Or

Aa

b

a

ARRAA

KMMOMMP

1

baetirtnwo

dd

h0

Sensors and Actuators A 242 (2016) 132–139

Contents lists available at ScienceDirect

Sensors and Actuators A: Physical

journa l homepage: www.e lsev ier .com/ locate /sna

ptical fiber array based simultaneous parallel monitoring ofesonant cantilever sensors in liquid

ref Mostafazadeh a,∗, Goksen G. Yaralioglu b, Hakan Urey a

Koc University, Electrical and Electronics Engineering Department, Istanbul, TurkeyOzyegin University, Electrical and Electronics Engineering Department, Istanbul, Turkey

r t i c l e i n f o

rticle history:eceived 18 December 2015eceived in revised form 29 February 2016ccepted 2 March 2016vailable online 5 March 2016

eywords:EMS sensors in liquidultichannel resonant frequency trackingptical fiber array readoutagnetic actuationultichannel lock-in amplifier

hase based resonance tracking

a b s t r a c t

This paper reports a novel method for simultaneous resonance monitoring of MEMS cantilevers usingphase based dynamic measurements without any electrical connections to the sensor array. MEMS can-tilevers are made of electroplated nickel and actuated remotely with magnetic field using an electro-coil.To our knowledge this is the first demonstration of simultaneous parallel optical monitoring of dynamicmode micro-cantilever array in liquid environment. Illumination is generated using a laser source anda diffractive pattern generator, which provides 500 �W laser power per channel. A compact fiber arraybased pick-up was built for optical readout. Its main advantages are easy customization to different sizeand pitch of sensor array, and good immunity to electrical noise and magnetic interference as the photodetectors are located away from the electro-coil. The resonant frequency of the cantilever is tracked witha custom multi-channel lock-in amplifier implemented in software. For demonstrating the stability andsensitivity of the system we performed measurements using glycerol solutions with different viscosi-ties. Measured phase sensitivity was 0.9◦/1% of Glycerol/DI-water solution and the standard deviation of

◦

measured phase was 0.025 . The resulting detection limit for Glycerol/DI-water solution was 280 ppm.The proposed method showed robust results with low laser power and very good noise immunity tointerference signals and environmental vibrations. The sensor technology demonstrated here is very sig-nificant as it is scalable to larger arrays for simultaneous and real-time monitoring of multiple biologicaland chemical agents during fluid flow.

© 2016 Elsevier B.V. All rights reserved.

. Introduction

Functionality and performance of MEMS resonant sensors haveeen demonstrated in wide variety of applications from chemicalnd physical detection to biological sensing. Operation in liquidnvironment is important for most of biological applications buthe mechanical quality factor reduces dramatically in liquids, mak-ng the dynamic and real-time measurements noisy. Simultaneouseadout of multiple sensors is desired for multiple analyte detec-ion and also to obtain reference measurements for eliminatingoise, drift, and ambient changes. The main focus of this researchas to demonstrate simultaneous dynamic mode measurements

f resonant cantilevers operated in liquids.

Cantilevers were introduced as a sensing element of various

etection systems due to their high sensitivity. Volatile chemicaletection in air has been demonstrated in resonant mode [1,2].

∗ Corresponding author.E-mail address: armostafazadeh@ku.edu.tr (A. Mostafazadeh).

ttp://dx.doi.org/10.1016/j.sna.2016.03.004924-4247/© 2016 Elsevier B.V. All rights reserved.

The high sensitivity stems from the fact that one can make can-tilevers with very small dimensions decreasing the mass of thesensitive element. Bio-agent detection in liquid environment havealso been demonstrated using micro cantilevers in static mode [3,4]where the spring constant of the cantilevers was made very small todetect atomic forces. Later, to increase the specificity of the sensorand to reduce the test time cantilever arrays have been introduced.The detection of multiple cantilevers in an array has increased thecomplexity of the measurement setups.

