A

La

b

a

ARRAA

KOCP

1

ftMmssslSpia

meooTimh

RT

0d

Sensors and Actuators B 155 (2011) 430–435

Contents lists available at ScienceDirect

Sensors and Actuators B: Chemical

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

CMOS optical detection system for point-of-use luminescent oxygen sensing

i Shena, Michael Rattermana, David Klotzkinb, Ian Papautskya,∗

School of Electronics and Computing Systems, University of Cincinnati, United StatesDepartment of Electrical Engineering, Binghamton University, United States

r t i c l e i n f o

rticle history:eceived 27 July 2010eceived in revised form 1 December 2010ccepted 3 January 2011

a b s t r a c t

This work describes a stable and sensitive optical oxygen sensor based on a consumer CMOS imagesensor array and polarization signal isolation. The consumer CMOS image sensor is inherently colordiscriminating, while the polarization is a wavelength-independent scheme for filtering excitation light.

vailable online 10 January 2011

eywords:xygen sensorMOS image sensorolarizerization signal isolation

Combination of these two components generates a compact, multi-color detection system applicableto luminescence-based sensing. The optical system is applied to point-of-use oxygen sensing based onquenching of platinum octaethylporphine (PtOEP) luminophore. Sensitivity of the demonstrated portableoxygen sensor is comparable to that of benchtop spectroscopy-based systems. Taking advantage of thespatial resolution of the CMOS image sensor, an oxygen insensitive reference can be integrated to improvethe reliability when powered by a battery. The low-cost and high sensitivity of the demonstrated optical

prom

sensing approach make it

. Introduction

Current research in point-of-care (POC) biotechnology hasocused on increased miniaturization and integration for higherhroughput, decreased reagent cost, and increased sensitivity.

ost of the mainstream biological analyses are based on opticalethods (i.e., luminescence, absorption). These optical detection

ystems typically require sophisticated instrumentation, such aspectrophotometers or photomultiplier tubes, and trained per-onnel. The need for portable and low-cost optical sensing hased to intense focus on miniaturization of sensor components.emiconductor and organic LEDs are gaining wide acceptance asortable light sources in such systems [1], and have been used in

mmunosensors [2,3] and chemical sensors [4,5]. Signal detectionnd isolation, however, remain active areas of research.

Filtering of excitation light from signal light is critical inicroscale fluorescent systems. A common system design places an

xcitation LED orthogonal to the interrogation region and detectorf the microfluidic system. This reduces system noise due to leakagef unabsorbed excitation light; however, such systems are bulky.

he systems using a stacked arrangement, where the light sources in-line with the interrogation region and detector, are much

ore compact and preferable in portable systems. In such a system,owever, the emission signal is drowned out by the leaking excita-

∗ Corresponding author at: School of Electronics and Computing Systems, 814hodes Hall, ML030, University of Cincinnati, Cincinnati, OH 45221, United States.el.: +1 513 556 2347; fax: +1 513 556 7326.

E-mail address: ian.papautsky@uc.edu (I. Papautsky).

925-4005/$ – see front matter © 2011 Elsevier B.V. All rights reserved.oi:10.1016/j.snb.2011.01.001

ising for point-of-use applications.© 2011 Elsevier B.V. All rights reserved.

tion light. A variety of techniques have been reported for filteringin these stacked micro fluorescence systems, including interfer-ence filters [6], color filters [7], and dye-based filters [8]. Theseapproaches however are wavelength-specific, making the systemapplication-specific and far less versatile. In this work, signal iso-lation is achieved by cross-polarizing the excitation and emissionlight; an approach we introduced recently [9,10]. Two perpendic-ularly placed polarizers, with extinction ratio of 28 dB, are able toseparate the emission light from the strong excitation light (Fig. 1a).Compared with using filters, this approach is inexpensive, easy tointegrate, and is wavelength independent.

In addition to filtering, a sensitive and reliable detector is criti-cal to developing a portable sensor. Conventional photomultipliertubes have been used for benchtop microfluidic systems [11,12].Silicon photodiodes [13,14], or avalanche photodiodes [15] can beused in POC or point-of-use systems. Integration of these externalcomponents however presents alignment and assembly challenges.In the past, we [9,10] and others [16,17] have used organic photo-diodes (OPDs) as low-cost integrated detection alternatives, beingflexible and miniaturizable [18]. However, OPDs suffer from lowpower conversion efficiency [19] and short lifetime [20,21].

