2282 IEEE SENSORS JOURNAL, VOL. 16, NO.

8, APRIL 15, 2016

Particle Sensor Using Solidly Mounted Resonators

Sanju Thomas, Member, IEEE, Farah H. Villa-López, Member, IEEE, Jan Theunis, Jan Peters,
Marina Cole, Member, IEEE, and Julian W. Gardner, Senior Member, IEEE

Abstract— This paper describes the development of a novel issued regulations in which threshold and target values of
particle sensing system employing zinc oxide based solidly PMs are defined to maintain them within safe exposure limits.
mounted resonator (SMR) devices for the detection of airborne Commercially available instruments for particulate matter
fine particles (i.e., PM2.5 and PM10). The system operates in
a dual configuration in which two SMR devices are driven detection are generally large in size, expensive and difficult
by Colpitts-type oscillators in a differential mode. Particles are to operate [6]. These instruments are based on several meth-
detected by the frequency shift caused by the mass of particles ods and techniques. As an example, filter-based gravimetric
present on one resonator with while the other acts as a reference samplers such as the PartisolTM Sampler (Thermo Scientific)
channel. Experimental validation of the system was performed are used in the UK monitoring network. They draw an air
inside an environmental chamber using a dust generator with the
particles of known size and concentration. A sensor sensitivity sample through a filter trapping certain particles, which needs
of 4.6 Hz per µg/m3 was demonstrated for the SMRs resonating to be weighted later in the laboratory [7]. Other automated
at a frequency of 970 MHz. Our results demonstrate that the mass measurement instruments for the continuous monitoring
SMR-based system has the potential to be implemented in CMOS of PM are also available such as the tapered element oscillating
technology as a low-cost, miniature smart particle detector for microbalance (TEOMTM ) and Beta gauges. Optical methods
the real-time monitoring of airborne particles.
mainly based on absorption and scattering of light are the most
Index Terms— Acoustic wave sensor, air quality monitoring, commonly used for particle detection, counting and size mea-
Colpitts oscillator, particle sensor, particulate matter, solidly surement. The Thermo ScientificTM 5030 SHARP monitor and
mounted resonator (SMR).
the GRIMM 1.107 monitor (GRIMM Technologies, Inc.) are
I. I NTRODUCTION examples of these type of instruments [8]. Personal sampling
instruments such as the the DataRAMTM pDR-1500 (Thermo
A IRBORNE particulate matter (PM10 and PM2.5) consists
of a mixture of chemical substances that can be found
in the air in the form of very small particles. PM10 refers
Scientific) are also currently on the market. Optical techniques,
however, are complex and costly because they require the
to those particles that have an aerodynamic diameter equal integration of several optical components [9].
to or smaller than 10 μm whereas PM2.5 are particles with For these reasons, a low-cost, real-time and portable particle
diameters of 2.5 μm or smaller. Particulate Matter (PM) has sensing device is desired and different approaches have been
been associated with adverse effects on human health and recently proposed. Lim et al. [10] reported a MEMS particle
the consequent increase in mortality and morbidity rates [1]. detector based on the corona discharge principle, whereas a
Cardiovascular diseases and respiratory problems such as heart MEMS electrometer was proposed by Jaramillo et al. for the
failure and reduced lung capacity have been linked to the counting of aerosol particles [11]. Park et al. [12] developed
exposure to airborne particulate pollution [2]. Other health a particle sensor using a paddle-type silicon cantilever and the
problems related to PM have been reported, such as diabetes, use of thin-film piezoelectric on silicon resonators has been
atherosclerosis, and their impact on birth outcomes [3]–[5]. proposed by Harrington et al. [13]. Thermally actuated MEMS
In order to reduce human exposure to PM and so minimize resonators were demonstrated for the mass measurement of
their adverse health effects, the U.S. Environmental Protection airborne particles [9], [14] and the use of such structures
Agency (EPA) and the European Commission (EC) have within aerosol impactors for the size separation of particles
have been proposed as well [15], [16]. The collaboration
Manuscript received July 22, 2015; revised October 3, 2015; accepted between the Institute of Semiconductor Technology (IHT) and
December 20, 2015. Date of publication December 24, 2015; date of current the Fraunhofer-Wilhelm-Klauditz-Institut led to the develop-
version February 24, 2016. This work was supported by the European
Commission 7th Framework Program through the Project entitled Multi- ment of silicon resonant cantilever sensors for the detection of
Sensor-Platform for Smart Building Management under Project 611887. The airborne nanoparticles [17], [18] and further work reported the
work of F. H. Villa-López was supported by the National Mexican Council development of portable cantilever-based detectors [19]–[21].
of Science and Technology. The associate editor coordinating the review of
this paper and approving it for publication was Dr. Sillas Hadjiloucas. Acoustic wave based devices have also been used as an
S. Thomas, F. H. Villa-López, M. Cole, and J. W. Gardner are alternative approach for particle sensing. These devices use a
with the Microsensors and Bioelectronics Laboratory, School of piezoelectric material in which a mechanical wave is generated
Engineering, University of Warwick, Coventry CV4 7AL, U.K. (e-mail:
sanju.thomas@warwick.ac.uk; f.h.villa-lopez@warwick.ac.uk; m.cole@ when an electrical field is applied. The mass loading onto the
warwick.ac.uk; j.w.gardner@warwick.ac.uk). resonator due to the particles deposited on the sensing area
J. Theunis and J. Peters are with the Flemish Institute for Tech- causes a shift in the resonant frequency of the device. Quartz
nological Research, Mol B-2400, Belgium (e-mail: jan.theunis@vito.be;
jan.peters@vito.be). Crystal Microbalances (QCMs) were used in the system pro-
Digital Object Identifier 10.1109/JSEN.2015.2512303 posed by Liang et al. for the measurement of particle mass
1558-1748 © 2015 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission.
See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.

