1. UDC 621.'375.9
LINE WIDTH OF A RUBY L A S E R W I T H A LIQUID SWITCH
S. I, Borovitskii, Yu. M. Gryaznov, and A. A. Chastov
Zhurnal Prikladnoi Spektroskopii, Vol. 5, No. 5, pp. 609-613, 1966
A Fabry-Perot interferometer is used in the wedge mode to examine the line width with a passive Q switch.
The width is found to fall from 0.2 cm -1 in free oscillation to 0.005 cm -1 in the single-pulse mode.
It is of interest to produce high-power laser radiation with the narrowest possible iine width. Q-switching provides
a considerabie increase in output power, but the line width is rather dependent on the method used. A rotating prism
gives a total width o f 0 . 6 - 1 . 5 c m -1, the pumping power then being the decisive factor [1]. A width of 0.02 c m -1 has
[2] been attained by the use of a crypmcyanin solution.
We have examined the spectrum of a ruby laser under various conditions with a liquid switch.
5 3 Z I q II 7 8 9
C.L
Fig. I. Block Diagram of the apparatus.
Experimental. Figure 1 shows the block diagram of the apparatus. The laser consists of a cavity of semiconfocal
type enclosing the ruby l a n d the liquid cell 2, the cavity being formed by the spherical mirror 3 (radius of curvature 2000
mm) and the set of p l a n e - p a r a l l e l plates 4 at the focal point of the mirror, The ruby rod was 120 m m long and 10 m m
in diameter, with p l a n e - p a r a l l e l ends and polished sides. The pumping was provided by two IFP-2000 lamps, pumping
energy 2 to 4 kJ. The filter was g a l l i u m phthalocyanin chloride in quinoline, which has its peak absorption at 6925
and a molar extinction coefficient of 2.4 • 105 at kma x. This solution was used in a 5 m m glass cell between the ruby
and the spherical mirror, the latter having a reflection coefficient of about 97% at 6943 ,~. This cell was placed at a
small angle to the axis. Tests were done with solutions whose transmissions (as measured with an SF-4 spectrophotometer)
were 25, 40, and 50%.
The number o f giant flashes was determined with the photomultiplier 5 (type FEU-28) and the oscilloscope 6. The
interferometer 8 was a standard Fabry-Perot operating in the wedge mode [8]; the plates had multilayer d i e l e c t r i c c o a t -
ings with R = 94~ at 6948 A. Tests were done with plate separations h = 10 and 30 m m . The resolution is [3] defined by
8~ --
1-R (1)
2a h 1/r~
and was 3.3 • 10 -3 c m -1 for h = 30 m m . The interferometer was illuminated by the parallel b e a m formed by the c o l l i -
m e t e r 7 taken from an OSK-3 o p t i c a l bench; parts l i a b l e to be damaged by the flashes were removed or replaced. The
output from the interferometer was recorded by a film placed in the c a m e r a 9.
It is very difficult to adjust the interferometer directly with a puised ruby laser, but a continuous-running laser
obviates this difficulty. The b e a m from the He-No laser 10 is taken via the b e a m splitter 11 and the stops 12. The
c o l l i m a t o r was adjusted to give roughly uniform intensity. The angle of the interferometer wedge was chosen to bring
rather more than one order of interference within the d i a m e t e r of the piates. The correct inclination of the plates to the
b e a m was determined from the change in the direction of motion of the fringes on turning the interferometer as a whole.
Adjustment of the length of the c o l l i m a t o r gave straight interference fringes. The pattern was tested for stability by a l -
tering the i n c l i n a t i o n of the laser b e a m within the limits allowed by the stops. Then the spectrum of the pulsed laser
was obtained on the ground glass when the b e a m passed through both stops.
445

2. Remits and Discussions. Free oscillation gave a total width of about 0.2 cm -1, as measured with h = 10 ram;
Fig. 2 shows the result and the result for Q switching with h = 30 mm. Free oscillation (Fig. 2a) led to overlap of orders, but
it is clear that there are many lines. Estimates of the frequency differences for the clearer parts indicate that these cor-
respond to modes differing in longitudinal parameters. Figure 2b shows that Q switching reduces the spectrum to a single
component, whose width (as measured with an MF-2 microphotometer) was 0.005 cm "~, which is close to the limit of
resolution of the interferometer. A similar spectral width for giant pulses has been reported [4] for Q switching by means
of a rotating prism, with mode selection by plane-parallel plates and a cell containing cryptocyanin. Figure 3 shows in-
terference patterns for this laser under various conditions.