Simultaneous monitoring of cantilevers in an array has beendemonstrated for atomic force microscopy operating in air. Thosesystems used electrical readout using fabricated piezo sensors[5,6], diffraction gratings with integrated photodiode array [7]. Inthese methods, all the cantilever deflections are read at the sametime using a dedicated electronics. This increases the number ofelectrical connections to the sensors and complicates the oper-

ation under water due to passivation requirements. Previouslyour group proposed a method that uses a single photodiode andfrequency-multiplexing for multi-analyte detection in air [8]. Sim-ilar frequency multiplexing approach was also reported for optical

A. Mostafazadeh et al. / Sensors and Actuators A 242 (2016) 132–139 133

tem (

latftccl

irmtomittisi

uwstmeTuncpoa

be obtained from the coil amplifier. We found that 33-gauge sin-gle layer 80-turn wire minimized parasitic capacitances as well asthe resistive losses. The diameter and height of the coil was 14 mmand 12 mm respectively. A low-distortion high-current broadband

Fig. 1. 3D sketch of the sys

ever readout method using PSD and operating in vacuum [9]. It islso possible to employ time multiplexing, which reads one can-ilever at a time using a single position sensitive detector (PSD),or static and dynamic mode measurements [10–14]. However,ime multiplexing methods cannot provide real-time noise can-ellation. The frequency-multiplexing methods are not suitable forantilevers in fluids where the quality factors are low, resulting inarge overlap in the frequency responses of the cantilevers.

For liquid operation, electrical readout methods are challeng-ng due to required electrical connections. Applying optical levereadout method for array of cantilevers in parallel for simultaneous

easurements requires complicated multiple PSD or quad detec-or fabrication with exact pitch of the cantilevers and also becausef the noise and detection limits of the proposed methods, paralleleasurements in liquids only reported in (i) Static mode by detect-

ng the amount of bending of the cantilever using a single PSD andime multiplexing [13,15], (ii) Using commercial devices with mul-iple four-Quadrant photodiode readouts [16,17], (iii) White-lightnterferometry and imaging based parallel readout also reported fortatic mode operation in liquids [18,19] and (iv) Single cantilevernterferometric readout [20,21].

As outlined above, simultaneous parallel measurements in liq-ids are challenging and there is no prior work in this area. In thisork, we used cantilevers with the same dimensions, which have

imilar resonant frequencies. Cantilevers are fabricated using elec-roplated and released Nickel films, and are actuated by a remote

agnetic actuator consists of two permanent magnets and anlectro-coil. They are driven open loop at the same drive frequency.he relative phase between the drive signal and the cantilevers aresed to monitor the dynamic changes of all the cantilevers simulta-eously in real-time with high precision. The measured phase shifts

an be used to monitor analytes in liquids or liquid properties. Highrecision measurements are obtained using a compact optical read-ut based on a fiber array and custom digital multi-channel lock-inmplifier for real-time monitoring. The system consists of a dis-

Fig. 2. MEMS chip inside the Plexiglas cartridge.

Distances are not in scale).

posable cartridge including the MEMS die and a reader unit thatcontains the optics and electronics for the magnetic actuator andthe optical sensors.

Section 2 and 3 of this paper describes magnetic actuator andconstruction of the fiber array, respectively. Section 4 shows mod-eling and optimization results of the readout optics. Section 5discusses the multichannel implementation of the detection elec-tronics. Finally, Section 6 presents experimental results obtainedfrom a multichannel system.

2. Magnetic actuation

The magnetic actuation engine of the system consists of anelectro-coil, two permanent magnets for magnetization of the can-tilevers and a high-current coil driver. The neodymium permanentmagnet bars (3 × 4 × 9 mm) were placed in parallel to each other insuch a way to form a DC magnetic field with magnetic lines in thedirection of the cantilevers. The magnetic field creates magneticdipole moments in the nickel cantilever along the DC field lines.When the cantilever is excited by the AC magnetic field createdby the electro-coil, the dipoles try to align themselves with the ACfield vibrating the cantilever at the frequency of the AC magneticfield. The electro-coil was optimized to create maximum AC mag-netic field over the cantilevers for the maximum current that can

Fig. 3. Illuminated MEMS cantilevers using diffractive pattern generator.

134 A. Mostafazadeh et al. / Sensors and Actuators A 242 (2016) 132–139

Pho todiod e

Beam Spli�er

He-NeLaser Source

Bi-Convex Lens(f=25.4mm)

Large Core Op�cal Fiber

Fluid In Fluid Out

Diffrac�onGra�ng

Microscope Lens (40x)

of the optical setup.

ccs1ug

seo

3

aa

da[soicaicc

tlllaacbtcttl

dfsN

Plexi glas Housing

Fig. 4. Side view

oil driver was designed to maintain a fixed amplitude sine-waveurrent to the coil. The driver input was fed with a fixed frequencyine-wave using a function generator. The coil current was set to

Ap − p during the experiments and the generated AC magnetic fieldnder the cartridge was measured around 5 mTrms which is in aood agreement with theoretical calculations.

