Professional Documents
Culture Documents
1. Introduction
The Fizeau wavemeter is one of the most promising devices
for measuring the wavelength of both c w and pulsed dye
lasers with an accuracy comparable to the linewidth produced
by most pulsed lasers. Most wavemeters for use with pulsed
lasers use a spatial filter with a metallic pinhole, which is
susceptible to damage caused by the high power density.
Diamond pinholes. which can withstand high powers. have
inherent drawbacks due to the translucence of the material
and to the limited range of available pinhole sizes (which are
rather large). The present design utilises a single-mode
optical fibre coupling which overcomes these difficulties as
well as the problem of input-beam alignment. The fibre
coupling is more convenient and easier to use than the spatial
pinhole as the input fibre can easily be aligned with the input
laser beam whereas the fixed pinhole design requires the laser
beam to be aligned with the wavemeter.
2. Principle of operation
The laser light entering the wavemeter is spatially filtered by a
single-mode fibre which creates a beam with a spherical
wavefront. This is made incident on an off-axis parabolic
mirror which produces a wide and well defined parallel beam
having a gaussian intensity profile. and directs it onto a wedge
consisting of two fused silica interferometer plates. After
refection from the wedge, the beam now consists of two
coherent parts almost parallel to each other, which produce a
Fizeau interference pattern; the fringes are registered by a
photodiode array interfaced to a computer. The wavelength
of the incident laser light is determined from the spacing and
phase of the fringe pattern. The approximate value of the
wavelength ,Iis (,obtained from the expression (Morris ef a1
1984)
I,,, = zpa (1)
where p is the fringe spacing and a is the wedge angle. The
optical path difference (OPD) A d , corresponding to a fringe
minimum chosen for reference, is calculated from the wedge
spacing and the wedge angle. and yields the interference
order :V of this fringe (Lu and Wu 1984): electronic data acquisition system is shown schematically in
figure 2 . The optical signal registered by the Reticon
RL1021H monolithic self-scanning linear photodiode array is
where N is rounded off to the nearest integer. The final value converted into a series of electric pulses, which are sampled
of the wavelength 2 can then be determined with an accuracy and held by a sample-and-hold amplifier. and are then digi-
limited by the error in the measurement of the wedge spacing: tised with 0.8,us conversion speed by an %bit A D C , loaded
into 1 k memory buffer and finally transferred to an IBM XT
computer. The fringe data processing and wavelength calcula-
tion are performed using the algorithm of Snyder (1980) in a
Turbo Pascal program at a rate of about 1 Hz. The power
3. Construction details requirement is about 0.01 mW for a TEM,),,c w laser source and
The optical layout of the wavemeter, which is shown in figure about 2,uJ energy per pulse for a pulsed laser source.
1, is similar to that proposed by Morris et a1 (1984). All At these energies one can expect the generation of
optical components, except the precision fibre coupler. are Stokes-shifted wavelengths in the fibre, which can lead to
mounted on a 40 mm thick aluminium plate placed inside an unexpectedly large errors in the calculated wavelengths. In
aluminium vacuum chamber maintained at 2 Pa. Laser light is order to estimate and minimise these errors we reduced the
spatially filtered by focusing it onto the 4 p m core of the F-SV power entering the fibre optic cable by a factor of about 50
Newport single-mode fibre and is formed into a parallel beam and allowed the photodiode array to accumulate and in-
by reflection from the surface of a 200" focal length 21" tegrate the signal of 50 pulses before each scan of the
off-axis parabolic mirror. The parallel beam is reflected from array. This procedure gives an average wavelength of 50 laser
the two inner uncoated Fizeau wedge surfaces and the result- pulses but with a negligibly small contribution from Stokes-
ing interfering components are focused by a cylindrical lens of shifted wavelengths created at higher power densities.
50 mm focal length onto a 25 mm long photodiode array of
1024 elements (Reticon RL1024H) placed at the zero-shear
position (Snyder 1981).
dF
797
\I/ Kedzierski er a1 J. Phys. E: Sci. Instrum. 21 (1988) 798-804. Printed in the UK
References
Lu D and W u Z 1984 A method for calculating the wedge
spacing of a Fizeau wavelength meter
Opt. Laser Technol. 16 206
Morris M B, McIlrath T J and Snyder J J 1984 Fizeau
wavemeter for pulsed laser wavelength measurement
Appl. Opt. 23 3862
A new class of multiple
Rogers J R 1982 Fringe shifts in multiple-beam Fizeau dispersion grating
interferometry
J . Opt. Soc. Am. 72 638
spectrometers
Snyder J J 1980 Algorithm for fast digital analysis of V Mazzacuratii, P Benassit and G RuoccoS
interference fringes t Dipartimento di Fisica, Universita’ degli studi ‘La
Appl. Opt. 19 1223 Sapienza’, Piazzale A Moro 2, 00185 Roma, Italy
Snyder J .J 1981 Fizeau wavemeter $ Dipartimento di Fisica, Universita’ di L’Aquila, Piazza
Proc. Soc. Photo-Opt. Instrum. Eng. 288 258 Annunziata 1, 68100 L’Aquila, Italy
1. Introduction
Inelastic light-scattering experiments requiring a resolution
better than a tenth of the wavenumber are generally per-
formed using Fabry-Perot (FP) interferometers or, in a few
cases, by instruments based on the coherent anti-Stokes
resonant scattering technique (CARS) (Lauebereau and Kaiser
1978). In current literature Brillouin spectroscopy has been
usually performed with single- or multi-pass FP interfer-
ometers, the main limit of which is represented by the low
ratio between the free spectral range (FSR) and the resolution
(6). Indeed the typical finesse ( F = F s R / value
~) ranges
between 80 and 100.
In particular, when we are interested in determining low-
frequency spectral shapes, a limited finesse can be a severe
shortcoming because of the superposition of adjacent spectral
orders. To avoid this problem the FSR needs to be increased,
thereby losing resolution or. alternatively, long and unreli-
able extrapolation procedures may be adopted (Mazzacurati
and Benassi 1987). Furthermore, a particularly serious
problem arises in analysing weak and smooth quasi-elastic
spectra, when high-intensity sharp lines as a Brillouin doublet
or an intense spurious elastic component are also present in
the spectrum. In this case the high contrast required is
achieved in Fabry-Perot interferometers using multi-pass
operation. This generally implies a loss of stability and lumi-
nosity, while the difficulties due to the low finesse are still
present, actually implying a limitation to the analysis of low-
frequency polarised and depolarised spectra in dense gases,
liquids, viscoelastic materials, anharmonic crystals and disor-
dered systems.
In this paper we present the first experimental results
obtained with a new grating multiple spectrometer which has
been designed and realised. in collaboration with the techni-
cal staff of the SOPRA company?, in order to simultaneously
fulfil the following requirements.