Authorized licensed use limited to: Central Electronics Engineering Research Institute. Downloaded on July 08,2022 at 08:39:35 UTC from IEEE Xplore. Restrictions apply.
 THOMAS et al.: PARTICLE SENSOR USING SMRs 2283

Fig. 1. Overall structure of the developed differential mode particle sensing

system based on Solidly Mounted Resonators.

concentration and size distribution [22] and a sensor based on

Surface Acoustic Wave Resonators (SAWR) was demonstrated
for the detection of fine particles by Thomas et al. [23].
The typical operating frequencies of QCMs are in the
range of 5 – 30 MHz [24] whereas the resonant frequency of
SAW devices is typically between 30 MHz and 1 GHz [25]. Fig. 2. Photograph of the developed particle sensing unit using SMRs and
working in a dual configuration. The overall dimensions are 49 mm × 44 mm.
Thin Film Bulk Acoustic Wave devices (TFBAW) make
use of thin film technology to operate at higher resonant
frequencies and therefore higher sensitivities are achieved
The differential signal is ported to a microcontroller for the
compared to other devices. TFBAW devices consist of a
measurement of the frequency and data are logged to a PC via
thin piezoelectric layer sandwiched between two electrodes
USB serial communication using National Instruments Lab-
and fabricated on top of a carrier substrate, typically silicon.
VIEW virtual instrumentation. The assembled particle sensing
The footprint of TFBAW devices is much smaller to that
unit is shown in Fig. 2. Dies containing a total of four SMRs
of SAWs and QCMs. Unlike SAW devices, TFBAWs are
were used but only one of the devices was connected. The
compatible with low-cost silicon technologies making them
SMRs were wire-bonded onto an LCC package and interfaced
suitable for monolithic integration.
to the oscillator boards. The reference SMR was covered
In TFBAW, acoustic isolation from the substrate must be
with a 3-D printed cap. The cap consisted of a thin layer
provided to the resonator structure in order to confine the
(0.15 mm thick) that completely covered the sensor preventing
wave energy and prevent wave dissipation into the substrate.
particles to fall onto its surface but also allowing the SMR
According to the way in which this is achieved, two differ-
to be exposed to the ambient conditions, i.e. temperature and
ent types of TFBAW devices can be differentiated namely,
humidity. In this way these common mode interferences can be
Film Bulk Acoustic Resonators (FBAR) and Solidly Mounted
suppressed by using the dual sensor configuration. The particle
Resonators (SMR).
sensing unit was enclosed in order to protect the electronic
The use of an FBAR device as the mass sensitive element
circuitry, having overall dimensions of 49 mm × 44 mm.
for the development of a portable particulate matter monitor
was proposed by researchers at the University of California,
Berkley [6], [26]. B. Solidly Mounted Resonator Structure
In this work, we present the development of a low-cost,
Zinc oxide based solidly mounted resonators were used
highly sensitive particle sensing unit employing zinc oxide
as the sensing element for developing the particle detector.
based Solidly Mounted Resonators working in a dual config-
A schematic of the employed SMRs is shown in Fig. 3(a).
uration and driven by Colpitts type oscillators.
The SMRs were fabricated at Cambridge University on a p-
type Si (100) substrate and consisted of a 2.96 μm thin film
II. D ESCRIPTION OF THE S YSTEM
of ZnO sandwiched between 200 nm thick Al electrodes and
A. Overall Outline deposited on top of an acoustic mirror formed by three pairs
The overall structure of the developed particle sensor system of alternating layers of 1.82 μm Mo and 1.65 μm SiO2 .
is shown in Fig. 1. The system operates in a dual mode The c-axis oriented ZnO layer was reactively sputtered from
configuration for the suppression of common mode interfer- a 4-inch Zinc target with an Ar/O2 mixture using a high
ences [27], such as temperature, humidity or pressure effects. target utilisation sputtering (HiTUS) system [28], obtaining
SMR devices resonating at ∼970 MHz are driven by a Colpitts a deposition rate of ∼20 nm/min.
type oscillator, one device is working as the reference channel The SMRs were fabricated using a 4 mask photolithography
whereas the second one is acting as the sensing device. The process for the patterning of the acoustic mirror, the bottom
output signal of the oscillators is sent to an interface board electrode, the top electrode and opening via through the
that includes an RF mixer, a low pass filter and a comparator. piezoelectric. The sensing area of the device determined by
The high frequency signals of the oscillators are mixed and the overlapping of the bottom and top electrode is 200 μm by
filtered obtaining a lower differential frequency output, which 200 μm while the footprint of a single device is 1 mm square.
is easier to measure at high resolution. The SMRs resonate at a frequency of ∼970 MHz. Details