The cantilever arrays used in this work are fabricated using aimple, single mask process utilizing sputter, photolithography,lectro-deposition and wet etching steps [22–24]. The dimensionsf the nickel cantilevers were 200 × 20 × 2 �m.

. Optical design

Fig. 1 shows 3D sketch of the system consists of the magneticctuation unit, the removable cartridge containing the MEMS diend the illumination and readout optics.

For illuminating multiple cantilevers with a single laser beam aiffractive pattern generator was used. Illuminating of cantileverrrays were also reported using fiber ribbon [10], VCSEL arrays11–13] and LED array [15]. Using the diffractive pattern generatorplits the laser beam to equally spaced laser spots where the pitchf the spots could be matched to cantilever pitches easily by adjust-

ng the distance between the diffraction pattern generator and theantilevers. Laser line generators were also reported for cantileverrray illumination [7,8] giving simplicity and ease of alignment butn this method only a small portion of the laser power reaches to theantilever tips and also the reflected laser power from the substrateould introduce disturbing interference signals.

The generated equally spaced laser beams were focused to theip of the cantilevers using a bi-convex focusing lens with focalength of 25.4 mm. An optional convex lens could be used to paral-elize the beams in order to maintain normal incidence angle usingarge diffraction angle pattern generators. Fig. 2 shows the remov-ble Plexiglas cartridge containing the MEM chip. The cartridge hasn inlet and outlet for liquids and a reservoir where the MEMShip is placed. In the figure, the MEMS chip (approximately 1 cmy 1 cm) had 10 rows of cantilevers where each row had 13 can-ilevers. Fig. 3 shows a zoomed view of the MEMS chip. Five nickelantilevers were illuminated using a 5 spot pattern generator. Inhe figure, every other cantilever was illuminated for demonstra-ion purposes. The cantilever pitch was 300 �m and pitch of theaser spots was set to 1.8 mm in the final optical design.

Fig. 4 shows the optical design of the experimental system in

etail. As a laser source, a low-power 5 mW He-Ne laser was pre-

erred because of its perfect Gaussian beam shape. To maintain amall enough laser spot on cantilevers, a 40 × microscope lens with.A. of 0.65 was used as a beam expander before focusing lens. A Bi-

Fig. 5. Two channel pick-up head and 3D sketch of the FR4 PCB pieces.

Convex lens with focal length of 25.4 mm was used for focusing thelaser beam on the cantilevers. The microscope lens creates a largelaser spot with diameter of 8 mm on the focusing lens. The distancebetween the focusing lens and the cantilevers is 10 cm. A five spotdiffractive pattern generator was placed at a proper distance fromthe cantilevers to match the spots pitch with cantilevers pitch andforms five laser spots with 8 �m spot size on cantilever tips. Forreflecting the laser spots to the cantilevers and transmitting thereflected beams, a beam splitter was used. We also implemented aCCD camera for optical alignment of the spots on the cantilevers.

Reflected beams from the cantilever tips were relayed to pho-todiodes using a custom made optical fiber array pick-up head.All the cantilevers were driven with a fixed frequency sine wavenear their resonances through the electro-coil. The cantilever vibra-tion changes the direction of the reflected laser beam, thereforeit modulates the light coupled into the fiber. We constructed thepick-up head using two FR4 pieces with carved parallel channelswith proper fiber pitch that matches the pitch of the illuminatedcantilevers. The optical fibers were sandwiched and glued betweenthese two FR4 pieces. The other end of the optical fibers was con-nected to two identical photodiodes with build in amplifiers. Usingthe optical fibers makes it possible to make pick-up heads with dif-ferent pitches for matching to different MEMS chip designs easily.Another important advantage of using optical fibers is its electricalnoise immunity. Since the modulated signals on the photodiodes

are in orders of few millivolts, the magnetic field generated by theelectro-coil could easily be coupled to photodiode signals. Usingoptical fibers prevented electrical interferences by keeping the pho-

A. Mostafazadeh et al. / Sensors and Actuators A 242 (2016) 132–139 135

e fiber

tp

dnoma

4

c

x

watoofi

P

Fig. 6. Analytical analysis of coupled optical power to th

odiodes far from the electro-coil. Fig. 5 shows the two channelick-up head and a 3D sketch of the FR4 PCB pieces.

A recent work reported optical fiber based readout for angularisplacement monitoring using two optical fibers, one for illumi-ating and the other for readout [25]. Need for a focusing micro lensn the tip of the fiber in this work makes the fabrication processuch more difficult. In our system, the fiber tips were simply cut

nd polished.

