Articleanalog.snu.ac.kr/publications/JBN_2013.pdf · Delivered by Publishing Technology to: Suhwan Kim IP: 147.46.61.45 On: Sun, 05 May 2013 01:59:52 Copyright American Scientific

Delivered by Publishing Technology to: Suhwan KimIP: 147.46.61.45 On: Sun, 05 May 2013 01:59:52

Copyright American Scientific Publishers

Copyright © 2013 American Scientific PublishersAll rights reservedPrinted in the United States of America

ArticleJournal of

BiomedicalNanotechnology

Vol. 9, 1164–1172, 2013www.aspbs.com/jbn

Detection and Identification of Breast CancerVolatile Organic Compounds Biomarkers UsingHighly-Sensitive Single Nanowire Array on a Chip

Yiwen Xu1, Hyunjoong Lee2, Yushi Hu1, Jiyong Huang1, Suhwan Kim2, and Minhee Yun1�∗1Yiwen Xu, Yushi Hu, Jiyong Huang, Minhee Yun, Department of Electrical and Computer Engineering,Swanson School of Engineering, University of Pittsburgh, Pittsburgh, PA 15261, USA2Hyunjoong Lee, Suhwan Kim, Department of Electrical Engineering and Computer Science,Seoul National University, Seoul 151-744, Republic of Korea

A single nanowire array on a chip with different materials of Palladium, Polypyrrole and Zinc Oxide has been fabricatedusing electrochemical deposition method. The fabricated single nanowire array has been demonstrated for highly sensitiveand specific diagnosis of breast cancer by detecting four volatile organic compound biomarkers: Heptanal, Acetophenone,Isopropyl Myristate and 2-Propanol. The demonstrated sensing limits for Heptanal, Acetophenone, Isopropyl Myristateand 2-propanol using individual Palladium, Polypyrrole and Zinc Oxide nanowires were 8.982 ppm, 798 ppb, 134 ppm,and 129.5 ppm, respectively, and the corresponding sensitivities of resistance change were in the range of 0.3%–5%which indicated excellent sensing performance of the single nanowires. The response time for Palladium, Polypyrroleand Zinc Oxide nanowires to achieve maximum conductance change was less than 200 seconds while also illustratingexcellent signal repeatability. With the principal component analysis of the resistance change versus time in each detectionperiod of the nanowire array, the smell prints for the four volatile organic compounds biomarkers of Breast Cancer arediscriminated in the 3-D plots.

KEYWORDS: Breast Cancer, Breath Analysis, Volatile Organic Compounds (VOC) Biomarkers, Sensitivity and Specificity, PrincipalComponent Analysis (PCA).

INTRODUCTIONBreast Cancer is the second most common cancer amongwomen in the United States. According to the latest survey,there were 207,090 women diagnosed with breast cancerand 39,840 women dying from it in the United States dur-ing 2010. Based on the National Cancer Institute’s (NCI’s)Black/White Cancer Survival Study, delays in diagno-sis of 4 weeks or more were reported by almost 40%of women patients, and almost 25% of women patientsreported delays in diagnosis of 8 weeks or more.1 Earlystage diagnosis and treatment are necessary and importantfor decreasing breast cancer mortality.The normal techniques for the screening test of breast

cancer are diagnostic mammogram, magnetic resonance

∗Author to whom correspondence should be addressed.Email: [email protected]: 25 July 2012Revised/Accepted: 6 January 2013

imaging (MRI) and breast ultrasound.2 Diagnostic mam-mogram which uses low-energy-X-rays for the examina-tion is the most common and effective method for thediagnosis of breast cancer now. Special images known ascone or spot views with magnification are applied for mak-ing a small area of abnormal breast tissue to be easier toevaluate. However, the false negative and positive rates forX-ray diagnostic mammogram are still high which resultsin delay in diagnosis or unnecessary biopsy examination.Furthermore, the risk of getting cancer from diagnosticX-ray also exists. According to the research of Gonzálezand Darby,3 the diagnostic X-ray used in USA causes 1.0%of the cumulative risk of cancer to age 75 years in women,which corresponds to 3122 cases per year. In 1981, Dolland Peto’ estimated that about 0.5% of cancer mortalityin USA was due to the attribution from diagnostic X-ray.As supplementary techniques of diagnostic mammogram,MRI and ultrasound are also applied for the screeningtest of breast cancer.3 MRI uses radio waves and strong