Recently, CMOS image arrays have attracted interest as detec-tors for POC/point-of-use systems because of the high-speedreadout and low power consumption. El Gamal and Eltoukhy [22]recently reviewed developments in CMOS image sensors and sug-

gested applications to lab-on-a-chip as one of the future directions.CMOS image sensors, commonly used today in digital cameras, arebuilt using the same process as ICs and consist of an array of pho-todetectors covered by a mosaic of microscale band-pass (Bayer)filters (Fig. 1b). Raw data taken by a CMOS image sensor are grids of

L. Shen et al. / Sensors and Actuat

Fig. 1. (a) Illustration of polarization filtering which permits isolation of opticalsignal (Rhodamine B emission centered at 625 nm) from high background signaldcdR

ttdwvsd

asOrgehgspopas

2

OoTopcscg

ct4ii

ue to leaking excitation light (530 nm). (b) Schematic of a photodetector arrayovered by Byer filters in a CMOS detector. (c) Arrangement of the light source (LED),etector (CMOS), polarizers, oxygen sensitive PtOEP film and oxygen insensitive filmhodamine B in the portable oxygen sensor.

hree primary colors: red, green and blue (RGB). Each pixel containshe intensity information of one color. The intensity of each color isescribed within a given range from a minimum and a maximumhich are dependent on the resolution of the analog to digital con-

erter. This approach of using CMOS detector as detector has beenuccessfully applied for bioassays [23] and fluorescence-based celletection [24–27].

In this work we describe an optical sensing system based onlow-cost consumer CMOS image sensor and explore the fea-

ibility of using it as a detector in a portable oxygen sensor.xygen detection is important in clinical diagnosis, biological

esearch and environmental monitoring. The conventional oxy-en sensors based on Clark electrodes consume oxygen and areasily poisoned by various organic compounds [28]. On the otherand, optical oxygen sensors offer quick response time, no oxy-en consumption, and high sensitivity [29]. The developed oxygenensor (illustrated in Fig. 1c) is based on oxygen sensitive metallo-orphyrin luminophore platinum octaethylporphine (PtOEP) andxygen insensitive Rhodamine B (RhB). The CMOS image sensorermits selective detection of the red emission from PtOEP and RhBs well as simultaneous monitoring of their emissions. The oxygenensor exhibits high sensitivity and good stability.

. Experimental methods

A CMOS image sensor with a USB interface (OmniVision,V9810) was used in this work. The sensor has a packaged footprintf 8.195 mm × 7.535 mm, with 6.160 mm × 4.606 mm imaging area.he active array size was 3488 × 2616 pixels (9 megapixel res-lution), each approximately1.75 �m × 1.75 �m. The sensor waslaced at the bottom of the device stack as the detector and wasontrolled by the supplied software. Captured raw images wereplit to three grayscale images (each representing a separate RGBolor) and quantitatively analyzed using ImageJ (ver. 1.43u). Eachrayscale image yielded an average pixel intensity value.

In the spectral response test, the white LED was biased at a

onstant potential of 3.0 V (∼16 mA forward current, based onhe manufacturer provided data sheet). Illumination light in the30–680 nm range was selected by the monochromatorin 10 nm

ncrements. Intensities in each of the RGB channels of the CMOSmage sensor were recorded. Spectral response of each channel

ors B 155 (2011) 430–435 431

was obtained when the recorded intensities were normalized tothe white LED emission spectrum.

The sensor is based on the oxygen-sensitive PtOEP encapsulatedin ethyl cellulose (EC). A green LED (Bivar R20GRN-F-0160) with acentral emission wavelength of �ex = 520 nm was used as the excita-tion light source (Fig. 1c). The excitation light was linearly polarizedby passing through polarizer 1 (NT45667, Edmund Optics). The redemission of the PtOEP film (�em = 644 nm) passed through polarizer2. The polarized excitation light not absorbed by the PtOEP film wasblocked by polarizer 2. When cross-polarized, the extinction ratioof excitation light to leakage light is 28 dB [30]. The oxygen sen-sitive film was prepared by completely dissolving 10 mg of PtOEP(PTO534, Frontier Scientific) in 10 mL tetrahydrofuran. The mixturewas added to a polymer solution prepared by dissolving 3 g EC (48%ethoxyl) in the mixture of toluene (24 mL) and ethanol (6 mL).

