Quantum cascade laser spectrometer for trace-gas detection of

Transplant Recipients with Acute Allograf Rejection”, J. of Heart and Lung. Transplant, 20(11), 1158 (2001) .... Serial Port. Transceiver. DAC. QCL Pulser.
344KB taille 0 téléchargements 287 vues
Quantum cascade laser spectrometer for trace-gas detection of exhaled Carbonyl Sulfide Gerard Wysocki, Stephen So, Matt McCurdy, Chad Roller, Damien Weidmann, Anatoliy A. Kosterev, J. Patrick Frantz, Robert F. Curl, and Frank K. Tittel Rice University, 6100 Main Street, Houston, TX, 77005, USA

http://www.ece.rice.edu/lasersci

QCL Sensor Wavelength Calibration

Motivation

1000 spectra averaged acquired within t = 4 s and fitted to 300 ppb OCS reference spectrum

S.M. Studer et.al.,“Patterns and Significance of Exhaled-Breath Biomarkers in Lung Transplant Recipients with Acute Allograf Rejection”, J. of Heart and Lung Transplant, 20(11), 1158 (2001)



As well as in patients with liver disease S.S. Sehnert et.al., “Breath biomarkers for detection of human liver diseases: preliminary study”, Biomarkers, 7(2), 174 (2002)



Application of quantum-cascade (QC) lasers allows the design of a compact sensitive, and selective trace-gas sensor



Goal: non-invasive rapid, in situ detection of trace-gases in exhaled human breath

-0.10 0

• • •

Absorption (sample)

-1 0.00 -0.05 100

200 300 points

400

Pulse repetition rate: 125 kHz Subthreshold current saw-tooth signal: Imax= 35 mA; t = 3.2 ms Range of a single frequency scan : ~ 0.3 cm-1

Data points Polynomial fit

-0.05

⎡ ⎤ ∆ν ⎥ δA ≈ σ ⎢ 2 ⎢ ∫ g (ν )dν ⎥ ⎣ ⎦

-3

tg(α) = 0.174 C OCS= 52.2 ppb

-3

2x10

-0.10

0 -2

0

-0.15

1x10

2x10

-2

3x10

-2

-3

Polynomial fit: -4 -7 2 Y = - 4.7 x 10 * X - 6.05 x 10 *X

-0.30

50

100

150

200

250

300

350

400

δA A

⋅ 52.2 ppb ≈ 0.27 ppb

4x10

∆ν - average data point spacing g(ν) – reference spectra recorded for 300ppb normalized by:

-3

2x10

1x10

-3

∫ g (ν )dν = 1

-4

σ = 2.52 x 10

0 -1x10

A - area under the spectral line

-3

2057.5

2057.6

2057.7

2057.8

-1

Wavenumber [cm ]

QC laser power optimization

= 2.2 × 10 −6

-3

0

0

Points

OCS Sensor Architecture

C (1σ ) =

52.2 ppb OCS + N2 @ 60 torr

6x10

12

Theoretical detection limit:

-2

4x10

Absorption (reference 300 ppb OCS) -3

8x10

-0.20 -0.25

• Compact for clinical/hospital use • Dimensions: 5.35 in x 3.00 in x 1.50 in • Ethernet, Serial, JTAG access for control and read out • Flash memory for long term storage • PC independent operation • Up to 12.5 MSPS 12-bit ADC

Standard error in the best-fit coefficient1:

4x10

Absorption

Elevated COS concentrations in exhaled breath have been reported in lung transplants recipients suffering from acute rejection

0.00

Measured data points Linear fit

-3

6x10

Fit residual



0.10 0.05

DSP System Controller Card

yi = A ⋅ g (ν i )

-3

8x10

Calibration curve

Etalon fringe pattern

Nitric Oxide (NO): Inflammatory and immune responses (e.g. asthma) and vascular smooth muscle response (6-100 ppb) Ethylene (H2C=CH2): Oxidative stress, cancer Carbon Monoxide (CO): Smoking response, CO poisoning, vascular smooth muscle response (400-3000 ppb)

• •

Relative wavenumber [cm ]



May 16 – 21, 2004

Detection Sensitivity of QCL sensor

fast frequency scan

Breath analysis is a non invasive way of human disease detection e.g.: detector signal [a.u.]



San Francisco, California, USA

1

A.A. Kosterev et al., Applied Optics, Vol. 40, No.30, p. 5522

OCS Concentration Calibration of QCL Sensor

DSP Fast Data Acquisition Architecture

60 cm

Multipass Cell PM

PC

20

-3

-2x10

0.2

2057.5

-19

6.0x10

-19

4.0x10

-19

2.0x10

0.0 2020

2040

2060

2080

2100 -1

O CS 150 ppb CO2 5 %

2 0 5 7 .0

Absorption

P 24

P 46

2 0 5 7 .5

2 0 5 8 .0 -1

W a v e n u m b e r [c m ]

2 0 5 8 .5

2

Sample Reference

0

200

300

400

100

200

Subthreshold current 20

0

5x10

-3

2x10

400

500

250

Subthreshold Current

300

30 0

10

0

20

30

250 ppb

200 ppb

150 ppb

40

Reference Cell Control / Ready

Signal Conditioner

80

100

200

300

400

COCS= 8.4 ppb

2057.5

500

2057.6

-1

2057.7

Wavenumber [cm ] 1x10

-2

5x10

-3

CO2 in patient's breath Fit by the refernce spectrum of 5% CO2

-3

-3

SPI

ADC

120

Measurement No.

