1

Sensors for Food Quality Safety: Detection and

Characterization

Sensors for Food Quality Safety: Detection and

Characterization

Joseph Irudayaraj, Associate ProfessorAgricultural and Biological Engineering

Purdue University

Joseph Irudayaraj, Associate ProfessorAgricultural and Biological Engineering

Purdue University

Key CollaboratorsKey CollaboratorsPenn State

Food Science: Steve Knabel, Richard Apenten, John Coupland, Koushik Seetharaman, Bob RobertsVet Sci: Chobi Debroy, Ali Demirci, Bhushan JayaraoChem/Materials: David Allara, Carlo Pantano

Purdue UniversityBindley Biosciences CenterDepts.: ABE, Food Sciences, Vet Sciences, ChemistryMike Ladisch, Cris Staiger, Garth Simpson, Sophie Liverve, Peixuan Guo, Kinam Park

IUPUI Cancer Center (Drs. Nakshatri, Sledge, MD)Mayo Clinic (Robert Jenkins, MD)

Penn StateFood Science: Steve Knabel, Richard Apenten, John Coupland, Koushik Seetharaman, Bob RobertsVet Sci: Chobi Debroy, Ali Demirci, Bhushan JayaraoChem/Materials: David Allara, Carlo Pantano

Purdue UniversityBindley Biosciences CenterDepts.: ABE, Food Sciences, Vet Sciences, ChemistryMike Ladisch, Cris Staiger, Garth Simpson, Sophie Liverve, Peixuan Guo, Kinam Park

IUPUI Cancer Center (Drs. Nakshatri, Sledge, MD)Mayo Clinic (Robert Jenkins, MD)

Several Others ….

Assessment of Food Quality and Safety Parameters

Assessment of Food Quality and Safety Parameters

Handled

Processed

Stored

Packaged

More than 90% of foodborne illnesses areattributed to bacteria and 6 of these cause over 50% of the illnesses

More than 90% of foodborne illnesses areattributed to bacteria and 6 of these cause over 50% of the illnesses

DoseDeathCasesBacteria

101 to 102400725,000E. coli O157:H7

104 to 1074,0003.8 * 106Salmonella

> 10810010,000C. perfringens

> 10612101.0 * 106S. aureus

400 t0 1034851,767L. monocytogenes

400 to 1065114.0 * 106C. jejuni

2

Non-Contact Ultrasound ImagingNon-Contact Ultrasound Imaging

DetectionFragments: minimum 3×3 mm2

Cylindrical objects: 1.5 mm in diameter

Cracks and sporadic porosities

DetectionFragments: minimum 3×3 mm2

Cylindrical objects: 1.5 mm in diameter

Cracks and sporadic porosities

Foreign object & internal disorder detection in cheese using Non-contact ultrasound imaging

Cho et al. (2000-2004)

Attenuation coefficient measurement

Attenuation coefficient measurement

Attenuation coefficient

= (IR1 – IR2) / (2*thickness)

Attenuation coefficient

= (IR1 – IR2) / (2*thickness)

Reflection

Dm

Air-coupled transducer 1

Air-coupled transducer 2

0

0.05

0.1

0.15

0.2

0.25

0.3

0.35

0.4

0.45

0.5

100 150 200 250

time (usec)

rela

tive

ampl

itude

(dB)

0

0.05

0.1

0.15

0.2

0.25

0.3

0.35

0.4

0.45

0.5

100 150 200 250

time (usec)

rela

tive

ampl

itude

(dB) IR1

IR2

NCU Imaging systemNCU Imaging system

To obtain NCU imageTo obtain NCU imageNon-contact air instability compensation transducer

Sample

X-Y position system

X

Y

NCU unit

Image processingImage processing

velocity relative attenuation

raw

processed

Glass fragment

3

Foreign object & internal disorderForeign object & internal disordermetal (5×3mm) glass (3×3mm) hole (d 4mm)

sporadic porosities

4.3

4.4

4.5

4.6

4.7

4.8

4.9

5

5.1

5.2

0 5 10 15 20

0

5

10

15

20

25

X Lateral distance [mm]

Y L

ater

al d

ista

nce

[mm]

Cheese Glass fragments in chicken breast

(Noncontact ultrasound image)

Food Quality – Non contact Food Quality – Non contact

ZnoseHoney authenticityHoney adulterationWine classificationApple quality

ZnoseHoney authenticityHoney adulterationWine classificationApple quality

λ

T

PiezoelectricsubstrateMetal

ITD

L

20 SECONDS SAMPLING TIME

895000

896000

897000

898000

899000

900000

901000

902000

903000

904000

905000

906000

0 5 10 15 20 25

TIME(SECONDS)

