quantum cascade laser based trace gas...
TRANSCRIPT
1
QUANTUM CASCADE LASER BASED TRACE GAS SENSORS
Christina YoungResearch Advisor: Prof. Dr. Boris Mizaikoff
Georgia Institute of TechnologySchool of Chemistry and Biochemistry
Atlanta, GA, USA
Boris [email protected] http://asl.chemistry.gatech.edu/
Christina Young [email protected]
2Why Mid-IR Technology for Trace Gas Sensing?
Analytes absorb with unique molecular signatures
Selectivity inherently based on wavelength
Single-mode, tunable QCLs promise sensitivity and miniaturization
Can be used for simultaneous quantitative monitoring of different analytes in air for environmental detection, process monitoring, biomarkers in breath, etc.
FIR
20 µm
500 cm-1
2.5 µm
4000 cm-1
800 - 400 nm
Maximum of black bodyemission at 300 KFINGERPRINT
(500 - 1300 cm )-1
NIR VISto UV
MID-INFRARED BAND
Boris [email protected] http://asl.chemistry.gatech.edu/
Christina Young [email protected]
3Outline
Quantum Cascade Laser Based Trace Gas Sensors
EC-QCL HWG Multianalyte Detection
Wavelength Selection by Cavity Length Variation
Boris [email protected] http://asl.chemistry.gatech.edu/
Christina Young [email protected]
EC = external cavity, HWG = hollow waveguide, QCL = quantum cascade laser
4Quantum Cascade Lasers (QCL) for Trace Gas SensingFundamentals
one period
active region
injector
minigap
probability density
miniband
Intersubband transitions between quantized conduction band states
Band Energy Diagram• Active region: Design of energylevels 3 and 2 to achieve:
• light amplification• desired laser frequency ν=(E3-E2)/h
• Injector: Supplies electrons
Young, C. et. al., Sensors and Actuators B: Chemical, 140(1), pp. 24-28.Boris [email protected] http://asl.chemistry.gatech.edu/
Christina Young [email protected]
5Widely Tunable EC-QCLs for Spectroscopic ApplicationsFundamental Principles
EL Characteristics
Cavity Response
Output Characteristics
EC-QCL
QCL100% HRcoating
95% HRcoating
λ 1λ 2λ 3
Location of the resonant peak shifts according to the angle of the grating.
Output Lens
QCL Gain Medium
Cavity Lens Cavity Filter
Scheme Courtesy of Daylight Solutions, Inc.• QCL tuning options: current or
temperature (few cm-1), external cavity
• EC provides ~ +/- 5% cm-1 tuning range of the central emission frequency
Young, C. et. al., Sensors and Actuators B: Chemical, 140(1), pp. 24-28.Boris [email protected] http://asl.chemistry.gatech.edu/
Christina Young [email protected]
6EC-QCL Emission CharacterizationCentral emission frequency of 1258 cm-1
Operating Conditions: Temp: 0 C, Pulse Width: 0.50 μsec, Frequency: 100.0 kHz, Duty Cycle: 5%, I: 1500 mA
EC-QCLMCT detector
BrukerIFS 66 FT-IR
Young, C. et. al., Sensors and Actuators B: Chemical, 140(1), pp. 24-28.Boris [email protected] http://asl.chemistry.gatech.edu/
Christina Young [email protected]
7EC-QCL Trace Gas Sensor Experimental SetupIndividual analytes measured
EC-QCL
Reference
Sample
Glass Stir Bar
Exponential Dilution Flask
Analyte InjectionNitrogen Carrier Gas
Gas Flow to SensorC = C0e - αt
WhereC = final concentrationC0 = initial concentrationα = flowrate/vol of EDFt = time elapsed between initial absorption and
signal regeneration
• Ethyl chloride at 1287.25 cm-1
• Dichloromethane at 1262 cm-1
• Trichloromethane at 1220 cm-1
Exponential dilution performed for three analytes to derive limit of detection (LOD):
Charlton, C., et. al., Appl. Phys. Lett. 86, 194102 (2005);Lovelock, J., Anal. Chem. 33, 163 (1961); Young, C.., et. al., Sens. Act. B. (2009)
Boris [email protected] http://asl.chemistry.gatech.edu/
Christina Young [email protected]
8Univariate ResultsEC-QCL precisely tuned to the Q-branch of CH2 wag vibrational mode for each analyte
50 60 70 80 901.80
1.95
2.10
2.25
2.40
2.55
Nor
mal
ized
Sig
nal (
V)
Time (min)30 35 40 45 50 55 60
0.14
0.16
0.18
0.20
0.22
Nor
mal
ized
Sig
nal (
V)
Time (min)40 50 60 70 80
2.3
2.4
2.5
2.6
2.7
2.8
2.9
3.0
Nor
mal
ized
Sig
nal (
V)
Time (min)
LOD
ClCl
= noise level
Ethyl ChlorideCl
LOD
13 ppb
Trichloromethane
LOD
Cl
ClCl
Variance
LODAssignment
Chemical Structure
Analyte
5 ppb
υ 7
7 ppm
5 ppm
11 ppb
15 ppb
Dichloromethane
υ 9 υ 4
Young, C. et. al., Sensors and Actuators B: Chemical, 140(1), pp. 24-28.Boris [email protected] http://asl.chemistry.gatech.edu/
Christina Young [email protected]
9Quantitative Measurement of An Analyte in MixturePartial Least Squares (PLS) model based on 11 training set mixtures
Cross Validation, 7 Latent Variables
35 40 45 50 55 6030
35
40
45
50
55
60
65
Concentration Measured
Con
cent
ratio
n P
redi
cted
Concentration Measured (ppm)
Con
cent
ratio
n Pr
edic
ted
(ppm
)
Quasi-unknownTraining Set Standards
ClCl
R2 = 0.977
1215 1230 1245 1260 1275 1290
1.0
1.2
1.4
1.6
1.8
2.0
2.2
Abs
orba
nce
(a.u
.)