Authorized licensed use limited to: Central Electronics Engineering Research Institute. Downloaded on July 08,2022 at 08:39:35 UTC from IEEE Xplore. Restrictions apply.
 2284 IEEE SENSORS JOURNAL, VOL. 16, NO. 8, APRIL 15, 2016

Fig. 4. Schematic of the Colpitts type oscillator circuitry designed for driving
the SMR devices.

interference between incident and reflected bulk waves occur

when the wavelength is an odd multiple of the double
substrate thickness. Thus a standing wave is created inside
the sensor boundaries as illustrated in the schematic drawing
shown in Fig. 3(b).

C. Oscillator Circuitry and Interface Board

Acoustic wave perturbations induced by particle deposition
present themselves as attenuation and velocity changes of the
bulk wave; however, the real-time detection of these changes
Fig. 3. (a) Three-dimensional representation of the ZnO based
Solidly Mounted Resonator with Al electrodes operating at ∼970 MHz.
require complex and bulky circuitry unsuitable for integrated
(b) Schematic diagram showing the longitudinal bulk wave propagation inside systems. It is possible to monitor the acoustic velocity changes
an SMR sensor illustrating the standing wave pattern created inside the indirectly with great measurement precision [32] by using
resonator thickness and the decay length into the adjacent medium associated
with the bulk wave.
the SMR sensor as a resonating element inside a simple
oscillator circuit. The dual SMR resonators, both operating
at a frequency of ∼970 MHz utilize the Colpitts oscillator
configuration with a grounded base configuration in order to
on the design, modelling and fabrication of these devices are obtain good frequency stability and sensitivity. Here the SMR
reported elsewhere [29]. input port is ac coupled to the transistor’s base and the output
An electrical signal applied between the electrodes gener- port is connected to the ground.
ates a mechanical wave that propagates along the bulk of the The Colpitts oscillator was chosen because it allows the
piezoelectric material. The total mass of the particles deposited SMR to operate in a 1-port configuration by grounding the
on the active area of the sensing SMR produces a shift in the output port as opposed to a Pierce oscillator in which it
resonant frequency, which needs to be measured. requires a 2-port configuration [33]. These oscillators offer
These SMRs have been designed to operate in a longitudinal good stability at frequencies above 500 MHz, lower harmon-
mode, which is characterised by particle displacements in the ics, lower component count and hence lower cost than other
same direction of the wave propagation. The application of an types of oscillator circuits [34]. Fig. 4 shows a schematic
alternating electric field between the two electrodes produces of our oscillator circuit. The commercial NPN Silicon RF
a longitudinal deformation through the thickness of the piezo- Transistor BFR92P used to provide gain to the active oscillator
electric material in a thin film resonator. Upon deformation, part, reduces the effect of parasitic capacitances considerably
the acoustic wave propagation through the bulk of the material when compared to an op-amp, allowing the circuits to operate
across the crystal and the particle displacement are both in at higher frequencies. The resonator behaves like an inductor
the same direction, normal to the sensor surface. Longitudinal between the series and parallel resonance regions of the SMR
mode SMRs have been used in this work for nanoparticle device. The radio frequency (RF) transistor along with the
detection as they have been found to be providing promising feedback capacitor C4 provides the gain to compensate for
results for sensing in air or gas [30], [31]. A constructive the resistive losses in the resonator [34].

Authorized licensed use limited to: Central Electronics Engineering Research Institute. Downloaded on July 08,2022 at 08:39:35 UTC from IEEE Xplore. Restrictions apply.
 THOMAS et al.: PARTICLE SENSOR USING SMRs 2285

Fig. 5. Spectrum of the SMR-Colpitts oscillator showing a resonant frequency

of 933 MHz.