FREQ

UEN

CY(

HZ)

Enose – abundant Literature

FTIR

PAS

Mirror

SampleMicrophone

KBr window

IR beam

Computer

IR spectrum

Diagram of FTIR-PASDiagram of FTIR-PAS

100 Hz 10 Hz

α = 0.001cm2/s

900 Hz

56 μm

18 μm6 μm

Depth

Phase modulation frequencies

4

3300

2929

1655 15

5514

5814

0713

2012

4211

1510

5692

685

5

3381

2931

1652

1559

1458

1420 11

54 1054

930

857 76

4

2959 29

23

1376

1162

996 97

3

839

719

4000 3500 3000 2500 2000 1500 1000 500

2843

1646

1457

Phot

oaco

ustic

Inte

nsity

Wavenumber (cm-1)

Starch

Protein

Polyethylen

Rapid-scan FTIR-PAS spectra of protein, starch, and polyethylene

Amide IIAmide I

857930

1457 1375Red color indicate polyethylene is above protein and starch

Asynchronous G2D correlation FTIR-PAS spectra of three-layer polyethylene/starch/protein

Fourier transform Raman spectroscopyFourier transform Raman spectroscopy

A near-infrared (1064 nm) laser is used as a probed beam to overcome the fluorescence effectThere is no sample preparationRaman spectrum is complementary to infrared spectrum

A near-infrared (1064 nm) laser is used as a probed beam to overcome the fluorescence effectThere is no sample preparationRaman spectrum is complementary to infrared spectrum

Direct determination of microorganisms on Food and package

Materials & Methods

Material: fungi (Aspergillus niger andFusarium verticilliodes), yeast (Saccharomyces cerevisiae), bacteria (Bacillus cereus, Lactobacillus casei, and E.coli (HB101, DH5α, JM107, JM101, K12, O157:H7)), and apple

Instrumentation: FTIR-PAS and FT-Raman

Data Analysis: Canonical variate analysis

5

FT-Raman spectra of uncontaminated apple and contaminated apple surfaces with different types of microorganisms

FT-Raman spectra of uncontaminated apple and contaminated apple surfaces with different types of microorganisms

Ram

an In

tens

ity

4000 3500 3000 2500 2000 1500 1000 500

Wavenumber (cm-1)

3379

2931

1658

1458 1304 1088 856

E.coli HB101

Aspergillus niger

Fusarium verticilliodes

Saccharomyces cerevisiae

Bacillus cereus

Lactobacillus casei

Apple skin

Discriminant canonical variate analysis based on the first two canonical variates from the spectra of whole apple

surface with/without microorganisms

Discriminant canonical variate analysis based on the first two canonical variates from the spectra of whole apple

surface with/without microorganisms

-6

-4

-2

0

2

4

6

-6 -4 -2 0 2 4 6 8 10

CV1

CV

2

Aspergillus nigerBacillus cereusE.coli HB101Fusaruim verticillioidesLactobacillus caseiSaccharomyces cerevisiaeApple skin

Bacteria 1. E. coli O157:H72. Salmonella Enteritidis3. Bacillus cereus4. Yersinia enterocolitis5. Shigella boydii

3 out of 5 bacteria were used to make 10 possible combination of mixtures:

m1: 1,2,3; m2: 1,2,4; m3: 1, 2, 5; m4: 1,3, 4; m5: 1, 3, 5; m6: 1,4,5; m7: 2,3,4; m8: 2, 3, 5; m9: 2 ,4, 5; m10: 3,4,5

Bacteria 1. E. coli O157:H72. Salmonella Enteritidis3. Bacillus cereus4. Yersinia enterocolitis5. Shigella boydii

3 out of 5 bacteria were used to make 10 possible combination of mixtures:

m1: 1,2,3; m2: 1,2,4; m3: 1, 2, 5; m4: 1,3, 4; m5: 1, 3, 5; m6: 1,4,5; m7: 2,3,4; m8: 2, 3, 5; m9: 2 ,4, 5; m10: 3,4,5

Testing and validation of the algorithm

Testing and validation of the algorithm

Differentiation of microorganisms in a cocktail of pathogens

Differentiation of microorganisms in a cocktail of pathogens

Example ‘fingerprint’ of each bacterium (for m2: 1, 2, 4)

Spectra after background interference removed

6

Discriminant analysis (m1: 1+2+3)

TestObservations

Centre of Group1

Centre of Group2

Centre of Group3

Centre of Group4

Centre of Group5

CV Score 1

CVScore

2

5

10

15

20

-5

-10

-15

-20

-25

10 20 30 40 50-10-20-30-40

Bacteria present1. E. coli O157:H72. Salmonella3. BacillusBacteria not present4. Yersinia5. Shigella100% accurate prediction for

this application !