1164 J. Biomed. Nanotechnol. 2013, Vol. 9, No. 7 1550-7033/2013/9/1164/009 doi:10.1166/jbn.2013.1651

http://www.aspbs.com/jbn



Xu et al. Detection and Identification of Breast Cancer VOC Biomarkers Using Highly-Sensitive Single Nanowire Array on a Chip

magnets instead of X-rays. The energy of the radio wavescan be absorbed by the body tissue and then released ina pattern by certain disease which translates into visualimages. MRIs are applied for those women who havealready been diagnosed with breast cancer as a diagnos-tic mammogram to have a better knowledge of the actualsize of the cancer part and search for the presence ofother cancers in the breast. However, the effectiveness ofMRI is still unclear, while the time and expense are fur-ther concerns of MRI. Breast ultrasound is another tech-nique using sound waves to outline part of the body. It ishelpful for women with dense breasts and no radiation isinvolved. But, this is only used as a complimentary testwith diagnostic mammograms. Since current techniquesare performing high false rates and risk, an accurate andnon-invasive diagnosis method is necessary for the screen-ing test of breast cancer.Breath analysis is a fast, non-invasive, convenient and

accurate method which utilizes the links between specificVolatile Organic Compounds (VOCs) and diverse medicalconditions. The composition of VOCs in human breath hasbeen investigated by many researchers.4�5 Based on therelationships between specific VOCs and diseases, severalapplications have been developed by taking advantage ofthe detection of specific VOC in breath for the diagnosisor monitoring of various diseases.6–10 Based on previousresult of Phillips et al. five VOC biomarkers of breast can-cer in breath were identified: Heptanal (Hep), 1-Phenyl-Ethanone (Ace), Isopropyl Myristate (IM), 2-Propanol and2,3-Dihydro-1-Phenyl-4(1H)-Quinazolinone.11�12 The pre-diction for breast cancer using these five VOC biomark-ers performs 93.8% sensitivity and 84.6% specificity inprevious research.11�12 Some patients with abnormal mam-mograms but no confirmed breast cancer on biopsy exam-ination can be eliminated using the breath analysis ofthese five VOC biomarkers. Therefore, it is desirable toutilize the detection of these VOC biomarkers in breathtest to develop a new sensing device for the diagno-sis of Breast Cancer. The traditional method for breathanalysis is using Gas Chromatography/Mass Spectrometry(GC-MS).13�14 GC-MS is precise in identifying differentVOCs and measuring their concentration in median level.However, since the concentration of the VOC biomark-ers in breath is down to ppm or ppb level, the normaltest using GC-MS is not able to complete accurate mea-surement and identification. Furthermore, complicated andexpensive pre-concentration procedures are required forbreast analysis based on GC-MS method, and the test timewill also be extended.Here we report a simple, cost-effective, sensitive and

most importantly—accurate single-nanowire array sensorfor the detections of various VOC biomarkers of BreastCancer. A Nanowire sensor is ideal for biomarker sens-ing because of its high surface to volume ratio, surfaceinteraction and conducting characteristics.15–21 Due to highsurface to volume ratio and nanoscale structure, the wires

shows quick response and high sensitivity to biomarkerswith comparatively low concentration. The interactionsbetween VOC biomarkers and nanowires will lead to theconductance change of the nanowire which can be utilizedas sensing feature for chemical and biomolecule sensing.In this report, a single nanowire array sensor on a chipfor the detection of Breast Cancer VOC biomarkers wasfabricated from electrochemically grown Palladium (Pd),Polypyrrole (PPy) and Zinc Oxide (ZnO) single nanowires.The electrochemical method produces consistent deposi-tion results with no interference between the nanowireswhich enables the fabrication of a nanowire array on thesame chip. Four Breast Cancer Biomarkers, including Hep-tanal (Hep), 1-Phenyl-Ethanone (Ace), Isopropyl Myris-tate (IM) and 2-Propanol were chosen as the target VOCbiomarkers for the sensing application. The nanowire arraysensor showed high sensitivity, good repeatability and lowsensing limits to the four VOC biomarkers of BreastCancer, and the required detection period was less than5 min. The principle component analysis (PCA) resultscollected from the simultaneous sensing test showed excel-lent discrimination between the smell prints of four VOCbiomarkers of Breast Cancer. These features indicated thecapability of this nanowire array sensor to be developedas a real-time, accurate and cost-effective sensor for thebreath analysis of Breast Cancer or other disease in thefuture. In addition, it is a promising replacement of tradi-tional diagnosis method for Breast Cancer.