For oxygen measurements, all automatic CMOS image sensorfunctions, such as exposure time control, gamma correction, andauto white balance (AWB), were set to manual. We used two flowmeters to control the ratio of oxygen and nitrogen supplied fromtanks. A mixing chamber was used to pre-mix the gases prior tointroduction to the gas testing chamber. Our sensor was located inthe center of the testing chamber; the chamber oxygen concentra-tion was indicated by a commercial oxygen sensor (Nuvair, Pro O2remote analyzer). During measurements, oxygen concentration inthe chamber was changed by adjusting input gas flow rates. Signalsand background were obtained by taking images of the PtOEP filmand a bare glass wafer. The values in the red channel were usedfor oxygen measurement. By subtracting the background from thesignal, the emission intensity of PtOEP at different oxygen concen-tration was obtained. To test stability of the sensor performanceunder an unstable light source, the forward voltage of the LEDwas manually varied by ±0.02 V during oxygen measurements. Theemission intensities of the RhB reference were standardized and theemission intensities of the PtOEP were calibrated accordingly.

3. Results and discussion

We first tested the spectral response of the CMOS chip RGBchannels to examine the color discriminating capability. We used amonochrometer to select specific wavelengths of illumination lightand recorded the spectral response in the 430–680 nm range in10 nm step increments. As illustrated in Fig. 2a, each channel ofthe CMOS detector exhibits a strong response to the correspond-ing wavelength range, as expected. The blue pixels exhibit a peakresponse at 460 ± 30 nm, while the green and red pixels peak at540 ± 40 nm and 600 ± 50 nm respectively. This test confirms thatthe CMOS detector we used is inherently discriminating to the RGBcolors and demonstrates the spectral response range of each colorpixel type. These results are helpful in the development of sensorsthat take advantage of the CMOS detector’s color discriminatingcapability.

While Bayer filters permit color discrimination in the CMOSdetector, response of each color pixel type does overlap. In a con-sumer camera this is desirable as the overlap permits capture ofcolors that fall between the standard RGB bins, such as yellowat 580 nm. In sensor applications, however, when individual colorchannels in the CMOS detector are used to distinguish betweenthe excitation and emission of analyte, the spectral overlap leadsto noise. For example, a sensor based on detection of fluores-cein isothiocyanate (FITC) – a fluorescent dye commonly used in

bioapplications – will detect emission signal in the green chan-nel (521 nm) and unabsorbed excitation light (488 nm) in the bluechannel. The overlap between the two channels (Fig. 2a) will leadto high background noise as a fraction of the excitation light willbe picked-up by the green channel.

432 L. Shen et al. / Sensors and Actuators B 155 (2011) 430–435

Fig. 2. (a) Responsivity of the CMOS detector in each of its color channels. Inset images takand 620 nm. (b) Response of the CMOS detector to 620 nm light as compared to a currendetector distorts the linear response.

Fe

wCofwlhop1

The gamma correction is given by:

Fa

ig. 3. The PtOEP absorption and emission spectra overlaid with the green LEDmission and the red channel response of the CMOS detector.

Optical signal isolation based on cross-polarization is a simple,avelength-independent approach that ideally complements theMOS detector. Cross-polarization offers a high extinction ratio,n the order of 28 dB, and is able to separate weak emission lightrom a strong background excitation in a stacked configurationhere the excitation light source and detector are positioned in-

ine (Fig. 1c). The approach is wavelength independent, as the

igh-efficiency linear polarizer films exhibit high-transmittancef ∼40% at wavelengths above 400 nm. We recently used cross-olarization to demonstrate quantitative measurements down to0 nM concentration of Rhodemine 6G fluorophore in a microflu-

ig. 4. (a) Emission intensity of PtOEP as a function of oxygen concentration. (b) Stern–Vnd high oxygen concentrations.

en by the CMOS detector illustrate response at center wavelengths 460 nm, 540 nmt measured by a silicon photodiode. The gamma correction function of the CMOS

idic lab-on-a-chip, a hundred-fold improvement from previouslypublished results [16,17]. Herein, the use of cross-polarization willpermit us to sensitively detect small changes in emission of PtOEPdue to quenching by oxygen in presence of the leaking excitationlight. However, before proceeding with discussion of the oxygensensor, we first describe further CMOS detector characterizationfocused on integrated detector functions capable of influencingsensor performance.