. Modeling and simulation

Beam traveling distance (x′) on the fiber’s tip plane could bealculated as follows:

’ (t) = 3zf ztipL’

sin (ωt) , (1)

here zf is the distance of fiber-tip plane to cantilever, ztip is themplitude of the tip deflection, ω is the angular frequency of actua-ion and L′ is the cantilever’s effective length including the distancef the laser spot from the tip of the cantilever.Assuming large coreptical fiber as a circular aperture, coupled optical power to theber can be expressed as

(x) =∫ ∫

A(x, y)G(x − x′, y)dx′dy =∫ ∫

A

G(x − x′, y)dx′dy, (2)

and the modulation rate as a function of beam location.

where,

A (x, y) = circ

(√x2 + y2

wcore

), (3)

G (x, y) = 2P�w2

e−2(x2+y2)

w2 . (4)

Here, A is the circular aperture’s amplitude transfer function withradius of wcore [26], G is the Gaussian laser beam intensity functionof radius w and x′ is the distance between fiber core center and theincident laser beam.

Fig. 3 shows the coupled optical power to the fiber as a functionof beam distance from the fiber center. In this case the coupled ACoptical power will be the derivative of the coupled optical poweralong x-axis.

pac (x) = ∂∂xP (x) . (5)

As seen in Fig. 6, there are always two locations correspondingto the two edges of optical fiber along x axis that gives the max-imum AC signal. These points are symmetric with respect to thefiber center and the responses are negative of each other at thesetwo points.

The amplitude of the readout signal is also highly dependent on

the spot size of the laser beam on the fiber plane. Smaller spot sizehas higher modulation rate and for a certain beam traveling dis-tance, smaller spot size causes higher modulation of coupled opticalpower. Tighter focusing creates larger spot on the fiber plane, there-

136 A. Mostafazadeh et al. / Sensors and Actuators A 242 (2016) 132–139

F(s

fttblmtotiiommistZ

sreWZTttbcc

5

patstcm

acsibus

5 10 15 20 250

5

10Bea m Traveling Distance (µm)

5 10 15 20 250

500

1000Spotsize (µm)

5 10 15 20 250

5

10Modulation Rate (ΔP/Δx)

5 10 15 20 251

1.5

2PD Vdc (V)

5 10 15 20 2510

15

20

25PD Vac (mV)

ig. 7. (a) Laser spot on the 20 × 200 �m cantilever aligned 25 �m from the edge.b) Reflected beam profile on the fiber plane. Location of the 500 �m optical fiberhown with dotted line.

ore to minimize the spot size on the fiber plate, the spot size onhe cantilever must be maximized. On the other side, it is impor-ant to illuminate the tips of the cantilevers with tightly focusedeams that reside completely on the cantilever surface. Having a

arger beam spot size than the cantilever surface area has threeain disadvantages. The first one is the clipping effect, which dis-

orts the Gaussian shape of the beam and it introduces side lobesn the fiber plane. The second problem is the reflected portion ofhe beam from the silicon surface of the substrate, which introducesnterfering signals on top of the main signal. The third disadvantages the lower reflected laser power which decreases the signal leveln the photo diode. Therefore, the beam spot size on the cantileversust be small enough with respect to the cantilever dimension toinimize these unwanted effects. Using optical simulations, min-

mum achievable spot size was calculated as 8 �m in our opticaletup. Fig. 7 shows the laser illumination on the 20 �m width can-ilever and on the PD plane. The laser spot size is 8 �m (FWHM).EMAX software was used for physical wave propagation.

Pick-up fiber head location was also optimized to maximize theignal. The beam traveling distance on the fiber plane has directelation to the distance of the fiber. Since laser beam also grows lin-arly with zf , the AC signal (vac) is expected to be remain constant.

e performed physical optics simulations using MATLAB® and alsoEMAX® to calculate the modulated signal on the photodiodes.hereby, we found that best sensitivity is obtained at 10 mm fromhe cantilevers. Fig. 8 shows the simulation results. In these simula-ions, cantilever’s tip vibration amplitude assumed to be 20 nmp-p

ased on our typical experimental results in DI-water. The electri-al properties of the amplified photo detector were considered foralculating the DC and AC signals.

. Multi-channel phase measurement

For measuring the phase of millivolt level signals from the out-uts of photodiodes with a noisy background, we implemented

PC-based multi-channel lock-in amplifier using a high-speedwo channel simultaneous sampling DAQ card for signal acqui-ition. The phase of each channel was measured with respect tohe phase of actuating coil current using a digital reference syn-hronous signal. Fig. 9 shows the block diagram of the implementedulti-channel lock-in amplifier.