EXPERIMENTAL DETAILSThe Ti/Au electrodes were fabricated on 4-in (100 mm)Si wafer using optical lithography and electron-beam(e-beam) evaporation. The unwanted Ti/Au was removedusing a lift-off process, while the remaining parts wereused for growth of single nanowires and became the con-tact electrodes for single nanowire sensors. After lift-off,100 nm thick PMMA layer was spin-coated on the surfaceof silicon wafer and 100 nm-wide channels were patternedusing e-beam lithography. After photoresist development,nanochannels with width of 100 nm were formed betweentwo Ti/Au electrodes. It was noted that narrow nanochan-nels with less than 100 nm width can be formed accord-ing to e-beam lithography process, and single nanowireswith diameters smaller than 100 nm could be grown withinnanochannel patterns. We use small slices (1 cm× 1 cm)for electrochemical deposition of single nanowires throughnanochannels between the electrodes. One Si wafer wasdesigned to contain 70 chips, and each single chip had 16pairs of electrodes.The electrochemical deposition process of single nano-

wire growth within the nanochannel was performed usingprobe station. First, connect two probes with the twoworking Ti/Au electrodes on both sides of the nano-channel. Then, a drop of electrolyte solution wasplaced on top of the nanochannel using micropipette.

J. Biomed. Nanotechnol. 9, 1164–1172, 2013 1165



Detection and Identification of Breast Cancer VOC Biomarkers Using Highly-Sensitive Single Nanowire Array on a Chip Xu et al.

The electrolyte solutions for single nanowire growth con-tained Pd(NH2�2(NO2�2 (Diamminepalladium Nitrite) andNH4SO3NH2 (Ammonia Sulfamate) with concentrations of10 g/L and 100 g/L for Pd nanowire, NaCl (Sodium Chlo-ride) and Pyrrole monomer (98%) with concentrations of0.2 mol/L and 0.1 mol/L for PPy nanowire, and 5 mmol/LZnCl2 (Zinc Chloride), 15 mmol/L Nacl, 50 mmol/LZn(OH)2 · 6H2O and 50 mmol/L C6H12N4 (HTMA) forZnO nanowire. With the two landed probes connected toa semiconductor analyzer (Agilent B150), a constant DCcurrent was then applied to the pair of electrodes withthe value of 400 nA, 700 nA and 700 nA for Pd, PPyand ZnO nanowire growth, respectively. The voltage sig-nal across the pair of electrodes was monitored duringthe growth. Once there was an obvious large drop to zeropotential of the monitored voltage signal, the nanowirewas formed successfully inside the nanochannel and natu-rally connected between source and drain electrodes. Theapplied current would be stopped in a few seconds in orderto stabilize the conductance of the nanowire and preventextra growth. The uniform property of the nanowire andthe excellent precision of the electrochemical depositionmethod have been verified through a transmission electronmicroscopy (TEM) picture of the nanowire with diameterof 100 nm after the fabrication in Figure 1. All the Pd, PPyand ZnO single nanowires were fabricated separately inparallel on the same chip using the electrochemical depo-sition method described above.In order to connect the chip with an external circuit,

the chip with grown nanowires was attached to a 44-pinchip carrier. Then wire-bonding process using a manualwedge wire-bonder (Kulicke and Soffa INC.) was appliedto connect the Ti/Au bonding pads (400 �m× 400 �m),to the pins on the chip carrier. After that, the chip carrierwith the nanowire chip was plugged into a PLCC socketwith 44 output pins.

Figure 1. Transmission Electron Microscopy (TEM) picture ofnanowire with diameter of 100 nm fabricated in electrochemi-cal deposition method.

The developed VOC biomarker sensing system in thisresearch is shown in Figure 2. The sensing system wascontrolled by Labview and included two gas lines witha MKS flow control system (MKS Multi Gas Controller647C, MKS Instruments Inc., USA), one plastic gas cham-ber with inner size of 29.4 mL, current amplificationinstrument (Keithley 428 Current Amplifier) and the dataacquisition system. One gas line was directly plugged intothe gas chamber with pure N2 as dilution gas, while theother gas line passed through a type of VOC biomarkerchemical solution first using N2 as carrier gas before enter-ing the gas chamber. The VOC biomarker gas would bemixed with dilution N2 in the gas chamber at room temper-ature. By setting a proper ratio of flow rates between dilu-tion N2 and carrier N2, any desired concentration of targetVOC biomarker gas was obtainable within the limit of thecontrol system. By altering the VOC biomarker chemicalsolution connected with the carrier N2 gas line, the singlenanowire sensing tests for all four VOC biomarkers werecompleted. Here, N2 was taken as the dilution gas, carriergas, and purging gas at the end of each detection periodto reset the sensing environment. N2 was chosen becausethere is no physical or chemical interaction between N2and Pd, PPy and ZnO nanowires.In order to detect the conductance change of the sin-

