Transcript Poster template: conference - A0 portrait

Demonstrating the detection and measurement of dissolved organic carbon in water using a new UV miniaturised Fourier Transform
Ali Hussain(1), Paul Scholefield(2), Hugh Spectrometer
Mortimer(1), Ed Tipping(2), Jessica Adams(2), and Don Monteith(2)
(1) RAL Space, STFC Rutherford Appleton Laboratory, Harwell, U.K. ([email protected])
(2) Lancaster Environment Centre, Centre for Ecology and Hydrology, Lancaster, U.K. ([email protected]),
Optical Configuration
Introduction
What and Why?
• DOC is a group of naturally occurring organic compounds found in catchments with peaty headwaters
• DOC removed before chlorination avoid production of toxic trihalomethanes
• EU legislation requires DOC monitored regularly and reported
Problem
• Current instruments are expensive,
Costs?
bulky, heavy, and power hungry
• DOC is gradually increasing with time
• Increased water treatment costs
Solution
• Companies need to make greater savings
• Develop prototype portable low
• Cost efficient monitoring required
cost low power spectrometer field
• Efficiency savings water treatment costs
deployable to measure DOC
Internal Optics
• Instrument uses 4 components: 2 concave mirrors, a beam splitter, a detector array
• Light entered is split, passes through a common path, and recombined at the focal
point onto a detector producing an interference pattern
• Mathematical transformation of the interference pattern extracts the light source’s
spectrum
Measurements
Spectrometer performance
Spectrometer is sensitive to measure fingerprint grease on the polished optical surface of the cuvette (left) and 2 stage
optical cleaning was used to return back to the original state. Known peaks and troughs from Potassium dichromate
were measured (right) and peaks below 300nm were harder to identify.
The Spectrometer
Design
Basic Attributes
• Novel, solid state Fourier Transform (FT)
spectrometer with compact common optical
path interferometry for stability against
vibration, humidity and temperature.
Large numerical aperture – maximum
signal for low light applications
• External and internal UV grade optics,
windowless linear detector array, Xenon flash
lamp, and a flow cell in the optical path
measuring UV absorption in fluids.
• On-board power (solar charging ability)
supplies flash lamp and dual wavefunction
generator to simultaneously trigger detector
and source with a delay.
Fast acquisition – all interferogram points
measured simultaneously, measurement
time detector limited
No moving parts – immune to vibration
and no maintenance required
Compact
and
lightweight
–
field
deployment suitable and integration onto
other instruments
External Optics
•
•
•
•
•
•
•
•
•
Xenon flash lamp (1mm arc, 2W, 1E9 pulses)
3mm iris reducing stray light into collimation lens
UV hot mirror blocks NIR radiation λ>750nm
½ inch diameter lens for light collimation (f~11mm)
Space and holder UV fused silica cuvette/flow cell
½ inch diameter lens for light focusing (f~19mm)
Variable iris spatial filter scattered light and adjusts input intensity
1 inch collimation lens produce top hat intensity beam profile into the spectrometer
Two neutral density filters reduce light pulse intensity by ~316
ND1.5
ND1.0
f 50mm
iris 2mm
DOC measurement
Water samples collected from known locations were filtered and showed low (left) to high (right) UV absorption
levels. The DOC was calculated using a two component model that separates strong and weak UV absorption as a
function of wavelength[1]. We used wavelengths 244nm, 350nm, modulated at 60Hz and 25s averaging time.
UV hot mirror
UV cuvette
iris 3mm
f 11mm
f 19mm
arc 1mm
Instrument Calibration and Corrections
Chirp Correction
UV Laser diode used to correct for chirp
along pixel array. Measured chirped and
de-chirped
interferograms
(left).
Corresponding chirped and de-chirped
Fourier transforms (right)
Spectral Calibration
Known narrowband (fwhm 10nm) filters
used to calibrate the spectral axis to a
few nm. Measured interferograms from
filters (left) and corresponding Fourier
transforms (right)
Xenon flash lamp
L12336-03
Wait for Noise
Averaging time
• Noise (error) as 1 standard deviation in the spectral signal for modulation
frequencies 10, 60, and 110Hz for 0.3-10s (left) and 3-100s (right)
• Entire 250-700nm spectrum used for single standard deviation value
• Best waiting time
30s for low noise
signal acquisition
Summary
Main outcomes
• Developed a prototype portable spectrometer for field deployment
measurement of DOC varying concentrations 2-15mg/L
• Demonstrated spectrometer calibration, chirp and cold pixel removal effect
on measured spectra
• Demonstrated for low noise applications optimum modulation frequency
60Hz, averaging time of 25s
• Measured spectra from Potassium dichromate and fingerprint grease on
sample surface
£$
Modulation frequency
Defect Pixel Correction
Identifying cold pixels and using software
signal averaging to eliminate unwanted
sharp dips in the interferogram signal
(left) introducing ringing in the Fourier
transform spectra (right)
[1] Heather T. Carter, Edward Tipping, Jean-Francois Koprivnjak, Mathew P. Miller, Brenda Cookson, John Hamilton-Taylor (2012). Freshwater DOM quantity and quality from a
two-component model of UV absorbance. Water Research, Vol. 46 (14), 4532-4542.
• Comparing standard deviation at all frequencies up to 300s averaging time
• Higher noise at 10Hz modulation, no gain modulating beyond 60Hz
• Increase flash lamp lifetime, reduce power consumption
Technology comparison
• Table below shows a quick comparison with existing instruments on the
market for measuring DOC in aquatic systems
Current instruments
Our lab prototype
Single wavelengths (λA, λB, λC)
Entire spectrum (λ250 – λ700)
Heavy (>20kg)
Power hungry (>20W)
Bulky system
Poor in low light levels
Light (<4kg)
Low power (<3W)
Portable unit
Independent of light levels