microFTS

Hugh Mortimer, Rutherford Appleton Laboratory (STFC)

Competition sponsored by:

Spectroscopy: What is it?

“The branch of science concerned with the investigation and measurement

of spectra produced when matter interacts with or emits electromagnetic

radiation.”

• Electromagnetic radiation is passed through or reflected from a sample.

• Some of the radiation is absorbed and some is transmitted or reflected by

the sample.

• Spectrometers separate (think glass prism) and measure the intensity of

the emerging radiation as a function of its wavelength – to enable

analysis of the optical, chemical, and physical properties of the

sample.

Spectrometers• Nowadays, two of the most commonly used spectrometer designs are

Michelson Fourier-transform spectrometers (FT-IRs) (high signal-to-

noise, but bulky and sensitive to disturbance) and miniature dispersive

spectrometers (small & rugged, but low signal-to-noise), which are suitable

to very different applications.

• Dispersive spectrometers disperse the incoming light into a frequency

spectrum, which is directly recorded by a detector array.

• Fourier-transform spectrometers split and recombine the incoming light to

generate an interference pattern which is recorded - the frequency spectrum

is subsequently generated by taking the Fourier-transform of the

interference pattern.

Over the next few slides I’ll explain the operating principles of each of

these spectrometer designs and compare them to our microFTS, a new

kind of Fourier-transform spectrometer.

Dispersive Spectrometers

ENTRANCE SLIT

MIRROR 1

GRATING

MIRROR 2

DETECTOR ARRAY

Light enters the spectrometer through a slit and

is reflected from a collimating mirror (MIRROR 1).

The collimated light is reflected from a

diffraction grating to disperse the light into

its separate wavelength components.

The dispersed light is reflected from a second

mirror (MIRROR 2) and refocused onto a detector

array.

The frequency spectrum of the incident radiation

is recorded directly by the detector array.

How they work

wavelength or frequency

inte

nsity

SPECTRUMOptical Bench

(Un-Folded Czerny-Turner)

MIRROR 1

(FIXED)

MIRROR 2

(MOVING)

SINGLE POINT DETECTOR

BEAMSPLITTER

Michelson FT-IR

Light enters the spectrometer and is split into two

perpendicular beams at the beamsplitter.

One beam of light is reflected from a static mirror

(MIRROR 1) and the other beam from a moving mirror

(MIRROR 2). The moving mirror introduces a time

delay to the second beam.

The two beams are brought back together at the

beamsplitter and interfere to form an interference

pattern that is temporally modulated.

Each data point in the interference pattern

corresponds to a time and thus a position of the

scanning mirror – the range and speed of the

scanning mirror determines the number of data

points in the interference pattern.

Taking the Fourier-transform (FT) of the interference

pattern yields the frequency spectrum of the incident

radiation.

FOURIER TRANSFORM

ALGORITHM

How it works

time

inte

nsity

wavelength or frequency

inte

nsity

INTERFERENCE PATTERN

SPECTRUM

Optical Bench

microFTS

MIRROR 1

DETECTOR ARRAY

BEAMSPLITTER

MIRROR 2

How it works

Light enters the spectrometer and is split into

two beams at the beamsplitter.

One beam of light is transmitted by the

beamsplitter and reflected from MIRROR

2, then MIRROR 1, before being again

transmitted by the beamsplitter and striking

the detector array. The other beam of light is

reflected by the beamsplitter and reflected

from MIRROR 1, then MIRROR 2, before

reflecting from the beamsplitter and striking

the detector array. The two beams are

focussed by the curved mirrors to combine

and interfere at the detector array.

An optical path difference between the two

beams is introduced by the different

distances the two beams travel around the

set-up. The resulting spatially modulated

interference pattern is spread across a

detector array.

Taking the Fourier-transform (FT) of the

interference pattern yields the frequency

spectrum of the incident radiation.