gle nanowire during the VOC biomarker sensing test, aconstant DC voltage (13.6 mV) was applied to the indi-vidual devices via a current amplifier, while the currentwas monitored and amplified into a voltage signal. Thesensing chamber was initially filled with pure dilution N2to clear the existing gas in the chamber and stabilize theconductance of the individual nanowire. After the resis-tance of the nanowire was stabilized as a constant refer-ence value, the gas line with carrier N2 was turned on,and the VOC biomarker gas then flew into the sensingchamber along with the carrier N2. The current throughthe individual nanowire was amplified and collected by thedata acquisition system as shown in Figure 2. The VOCbiomarker gas flow was stopped after the detection periodof 200 seconds. The dilution N2 would clear the absorbedVOC biomarkers on the surface of the nanowire, and theresistance of the nanowire would be recovered to originalreference value. The above steps were repeated for differ-ent concentration sensing tests. Finally, the collected datawas converted into resistance change ratio �R−R0�/R0versus time. Here, R represents the resistance of individ-ual nanowire during detection, and R0 represents the initialresistance as reference value.An additional control circuit shown in Figure 3 was

added into the sensing system to perform simultaneoussensing test. A multiplexing circuit was inserted betweenthe sensor chip in the gas chamber and the current ampli-fier. The circuit sequentially selects one of 3 nanowiresensors from the array and connects the selected sensorto the current amplifier with CMOS switch IC by turn-ing on only one switch at a time. Control signal for this

1166 J. Biomed. Nanotechnol. 9, 1164–1172, 2013




Figure 2. Schematic of single nanowire sensor sensing system including two gas lines with a MKS flow control system (MKSMulti Gas Controller 647 C, MKS Instruments Inc., USA), one plastic gas chamber with inner size of 29.4 mL, current amplificationinstrument (Keithley 428 Current Amplifier) and data acquisition system.

switch was generated using a function generator, binarycounter IC, and 2:4 decoder circuit. The function generatordrives the binary counter with repeating clock pulse. Thecounter is configured to generate binary number from 0 to2 sequentially with the clock. Then the 2:4 decoder circuit,composed of several logic gate ICs, converts the binarynumber input into one selecting signal for the switch.This procedure was repeated automatically and the currentthrough each nanowire type was collected simultaneouslyand continuously with time-multiplexing manner describedabove. The collected data was then converted into resis-tance value via LabView, and the specific value for eachnanowire was separated by setting a different resistancerange.

1 0 0

1 00

100

CMOS Switch IC

ISen

RGai

VBiasVSen

Current AmplifierSensor Array

in Gas Chamber

VO

To DataAcquisition

System

2:4 Decoder

74193

Binary Counter IC

QA

QBUP

Function Generator

Pd

PP

ZnO

0123

0 1 0

0 0 0

0

1

1 020 …

…

…

…

…

…

Figure 3. Control circuit for a simultaneous sensing test. It consists of a function generator, binary counter IC, 2:4 decoder,and CMOS switch IC. The circuit sequentially connects one of 3 nanowire sensors from the array to the current amplifier withCMOS switch IC by turning on only one switch at a time.

Finally, Principle component analysis is employed todisplay the slope of the resistance change versus timeduring each detection period in this research. PCA isa well-known mathematical method for an orthogonaltransformation to convert a set of observations of possiblecorrelated variables into a set of values of linearly uncor-related variables.21 PCA analysis offers an effective tool tocompress the data by reducing the number of dimensionsand eliminating the redundant information without losingmuch information.21 By using PCA for the nanowire arraysensing, the sensing signals for different VOC biomarkerswere successfully separated in 3-D plots. Compared withthe PCA sensing database of the single nanowire array,the target VOC biomarkers can be identified through the





location of the sensing signal in the PCA plots. Therefore,the PCA result indicated the capability of the nanowirearray to be developed as a real time diagnostic sensor forBreast Cancer.

RESULTS AND DISCUSSIONSingle nanowire sensing test was completed for eachkind of nanowire in order to set up relationship betweennanowire material and the VOC biomarkers of BreastCancer, and clarify the sensing feature for each kind ofnanowire. The sensing results of single nanowires of Pd,PPy and ZnO to four VOC biomarkers of Breast Cancer arepresented in Figures 4(a)–(d), separately. Due to the limita-tion of the flow rate control system, the ratio of flow ratesbetween dilution N2 and carrier N2 varied from 500:10 to500:1, and the corresponding concentration ranges for Hep,Ace, IM and 2-Propanol including the sensing limits were88.24 ppm–8.98 ppm, 7.84 ppm–798. ppb, 1315.69 ppm–134.00 ppm, and 1272.55 ppm–129.54 ppm, respectively.