While CCDs offer better low-light performance, a CMOS imagesensor was selected for this work based on the overall sensor needs.The two image sensor types were recently compared by Gamaland Eltoukhy [22], who suggested that CMOS sensors are bettersuited for low cost portable sensors. Most important to our appli-cation, since most camera control functions of a CCD are integratedoff-chip, the CCD-based detection systems tend to increase in sizeand cost very rapidly. A CMOS array integrates sensing with A/Dconverter and gain amplifier at the pixel level, making it highlycustomizable [22]. Although these features make CMOS sensorspopular with the consumer market, their impact must be fullyunderstood for application to the light intensity-modulated opti-cal sensing. Thus, in the next set of experiments we manuallycontrolled these functions (by writing the registers through thestandard serial camera control bus interface) to investigate capa-bilities of a consumer CMOS detector.

In photography, gamma correction is used to compensate non-linear decoding process of displays by encoding the signal inversely.

Iout = Iin

(Iin

Imax

)�

(1)

olmer plot of the oxygen sensor. Inset images illustrate emission of PtOEP at low

Actuat

wtlOvtcpt6dbdtmtsist

osT6iarrtrima

syPvtitwa

Fa(

L. Shen et al. / Sensors and

here Iout is the displayed intensity, Iin is the detected intensity, andhe exponent � is the gamma correction index. The Imax is the upperimit which depends on the color depth of the image sensor. TheV9810 is a 24-bit color image sensor so its Imax is 255. The defaultalue of � is 1/2.2 which makes the response of the CMOS detec-or non-linear to the light intensity [31]. Turning off the gammaorrection by setting � = 1 leads to the CMOS detector exhibitingerfect linearity satisfying a fundamental requirement as a detec-or. Fig. 2b illustrates the red channel response to monochromatic20 nm light generated by biasing LED from 2.6 V to 2.88 V to yieldifferent intensities directly measured by a silicon photodiode. Thelue and green channels were tested in the same way and theyemonstrated good linearity as well. The effect of gamma correc-ion is clearly evident. As these results indicate, gamma correction

ust be turned off for the detector to yield a linear response overhe entire sensor range. However, gamma could provide additionalignal amplification for low-light conditions if sensor range is lim-ted. The auto white balance (AWB) and auto exposure were testedimilarly. They were also manually controlled to keep the consis-ency during the following experiments.

Having examined the CMOS detector features, in the next setf experiments we explored application to luminescent oxygenensing using PtOEP, which is known to be sensitive to oxygen.he absorption and emission of PtOEP are centered at 535 nm and40 nm when encapsulated in EC [32]. As Fig. 3 illustrates, there

s a significant overlap between the green LED emission (520 nm)nd the absorption of PtOEP. The red channel, which has strongesponse to red light, was used to detect the PtOEP emission. Theesponse of red channel cuts off at around 580 nm, resulting in fur-her filtering of the excitation light. Moreover, the green channelesponds to the LED emission well but does not respond to thentensity variation of the red PtOEP emission. Thus, it is possible to

onitor the intensity of the light source using the green channelnd calibrate the measured result accordingly.

This oxygen sensor based on CMOS detector exhibited highensitivity as well as good linearity in the Stern–Volmer anal-sis. Fig. 4a illustrates the steady state emission intensities oftOEP sensor subject to different oxygen levels. The red pixelalues decrease significantly with the increasing oxygen concen-ration as the quenching effect becomes stronger. The intensities

n green and blue channels are much more stable because thesewo channels are inert to the red emission of PtOEP which variesith the oxygen concentration. The standard approach to char-

cterizing luminescence quenching based oxygen sensors is using

ig. 5. Response of the oxygen sensor exposed to an alternating streams of oxygennd nitrogen gases. The arrows indicate the time when the gases were switchedfilled arrows indicate nitrogen; hollow arrows indicate oxygen).