Fig. 6. Block diagram of the setup used to perform particle testing with the
For the circuit to oscillate, the Barkhausen criterion needs developed SMR-based unit.
to be satisfied when the SMR is connected between the
RF transistor base and ground. The LC tank circuit and
the grounding conditions of the transistor base provided tions and the output frequency signal range. The comparator
through the SMR create the initial startup oscillator frequency, converts this differential frequency output into a digital signal
which is a few MHz above the steady state frequency. The with voltage level compatible to the voltage tolerance of the
LC oscillation noise and the wideband noise energy at the microcontroller digital input pins (3.3 V), where the frequency
SMR’s resonance frequency will get stored in the device [35]. counting takes place. The interface board is powered by the
The following two conditions including unity total loop gain 5 V supply of the microcontroller, which in turn is powered
magnitude and 0° phase shift for the entire loop [36] will via a USB connection to the PC.
be fulfilled when a standing electrical wave is created inside A low-cost microcontroller Teensy 3.1 was used in order
the resonator. As the energy builds up in the SMR, more to accomplish the frequency counting of the square output
current flows through the device, which results in shifting signal and log the data to a computer through USB serial
of the oscillation frequency to the resonance frequency of communication. These data were recorded with virtual instru-
the SMR. The output spectrum from the Colpitts based SMR ment developed using LabVIEW software which also allows
oscillator showing a resonant frequency of 933 MHz obtained the real time visualization of the data.
by an RF oscilloscope (Tektronix MDO3012 Mixed Domain
Oscilloscope) is shown in Fig. 5.
When the particles are deposited on to the sensor surface, III. E XPERIMENTAL S ETUP
the total phase of the feedback loop will get shifted by a Characterization of the developed particle sensor based
certain amount due to attenuation and velocity changes of upon solidly mounted resonators took place inside a sealed
the bulk wave. As a result, the Barkhausen criterion will be environmental chamber at VITO, Belgium. A schematic of
satisfied at a lower frequency and thus the resonance of the the experimental setup is shown in Fig. 6. The SMRs were
SMR will change from the initial resonant frequency to a placed inside the test chamber together with a range of
lower resonant frequency, resulting in a shift in frequency. reference commercial instruments for real-time monitoring of
In order to measure the frequency shift caused by the particles particle deposition within the test chamber. The commercial
deposited onto the sensing SMR device, an interface board sensors include an acoustic based Quartz Crystal microbalance
was designed that consists of a double balanced RF signal (Vitrocell® Systems), and two different optical particle coun-
mixer (Hittite Microwave Corporation), an RF low pass filter ters (Dylos Corporation, Grimm Technologies Inc.), which
(Mini-circuits®), a comparator (Analog Devices) and a dual were placed adjacent to the SMR based research sensors.
linear voltage regulator (Micrel® ). The oscillator boards were The test rig consisted of a dust generator (TOPAS® ), a suc-
connected to the interface board and powered by the 2.5 V tion pump for controlling the dust flow into the chamber and a
output of the low-dropout regulator. The reference and sensing humidity control unit. Photographs of the experimental setup
oscillator frequencies were mixed using a heterodyne down- are shown in Fig. 7. Typical conditions inside the chamber
conversion technique and the low pass filter was used to output with an internal volume of 0.72 m3 were 24 °C and 22% RH.
only the difference frequency. In this way, the mixing circuit The SMR sensors and the commercial instruments were placed
helps in reducing both the effects of common mode varia- at one of the corners inside the chamber as shown in Fig. 8.

Authorized licensed use limited to: Central Electronics Engineering Research Institute. Downloaded on July 08,2022 at 08:39:35 UTC from IEEE Xplore. Restrictions apply.
 2286 IEEE SENSORS JOURNAL, VOL. 16, NO. 8, APRIL 15, 2016

Fig. 8. Experimental setup: SMR based particle detector and reference

instruments inside the test chamber.

Fig. 7. Experimental setup: (a) environmental chamber with data acquisition

hardware and (b) humidity control and dust generator setup.

Fig. 9. Frequency shift of the SMR particle sensor due to the injection of a
The PM concentration readings from all the instruments were known concentration of Arizona dust.
continuously logged to the PCs.
The target particles during these measurements included
Arizona test dust (Powder Technology Inc.) with nominal concentration inside the chamber in units of μg/m3 . On the
particle diameter of 0-3 μm. The particle size distribution other hand, as gravitational sedimentation is used to collect
of these test particles had a median, d50 , of 0.927 μm. A the particles onto the SMR-based sensor and the commercial
90% of the particles were found to be below a diameter QCM device, the sensors will show a frequency shift only after
of 1.526 μm (d90 ) whereas only 10% of the particles were particles have settled on their respective surfaces. The greater
below 0.712 μm (d10 ). The free-settling velocity [37] of the amount of particles that were injected, the greater amount
Arizona dust particles with 1.5 μm in size would be ∼81 μm/s of particles that will deposit onto the resonator.
whereas finer particles with a diameter of 0.5 μm fall at
∼9 μ/s. Therefore, only the bigger particles will fall onto the IV. R ESULTS AND D ISCUSSION
resonator whereas the very small particles are unlikely to settle After the injection of the test particles into the environmen-
on the device. Hence, UPFs are not detected with this method. tal chamber, a change in the oscillating frequency of the SMR
Particles were injected into the chamber for a certain period based sensor was observed. Fig. 9 shows a typical frequency
of time ranging from a couple of seconds to up to 10 minutes shift measurement of the SMR system when exposed to a
obtaining dust concentrations as low as 20 μg/m3 and as predefined amount of dust concentration. There was a decrease
high as 30,000 μg/m3 as measured by the optical particle in the resonant frequency of the SMR sensor as shown in
counters (OPCs); namely the Grimm monitor. The longer the Fig. 9, due to the addition of nanoparticles on to the sensor
particle injection period, the greater the amount of particles surface. The environmental chamber was completely sealed
inside the chamber. As the particles are injected, the commer- to ensure ambient temperature stability, to avoid any external
cial optical sensor draws a sample air and measures the particle wind effects and the deposition of any foreign material onto the

Authorized licensed use limited to: Central Electronics Engineering Research Institute. Downloaded on July 08,2022 at 08:39:35 UTC from IEEE Xplore. Restrictions apply.
 THOMAS et al.: PARTICLE SENSOR USING SMRs 2287

Fig. 10. Real time frequency shift measurement of the SMR sensor response Fig. 11. Measured frequency shifts of the SMR sensor due to the different
to the deposition of Arizona dust compared to the commercial monitors concentrations of Arizona dust.
including QCM device and Grimm OPC.