The digital input (Sync) synchronized the generation of sinend cosines waves numerically at the excitation frequency of theantilevers. Analog signals from the photodiodes were sampledimultaneously at the rate of 5 MSps. Because of the limitations

n real time data transferring, samples were transferred as datalocks of 100 k samples and were recombined to achieve a contin-ous signal. After multiplications with quadrature sine and cosineignals, a moving average low-pass filter demodulated the quadra-

Fibe r t o Can til ever Distan ce (mm )

Fig. 8. Fiber distance optimization results.

ture components of the input signals. Dividing the in-phase signalby the quadrature and calculating the inverse tangent of the resultgave the phase of input signals with respect to phase of the excita-tion signal. The optimum stability was achieved at moving averagelength of 10 M samples corresponding to 2 s in viscosity experi-ments. While the first demonstration was limited to two channelsdue to the DAQ card, the architecture is easily scalable to morechannels using a DAQ card with larger channel count.

In our application we focused on the relative phase value. How-ever, if the resonant frequency is needed, one can use the followingequation

ωn (ϕ) = 2ωext1

Qtan(ϕ) +√

1(Qtan(ϕ))2 + 4

. (6)

In this equation, ωext is the excitation or actuating angular fre-quency, ωn is the angular natural frequency and Q is the qualityfactor. Above equation requires a precise value of the quality fac-tor for each cantilever. Quality factor can be calculated from thefrequency sweep results.

6. Experimental results

We calculated the Q-factor and natural frequency of the can-tilevers by fitting the frequency sweep data with a Lorentzianfunction. Fig. 10 shows the frequency and phase response of twoMEMS cantilevers in DI-water. The resonant frequencies were cal-

A. Mostafazadeh et al. / Sensors and Actuators A 242 (2016) 132–139 137

Fig. 9. LabVIEW based multi-channel Lock-In Amplifier.

0.0

2.0

4.0

6.0

8.0

Out

put

(mV

rms)

-180

-135

-90

-45

0

0 10 20 30 40 50

Pha

se (°)

Frequency (kHz)

CH1 CH2

Fp

ca

vud

itbcw

m2cvtom

-13.2

-6.6

0.0

6.6

13.2

-0.2

-0.1

0

0.1

0.2

0 5 10 15 20 25 30

Fre

quen

cy D

rift

(Hz)

Ph a

se D

rift

(°)

Time (min)

CH1 CH2

Fig. 11. Phase and equivalent frequency stability.

-120

-110

-100

-90

-80

-70

-60

Pha

se (°)

CH1

DI

10%

20%

30%

40%

50%

DI DI DI DI DI

-110

-100

-90

-80

-70

-60

-50

0 10 20 30 40 50

Pha

se (°)

Time (min)

CH2

DI

10%

20%

30%

40%

50%

DI DI DI DI DI

ig. 10. Frequency and phase response of the two MEMS cantilevers measured inarallel.

ulated as 32.1808 and 34.7634 kHz, and the quality factors as 4.67nd 4.85 respectively for CH1 and CH2.

For demonstrating the performance and stability of the system, aiscosity monitoring experiment was performed on two cantileverssing different glycerol solutions. Based on the frequency sweepata we set the actuation frequency to 32 kHz.

Fig. 11 shows the stability of the system in DI-water. The stabil-ty experiment was performed in DI-water after 30 min warm-upime in room temperature (25 ± 0.5 ◦C). The results were shownoth as relative phase with respect to average phase value for eachantilever and as equivalent natural frequency drift for comparingith frequency based methods [23].

Fig. 12 shows the viscosity experiment result. In this experi-ent, we used DI-water and Glycerol solutions in DI-water of 10%,

0%, 30%, 40% and 50% weight per volume. The demonstrated vis-osity phase results are un-interrupted data and the system showed

ery robust results in terms of stability and repeatability even whenhe liquid flow occurred. Fig. 13 shows the relative phase shiftf each channel versus the Glycerol/DI-water weight percent. Theeasured phase data is insensitive to amplitude changes and other

Fig. 12. Experimental results where phase shift is due to the various Glycerol solu-tions in DI water.

disturbances in the fluid allowing measurements even during thefluid flows in the cell.