ors B 155 (2011) 430–435 433

Stern–Volmer analysis in which,

I0I

= 1 + KSV[O2] (2)

where I0 and I are the emission intensities in the absence and pres-ence of oxygen at concentration of [O2], respectively,and KSV isStern–Volmer constant. Fig. 4b illustrates a typical Stern–Volmerplot of the developed oxygen sensor. The high correlation coeffi-cient (0.9982) indicates a linear relationship between the oxygenconcentration and the ratio of emission intensities in the absence(I0) and presence (I1 0 0) of oxygen. The ratio I0/I1 0 0 as sensitivity ofthe oxygen sensor is estimated to be ∼41. This result is comparableto the ∼50 values reported by others using an external spectropho-tometer [33,34] and is much higher than that of an integratedlab-on-a-chip sensor (∼1.4) reported recently [35]. The insets inFig. 4b illustrate images taken at low and high oxygen concentra-tions. As emission of PtOEP is strongly quenched by oxygen, littleemission light is observed and the corresponding image was dark(right inset). In low oxygen conditions, however, the emission ofPtOEP film increases and the image becomes red.

The oxygen sensor exhibited reversible quenching and goodrepeatability when exposed to an alternating atmosphere of oxy-gen and nitrogen. A typical dynamic response of the sensor isillustrated in Fig. 5. The response time of an optical oxygen sen-sor is considered as the time of 90% total intensity change whenswitching between oxygen and nitrogen conditions. The responsetime of the oxygen sensor in this work was ∼4 s upon switch-ing from nitrogen to oxygen and the reverse switch time, fromoxygen to nitrogen, was ∼80 s. The asymmetric response curvewith a slow O2 to N2 response compared to N2 to O2 responseis typical of the luminescent thin film oxygen sensors [36]. Theresponse is comparable to the reported sensor using polystyreneas luminophore-encapsulating matrix (18 s/60 s) [33]. We believethe response time is limited by the mass-transport in our gas testchamber (∼3 L volume), which can take 22–60 s to fill dependingon input gas flow rate (up to 8 l per min). Reducing the test cham-ber volume will improve sensor response time in the gas cyclingtest. A number of methods may also be used to improve the filmresponse time, such as using Pt porphyrins bearing side groupsdesigned to increase solubility at higher concentration in a thin-ner layer [30] or use a material with high diffusivity to oxygen

[34].

The oxygen sensor showed excellent reliability after 30 h expo-sure to a green LED in ambient air (21% oxygen condition). Fig. 6illustrates intensities of the RGB channels during the test. At 21%O2, emission of PtOEP was not strong but very stable during the

Fig. 6. Stability of the RGB channels in the CMOS detector of the oxygen sensor inambient air.

434 L. Shen et al. / Sensors and Actuators B 155 (2011) 430–435

Feda

1dLwtoooc

ssamocsigR

tesooifisoR

Fig. 8. PtOEP and RhB emission at different oxygen concentrations under fluc-

ig. 7. (a) Emission of PtOEP is quenched by oxygen while emission of the refer-nce Rhodamine B remains unchanged. Light emission in the red channel of CMOSetector at (b) low oxygen and (c) high oxygen conditions. Dashed lines indicatepproximate boundaries of each film.

800-min test. The signals fluctuated slightly initially but stabilizeduring the first ten minutes. The fluctuation may be caused by theED or the power supply. This result suggests that a ten-minutearm-up time is necessary for the system to reach a stable condi-

ion for accurate measurement. After the warm-up, the mean valuef the intensities in red channels is 16.5 with a standard deviationf 0.33, resulting in a coefficient of variation of only 0.02. Basedn these results, the oxygen sensor is reliable and can be used forontinuous oxygen sensing applications.

The oxygen sensor exhibited high sensitivity when the lightource was powered by the stable external power supply. However,uch a power supply is not always available outside the laboratorynd batteries are used in most portable devices. While batteries areuch smaller in size and lighter in weight, their output decreases

ver time with use or even due to storage. The lower driving voltageauses lower light source emission; as a result, the PtOEP emis-ion is a function of the oxygen concentration and the light sourcentensity. Thus, an oxygen insensitive reference should be inte-rated to monitor the light source intensity. In the next experiment,hB was integrated as a reference.

