Longwave Infrared LWIR Coded Aperture Dispersive Spectrometer

1
Longwave Infrared (LWIR) Coded Aperture
Dispersive Spectrometer
C. A. Fernandez, B.D. Guenther, M. E. Gehm, and
D. J. Brady Duke University Fitzpatrick Institute
for Photonics and Department of Electrical and
Computer Engineering
M.E. Sullivan Centice Corp, 215 Southport
Dr, Suite 1000, Morrisville, NC 27560
System Design
Introduction and Motivation
Optical Design
Device Specifications
Family of Aperture-coded Spectrometers
Spectrometers in the LWIR
We have been involved in the development of
various coded aperture spectrometers. Our systems
replace the input slit with a mask containing a
well-defined code.

• Traditionally, Fourier transform (FTIR) based
  spectrometers are known for their high throughput
  (Jacquinot) and multiplex (Felgett) advantages.
  Fourier transform systems face mechanical
  challenges.
• Fourier-transform systems contain movable parts
• Can require long acquisitions times to record
  maximum signal strength
• High precision FTIR instruments are costly
• We have designed and fabricated a static
  aperture-coded, dispersive longwave infrared
  spectrometer with a spectral band of 9µm-11µm
  that uses a microbolometer array at the detector
  plane.
• Our dispersive spectrometer is static
  mechanically robust.
• Mask exhibits both throughput and multiplex
  advantages for the dispersive-based spectrometer.
• Low cost system

• Coded aperture at input
• Based on Hadamard code N12
• 3.54mm x 1.08mm rectangular aperture
• Pattern electroformed onto 11mm nickel disc 50µm
  thick
• Microbolometer array
• Detects change in heat flux through internal
  resistor
• Uncooled detector array
• Low resolution
• Low fidelity
• 160 x 120 pixels (30µm pitch)
• Spectral response 7µm-14µm
• Custom Optical Design
• f/2 system

L2
L1
MIRROR
Focal Plane
Slit vs. Mask
MASK

• Throughput is limited by aperture size
• Widening slit sacrifices spectral resolution
• Mask (coded aperture) is a regular pattern of
  many, small transmissive and opaque openings
• Mask yields high spectral resolution and high
  throughput

FOCAL PLANE
Dispersive Element
GRATING
L3
L4
Device Layout
Coded aperture High throughput, high resolution.
Narror Slit Poor throughput, high resolution
Wide Slit High throughput, poor resolution.
Experimental Results
Mechanical Design
Spectral Reconstructions
SNR Comparison between Slit and Mask
SNRMASK
SNRMASK
SNRMASK
1.69
3.49
1.82
SNRSLIT
SNRSLIT
SNRSLIT
Spectral Reconstruction
Slit
Mask
IMAGE AT DETECTOR
Description of Experiment

• We have performed a proof of concept experiment
  by recording the spectral emission from a tunable
  CO2 laser.
• Reconstructions from a slit were compared to that
  from the mask
• The mask feature size was equal to the width of
  the slit.
• The height of the slit was comparable to that of
  the mask.
• Signal-to-noise (SNR) measurements were compared
  for both slit and mask at various wavelengths.
• SNR is obtained by taking the ratio of the
  spectral peak and the standard deviation of the
  spectral noise (noise floor)
• Ratio of deviation of the pixel value from the
  mean was taken for a single background image, as
  well as, 100 background images.
• For all pixels in a single background image the
  deviation was about 14.49.
• Averaging 100 background images yielded a
  deviation between 0.01 and 0.06.

Emission Spectral Calibration
SHUFFLED HADAMARD N12
BINNING
Emission Spectral Calibration

• A tunable CO2 laser was used for calibration.
• Tunable, coherent source.
• Provides a well characterized spectral feature.
• Data at various wavelengths are saved for use in
  the calibration of the x-axis in future
  reconstructions.

BINNED DATA
NNLS ESTIMATES
BINNING
Conclusions
Experimental Overview
RECONSTRUCTION
FTIR vs. Coded Aperture Overview

• We have built a static coded aperture LWIR
  spectrometer prototype that can reconstruct a
  coherent source.
• We have modeled the background noise
  characteristics of our optical system.
• Finally, we have also compared the performance of
  a slit to that of a coded aperture at the input.
• With a N12 Hadmard mask, an improved SNR factors
  are measured over the slit.
• While there have been shortcomings with this
  device, it is the first aperture-coded dispersive
  system to use a microbolometer at the focal plane.

• FTIR systems have been traditionally used to
  examine emission spectra in the longwave infrared
  region. High cost and precision are some
  drawbacks associated with these systems.
• Use of a static coded-aperture dispersive system
  obtains comparable throughput from the sample.
• Making the system static solves the problem of
  high precision required for FTIR systems for
  resolution constraints.
• In general, dispersive systems have lower costs
  associated with them unlike FTIR systems.