(a) (b)

(c) (d)

Figure 4. Sensing results of single nanowire sensing test for Pd, PPy and ZnO nanowires versus (a) Hep, (b) Ace, (c) IM,(d) 2-Propanol with seven detection periods of 200 seconds. The ratios of flow rates between the carrier N2 and dilution N2 foreach detection period were 10–500, 8–500, 6–500, 4–500, 2–500, 1–500 and 5–500 seconds.

The single nanowires were exposed to each concentra-tion of the VOC biomarkers for 200 seconds which wastaken as the detection period. As illustrated in Figure 4,the resistance increased gradually along with the afflux ofVOC biomarkers due to the absorption onto the nanowiresurface. When the VOC biomarkers were pumped out,the resistance decreased rapidly at the beginning and ulti-mately returned to the reference value. From the curvesin Figure 4, the response time of the single nanowiresincreased with decreased VOC biomarker concentrationbecause the decreased concentration slowed down the dif-fusion rate of gas molecules on the nanowire surface.Besides, the last test period with ratio of flow rates of500:5 performed expected sensitivities compared with thetest period with ratio of flow rates of 500:6 and 500:4,which indicated the excellent repeatability of the singlenanowires without degradation during the sensing tests.Also, as shown in Figures 4(a)–(d), the sensitivities ofPd and PPy nanowires versus four VOC biomarkers ofBreast Cancer were close, which were comparatively lower





than the sensitivities of ZnO nanowire. The sensitivitiesof PPy nanowire versus the sensing limits of Hep, Ace,IM and 2-Propanol were 1.128%, 0.236%, 0.663% and0.561%, respectively. However, the sensitivities of ZnOnanowire versus the same sensing limits were 4.594%,3.110%, 5.460% and 2.672%, respectively. The sensingperformance of Pd, PPy and ZnO nanowires with dis-tinct sensitivities, similar excellent repeatability and quickresponse indicated the great capability of single nanowirearray sensor for the detection of VOC biomarkers of BreastCancer.The sensing mechanism for the nanowires versus VOC

biomarkers is normally explained with doping or dedopingeffect. Here, the sensing mechanism of PPy nanowire ver-sus four VOC biomarkers of Breast Cancer is introducedin detail. Since most of VOCs are not reactive at roomtemperature, it was difficult to detect them by their chem-ical reactions with conducting polymers. However, theymight have weak physical interactions with the sensingpolymers, to produce either absorption or swelling of thepolymer nanowires.22 According to the model developedby Hwang and Lin based on Langmuir isotherm,23–25 activesites for VOC absorption were disputed on the surface ofthe conductive polymer nanowire. The overall resistanceof the nanowire was taken as m resistances of R in par-allel and each R was considered as n resistances of r0 inseries as shown in the following equation. Here, m, R, n,r0 represented the number of conduction paths, the resis-tance of one conduction path, number of active site in oneconduction path, and the resistance of original active site.

R= nr0 (1)When VOC biomarker gas was injected into the chamber,some of the VOC molecule would be absorbed onto theactive sites. Due to the doping and dedoping effect, thesefour VOC biomarkers would donate electron as electron-donator when they were absorbed onto the active site.Since PPy nanowires are considered as a p-type conduc-tive polymer, the doping level was decreased and the workfunction was increased during the exposure as shown inFigure 5. Finally, the partial resistance of the active sitewith absorbed VOC molecule would be increased to r1.The site coverage of absorption was represented by �.Therefore the resistance R would be:

R= n�r1+n�1−��r0 (2)As a combination of the empty and occupied active sites,the overall resistance of the PPy nanowire would beincreased. For different VOC biomarkers, the site coverageof absorption and the extent of dedoping effect would bedifferent. Therefore, the resistance change and the sensi-tivity are distinct for different VOC biomarkers. Since thephysical absorption was dominant between PPy nanowireand VOC biomarkers of breast cancer, the absorbed VOCmolecule would be able to be desorbed by the dilution

Figure 5. Sensing mechanism expression of PPy nanowireversus VOC biomarkers with the structure and the dedopingeffect before and after the absorption of VOC on the active siteof the conductive polymer nanowire.