Wavenumbers (cm-1)
Dichloromethane
Trichloromethane
Ethyl Chloride
EC-QCL HWG gas spectrum, 1 cm-1 resolution PLS Model Validation
(MATLAB (Mathworks ©)
Young, C. et. al., Sensors and Actuators B: Chemical, 140(1), pp. 24-28.Boris [email protected] http://asl.chemistry.gatech.edu/
Christina Young [email protected]
10Outline
Quantum Cascade Laser Based Trace Gas Sensors
EC-QCL HWG Multianalyte Detection
Wavelength Selection by Cavity Length Variation
Boris [email protected] http://asl.chemistry.gatech.edu/
Christina Young [email protected]
EC = external cavity, HWG = hollow waveguide, QCL = quantum cascade laser
11Mounting and Measuring the QCL
FP-QCL chip
Cleave in the plane of the crystal lattice with diamond knife and microscope
Cleaved single facet
Prepare sub-mount by gluing a insulated gold pad onto a copper block
Bond gold wires from gold pad to QCL chip to transport current
Apply In0.97Ag0.3 to Cu block and align laser on top with micro-tweezers. Keep aligning as In0.97Ag0.3 is heated from 130 C to 180 C and back to 130 C, chemically bonding chip to block
Cu block with Au pad
Laser sub-mount before bonding
Laser sub-mount after bonding
CryostatFT-IR
XYZ positioner
Laser Sub-Mount
LN2
Collaboration with Mid-Infrared Photonics Group, MIRTHE, Princeton University
Photo courtesy of C. Gmachl
Young, C. et. al., Applied Physics Letters, 94(9), pp. 091109.Boris [email protected] http://asl.chemistry.gatech.edu/
Christina Young [email protected]
12QCL Emission Tuning by Δ Cavity LengthTheoretical
QCL Band Energy Diagram
Theoretical Calculations
L
Young, C. et. al. Applied Physics Letters, 94, 091109, (2009).Boris [email protected] http://asl.chemistry.gatech.edu/
Christina Young [email protected]
13Towards Precise Overlap of QCL and Analyte Absorption
Frequency emission shifts with respect to cavity length due to changes in n and E
Young, C. et. al. Applied Physics Letters, 94, 091109, (2009).Boris [email protected] http://asl.chemistry.gatech.edu/
Christina Young [email protected]
14
Experimental Results Confirm Theoretical Calculations Cavity Length Dependence: Δ Efield Causes Δ λem
Shorter cavity length:- Increase in Efield (Vth)- Increase in λem
Emission Frequency Vs. Vth
Vth Vs. Cavity Length
Young, C. et. al. Applied Physics Letters, 94, 091109, (2009).Boris [email protected] http://asl.chemistry.gatech.edu/
Christina Young [email protected]
15Acknowledgments
• Applied Sensors Laboratory, Georgia Tech, Atlanta, GA
• Daylight Solutions, Inc.
• Mid-IR Photonics Group at Princeton University
• Mark Disko, Andy Riley, John Szobota, John Martin and Neil Brons atExxonMobil Research and Engineering Company, Annandale, NJ
• Adtech Optics, Inc. for providing a507BH QCL
Many thanks to colleagues and co-workers at…
Boris [email protected] http://asl.chemistry.gatech.edu/
Christina Young [email protected]
16Questions?
Boris [email protected] http://asl.chemistry.gatech.edu/
Christina Young [email protected]