Fig. 2. Interferograms for laser radiation: a) free oscillation;
b) with Q-swkching by phrhalocyanin (50% transmission,
pumping energy 2300 J).
A single component is seen in each order (Fig. 3a) in single-pulse working, so only one longitudinal mode is gener-
ated. More lines appear as the pumping energy is raised; in the two-pulse mode we get two lines (Fig. 31>), and we may
mppose that each pulse corresponds to its own line. The frequency shift of about 0.03 cm -1 between pulses (Fig. 31>)may
be due to change in optical length of the cavity caused by heating of the ruby. Three pulses occur at higher pumping
energies, and three lines are seen.
Fig. 3. Interferograms for pumping energies of: a) 2380 J; b) 3100 J; c) 3420 J (40%
transmission); d) 3600 J (26% transmission).
The number of lines does not always equal the number of pulses, e . g . , for the two-pulse mode of Fig. 3c, where
the pumping energy was about 10% above that for Fig. 3b, there are five lines. Here the heating effect is accompanied
by the occurrence of several axial modes separated by several times the interval between types of oscillation.
Figure &t shows the pattern in the single-pulse condition with a cell of 25% transmission; the pumping energy is
about 1.5 times that for Fig. 3a, which causes excitation of two modes whose longitudinal parameters differ by one unit,
while the transverse ones coincide, as is evident from the line doubling, which is clear in the right-hand line. The com-
ponents of the double line have a separation of 0.0043 cm "I. The frequency difference between two modes whose axial
parameters differ by unity is
Av = {2 [L -t- la (n~ - - 1 ) + l~ (n2 - - 1)l }-x, (2)
in which L is the distance between the mirrors; l, and n, are the length and refractive index of the rod, while 12 and n2
are the same for the cell. The result is Z~u = 0.0045 cm -t, which agrees with the result from Fig. 3d.
Different ruby rods gave no essentially different interference pattern; but the rods did differ considerably when a
plane-parallel cavity was used. Even the single-pulse condition gave several widely separated lines with certain rods.
Here crystal inhomogeneity distorts the cavity and produces large differences of Q between modes, whereas the crystal
produces only small perturbations in a spherical cavity [5].
446

3. If a single mode illuminates the interferometer unevenly, the lines should vary in intensity along their length and
as between orders (Fig. 2b and 3a). Uniform illumination overcomes these defects if a single mode is generated, but the
problem becomes complicated if there are many modes. The line intensities are not generally proportional to the ener-
gies in the various modes, so it is largely impossible to use a wedge interferometer when mode intensities are to be com-
pared.
The formula applies strictly only to a Fabry-Pemt interferometer and is usable for a wedge only when the addition-
al path difference between the extreme interfering beams (due to the wedge and to divergence of the beam) does not ex-
ceed k/2.
All the same, this system gives a simple interference pattern, which is easily interpreted: it may be used in ob-
taining time-resolved spectra in conjunction with an SFR unit.
We are indebted to O. L. Lebedev and V. K. Kolesnikova for assistance in this work.
REFERENCES
1. T. V. Gvaladze, L K. Krasyuk, P. P. Pashinin, A. V. Prokhindeev, and A. M. Prokhorov, ZhETF, 48, 106,
1965.
2. B. H. Softer, J. Appl. Phys., 85, 2551, 1964.
3. F. A. Korotev, Vestnik Mosk. Univ., set. mekh., m a t . , astron., fiz., khim., no. 7, 101, 1953.
4. F. J. McClung, and D. Weiner, IEEE J. Quant. Electron., 1, no. 2, 94, 1965.
5. A. M. Leontovich and A. P. Veduta, ZhETF, 46, 71, 1964.
6. G. Bret, and F. Gires, Appl. Phys. Letters, 4, no. 10, 175, 1964.
28 December 1965
447