N2. Hence, the doping level of the conductive polymerwould be recovered to original level. This desorption ofthe biomarkers was evidenced by resistances returning tothe original level when biomarker gases were turned off.This sensing mechanism was applied for all four VOCbiomarkers of Breast Cancer detected in this research.As illustrated in Figure 4(d), the highest concentration

sensing test for 2-Propanol shows different sensing curvescompared with other VOC biomarkers. The suspected rea-son for these phenomena might be the chemical reactionbetween 2-Propanol and the nanowire. The swelling effectbetween 2-Propanol and the conductive polymers was alsosuspected as another significant factor for the abnormalsensing feature.23 For pure conductive polymer, due to theswelling effect, inserting analyte molecules into the con-ductive polymer nanowire generically increases interchaindistance, which affects the electron hopping between dif-ferent polymer chains. This is controlled by the vapormolecular volume, and the affinity of the vapor to the sens-ing polymer and the physical state of the polymer. Boththe chemical reaction and the swelling effect would leadto the permanent change of the composition and structureof the nanowire. Therefore, the resistance would not beable to be recovered to its original reference value in thecase of 2-Propanol sensing.To characterize the determining factors for sensitivity

and specificity of Pd, PPy and ZnO nanowires, sensingtest of single nanowires with different resistances versusIM using a single nanowire sensing system proceeded inthe following step, and the results are shown in Table I.The detection period was applied as 200 seconds, and theratio of flow rates between dilution N2 and carrier N2was fixed as 500–1 with a concentration of 134 ppm.As described in the previous research,26 due to the vari-ance of the deposition parameters, the deposition result ofnanowire growth leads to the differences in the nanowirestructure and distinct resistances. The physical propertiesof the nanowire structure affect the sensing performanceof the nanowire, and the sensing features of the nanowireswill be different. Since the resistance reflects the feature





Table I. Specificity sensing test for Isopropyl Myristate (IM)with Pd, PPy and ZnO nanowires with different resistances atratio of flow rate of 500–1 and concentration of 134 ppm.

Pd Ppy ZnO

Sensitivity (%) 0.5009 0.0108 0.0568 0.294 0.2666 0.2503Resistance (�� 1,055 184 7,824 1,300 2,087 2,572

of the nanowire structure, it is important and necessary tocharacterize the effect of resistance for the sensing resultversus VOC biomarkers of Breast Cancer. As shown inTable I, the Pd nanowire with resistance of 1,055 � per-formed sensitivity of 0.50% which was 50 times of thesensitivity of the Pd nanowire with the resistance of 184�.Besides, the PPy nanowire with the resistance of 1,300 �showed higher sensitivity compared with the PPy nanowirewith resistance of 7,824 �. Therefore, the variance of theresistance and structure will definitely affect the sensingfeature of the nanowires, and the resistance range around1 k� might be the effective conductance range with bet-ter sensing performance for Pd and PPy nanowires. As aresult, the deposition process required being consistent andthe resistance of the nanowire should be limited withinthe effective range around 1 k�–2 k� in order to achievebetter sensitivity. For the specificity characterization, thebest sensitivities for each kind of nanowire were taken intocomparison. The compared result showed that Pd nanowireproduced the highest sensitivity of 0.50%, while PPy andZnO nanowires showed sensitivities of 0.29% and 0.26%.The obvious difference of the sensitivities indicated theexcellent specificity of Pd, PPy and ZnO nanowires whichcan be utilized to develop single nanowire arrays to dis-criminate the VOC biomarkers of Breast Cancer.In order to test the performance of different nanowires

simultaneously, and to develop the nanowire array sen-sor on a chip for the detection of VOC biomarkers ofBreast Cancer, all three different materials of nanowireswere deposited as a nanowire array on the same chip andtested using simultaneous sensing system. By using theadditional control circuit, Pd, PPy and ZnO nanowireswere connected to the current amplification instrumentconcurrently, and the sensing signals were collected simul-taneously. The ratios of flow rates for the detection peri-ods of 200 seconds varied from 500–10 to 500–1 whichwas applied for single nanowire sensing tests. The sens-ing curves for the nanowires versus VOC biomarkers werethe same as the results during the single nanowire sens-ing test as shown in Figure 4. The slope of the resis-tance change versus time was calculated and applied forPCA analysis. Figure 6 shows the experimental resultsof a 3-D dimensional plot using PCA analysis. Here, thedata point in the same color represents the same VOCbiomarker, and different points with the same color rep-resent different concentrations. To clarify the sensing datafor different concentration tests, the size of the data point

(a)

(b)

Figure 6. 3-D PCA plots in different angles (a) and (b) for Pd,PPy and ZnO nanowires versus four VOC biomarkers of BreastCancer. The ratio of flow rates varies from 10–500 to 1–500.