The RhB emission does not change with oxygen concen-ration while the PtOEP emission is sensitive to oxygen. Tonhance RhB stability it was sealed in a glass capillary. Emis-ion of both luminophores was recorded by taking advantagef the CMOS sensor’s spatial capabilities. The intensity profilesf two images taken at high and low oxygen conditions arellustrated in Fig. 7. The PtOEP emission decreases significantly

rom 190 to 20 a.u. as expected when the oxygen concentrationncreases. On the other hand, the emission of RhB remains con-tant. The sensor images clearly illustrate the quenching effectf oxygen on PtOEP. The black tape between the PtOEP andhB blocks the bleeding light from PtOEP at low oxygen con-

tuating LED bias. (a) Uncompensated and (b) compensated. Triangles representRhB signal, while circles represent PtOEP intensity. Insets illustrate correspondingStern–Volmer plots.

centrations and thus enhances the stability of the RhB referencesignal.

The integrated reference makes the oxygen sensor stable andaccurate because any fluctuation of the power supply can beobserved from the emission intensity of RhB and then negated.To demonstrate this, the LED bias was intentionally varied dur-ing oxygen measurement by ±0.02 V, which corresponds to lessthan 1% change. The oxygen sensor was first characterized with LEDbiased at 3.0 V using a Keithley 6487 voltage source. Fig. 8a illus-trates the resulting RhB and PtOEP emission intensities that deviatefrom the calibration curve (dotted lines). After the RhB intensi-ties were used to compensate fluctuations in the PtOEP intensities,the sensor exhibited good linearity in response (Fig. 8b). The insetimages in Fig. 8 illustrate the Stern–Volmer plots based on the cal-ibrated PtOEP intensities. The coefficient of determination (R2) of0.88 demonstrates the good linearity of the sensor. Without theRhB reference, the R2 decreases to 0.54 under less than 1% of powersupply output fluctuation.

4. Conclusions

In summary, a portable optical oxygen sensor using a low-costCMOS image sensor as a detector and a cross-polarization schemeas signal isolation was demonstrated. The sensor exhibited sensi-tivities comparable to that of macroscale benchtop sensor systems.

Actuat

Tttaoapsa

A

E

R

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[[

[

[

[

[

L. Shen et al. / Sensors and

he integrated oxygen insensitive reference was monitored simul-aneously by the CMOS sensor, permitting compensation whenhe light source is not stable. We are currently in process ofpplying the system to multi-analyte sensing by taking advantagef the detector’s inherent color discriminating capabilities. Over-ll, the described approach of using the wavelength-independentolarizer scheme and color-recognizable CMOS image sen-or clearly demonstrate suitability for reliable, on-site sensorpplications.

cknowledgements

This work was supported by the NSF awards ECCS-0725812 andCCS-0930305.

eferences

[1] F.B. Myers, L.P. Lee, Innovations in optical microfluidic technologies for point-of-care diagnostics, Lab Chip 8 (2008) 2015–2031.

[2] B.C. Heinze, J.-Y. Song, C.-H. Lee, A. Najam, J.-Y. Yoon, Microfluidic immunosen-sor for rapid and sensitive detection of bovine viral diarrhea virus, Sens.Actuators B 138 (2009) 491–496.

[3] D.B. Papkovsky, T.C. O’Riordan, G.G. Guilbault, An immunosensor based on theglucose oxidase label and optical oxygen detection, Anal. Chem. 71 (1999)1568–1573.

[4] R. Irawan, S.C. Tjin, X. Fang, C.Y. Fu, Integration of optical fiber light guide,fluorescence detection system, and multichannel disposable microfluidic chip,Biomed. Microdev. 9 (2007) 413–419.

[5] M. Hajj-Hassan, T. Gonzalez, E. Ghafar-Zadeh, H. Djeghelian, V. Cho-davarapu, M. Andrews, D. Therriault, Direct-dispense polymericwaveguides platform for optical chemical sensors, Sensors 8 (2008)7636–7648.