The phase shift in this experiments, is a result of changes inboth viscosity and density [23]. Viscosity of the liquids used in this

experiment, changes from 0.95 cP for DI-water up to 4.10 cP for 50%Glycerol solution (measured using Anton Paar AMVn commercialviscometer at room temperature of 23 ◦C), and the density changefrom 1.00 to 1.12 kg/m3. The relative impact of viscosity and den-

138 A. Mostafazadeh et al. / Sensors and A

0

10

20

30

40

50

60

0 10 20 30 40 50

Rel

ativ

e P

hase

Dri

ft (°)

Glycerol / DI-water weight (%)

CH1

CH2

Linear (C H1)

Linear (C H2)

saitftdiw

7

pttctttdttdfa

A

KfMig

R

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

Fig. 13. Relative phase shift versus Glycerol/DI-water weight (%).

ity on the phase shift depends of the cantilever geometry as wells other parameters [27]. The resulting phase shifts in our exper-ments are mainly due to the viscosity changes, and the impact ofhe density change is small. Nevertheless, in our experiments weocused on the phase shift as a function of the Glycerol concentra-ion in DI-water. Based on the stability experiment, the standardeviation of around 0.025◦ achieved. Using the slope of the sensitiv-

ty graphs, the minimum detectable limit of the system for Glyceroleight in DI-water was calculated as 280 ppm.

. Conclusion

This method showed robust measurement results with low laserower of 500 �W/channel and very good noise immunity betweenhe cantilevers. The demonstration was limited two cantilevers dueo data acquisition card limitations and the number of channelsan be increased. The measured phase data is insensitive to can-ilever displacement amplitude changes and other disturbances inhe fluid allowing measurements even during the fluid flows inhe cell. As an application, a viscosity experiment was done withifferent Glycerol in DI-water solutions. The sensitivity of the sys-em was 0.9◦ phase drift per Glycerol/DI-water weight percent andhe minimum detectable limit was 280 ppm. The sensor technologyemonstrated here is very significant as it is scalable to larger arrays

or simultaneous and real-time monitoring of multiple biologicalnd chemical agents during fluid flow.

cknowledgements

The authors acknowledge Dr. Onur C akmak and Dr. Necmettinılınc for his support in fabrication of the MEMS chips, Fehmi C ivitc i

or invaluable discussions about the optics design and Prof. Dr. Aliostafazadeh for his help in mathematical derivations and numer-

cal calculation methods. This research is supported by TÜBITAKrant No. 111E184 and 113S074.

eferences

[1] E.A. Wachter, T. Thundat, Micromechanical sensors for chemical and physicalmeasurements, Rev. Sci. Instrum. 66 (1995) 6, http://dx.doi.org/10.1063/1.1145484.

[2] T.P.I. Thundat, R.J.W. Oden, Microcantilever sensors, Microscale Thermophys.Eng. 1 (1997) 185–199, http://dx.doi.org/10.1080/108939597200214.

[3] J.K. Gimzewski, C. Gerber, E. Meyer, R.R. Schlittler, Observation of a chemicalreaction using a micromechanical sensor, Chem. Phys. Lett. 217 (1994)589–594, http://dx.doi.org/10.1016/0009-2614(93)E1419-H.

[[

ctuators A 242 (2016) 132–139

[4] U. Dammer, M. Hegner, D. Anselmetti, P. Wagner, M. Dreier, W. Huber, et al.,Specific antigen/antibody interactions measured by force microscopy,Biophys. J. 70 (1996) 2437–2441, http://dx.doi.org/10.1016/S0006-3495(96)79814-4.

[5] S.C. Minne, S.R. Manalis, C.F. Quate, Parallel atomic force microscopy usingcantilevers with integrated piezoresistive sensors and integratedpiezoelectric actuators, Appl. Phys. Lett. 67 (1995) 3918, http://dx.doi.org/10.1063/1.115317.

[6] S.C. Minne, G. Yaralioglu, S.R. Manalis, J.D. Adams, J. Zesch, a. Atalar, et al.,Automated parallel high-speed atomic force microscopy, Appl. Phys. Lett. 72(1998) 2340–2342, http://dx.doi.org/10.1063/1.121353.

[7] T. Sulchek, R.J. Grow, G.G. Yaralioglu, S.C. Minne, C.F. Quate, S.R. Manalis, et al.,Parallel atomic force microscopy with optical interferometric detection, Appl.Phys. Lett. 78 (2001) 1787–1789, http://dx.doi.org/10.1063/1.1352697.