with lower concentration is larger than the one with higherconcentration. The variances possessed by the principalcomponents were 73.25% (PC1), 22% (PC2), 4.75% (PC3),and all the sensing information from the nanowire arraysensor is applied. As illustrated in Figure 6, with differ-ent angle of view, the clustering of data points in the samecolor was desirable because the data points representing thesame target gases tend to occupy specific areas as smellprints on the PCA plot, and different smell prints should notmerge. Another simultaneous sensing test was completedwith fixed concentration of four VOC biomarkers in orderto further identify the specificity of the single nanowirearray. By applying different ratios of flow rates, the sameconcentration of four VOC biomarkers in the range of107.19 ppm–267 ppm was achieved. The applied detectionperiod was kept as 200 seconds and the calculated slopeof resistance change versus time was applied for the PCA.The PCA results were shown in Figure 7 and colors for fourVOC biomarkers were the same as Figure 6. As expected,the smell prints matched the previous simultaneous test.By rotating the angle of view for the 3-D PCA plots, all





(a)

(b)

Figure 7. 3-D PCA plots in different angles (a) and (b) forPd, PPy and ZnO nanowires versus four VOC biomarkers ofBreast Cancer. The concentration varies from 1107.19 ppm to267 ppm.

four VOC biomarker smell prints were distinguishable inthe same low concentrations. Both simultaneous sensingtests indicated success in specifying individual smell printsfor four VOC biomarkers of Breast Cancer down to a ppbconcentration level. Therefore, this single nanowire arrayconsisting of Pd, PPy and ZnO nanowires is promising forthe application of real time diagnostic sensing test of VOCbiomarkers in human breath.

CONCLUSIONSNanowire array sensors with Pd, PPy and ZnO nanowiresfabricated in parallel on a chip was successfully demon-strated using electrochemical method. Our results indi-cated that rapid response during the detection period forfour kinds of VOC biomarkers of Breast Cancer in lowconcentration was feasible, and no pre-concentration pro-cedure was required. The sensing limits for the nanowirearray sensor were around 1 ppm–100 ppm, which wereclose to the normal concentration level of VOC biomarkersin human breath, and the sensitivities for the sensing limits

were around 0.3%–5.0%. The smell prints for four VOCbiomarkers of Breast Cancer were completely separatedafter PCA of the simultaneous sensing data, which indi-cated great specificity of this single nanowire array sensor.The sensing results of this single nanowire array sensorshowed great sensitivity, reproducibility, specificity andhigh accuracy. Also of note, the electrochemical deposi-tion procedure was simple and consistent, and the sensingtest was able to be completed within 5 min in real timesensing process. All these features of the single nanowirearray sensor indicate the great capability as a replacementfor the diagnosis of Breast Cancer and bright economicfuture for real commercial application.

Acknowledgments: This work was partially supportedby the National Science Foundation (NSF-ECCS 0824035)and National Institutes of Health (NIH 1R21EB008825).

REFERENCES1. L. S. Caplan, K. J. Helzlsouer, S. Shapiro, L. S. Freedman, R. J.

Coates, and B. K. Edwards, System delay in breast cancer in whitesand blacks. Am. J. Epidemiol. 142, 804 (1995).

2. C. K. Kuhl, S. Schrading, C. C. Leutner, N. Morakkabati-Spitz,E. Wardelmann, R. Fimmers, W. Kuhn, and H. H. Schild, Mam-mography, breast ultrasound, and magnetic resonance imaging forsurveillance of women at high familial risk for breast cancer. J. Clin.Oncol. 23, 8469 (2005).

3. A. B. de González and S. Darby, Risk of cancer from diagnosticX-rays: estimates for the UK and 14 other countries. The Lancet363, 345 (2004).

4. M. Phillips, Method for the collection and assay of volatile organiccompounds in breath. Anal. Biochem. 247, 272 (1997).

5. M. Phillips, J. Herrera, S. Krishnan, M. Zain, J. Greenberg, and R. N.Cataneo, Variation in volatile organic compounds in the breath ofnormal humans. J. Chromatogr. B Biomed. Sci. Appl. 729, 75 (1999).

6. G. Peng, U. Tisch, O. Adams, M. Hakim, N. Shehada, Y. Y. Broza,S. Billan, R. Abdah-Bortnyak, A. Kuten, and H. Haick, Diagnosinglung cancer in exhaled breath using gold nanoparticles. Nat. Nan-otechnol. 4, 669 (2009).

7. P. J. Mazzone, Analysis of volatile organic compounds in the exhaledbreath for the diagnosis of lung cancer. J. Thorac. Oncol. 3, 774(2008).

8. A. Wehinger, A. Schmid, S. Mechtcheriakov, M. Ledochowski,C. Grabmer, G. A. Gastl, and A. Amann, Lung cancer detection byproton transfer reaction mass-spectrometric analysis of human breathgas. Int. J. Mass Spectrom. 265, 49 (2007).

