Re: dye laser details

From: Emmanuel Gustin (Emmanuel.Gustin@skynet.be)
Date: Fri Jan 18 2002 - 19:29:33 EST


> Could someone explain how the tuneable part of the dye laser works ?
> What is being altered such that different wavelengths result ?

Basically, what is being altered is the gain at a given wavelength,
but there are several ways to do that, and maybe it is less than
obvious why that actually changes the laser wavelength.

So allow me to start with the fundamentals. To get laser activity
you must achieve a population inversion in your system; this is
equally true for dye molecules, ionised gas atoms, or impurities
in a ruby, saphhire or YAG crystal, and other lasers. What you
need is an efficient fluorescence transition which can be 'pumped'
high enough so that there are more molecules in the excited state
than in the ground state.

In a gas laser the different wavelengths are different transitions,
all quite narrow, which can be pumped high enough to achieve
inversion. In a dye laser the different wavelengths are found
within one broad transition; generally corresponding to its
fluorescence band. (Rhodamine 6G, for example, can be made
to lase between approx. 570 and 640 nm.) That means that
the possible laser wavelengths in a dye laser are in direct
competition with each other; which in a gas laser is not necessarily
the case. With suitable mirrors you can run a gas laser at several
wavelengths simultaneously, but in a dye laser there is
a winner-takes-all competition because they all exploit the same
excited population.

Remember LASER = 'light amplification by stimulated emission
of radiation'. Stimulated emission basically makes copies
of the fotons you put in, so your amplification is proportional
to your input, with a factor that is set by the gain of the dye
and the laser cavity. And all wavelengths in a dye laser 'eat'
from the same limited supply of fotons. Thus the wavelength
with the best gain takes the largest share of new fotons, grows,
and then takes an even larger share; wavelengths with a smaller
gain lose at every pass until they die out. (The rich get richer,
the poor get poorer.) The process is very nonlinear, so a small
difference in gain will tune the laser to a single wavelength.
To select laser wavelength, dye laser cavities contain an optical
element which has just a little more transmission at one
wavelength than at all others, and is tunable.

There are several approaches. A simple one is a thin element,
a so-called tuning etalon, with a slight wedge-shape. This is
positioned on a slide and acts as a very simple interference
filter. The transmission peak of an interference filter depends
on its thickness, so it varies along the length of the wedge;
by putting the right part of it in the beam you can change
the laser wavelength. A potential disadvantage of an etalon is
that it can still several very closely-spaced lines to operate
simultaneously, so often several are used in combination.

Very popular are birefringent filters, usually in round rotating
holders that contain one or more stacked thin plates of birefringent
material. Generally, the speed of light in a material is dependent
on  wavelength; in a birefringent material the speed of light is
also different for light polarised along two different axes of the
crystal. The result is that by rotating the filter to a certain angle,
light at wavelengths other than the selected one is given an
elliptical polarisation and thus part of it is 'coupled out' from
the polarisation that is lasing. (In most gas and dye lasers the
laser light has vertical polarisation, because windows and
filters are set at the Brewster angle from the vertical, so that
this polarisation has a higher gain.)

In practice it is not simple to design a suitable filter for a laser.
Bandwidth and line stability also depend on the choice of the filter.
Narrow-band filters produce stable and narrow lines but often with
a loss of efficiency (sometimes they are also used on gas lasers
to get very narrow lines) while wide filters can be more efficient
but sometimes allow the laser line to drift a few nanometers.

Emmanuel Gustin



This archive was generated by hypermail 2b29 : Wed Apr 03 2002 - 11:59:21 EST