Signal Conditioner

SH B

SCI

ƒ Sample was taken from lung transplant patient suffering from bronchiolitis* ƒ Sampling was performed using chemically inert 1 liter tedlar sampling bags and analyzed within 2 hours after collection ƒ Spectrum was measured at a total pressure of 60 torr

100

200

300

400

500

Points

• Spectral noise level is inversely proportional to laser power • Modulation of laser bias produce fluctuations of laser power which limits the effective range of frequency scans • Effect is reduced by amplitude modulation of laser current pulses

CCO = 5.1 % 2

* The authors wish to thank Dr. Remzi Bag and Carolyn M. Paraguaya from Baylor College of Medicine, Houston, TX for supplying breath samples

0 2055.8

2055.9

-1

Wavenumber [cm ]

2056.0

Reference Detector Delay A

Sample Detector

SPI Interface

100 spectra averaged acquired within t = 0.4 s and fitted to 300 ppb OCS reference spectrum

OCS 50 ppb @ 60 torr 0

SH A

F2812 DSP

DAC

40

Signal Conditioner

300 ppb

OCS in patient's breath Fit by the reference spectrum of 50 ppb OCS

0

0

-5x10

300

200

50

OCS and CO2 Concentration Measurements in Exhaled Breath

Points

OCS 50 ppb @ 60 torr

150

1x10 1

500

0

100

1000 spectra averaged acquired within t = 4 s and fitted to 300 ppb OCS reference spectrum

3

0 100

50

-3

Points

-3

0

2

0

Points

2 0 5 6 .5

1

40

Subthreshold current 20

0

0 .0

3

40

0

Data points Linear fit

50

300 250 200 150 100 50 0

Optics

Reference Cell

SCI Interface

Thermoelectric Cooler

1

-3

2057.8

100

Laser peak power [a.u.]

Sample Reference

5x10

2057.7

1/cm

0 0

2

-5x10

P 10

-2

P 49

1.0x1 0

P 11

-2

P 26

-2

2.0x1 0

P 12

3.0x1 0

P 48

4.0x1 0

-2

P 13

-2

P 14

A b s o r b tio n

wavenumber [cm ]

5.0x1 0

2057.8

3

Detector signal [V]

R - branch

[mA]

-1

-2

Line strength [cm /molec. cm ]

-18

1.0x10

2057.6

Reference concentration [ppb]

QCL Pulse Amplitude Modulation

ƒ Line intensity: 7.49·10-19 cm-1/molecule⋅cm-2 ƒ Minimal spectral interference by nearby CO2 and H2O absorption lines ƒ Availability of a CO2 line within the fast tuning range of the QCL for ventilation monitoring simultaneously with an OCS measurement

-18

P - branch

2057.7 -1

1.2x10

-19

2057.6

2057.5

QCL

Pulse trigger

0.00

150

QCL Pulser

Delay B

wavenumber [cm ]

OCS ro-vibrational Spectrum

8.0x10

0.4

10 5

RB – reference beam M – mirror BS – beam splitter PM – off-axis parabolic mirror

QCL – quantum cascade laser chip LH - laser housing CL – collimating lens SB – sample beam

0.6 15

Scattering of the concentration measurement: σ = 1.2 ppb

100 ppb

-3

2x10

0

DAQ CARD NI 6024E

PCMCIA

0.8

200

300 ppb 200 ppb 100 ppb 50 ppb 30 ppb

0.02

75 ppb

25

250

50 ppb

4x10

0.27 ppb ⋅ 1000 100 = 0.85 ppb

0.04

300

40 ppb

-3

t

FUNCTION GENERATOR

Laser line width: FWHM: ~0.04 cm-1

OCS P(11) pressure =1.2 torr COCS ≈ 1.6 ppm

6x10

[mA]

TRIGGER PULSE GENERATOR

SAMPLE

t REFERENCE

PULSE AMPLITUDE CONTROL

REFERENCE SAMPLE

Absorption

GATE GENERATOR

APEAK/σNOISE

absorption

-3

TRACK & HOLD’s

t

PULSED QCL DRIVER

Theoretical sensitivity:

30 ppb

-3

8x10

MCT DETECTOR

absorption

M M

Absorption

CL

Absorption

QCL

Measured concentration [ppb]

-2

1x10

BS LH

Calibration curve

laser peak power [a.u.]: 2.5 1.5 0.75

Concentration [ppb]

M

RB

FWHMOCS/FWHMQCL

SB

Detector signal [V]

MULTIPASS HERRIOTT CELL

Pulse Trigger

Serial Port Transceiver

PC Workstation

Signal Conditioner

Analog Delay Generator

Conclusions • Sensitivity of current QC laser based OCS sensor is at the ~ 1.2 ppb level • Sensor has capability of simultaneous COS and CO2 concentration measurements • Conversion to autonomous, compact, high speed processing and control electronics demonstrated • Studies to date of exhaled breath in lung transplant recipients show low OCS content which suggests the following: ƒ Further increase of sensor sensitivity ƒ Develop chemical “amplification” methodology ƒ Further optimization of breath collection technique