Fig. 1 is an opti-
cal microscope image illustrating the morphological changes
observed in a methylphenylsilicon polymer (OV-25) coating
after 1 h. OV-25 is commonly used for the detection of non-
polar aromatic molecules [8]. Initially, the transducer surface
is coated with a smooth polymer film, but after 1 h and under
standard atmospheric conditions, the coating agglomerates into
an ensemble of small droplets. Crosslinking has been proposed
as a potential method to prevent these kinds of unwanted mor-
phological changes that are frequently observed in SAW sensor
coatings. Unfortunately, it is very challenging to perform chem-
ical reactions on the surface of a SAW device without interfering
with the SAW resonance condition [4]. In addition, rigid poly-
mer films were found to exhibit lower vapor sensitivities and are
susceptible to fragmentation [4].
Our research group has been investigating the potential
advantages of using particulate polymer coatings instead of
continuous films on SAW devices. Previously, we demonstrated
that particulate coatings can significantly improve the sensitivity
of SAW devices by providing a higher surface area for vapor
interactions [9]. However, it is well known that particulate
coatings can act as scattering centers and interfere with the
propagation of acoustic waves, thereby compromising the
acoustic resonance condition. The effect of coating morphology

0925-4005/$ – see front matter © 2007 Elsevier B.V. All rights reserved.
doi:10.1016/j.snb.2007.04.010
558 D. Pestov et al. / Sensors and Actuators B 126 (2007) 557–561
Fig. 1. Film disintegration of OV-25 polymer on the SAW device surface: (A) 5 min after coating and (B) 1 h after coating. The light regions are gold interdigitated
electrodes deposited onto the underlying dark quartz substrate.
on device performance can be evaluated by quality factor (Q0 )
measurements. The quality factor is a quantitative representa-
tion of the sharpness of the resonance curve. In general, SAW
devices with a low quality factor have lower sensitivities and
are not desirable as chemical sensors [10]. In this paper, we
report a method for increasing the stability of SAW coatings
without degrading the acoustic resonance condition of the
SAW transducer. Submicron polymer particles were deposited
and polymerized directly on the SAW device surface, and the
resonance condition and vapor affinity of the SAW sensor were
monitored over time. It is shown that the vapor affinity of the
coating is stable and can be tuned through the introduction of
a template vapor during the polymerization step.
2. Experimental
2.1. Materials and equipment
HPLC-grade pentane, hexane, heptane, octane, isooctane,
nonane, decane and 1-butanol were obtained from Fisher
Scientific. Freon 22 (chlorodifluoromethane) was obtained
from DuPont. The diethyl p-phenylenediacrylate (EPA) was
synthesized using a procedure described previously [11].
Methylphenylsilicon OV-25 was obtained from Sigma–Aldrich.
Nitrogen was from National Welders and has grade 5.5. SAW
250 MHz resonators were from Microsensors Systems, Inc. For
polymerization of EPA particles, a UV lamp from American
Ultraviolet was used. This lamp has 100 W power and emits
UV within the range of 325–382 nm. Evaluation of the electrical
properties of the SAW devices was performed with an HP 8353 D
network analyzer. Unloaded quality factors (Q0 ) were calculated
with the Qzero computer program developed by Kajfez [12].
2.2. SAW device preparation
2.2.1. Particle development
A schematic representation of the RESS process is depicted
in Fig. 2. The solute material (EPA monomer) was dissolved
in supercritical Freon 22 in a pressure vessel. The temperature
of the pressure vessel was controlled using a tube furnace. The
single-phase supercritical solution was expanded through a cap-
illary nozzle into atmospheric conditions, and phase separation
occurred as the expanding solution crossed into the two-phase
region. Solvent-free submicron EPA monomer particles were
precipitated and deposited directly onto the surface of SAW
resonators. The coating thickness was carefully controlled and
corresponded to a SAW frequency shift of about 400 kHz. A
more detailed description of the apparatus and experimental pro-
cedure is provided in a previous article [9]. In order to study the
morphology and the structure of the monomer and polymer parti-
cles, the particulate coatings on the SAW devices were replicated
on glass slides and were examined by optical and SEM (JEOL
JSM-820) microscopy.
2.2.2. Polymerization of EPA particles
The EPA coatings on the SAW devices and glass slides were
polymerized in air and in the presence of 1-butanol vapor, which
was chosen as a model template molecule. The SAW devices
were placed in a sealed UV-transparent chamber and the cham-
ber was filled either with air or air with saturated template vapor.
The chamber was kept in the dark until polymerization initiation.
In order to facilitate polymerization, the samples were exposed
to UV light for 15 min from a distance of 22 cm. After polymer-
ization, the coatings were flushed with dry nitrogen for 20 min
to remove the template.
2.2.3. Sensitivity evaluation
The resulting SAW sensor response was tested upon exposure
to analyte vapors at various concentrations using a custom-built
sensor testing system. The details of this system were described
previously [9]. The system is equipped with a vapor generation
unit. This unit produces saturated organic vapors at 10 ± 0.1 ◦ C.
The organic vapors are diluted with nitrogen and delivered to a
pair of SAW devices at a constant temperature of 30 ◦ C. The pair
of SAW devices includes one reference device and one coated
device, and the frequency difference is measured. The reference
device is uncoated and is exposed to pure nitrogen at the same
flow rate and temperature as the analyte stream directed to the
Fig. 2. Schematic representation of the RESS process: (1) high pressure vessel
with the supercritical solution, (2) capillary nozzle and (3) SAW transducer.
 D. Pestov et al. / Sensors and Actuators B 126 (2007) 557–561 559