FOURIER TRANSFORM

ALGORITHM

distance

inte

nsity

wavelength or frequency

inte

nsity

INTERFERENCE PATTERN

SPECTRUM

Optical Bench

Comparison TableInstrument Format Michelson Dispersive microFTS

Also Known As FT-IR or FT-NIR spectrometersGrating-based, diode array, or

CCD spectrometers

Static-Imaging Fourier

Transform Spectroscopy

(SIFTS)

Moving Parts? Yes No No

Relative Cost High Low Low

Data Acquisition SpeedSlow ~ 10s

(limited by mirror scan speed)

Fast ~ 100ms

(limited by detector read-out

rate)

Fast ~100ms

(limited by detector read-out

rate)

Signal-to-Noise Ratio /

SensitivityHigh Low Medium

Internal Wavelength

Calibration Possible?Yes No Yes

Size And Weight Large and heavy Compact and light Compact and light

Spectral Region Of

OperationMIR to NIR NIR , Vis and UV MIR, NIR, Vis and UV

Examples of Market

Leading Brands

Bruker, PerkinElmer,

ThermoFisher Scientific, ABB,

FOSS

Ocean Optics, Avantes,

ThermoFisher ScientificN/A

Basically, it’s the best of both worlds: Great signal (almost) like the bulky lab setups

we all know but fast, small & rugged like miniature dispersive spectrometers.

Early Prototype microFTS

A an early prototype

instrument, operating in the visible

spectral region.

The instrument is fibre-optic fed

and uses a off-the-shelf optical

components and a CCD detector

array.

35m

m

BEAMSPLITTER

MIRROR 2

MIRROR 1

DETECTOR ARRAY

FIBRE OPTIC INPUT

Early Performance Specifications

Note that these are indicative specifications, based on laboratory-measurements

i. The system will incorporate alternative detector technologies, such as linear PbSe (lead salt) arrays, 2d VOx

microbolometer arrays, and others.

ii. Wavelength range is dependent on beamsplitter material and detector type. For the IR model the ranges quoted

are for ZnSe or KBr beamsplitters.

iii. Instrument resolution is determined by the maximum optical frequency (or minimum wavelength) in the recorded

spectrum. The instrument resolution was measured as the FWHM of a laser source.

iv. Measurement of instrument stability was limited by the experimental set-up. It is anticipated that it will be an

order of magnitude better than the value quoted. A calibration laser can be incorporated into the instrument for

enhanced stability (equivalent to the use of a HeNe calibration laser in a Michelson-interferometer).

v. SNR measurements were made as the ratio of the central peak of a laser line to an average background noise

level.

Spectral Region UV-Vis IR

Detector(i) 2d CCD array Linear PZT pyroelectric array

Wavelength

Range(ii)200 to 1 050 nm

2 to 17 µm

(5 100 to 600 cm-1)

or

2 to 20+ µm

(7 000 to 350 cm-1)

Resolution(iii)

@ 200 nm

(50 000 cm-1)

< 1.2 nm

(300 cm-1)

@ 10 µm

(1 000 cm-1)

0.1 µm

(10 cm-1)

@ 500 nm

(20 000 cm-1)

< 2.9 nm

(120 cm-1)

@ 20 µm

(500 cm-1)

0.2 µm

(5 cm-1)

Stability(iv) Better than 1 part in 25 000 per ⁰CEquivalent to: 0.04 nm per ⁰C @ 1 000 nm (0.4 cm-1 per ⁰C at 10 000 cm-1)

SNR(v) 500:1 100:1

Size ~ 160 x 115 x 45 mm

Mass ~ 0.5 kg

How can you help?

We’ve developed what we think is a really cool instrument, a real step forward in spectrometer design. While it was developed for atmospheric measurements on Mars, we’d like to find out how it could be useful for us Earthlings.

• Do you have some thoughts on how to implement the proposed applications listed in the challenge page below?

• Can you think of any new applications?

• It’s a pretty modular setup – How could we modify the instrument to be even more useful or useful in a different setting?

Thanks a lot guys! Remember to use those marbles…