[8] S.Z. Lulec, C. Sagiroglu, A. Mostafazadeh, E. Ermek, E. Timurdogan, Y. Leblebici,et al., Simultaneous self-sustained actuation and parallel readout with MEMScantilever sensor array, Micro Electro Mech. Syst. (MEMS), 2012 IEEE 25th Int.Conf. (2012) 644–647, http://dx.doi.org/10.1109/MEMSYS.2012.6170269.

[9] W.W. Koelmans, J. van Honschoten, J. de Vries, P. Vettiger, L. Abelmann, M.C.Elwenspoek, Parallel optical readout of cantilever arrays in dynamic mode,Nanotechnology 21 (2010) 395503, http://dx.doi.org/10.1088/0957-4484/21/39/395503.

10] H.P. Lang, R. Berger, C. Andreoli, J. Brugger, M. Despont, P. Vettiger, et al.,Sequential position readout from arrays of micromechanical cantileversensors, Appl. Phys. Lett. 72 (1998) 383–385, http://dx.doi.org/10.1063/1.120749.

11] M.K. Baller, H.P. Lang, J. Fritz, C. Gerber, J.K. Gimzewski, U. Drechsler, et al., Acantilever array-based artificial nose, Ultramicroscopy 82 (2000) 1–9, http://dx.doi.org/10.1016/S0304-3991(99)00123-0.

12] F.M. Battiston, J.P. Ramseyer, H.P. Lang, M.K. Baller, C. Gerber, J.K. Gimzewski,et al., A chemical sensor based on a microfabricated cantilever array withsimultaneous resonance-frequency and bending readout, Sens. Actuators BChem. 77 (2001) 122–131, http://dx.doi.org/10.1016/S0925-4005(01) 00683-9.

13] H. Lang, M. Hegner, C. Gerber, Cantilever array sensors, Mater. Today 8 (2005)30–36, http://dx.doi.org/10.1016/S1369-7021(05)00792-3.

14] N. Nugaeva, K.Y. Gfeller, N. Backmann, H.P. Lang, M. Düggelin, M. Hegner,Micromechanical cantilever array sensors for selective fungal immobilizationand fast growth detection, Biosens. Bioelectron. 21 (2005) 849–856, http://dx.doi.org/10.1016/j.bios.2005.02.004.

15] M. Toda, A.N. Itakura, S. Igarashi, K. Büscher, J.S. Gutmann, K. Graf, et al.,Surface stress, thickness, and mass of the first few layers of polyelectrolyte,Langmuir 24 (2008) 3191–3198, http://dx.doi.org/10.1021/la7028214.

16] H. Hou, X. Bai, C. Xing, N. Gu, B. Zhang, J. Tang, Aptamer-based cantilever arraysensors for oxytetracycline detection, Anal. Chem. 85 (2013) 2010–2014,http://dx.doi.org/10.1021/ac3037574.

17] K. Gruber, T. Horlacher, R. Castelli, A. Mader, P.H. Seeberger, B. a. Hermann,Cantilever array sensors detect specific carbohydrate-protein interactionswith picomolar sensitivity, ACS Nano 5 (2011) 3670–3678, http://dx.doi.org/10.1021/nn103626q.

18] S. Kelling, F. Paoloni, J. Huang, V.P. Ostanin, S.R. Elliott, Simultaneous readoutof multiple microcantilever arrays with phase-shifting interferometricmicroscopy, Rev. Sci. Instrum. 80 (2009) 093101, http://dx.doi.org/10.1063/1.3212667.

19] J. Polesel-Maris, L. Aeschimann, A. Meister, R. Ischer, E. Bernard, T. Akiyama,et al., Piezoresistive cantilever array for life sciences applications, J. Phys.Conf. Ser. 61 (2007) 955–959, http://dx.doi.org/10.1088/1742-6596/61/1/189.

20] T. Sulchek, R. Hsieh, S.C. Minne, C.F. Quate, S.R. Manalis, Interdigital Cantileveras a Biological Sensor, (2001) 562–566.

21] S.T. Koev, W.E. Bentley, R. Ghodssi, Interferometric readout of multiplecantilever sensors in liquid samples, Sens. Actuators B Chem. 146 (2010)245–252, http://dx.doi.org/10.1016/j.snb.2010.02.038.

22] O. Cakmak, N. Kilinc, E. Ermek, A. Mostafazadeh, C. Elbuken, G.G. Yaralioglu,et al., LoC sensor array platform for real-time coagulation measurements,Micro Electro Mech. Syst. (MEMS), 2014 IEEE 27th Int. Conf. (2014) 330–333,http://dx.doi.org/10.1109/MEMSYS.2014.6765643.