[38], [39] that is given by (1):


2η
δ= (1)
ωρ
sensors within the laboratory setting. Additionally, the effects
of any temperature or humidity changes are also eliminated where ω = 2π f is the angular frequency, η is the viscosity
by operating the sensor system in the differential mode. and ρ is the density of the lossy adjacent medium.
As the deposition method is based on the sedimentation of The frequency response of the SMR particle sensor was
the particles due to gravity, an average particle settling period measured to various levels of dust concentrations to establish
of 2 minutes was observed for Arizona dust defined by the the sensitivity of the particle sensor. Based on the data col-
time in which the frequency of the SMR sensor started shifting lected from the experimental measurements, the relationship
due to the added mass of the settled particles. The real time between the SMR sensitivity and particle concentrations has
measurements of the SMRs are comparable to the response been found to be a linear response as shown in Fig. 11, with a
of the commercial QCM device. The operating frequency of sensitivity of 4.6 Hz per μg/m3 confirming to have the SMR
both acoustic wave based devices (SMR and QCM) decreased sensor to be operating in the non-saturated regime. In addition,
due to the added mass of particles falling onto their sensing the minimum detectable particle concentration of the SMR
surface at very similar times. However, the mass sensitivity based system was found to be about 20 μg/m3 .
demonstrated by the SMR device is orders of magnitude higher High frequency solidly mounted resonators will only
than the reference QCM sensor (thousands of kHz for the SMR respond to the surface interactions that occur within the close
compared to less than 1 Hz for the QCM) as can be observed proximity of the sensor surface. An 970 MHz SMR based
from Fig. 10. sensor will have a decay length that is up to the range of 1 μm,
The mass deposited at the surface of the SMR sensor depending on the type of the adjacent lossy medium probe.
follows the thickness extensional vibration of the piezoelectric As a result, those particles that are having diameters smaller
material and hence the loaded SMR would simply behave than the decay length, will only get probed by the SMR device
as if it were thicker. The effective wavelength of the bulk by allowing acoustic coupling of the entire particle volume.
wave is thus increased and consequently its resonant frequency Otherwise, the sensor only probes the particle partially near
decreases. This explains the frequency shift produced by its surface, resulting in the non-response of the sensor due to
the deposition of nanoparticles on the active area of the larger sized particles. Hence there is a need for tailoring of the
solidly mounted resonator, assuming that there is no energy decay length to suit the size of the particles to be detected.
dissipation and hence will only be valid for thin, rigid and The decay length of an acoustic wave resonator device
uniform films [37] having similar acoustic behaviour as that always depends on the sensor resonance frequency. For an
of the bulk piezoelectric material. SMR device, the amplitude and decay length of the lon-
However, when the acoustic properties of the deposited gitudinal acoustic wave transmitted to the adjacent lossy
viscoelastic nanoparticle probe layer differ significantly from medium decrease with increasing sensor resonance frequency.
those of the SMR, it experiences viscous coupling primarily This enables the particle sensor described here to be capable
due to the lossy surrounding media causing the shear longitu- of detecting submicron-sized particles with picogram mass
dinal waves to dissipate into the adjacent media, resulting in range, using sensitivity-tailored frequency-dependent designs
the degradation of the Q-factor. The longitudinal shear wave of acoustoelectric sensors. During the performed experiments,
propagating into the adjacent lossy medium depends on the we have not observed saturation of the sensors which suggest
resonant frequency of the SMR and the signal amplitude is that the particles may have not form multilayers on the
degraded exponentially with a characteristic decay length (δ) resonator surface.

Authorized licensed use limited to: Central Electronics Engineering Research Institute. Downloaded on July 08,2022 at 08:39:35 UTC from IEEE Xplore. Restrictions apply.
 2288 IEEE SENSORS JOURNAL, VOL. 16, NO. 8, APRIL 15, 2016