9. M. Righettoni, A. Tricoli, and S. E. Pratsinis, Si:WO3 Sensors forhighly selective detection of acetone for easy diagnosis of diabetesby breath analysis, Anal. Chem. 82, 3581 (2010).

10. Y. Zhang, G. Gao, Q. Qian, and D. Cui, Chloroplasts-mediatedbiosynthesis of nanoscale Au-Ag alloy for 2-butanone assay basedon electrochemical sensor. Nanoscale Research Letters. 7, 475(2012).

11. M. Phillips, R. N. Cataneo, B. A. Ditkoff, P. Fisher, J. Greenberg,R. Gunawardena, C. S. Kwon, F. Rahbari-Oskoui, and C. Wong,Volatile markers of breast cancer in the breath. The Breast Journal,9, 184 (2003).

12. M. Phillips, N. Altorki, J. H. M. Austin, R. B. Cameron, R. N.Cataneo, J. Greenberg, R. Kloss, R. A. Maxfield, M. I. Munawar,H. I. Pass, A. Rashid, W. N. Rom, and P. Schmitt, Prediction ofbreast cancer using volatile biomarkers in breath. Cancer Biomark-ers, 3, 95 (2007).





13. J. M. Sanchez and R. D. Sacks, GC analysis of human breath with aseries-coupled column ensemble and a multibed sorption trap. Anal.Chem. 75, 2231 (2003).

14. A. Hryniuk and B. M. Ross, Detection of acetone and isoprenein human breath using a combination of thermal desorption andselected ion flow tube mass spectrometry. Int. J. Mass Spectrom.285, 26 (2009).

15. W. Lu and C. M. Lieber, Semiconductor nanowires. J. Phys. D: Appl.Phys. 39, R387 (2006).

16. Y. Wu, J. Xiang, C. Yang, W. Lu, and C. M. Lieber, Single-crystalmetallic nanowires and metal/semiconductor nanowire heterostruc-tures. Nature 430, 61 (2004).

17. S. C. Hernandez, D. Chaudhuri, W. Chen, N. V. Myung, andA. Mulchandani, Single polypyrrole nanowire ammonia gas sensor.Electroanalysis 19, 2125 (2007).

18. L. Dong, T. Hollis, S. Fishwick, B. A. Connolly, N. G. Wright, B. R.Horrocks, and A. Houlton, Synthesis, manipulation and conductivityof supramolecular polymer nanowires. Chem. -Eur. J. 13, 822 (2007).

19. H. Yan, L. Zhang, J. Shen, Z. Chen, G. Shi, and B. Zhang, Synthe-sis, property and field-emission behaviour of amorphous polypyrrolenanowires. Nanotechnology 17, 3446 (2007).

20. Y. W. Heo, D. P. Norton, L. C. Tien, Y. Kwon, B. S. Kang, F. Ren,S. J. Pearton, and J. R. LaRoche, ZnO nanowire growth and devices.Mater. Sci. Eng. R Rep. 47, 1 (2004).

21. I. T. Jolliffe, Principal Component Analysis, Series: Springer Seriesin Statistics, 2nd edn., Springer, NY (2002).

22. H. Bai and G. Shi, Gas Sensors Based on Conducting Polymers.Sensors 7, 267 (2007).

23. C. W. Lin, B. J. Hwang, and C. R. Lee, Characteristics and sensingbehavior of electrochemically codeposited polypyrrole-poly(vinylalcohol)thin film exposed to ethanol vapors. J. Appl. Polym. Sci.73, 2079 (1999).

24. B. J. Hwang, J.-Y. Yang, and C. W. Lin, A microscopic gas-sensingmodel for ethanol sensors based on conductive polymer compos-ites from polypyrrole and poly(ethylene oxide). J. Electrochem. Soc.146, 1231 (1999).

25. C. W. Lin, S. S. Liu, and B. J. Hwang, Study of the actions of BTEXcompounds on polypyrrole film as a gas sensor. J. Appl. Polym. Sci.82, 954 (2001).

26. Y. Hu, D. Perello, U. Mushtaq, and M. Yun, A single palladiumnanowire via electrophoresis deposition used as a ultrasensitivehydrogen sensor. IEEE Trans. Nanotechnol. 7, 693 (2008).


Documents

Articleanalog.snu.ac.kr/publications/JBN_2013.pdf · Delivered by Publishing Technology to: Suhwan Kim IP: 147.46.61.45 On: Sun, 05 May 2013 01:59:52 Copyright American Scientific