23] O. Cakmak, C. Elbuken, E. Ermek, A. Mostafazadeh, I. Baris, B. Erdem Alaca,et al., Microcantilever based disposable viscosity sensor for serum and bloodplasma measurements, Methods 63 (2013) 225–232, http://dx.doi.org/10.1016/j.ymeth.2013.07.009.

24] O. Cakmak, E. Ermek, N. Kilinc, S. Bulut, I. Baris, I.H. Kavakli, et al., A cartridgebased sensor array platform for multiple coagulation measurements fromplasma, Lab Chip 15 (2015) 113–120, http://dx.doi.org/10.1039/C4LC00809J.

25] J.M.S. Sakamoto, C. Kitano, G.M. Pacheco, B.R. Tittmann, High sensitivity fiberoptic angular displacement sensor and its application for detection ofultrasound, Appl. Opt. 51 (2012) 4841–4851, http://dx.doi.org/10.1364/AO.51.004841.

26] J.W. Goodman, Introduction to Fourier Optics, 2nd ed., McGraw-Hill, 1996.27] O. Cakmak, E. Ermek, N. Kilinc, G.G. Yaralioglu, H. Urey, Precision density and

viscosity measurement using two cantilevers with different widths, Sens.Actuators A Phys. 232 (2015) 141–147, http://dx.doi.org/10.1016/j.sna.2015.

05.024.

and A

B

tLtmHsLt

of SPIE, IEEE, and OSA. He received an advanced grant from the European ResearchCouncil (ERC-AdG) in 2013, Outstanding Faculty Award from Koc University in 2013,TÜB˙ ITAK-Encouragement Award in 2009, Outstanding Young Scientist Award fromthe Turkish Academy of Sciences (TÜBA) in 2007, and Werner Von Siemens Excel-lence Award in 2006.

A. Mostafazadeh et al. / Sensors

iographies

Aref Mostafazadeh was born in 1978 in Tabriz, Iran. Hereceived the B.S. degree in electrical and electronics engi-neering from Tabriz Azad University (TAU), Tabriz, Iran in2001, M.A. degree in photography from Fine Arts Depart-ment of Marmara University, Istanbul, Turkey in 2008, andthe M.S. degree in electrical and computer engineeringfrom Koc University (KU), Istanbul, Turkey in 2010. He iscurrently pursuing Ph.D. degree in electrical and electron-ics engineering at Koc University. After graduation fromTAU, he worked in a private research firm in Tabriz. In2006, He moved to Istanbul to enroll in the MA programin Photography of Marmara University where he workedon the imaging optics and quality factors of imaging sys-

ems. From 2008 to 2010 he was research and teaching assistant in KU Laser Researchab. and from 2010 to 2015 in KU Optical Microsystems Lab. In 2015, he foundedhe Electronics Research Lab. in GAMAK Motors Company, the largest electric motor

anufacturer in Turkey and started as chief design engineer and research scientist.is research interests also include imaging systems, optical MEMS, resonant MEMS

ensors and actuators, development of resonance tracking systems, Phase-Lockedoops, high precision measurement systems, power electronics and advanced con-rol methods.

Goksen G. Yaralioglu received his B.Sc., M.Sc., and Ph.D.degrees from Bilkent University, Ankara, Turkey, in 1992,1994, and 1999, respectively, all in electrical engineering.He is now working as an Associate Professor in the Elec-trical and Electronics Engineering Department, Ozyegin

University, Istanbul, Turkey. His current research interestsinclude design, modeling and applications of microma-chined ultrasonic transducers, atomic force microscopy atultrasonic frequencies, microfluidic channels, MEMS sen-sors and inertial sensors.

ctuators A 242 (2016) 132–139 139

Hakan Urey received the B.Sc. degree from Middle EastTechnical University, Ankara in 1992, and M.Sc. and PhD.degrees from Georgia Institute of Technology in 1996 andin 1997, all in Electrical Engineering. After completing hisPh.D., he joined Microvision Inc.-Seattle as Research Engi-neer and he played a key role in the development of theRetinal Scanning Display technology. He was the Princi-pal System Engineer when he left Microvision to join thefaculty of engineering at Koc University in 2001. He waspromoted to Associate Professor in 2007 and Professor in2010. He published about 50 journals and 100 interna-tional conference papers, 7 edited books, 4 book chapters,and has more than 25 issued and several pending patents.

His research interests are in the area of optical MEMS, micro-optics and optical sys-tem design, 2D/3D display and imaging systems, and biosensors. He is a member