Concept has been extended to a cocktail of up to 12 pathogens in apple juice

Food SafetyFood SafetyCharacterization

DetectionCharacterization

Detection

SpectroscopySpectroscopy BiosensorsBiosensors

FTIRRaman

Confocal-Raman

Single particleIR flim based

Biomimetic, SPR

BiosecurityBiosecurity

Nanoscalemethods

Single molecule methods

Chemistry, Physics,Engineering

SPR based pathogen detection

Sensor surface setup Flush

Ethanol D.Water HCl NaOH D.Water

SAM ActivationEDC & NHS mixture

Ab ImmobilizationAb in SA

Establish baseline: PBSTInject Ag: Ags in PBSTDissociation: 20 mM NaOHEstablish baseline: PBSTReuse the chip

Sensor surface setup Flush

Ethanol D.Water HCl NaOH D.Water

SAM ActivationEDC & NHS mixture

Ab ImmobilizationAb in SA

Establish baseline: PBSTInject Ag: Ags in PBSTDissociation: 20 mM NaOHEstablish baseline: PBSTReuse the chip

Assay Development

Sensitivity and SpecificitySensitivity and SpecificityO157:H7

7

Mid-Infrared BiosensorsMid-Infrared Biosensors

Chalcogenide film (GeSbSe)

Silicon substrate

Incoming IR beam To

detector

antibody Target (bacterial cells)

Gold coating

GeSe2 cladding layer

700 800 900 1000 1100 1200 1300 1400Wave numbers, cm-1

abso

rban

ce, a

rbitr

ary

unit

treated with E. coli K12treated with S. enteriditistreated with E. coli O157:H7treated with PBS buffertreated with E. coli cocktailtreated with all-three cocktail

Leigh, PSU, Purdue

Gold Islands on Chalcogenide films

50 % 50 % DiynoicDiynoic acidacid

50 % 50 % Natural LipidsNatural Lipids

++1. Sonication2. UV irradiation

Biomimetic color nanocompositesBiomimetic color nanocomposites

50 – 300 nm

Ben Gurion Univ (Jelenik group)

Professor Raz Jelinek, Ben Gurion University

Color sensor for pathogen detectionProfessor Raz Jelinek, Ben Gurion University

Color sensor for pathogen detection

Platform for pathogen detection through visible color changes.

Platform is based on films composed of lipids and chromatic polymer (polydiacetylene)

Platform for pathogen detection through visible color changes.

Platform is based on films composed of lipids and chromatic polymer (polydiacetylene)

control 1 ppb 4 ppb 20 ppb 40 ppb 4 ppm

The system detects the presence of substances that are active on biological membranes. Very high intrinsic sensitivity

Visible detection of Salmonella typhi. Less than 100,000 bacterial particles detected by eye. Higher sensitivity obtained using fluorescence microscopy

Positive control

109

107

105 106

108

medium

Salmonella typhimurrium E. coli

Lipid/PDA film

Agar growth matrix

Bacterial colony

Another color-based sensing assembly

8

Possible sensor productsPossible sensor products

Pathogen diagnosis patch for foodstuff

Generic food productPathogencontamination

Nanoparticle-based sensorsNanoparticle-based sensors

COOH

COOH

OHHNO3

(OH2)+(NO3)- + NH2CHCH2COOH

11-mercaptopropionic acid

ethanol

pH 2(O2C)CH2CHNH2

R

O

C COHOH

L-aspartic acid

(a)

(b)

(c)

O O

C OH +

H2O

C

N

N R1

R2

O

R C O C

NH

NHR1

R2

NHS+R C O

O

N

O

O

C

N

N R1

R2

O+

NH2 -RhPrPc R C

O

NH -streptavidin

(d)EDC

R C O

O

N

O

O

+ NH2 -streptavidin NHO

O

O

+R C

O

NH -streptavidin

R C

O

NH -streptavidin + biotin aptamer

R C

O

NH -streptavidin biotin aptamer

R C

O

NH -streptavidin biotin aptamer

+

biotinaptamerRhPrPc

(e)