[6] M.A. Burns, B.N. Johnson, S.N. Brahmasandra, K. Handique, J.R. Webster, M.Krishnan, T.S. Sammarco, P.M. Man, D. Jones, D. Heldsinger, C.H. Mastrangelo,D.T. Burke, An integrated nanoliter DNA analysis device, Science 282 (1998)484–487.

[7] S.J. Hart, R.D. Jiji, A simple, low-cost, remote fiber-optic micro volume fluores-cence flowcell for capillary flow-injection analysis, Anal. Bioanal. Chem. 374(2002) 385–389.

[8] O. Hofmann, X. Wang, A. Cornwell, S. Beecher, A. Raja, D.D.C. Bradley, A.J.DeMello, J.C. DeMello, Monolithically integrated dye-doped PDMS long-passfilters for disposable on-chip fluorescence detection, Lab Chip 6 (2006)981–987.

[9] A. Pais, A. Banerjee, D. Klotzkin, I. Papautsky, High-sensitivity, disposable lab-on-a-chip with thin-film organic electronics for fluorescence detection, LabChip 8 (2008) 794–800.

10] A Banerjee, A. Pais, I. Papautsky, D. Klotzkin, A polarization isolation methodfor high-sensitivity, low-cost on-chip fluorescence detection for microfluidiclab-on-a-chip, IEEE Sens. J. 8 (2008) 621–627.

11] P. Hartmann, M.J.P. Leiner, M.E. Lippitsch, Luminescence quenching behaviorof an oxygen sensor based on a Ru(II) complex dissolved in polystyrene, Anal.Chem. 67 (1995) 88–93.

12] H. Chuang, M.A. Arnold, Radioluminescent light source for optical oxygen sen-sors, Anal. Chem. 69 (1997) 1899–1903.

13] C. McDonagh, C. Kolle, A.K. McEvoy, D.L. Dowling, A.A. Cafolla, S.J. Cullen, B.D.MacCraith, Phase fluorometric dissolved oxygen sensor, Sens. Actuators B 74(2001) 124–130.

14] A.N. Watkins, B.R. Wenner, J.D. Jordan, W. Xu, J.N. Demas, F.V. Bright, Portable,low-cost, solid-state luminescence-based O2 sensor, Appl. Spectrosc. 52 (1998)750–754.

15] Y. Kostov, P. Harms, L. Randers-Eichhorn, G. Rao, Low-cost microbioreactor forhigh-throughput bioprocessing, Biotechnol. Bioeng. 72 (2001) 346–352.

16] O. Hofmann, P. Miller, P. Sullivan, T.S. Jones, J.C. Demello, D.D.C. Bradley, A.J.Demello, Thin-film organic photodiodes as integrated detectors for microscalechemiluminescence assays, Sens. Actuators B 106 (2005) 878–884.

17] X. Wang, O. Hofmann, R. Das, E.M. Barrett, A.J. DeMello, J.C. DeMello, D.D.C.Bradley, Integrated thin-film polymer/fullerene photodetectors for on-chipmicrofluidic chemiluminescence detection, Lab Chip 7 (2007) 58–63.

ors B 155 (2011) 430–435 435

18] T.N. Ng, W.S. Wong, M.L. Chabinyc, S. Sambandan, R.A. Street, Flexible imagesensor array with bulk heterojunction organic photodiode, Appl. Phys. Lett. 92(2008) 213303.

19] P. Peumans, S.R. Forrest, Very-high-efficiency double-heterostructure copperphthalocyanine/C60 photovoltaic cells, Appl. Phys. Lett. 79 (2001) 126–128.

20] P. Vivo, J. Jukola, M. Ojala, V. Chukharev, H. Lemmetyinen, Influence of Alq3/Aucathode on stability and efficiency of a layered organic solar cell in air, Sol.Energy Mater. Sol. Cells 92 (2008) 1416–1420.

21] Q.L. Song, M.L. Wang, E.G. Obbard, X.Y. Sun, X.M. Ding, X.Y. Hou, C.M. Li, Degra-dation of small-molecule organic solar cells, Appl. Phys. Lett. 89 (2006) 251118.

22] A. El Gamal, H. Eltoukhy, CMOS image sensors, IEEE Circuits Devices Mag. 21(2005) 6–20.