Fig. 3. Polymerization of diethyl ester of p-phenylenediacrylic acid (EPA).

Fig. 4. SEM images of EPA polymer particles prepared by RESS deposition and UV irradiation: (A) immediately after RESS and (B) after exposure to 1-butanol
and photopolymerization.

coated device. The frequency difference versus time is recorded

for increasing concentrations of each analyte gas. Before each
test, the sensors are preconditioned by flushing with nitrogen.
The testing station is automated using LabView Software.

3. Results and discussion

Diolefinic monomer EPA was chosen in this study because

it allowed the direct conversion of monomer particles to poly-
mer particles by a topochemical reaction [13–15]. The reaction
is a [2 + 2] cycloaddition and it produces a linear polymer with
cyclobutane rings in the main chain as shown in Fig. 3 [16,17].
Additional promoters, such as initiators are not required; the Fig. 5. Resonance curves for SAW devices: (1) coated with EPA polymer par-
polymerization is self-consistent by UV irradiation. Fig. 4 shows ticles and (2) without any coating.
scanning electron microscope images of the polymer particles as
factor calculations showed that Q0 varied between 7000–10,000
prepared by RESS and after photopolymerization. Initially the
and 11,000–12,000 for the coated and uncoated SAW devices,
particles appear as elongated rods with a length near 1 ␮m and a
respectively. After 1380 days of aging, the change in the vapor
diameter around 200 nm. After vapor exposure and photopoly-
sensitivity of the coated SAW devices was within 20% of the
merization, the particles appear more rounded with a maximum
original value for all tested devices and vapors. Typical SAW
diameter near 1 ␮m. At this time, we do not have a complete
sensitivity changes over time are depicted in Fig. 6. Our present
explanation for the morphological changes observed in the EPA
study has concluded that these SAW devices show high quality
particles, but believe that the organic template vapor permeates
factors and high stability over time and are therefore compatible
into the EPA particles inducing these changes. During the RESS
with the requirements for long-term sensing applications.
process, the particle nucleation and growth occur within a very
short time period, yielding small submicron particles. These par-
ticles were found to adhere well to the sensor surface and would
not detach unless washed off with liquid solvents [9]. A network
analyzer was used to measure the return loss of each coated SAW
device and we found that the small EPA particles did not signifi-
cantly degrade the resonance condition. Fig. 5 shows the typical
resonance curves for the coated and uncoated SAW devices. We
believe that the excellent acoustic properties of the SAW devices
coated with EPA particles can be attributed to the uniformity of
mass loading per area for each interdigital finger on the trans-
ducer. This is due to small particle size and their random but Fig. 6. SAW sensitivity change to heptane vapor: (1) initial sensitivity, (2)
uniform spatial distribution on the transducer surface. Quality sensitivity after 1.3 years and (3) after 3.8 years.
560 D. Pestov et al. / Sensors and Actuators B 126 (2007) 557–561
Fig. 7. Relative averaged sensitivities of 1-butanol imprinted and non-imprinted
SAW resonators to a series of alkanes ranging from n-pentane (C5 ) to n-decane
(C10 ) including iso-octane. The baseline at 100% corresponds to the sensitivity
of non-imprinted SAW device to C5 –C10 alkanes.
Previously, we found that solid-state polymerization provided
a unique opportunity to prepare polymers with a molecular
porosity specific to the size of a given template molecule
[11,18,19]. For instance, it was shown that when the polymer-
ization was carried out in the presence of heptane vapor, the
polymer coating exhibited an affinity to alkanes with molec-
ular dimensions equal to or less than heptane. This suggests
that the polymer coating discriminates molecules based on a
size-exclusion mechanism. This approach has similarities to
molecular imprinting [20]. The proposed size-exclusion mech-
anism motivated us to test different template molecules with
equivalent molecular dimensions. We chose a template vapor
from a different chemical family to test against alkane analytes.
In this paper, we present our results on SAW device performance