(f)

COOH

COOH

OHHNO3

(OH2)+(NO3)- + NH2CHCH2COOH

11-mercaptopropionic acid

ethanol

pH 2(O2C)CH2CHNH2

R

O

C COHOH

L-aspartic acid

(a)

(b)

(c)

O O

C OH +

H2O

C

N

N R1

R2

O

R C O C

NH

NHR1

R2

NHS+R C O

O

N

O

O

C

N

N R1

R2

O+

NH2 -RhPrPc R C

O

NH -streptavidin

(d)EDC

R C O

O

N

O

O

+ NH2 -streptavidin NHO

O

O

+R C

O

NH -streptavidin

R C

O

NH -streptavidin + biotin aptamer

R C

O

NH -streptavidin biotin aptamer

R C

O

NH -streptavidin biotin aptamer

+

biotinaptamerRhPrPc

(e)

(f)

DNA sensors (Anal Chem); protein-aptamer (JACS), Enzyme Kinetics (NanoBio Tech J)

Rod shaped gold nanoparticles have unique optical properties

Rod shaped gold nanoparticles have unique optical properties

Gold nanoparticles

Gold nanorods

The red axis representsspherical shaped goldnanoparticles and appears atabout 500nm.

The green axis representsrod shaped goldnanoparticles and appearsat 600nm or above depends upon the aspect ratio of the rods

Au nanoparticles

Au nanorods

+ HAuCl4, AgNO3,Ascorbic acid +

Seed-mediated growth of gold nano rods

Yu and Irudayaraj [90% yield]

CTAB bilayer

anchor

targets

Capturing agents

9

Functionalization and Target detectionProbe: human IgG; Target: goat anti-human IgG

400 450 500 550 600 650 700 750 800 8500

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

Wave length, nm

Abs

orba

nce,

AU

RodsRods with SAM

Rods woth SAM and IgG

IgG/antiIgG binding

520 nm

679.5 nm

687 nm699.5 nm

710.5 nm

400 450 500 550 600 650 700 750 800 850 9000

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

Wave length, nm

Abs

orba

nce,

AU

RodsRods with SAM & IgG

Rods with SAM

IgG/anti IgG binding

514.5 nm

762.5 nm775 nm

797 nm839.5 nm

Multiplexing using nanorods

Bioaffinity SensorsBioaffinity Sensors

Receptor RDyeLectinEnzymeApoenzymeAntibodyReceptorTransport system

Receptor RDyeLectinEnzymeApoenzymeAntibodyReceptorTransport system

Chemical signal SProteinSaccharideHarmoneSubstrateAntigenSubstrate analogue

Chemical signal SProteinSaccharideHarmoneSubstrateAntigenSubstrate analogue

S + R SR

TransducersTransducersThermistors [Thermometric indication]

A change in enthalpy due to enzyme-catalyzed reactionOne reaction step and no final product

Thermistors [Thermometric indication]A change in enthalpy due to enzyme-catalyzed reactionOne reaction step and no final product

An enzyme attached thermistor is dipped in a sample Enzyme : Glucose oxidase, Urease, Trypsin

Substrate : Glucose, Urea, Cholesterol

Electrochemical TransductionPotentiometricAmperometricConductometric

Electrochemical TransductionPotentiometricAmperometricConductometric

Selected Biosensors (sugars)

Selected Biosensors (sugars)

milk0.002-350AmperoLactose

Brewer’s yeast frem

0.03-2.57AmperoMaltose

On-line< 8g/lpH-electrode

Glucose

cocoa0.2-2.820AmperoGlucose

ApplicationRange(mM)Life(d)PrincipleAnalyte

10

Other ExamplesOther Examples

0.01-160chemuLum

Lysine

E.Coli(culture)

0.025-1.090AmperoTryptophan

oils0.1-1.210AmperoFatty acids

juices0.001-1.018-21AmperoCitrate

ApplicationRange(nM)Life(d)PrincipleAnalyte

Questions in DevelopmentQuestions in DevelopmentMarketAlternate methods and costSensitivity and SpecificitySample matrix – is this changing?Measurement conditionsApplication – in/at/on/off -lineMeasurement range, time, life

MarketAlternate methods and costSensitivity and SpecificitySample matrix – is this changing?Measurement conditionsApplication – in/at/on/off -lineMeasurement range, time, life

Thank youThank you