23] D.M. Vykoukal, G.P. Stone, P.R.C. Gascoyne, E.U. Alt, J. Vykoukal, Quantita-tive detection of bioassays with a low-cost image-sensor array for integratedmicrosystems, Angew. Chem. Int. Ed. 48 (2009) 7649–7654.

24] Y. Hosseini, K.V.I.S. Kaler, Integrated CMOS optical sensor for cell detection andanalysis, Sens. Actuators A 157 (2010) 1–8.

25] X. Heng, D. Erickson, L.R. Baugh, Z. Yaqoob, P.W. Sternberg, D. Psaltis, C. Yang,Optofluidic microscopy—a method for implementing a high resolution opticalmicroscope on a chip, Lab Chip 6 (2006) 1274–1276.

26] G. Medoro, N. Manaresi, A. Leonardi, L. Altomare, M. Tartagni, R. Guerrieri, A lab-on-a-chip for cell detection and manipulation, IEEE Sens. J. 3 (2003) 317–325.

27] J.M. Song, M. Culha, P.M. Kasili, G.D. Griffin, T. Vo-Dinh, A compact CMOS biochipimmunosensor towards the detection of a single bacteria, Biosens. Bioelectron.20 (2005) 2203–2209.

28] J.R. Bacon, Determination of oxygen concentrations by luminescence quench-ing of a polymer-immobilized transition-metal complex, Anal. Chem. 59 (1987)2780–2785.

29] P. Douglas, K. Eaton, Response characteristics of thin film oxygen sensors, Pt andPd octaethylporphyrins in polymer films, Sens. Actuators B 82 (2002) 200–208.

30] D.B. Papkovsky, G.V. Ponomarev, W. Trettnak, P. O’Leary, Phosphorescent com-plexes of porphyrin ketones: optical properties and application to oxygensensing, Anal. Chem. 67 (1995) 4112–4117.

31] G. Sharma, Digital Color Imaging Handbook, 2003.32] S.R. Ricketts, P. Douglas, A simple colorimetric luminescent oxygen sensor

using a green LED with Pt octaethylporphyrin in ethyl cellulose as the oxygen-responsive element, Sens. Actuators B 135 (2008) 46–51.

33] S.-K. Lee, I. Okura, Porphyrin-doped sol–gel glass as a probe for oxygen sensing,Anal. Chim. Acta 342 (1997) 181–188.

34] T.-S. Yeh, C.-S. Chu, Y.-L. Lo, Highly sensitive optical fiber oxygen sensor usingPt(II) complex embedded in sol–gel matrices, Sens. Actuators B 119 (2006)701–707.

35] E. Kraker, A. Haase, B. Lamprecht, G. Jakopic, C. Konrad, S. Köstler, Integratedorganic electronic based optochemical sensors using polarization filters, Appl.Phys. Lett. 92 (2008) 033302.

36] R. Herne, J. Brocas, E. Vander Donckt, The response time of optical sensors basedon luminescence quenching, Anal. Chim. Acta 364 (1998) 131–141.

Biographies

Li Shen received the BS and MS degrees in electrical engineering from SoutheastUniversity, Nanjing, China. He is currently pursuing the PhD degree in electricalengineering at the University of Cincinnati. His research interests include opticalgas sensors, lab-on-a-chip, and biosensors for point-of-care applications.

Michael Ratterman is currently pursuing a PhD at the University of Cincinnati, afterreceiving his BS in electrical engineering in 2009. During that time, his senior projectdevelopment team won the “Best of EE” award at the Cincinnati Tech Expo. Hisresearch interests include microfluidics, optical-gas sensing, embedded firmwaredesign, and image processing.

David Klotzkin is an Associate Professor in the Electrical and Computer EngineeringDepartment at Binghamton University. He received a PhD from the University ofMichigan in 1998. His current research interests include optical sensing and opticalcommunications.

Ian Papautsky is an Associate Professor in the School of Electronics and Comput-ing Systems at the University of Cincinnati. He received a PhD in bioengineeringfrom the University of Utah, Salt Lake City, UT. His research interests focuson developing disposable polymer microfluidic lab-on-a-chip (LOC) systems forpoint-of-care (POC) applications and bio/chemical sensors for in situ sensing andanalysis.