V. C ONCLUSIONS AND F URTHER W ORK [12] B. Park, J. Hong, and S.-B. Lee, “Real-time detection of airborne dust
particles using paddle-type silicon cantilevers,” J. Vac. Sci. Technol. B,
A particle sensor has been developed based upon two solidly vol. 27, no. 6, pp. 3120–3124, 2009.
mounted resonators interfaced to Colpitts type oscillators. [13] B. P. Harrington, R. Abdolvand, A. Hajjam, J. C. Wilson, and
The sensor works in a differential mode compensating for S. Pourkamali, “Thin-film piezoelectric-on-silicon particle mass sen-
sors,” in Proc. IEEE Int. Freq. Control Symp. (FCS), Newport Beach,
common temperature and humidity effects. The SMR based CA, USA, Jun. 2010, pp. 238–241.
system was characterized inside an environmental chamber, [14] A. Hajjam, J. C. Wilson, A. Rahafrooz, and S. Pourkamali, “Fabrication
where particles of 0-3 μm in size were injected. The added and characterization of resonant aerosol particle mass sensors,” in Proc.
IEEE 23rd Int. Conf. Micro Electro Mech. Syst. (MEMS), Hong Kong,
mass of the particles deposited on the resonator caused a Jan. 2010, pp. 863–866.
frequency shift on the SMR device. Real-time measurements [15] E. Mehdizadeh, V. Kumar, S. Pourkamali, J. Gonzales, and
of the frequency shift demonstrated the capability of these R. Abdolvand, “A two-stage aerosol impactor with embedded MEMS
resonant mass balances for particulate size segregation and mass con-
acoustic wave based devices to detect fine particles with a centration monitoring,” in Proc. IEEE SENSORS, Baltimore, MD, USA,
sensitivity of 4.6 Hz per μg/m3 . Particles were collected onto Nov. 2013, pp. 1–4.
the resonator surface by gravitational sedimentation. However, [16] E. Mehdizadeh, J. C. Wilson, A. Hajjam, A. Rahafrooz, and
S. Pourkamali, “Aerosol impactor with embedded MEMS resonant
by using this method ultrafine particles are unlikely to settle mass balance for real-time particulate mass concentration monitoring,”
on to the device. Therefore, for the development of the full in Proc. 17th Int. Conf. Solid-State Sens., Actuators, Microsyst.
sensor system, other particle deposition methods will have to (TRANSDUCERS EUROSENSORS), Barcelona, Spain, Jun. 2013,
pp. 661–664.
be implemented. Further work is currently being carried out
[17] H. S. Wasisto et al., “A resonant cantilever sensor for monitoring
towards the development of a low-cost particle sensor system airborne nanoparticles,” in Proc. 16th Int. Solid-State Sens., Actuators,
in a package (SiP) based on SMR devices by interfacing Microsyst. Conf. (TRANSDUCERS), Beijing, China, Jun. 2011,
the resonators to standard CMOS circuitry. Ultimately full pp. 1116–1119.
[18] H. S. Wasisto S. Merzsch, A. Waag, E. Uhde, T. Salthammer, and
integration of the SMRs with CMOS technology will be E. Peiner, “Airborne engineered nanoparticle mass sensor based on
realized for the development of a low-power, low-cost smart a silicon resonant cantilever,” Sens. Actuators B, Chem., vol. 180,
miniature sensor for the real-time monitoring of particulate pp. 77–89, Apr. 2013.
[19] S. Merzsch et al., “Low-weight electrostatic sampler for airborne
matter. nanoparticles,” in Proc. IEEE Sensors, Limerick, Ireland, Oct. 2011,
ACKNOWLEDGMENT pp. 1177–1180.
[20] H. S. Wasisto S. Merzsch, A. Waag, E. Uhde, T. Salthammer, and
The authors acknowledge Prof. Andrew Flewitt and E. Peiner, “Portable cantilever-based airborne nanoparticle detector,”
Mr. Girish Rughoobur at Cambridge University for fabrication Sens. Actuators B, Chem., vol. 187, pp. 118–127, Oct. 2013.
[21] H. S. Wasisto, S. Merzsch, E. Uhde, T. Salthammer, A. Waag, and
of the SMR devices. E. Peiner, “Handheld micromechanical cantilever mass sensor for early
R EFERENCES detection of carbon nanoparticles,” in Proc. TG/GMA Symp. Sensors
Meas. Syst., Nuremberg, Germany, Jun. 2014, pp. 1–6.
[1] R. Rückerl, “Health effects of particulate air pollution: A review of
[22] D. Liang, W.-P. Shih, C.-S. Chen, and C.-A. Dai, “A miniature system
epidemiological evidence,” Inhalation Toxicol., Int. Forum Respiratory
for separating aerosol particles and measuring mass concentrations,”
Res., vol. 23, no. 10, pp. 555–592, 2011.