using the model template molecule 1-butanol.
In order to compare the sensitivity of the SAW resonators,
two sets of SAW devices were prepared. The first set had a non-
imprinted coating (polymerized in air), and the second set had
an imprinted coating (polymerized in the presence of 1-butanol
vapor). By comparing the vapor affinity of the SAW devices
with the non-imprinted coating to the ones with the imprinted
coating, it is possible to measure the relative vapor affinity of
these devices [21]. The sensitivity for each SAW device was
calculated as the slope of frequency shift versus vapor concentra-
tion curve. The sensitivity difference between the non-imprinted
and 1-butanol-imprinted SAW devices when exposed to C5 –C10
alkanes is shown in Fig. 7. The SAW device imprinted to 1-
butanol shows the highest affinity to the C5 alkane, because
these two molecules have similar molecular sizes. The sensi-
tivity difference between the “small” C5 and the “big” C10 is
on the order of 400%. By comparison, the highest sensitiv-
ity difference we observed with alkane imprinted coatings was
about 25% [11]. The higher sensitivity improvement observed in
the present study may be attributed to intermolecular hydrogen
bonding between 1-butanol and EPA during particle polymeriza-
tion. The stronger bonding between the template molecule and
the polymer results in a higher degree of molecular porosity.
The 1-butanol-imprinted SAW devices also showed different
affinity to the linear versus branched isomers of various alkanes.
Fig. 7 shows this difference for n-octane and iso-octane. This
may suggest that the geometry of the analyte molecule plays an
important role in the uptake of vapors into the sensor coating.
Isomers of other alkanes were tested and the results of this study
will be presented in a future publication.
4. Conclusion
RESS-coated SAW devices modified by solid-state polymer-
ization exhibited very good electrical/acoustical properties. The
device performance remained stable after 3.8 years. The use of
rigid polymer particles instead of rubbery polymer films may
provide advantages in terms of sensor long-term stability and
tunability. The RESS micronization process appears to eliminate
the disintegration problem that can occur in the polymerization
of thin films. In addition, the SAW resonance condition is not
significantly degraded if the particles are small and uniformly
distributed over the surface of the device.
The coatings prepared by solid-state polymerization of
diolefinic monomers in the presence of a template molecule
exhibited a high affinity to analytes of equal or smaller molecular
dimensions [11,18,19]. In this paper, a new template vapor from
a different chemical family was tested against alkane analytes.
The model template vapor, 1-butanol, is approximately equiv-
alent to the C5 alkane in size. The SAW device imprinted to
1-butanol showed a sensitivity enhancement of 400% between
the C5 and the C10 alkanes.
References
[1] H. Wohltjen, R. Dessy, Surface acoustic-wave probe for chemical-analysis.
1. Introduction and instrument description, Sens. Actuators, B: Chem. 51
(1979) 1458–1464.
[2] R.D.S. Yadava, R. Chaudhary, Solvation, transduction and independent
component analysis for pattern recognition in SAW electronic nose, Sens.
Actuators, B: Chem. 113 (2006) 1–21.
[3] W.A. Groves, A.B. Grey, P.T. O’shaughnessy, Surface acoustic wave (SAW)
microsensor array for measuring VOCs in drinking water, J. Environ.
Monit. 8 (2006) 932–941.
[4] I.D. Avramov, S. Kurosawa, M. Rapp, P. Krawczak, E.I. Radeva, Investi-
gations on plasma-polymer-coated SAW and STW resonators for chemical
gas-sensing applications, IEEE Trans. Microw. Theory Tech. 49 (2001)
827–837.
[5] J.W. Grate, E.T. Zellers, The fractional free volume of the sorbed vapor in
modeling the viscoelastic contribution to polymer-coated surface acoustic
wave vapor sensor responses, Anal. Chem. 72 (2000) 2861–2868.
[6] D.S. Ballantine Jr., H. Wohltjen, in: R.W. Murry, R.E. Dessy, W.R.
Heinemann, J. Janata, W.R. Seitz (Eds.), Chemical Sensors and Microin-
strumentation, American Chemical Society, Washington, DC, 1989, pp.
222–236.
[7] C. Hartmann-Thompson, J. Hu, S.N. Kaganove, S.E. Keinath, D.L. Keeley,
P.R. Dvornic, Hydrogen-bond acidic hyperbranched polymers for surface
acoustic wave (SAW) sensors, Chem. Mater. 16 (2004) 5357–5364.
[8] J. Park, E.T. Zellers, Temperature and humidity compensation in the deter-