Sensors, vol. 10, no. 4, pp. 3641–3654, 2010.
[2] F. Dominici et al., “Fine particulate air pollution and hospital admission
[23] S. Thomas, Z. Racz, M. Cole, and J. W. Gardner, “Dual high-frequency
for cardiovascular and respiratory diseases,” J. Amer. Med. Assoc.,
Surface Acoustic Wave Resonator for ultrafine particle sensing,” in Proc.
vol. 295, no. 10, pp. 1127–1134, 2006.
[3] N. Künzli et al., “Ambient air pollution and atherosclerosis in IEEE SENSORS, Nov. 2013, pp. 1–4.
Los Angeles,” Environ. Health Perspect., vol. 113, no. 2, pp. 201–206, [24] B. Drafts, “Acoustic wave technology sensors,” IEEE Trans. Microw.
2005. Theory Techn., vol. 49, no. 4, pp. 795–802, Apr. 2001.
[4] J. F. Pearson, C. Bachireddy, S. Shyamprasad, A. B. Goldfine, and [25] J. W. Gardner, V. K. Varadan, and O. O. Awadelkarim, Microsensors,
J. S. Brownstein, “Association between fine particulate matter and MEMS, and Smart Devices, New York, NY, USA: Wiley, 2001.
diabetes prevalence in the U.S.,” Diabetes Care, vol. 33, no. 10, [26] I. Paprotny, F. Doering, P. A. Solomon, R. M. White, and L. A. Gundel,
pp. 2196–2201, 2010. “Microfabricated air-microfluidic sensor for personal monitoring of air-
[5] P. S. Shah and T. Balkhair, “Air pollution and birth outcomes: borne particulate matter: Design, fabrication, and experimental results,”
A systematic review,” Environ. Int., vol. 37, no. 2, pp. 498–516, 2011. Sens. Actuators A, Phys., vol. 201, pp. 506–516, Oct. 2013.
[6] J. P. Black, A. Elium, R. M. White, M. G. Apte, L. A. Gundel, and [27] S. Thomas, S. L. T. Leong, Z. Racz, M. T. Cole, and J. W. Gardner,
R. Cambie, “MEMS-enabled miniaturized particulate matter monitor “Design and implementation of a high-frequency surface acoustic wave
employing 1.6 GHz aluminum nitride thin-film bulk acoustic wave sensor array for pheromone detection in an insect-inspired infochemical
resonator (FBAR) and thermophoretic precipitator,” in Proc. IEEE communication system,” in Proc. 14th Int. Meeting Chem. Sensors,
Ultrason. Symp., New York, NY, USA, Oct. 2007, pp. 476–479. Nuremberg, Germany, 2012, pp. 11–14.
[7] Air Quality Expert Group, “Particulate matter in the United Kingdom,” [28] L. García-Gancedo, “Room-temperature remote-plasma sputtering of
Dept. Environ., Food, Rural Affairs Rep., 2005. c-axis oriented zinc oxide thin films,” J. Appl. Phys., vol. 112, no. 1,
[8] H. Grimm and D. J. Eatough, “Aerosol measurement: The use of optical p. 014907, 2012.
light scattering for the determination of particulate size distribution, [29] F. H. Villa-López, G. Rughoobur, S. Thomas, A. J. Flewitt, M. Cole, and
and particulate mass, including the semi-volatile fraction,” J. Air Waste J. W. Gardner, “Design and modelling of solidly mounted resonators for
Manage. Assoc., vol. 59, no. 1, pp. 101–107, 2009. low-cost particle sensing,” Meas. Sci. Technol, vol. 27, no. 2, p. 025101,
[9] A. Hajjam, J. C. Wilson, and S. Pourkamali, “Individual air-borne 2015.
particle mass measurement using high-frequency micromechanical res- [30] R. Gabl et al., “Novel integrated FBAR sensors: A universal technology
onators,” IEEE Sensors J., vol. 11, no. 11, pp. 2883–2890, Nov. 2011. platform for bio- and gas-detection,” in Proc. IEEE Sensors, Oct. 2003,
[10] H. H. Lim, D. Park, J. Y. Maeng, J. Hwang, and Y. J. Kim, “MEMS pp. 1184–1188.
based integrated particle detection chip for real time environmental [31] I. Paprotny, F. Doering, and R. M. White, “MEMS particulate mat-
monitoring,” in Proc. 19th IEEE Int. Conf. Micro Electro Mech. ter (PM) monitor for cellular deployment,” in Proc. IEEE Sensors,
Syst. (MEMS), Istanbul, Turkey, Jan. 2006, pp. 62–65. Nov. 2010, pp. 2435–2440.
[11] G. Jaramillo, C. Buffa, M. Li, F. J. Brechtel, G. Langfelder, and [32] Z. Rácz, M. Cole, J. W. Gardner, M. F. Chowdhury, W. P. Bula,
D. A. Horsley, “MEMS electrometer with femtoampere resolution for J. Gardeniers, S. Karout, A. Capurro, and T. C. Pearce, “Design and
aerosol particulate measurements,” IEEE Sensors J., vol. 13, no. 8, implementation of a modular biomimetic infochemical communication
pp. 2993–3000, Aug. 2013. system,” Int. J. Circuit Theory Appl., vol. 41, no. 6, pp. 653–667, 2013.

Authorized licensed use limited to: Central Electronics Engineering Research Institute. Downloaded on July 08,2022 at 08:39:35 UTC from IEEE Xplore. Restrictions apply.
 THOMAS et al.: PARTICLE SENSOR USING SMRs 2289

[33] A. T. Nimal, M. Singh, U. Mittal, and R. D. S. Yadava, “A comparative Jan Theunis received the M.Sc. degree in
analysis of one-port Colpitt and two-port Pierce SAW oscillators for bioscience engineering from the Katholieke Univer-
DMMP vapor sensing,” Sens. Actuators B, Chem., vol. 114, no. 1, siteit Leuven, Leuven, in 1988. He joined the Flem-
pp. 316–325, 2006. ish Institute for Technological Research in 1999.
[34] S. Thomas, Z. Rácz, M. Cole, and J. W. Gardner, “High-frequency From 2004 to 2008, he was the Head of the Air
one-port Colpitts SAW oscillator for chemical sensing,” in Proc. 6th Quality Measurements Team.
Int. Conf. Adv. Circuits, Electron., Micro-Electron. (CENICS), 2013, He is currently in charge of the strategic research
pp. 13–17. program on sensor networks and mobile sensing
[35] EPCOS. (2014). Design-Guide for the SAW Oscillator Optimisation for air quality. The research program is developing
for Best Frequency Stability, Fast Start Up Time and Ghost-Less novel air quality monitoring strategies based on low-
Working. [Online]. Available: http://en.tdk.eu/blob/530946/download/3/ cost sensors and portable monitoring devices. The
pdf-an25.pdf research group has unique lab test facilities to test and validate measurement
[36] R. F. Schmitt, J. W. Allen, and R. Wright, “Rapid design of SAW equipment and methods.
oscillator electronics for sensor applications,” Sens. Actuators B, Chem.,
vol. 76, nos. 1–3, pp. 80–85, 2001.
[37] A. Janshoff, H.-J. Galla, and C. Steinem, “Piezoelectric mass-sensing
devices as biosensors—An alternative to optical biosensors?” Angew. Jan Peters received the M.Sc. degree in
Chem. Int. Ed., vol. 39, no. 22, pp. 4004–4032, 2000. bioscience engineering from the Katholieke
[38] G. N. M. Ferreira, A.-C. da-Silva, and B. Tomé, “Acoustic wave Universiteit Leuven in 2002, and the Ph.D. degree
biosensors: Physical models and biological applications of quartz crystal in bioscience engineering from Ghent University
microbalance,” Trends Biotechnol., vol. 27, no. 12, pp. 689–697, 2009. in 2008. He joined the Laboratory of Air Quality
[39] M. Nirschl, M. Schreiter, and J. Vörös, “Comparison of FBAR and Measurements, Flemish Institute for Technological
QCM-D sensitivity dependence on adlayer thickness and viscosity,” Research, in 2011. His main research interests
Sens. Actuators A, Phys., vol. 165, no. 2, pp. 415–421, 2011. are development and testing of new sensing
technologies for air quality monitoring, monitoring
of air pollution at industrial and urban environment,
and implementation of air pollution reduction
technologies.

Marina Cole (M’98) received the B.Sc. degree from

the University of Montenegro (formerly Yugoslavia),
and the Ph.D. degree from Coventry University,
Sanju Thomas (M’12) received the B.Eng. degree Coventry, U.K. She joined the School of Engineer-
from Visveswaraiah Technological University, ing, University of Warwick, Warwick, U.K., in 1996,
Karnataka, India, and the M.Sc. degree in as a Post-Doctoral Research Assistant, where she
biomedical engineering from the University of was appointed as a Lecturer in Electronic Engi-
Warwick, Coventry, U.K., where he is currently neering in 1998. Her main research interests are
pursuing the Ph.D. degree. He is a Research Staff integrated silicon-based sensors, SAW-based sen-
Member with the School of Engineering, University sors, analog and mixed-signal ASICs, smart sensors,
of Warwick. His main research interests are actuators, and microsystems.
chemical sensing and artificial olfaction, acoustic
wave sensors (SAW/FBAR), analog CMOS circuit
design, smart sensors, and microsystems.
Julian W. Gardner (M’91–SM’02) received
the B.Sc. degree in physics from the Univer-
sity of Birmingham, Birmingham, U.K., in 1979,
the Ph.D. degree in physical electronics from
Cambridge University, Cambridge, U.K., in 1982,
and the D.Sc. degree in electronics engineering from
Warwick University, Coventry, U.K., in 1997.
He is currently a Professor of Electronic Engi-
neering with the School of Engineering, University
of Warwick. He is also the Head of Electrical and
Farah H. Villa-López (M’12) received the B.Sc. Electronic Engineering and the Microsensors and
(Hons.) degree in mechanical and electrical engi-
Bioelectronics Laboratory. He has authored or co-authored over 500 technical
neering from Universidad Veracruzana, Xalapa, papers and patents, and six technical books in the area of microsensors and
Mexico, in 2012. She is currently pursuing the machine olfaction. His research interests include the modeling of silicon
Ph.D. degree with the Micronsensors and Bio- microsensors, chemical sensor array devices, biomimetic MEMS devices, and
electronics Laboratory, University of Warwick, electronic noses.
Coventry, U.K. Her research interests include Dr. Gardner is a Fellow of the Institute of Engineering and Technol-
acoustic wave sensors and microsystems. ogy (U.K.) and was elected as a Fellow of the Royal Academy of Engineering
in 2006. He received the J. J. Thomson Medal for Outstanding Achievement
in Electronics by the Institute of Engineering and Technology in 2007.

Authorized licensed use limited to: Central Electronics Engineering Research Institute. Downloaded on July 08,2022 at 08:39:35 UTC from IEEE Xplore. Restrictions apply.