mination of solvent vapors with a microsensor system, Analyst 125 (2000)
1775–1782.
[9] N. Levit, D. Pestov, G. Tepper, High surface area polymer coatings for
SAW-based chemical sensor applications, Sens. Actuators, B: Chem. 82
(2002) 241–249.
 D. Pestov et al. / Sensors and Actuators B 126 (2007) 557–561
[10] R.A. Mcgill, R. Chung, D.B. Chrisey, P.C. Dorsey, P. Matthews, A. Pique,
T.E. Mlsna, J.L. Stepnowski, Performance optimization of surface acoustic
wave chemical sensors, IEEE Trans. Ultrason. Ferr. Freq. Cont. 45 (1998)
1370–1380.
[11] D. Pestov, N. Levit, V. Maniscalco, B. Deveney, G. Tepper, Molecular
imprinting using monomers with solid-state polymerization, Anal. Chim.
Acta 504 (2004) 31–35.
[12] D. Kajfez, Q Factor, Vector Fields, Oxford, MS, 1994, p. 43.
[13] S. Takahashi, H. Miura, H. Kasai, S. Okada, H. Oikawa, H. Nakanishi,
Single-crystal-to-single-crystal transformation of diolefin derivatives in
nanocrystals, J. Am. Chem. Soc. 124 (2002) 10944–10945.
[14] R. Balaji, D. Grande, S. Nanjundan, Photoresponsive polymers having
pendant chlorocinnamoyl moieties: synthesis, reactivity ratios and pho-
tochemical properties, Polymer 45 (2004) 1089–1099.
[15] P.R. West, B.F. Griffing, Multilayer photoresist process utilizing cinnamic
acid derivatives as absorbant dyes, US Patent 4,535,053 (1985).
[16] F. Suzuki, Y. Suzuki, H. Nakanishi, M. Hasegawa, Four-center type
photopolymerization in the solid state. III. Polymerization of phenylene-
diacrylic acids and their derivatives, J. Polym. Sci., Part A-1 7 (1969)
2319–2332.
[17] M. Hasegawa, Four-center photopolymerization in the crystalline state,
Adv. Polym. Sci. 42 (1982) 1–49.
[18] D. Pestov, N. Levit, V. Maniscalco, G. Tepper, Molecular imprinting
using monomers with solid-state polymerization, in: Second International
Workshop on Molecular Imprinting, La Grande Motte, Laguedoc, France,
September 15–19, 2002.
[19] D. Pestov, N. Levit, V. Maniscalco, G. Tepper, Molecularly imprinted poly-
mers for SAW sensors, in: 202nd Meeting of the Electrochemical Society,
Salt Lake City, October 20–24, 2002.
[20] L. Ye, K. Mosbach, Molecularly imprinted materials: towards the next
generation, Mater. Res. Soc. Symp. Proc. 723 (2002) 51–59.
[21] J.W. Grate, Acoustic wave microsensor arrays for vapor sensing, Chem.
Rev. 100 (2000) 2627–2647.
561
Biographies
Dmitry Pestov is a research associate in mechanical engineering at Virginia
Commonwealth University in Richmond. He obtained his MS (chemical engi-
neering) and later PhD (organic chemistry) from the Saint-Petersburg State
Institute of Technology, Russia. His main research interests are on molecular
imprinting, materials for chemical sensors, SAW sensors and fluorescence-based
chemical sensors.
Ozge Guney-Altay received her BS and MS in chemical engineering from
Bogazici University, Istanbul, Turkey in 1993 and 1995, respectively. She earned
her PhD in chemical engineering from Texas A&M University, College Station,
Texas in 2001 and joined to the School of Engineering at Virginia Common-
wealth University as a research faculty and instructor. Her present research
interests are in advanced polymer coatings for highly selective sensors and
membranes, preparation of nano-structured materials using supercritical fluid
technology, controlled and targeted delivery of therapeutic agents and processing
biomaterials for improved biocompatibility.
Natalia Levit received her MS in polymer and textile engineering from Omsk
Technological Institute, Russia in 1985 and PhD in materials science of poly-
meric materials from Saint-Petersburg State University of Technology and
Design, Saint-Petersburg, Russia in 1989. She has extensive research expe-
rience in materials science of polymers, fibers and polymer composites. She
worked at VCU as a research assistant professor until 2005 working on engi-
neering of nanoscale polymer surfaces and nanocomposites for chemical sensing
application.
Gary Tepper is a professor of mechanical engineering and Director of the
Microsensor and Radiation Detector Research Laboratories at Virginia Com-
monwealth University in Richmond. He is also the founder and President of
Sentor Technologies Inc. in Glen Allen, Virginia. He obtained his PhD from
the University of California in San Diego in engineering sciences (engineering
physics) and his undergraduate degree from Penn State.