SOME POTENTIALITIES OF OPTICAL MASERS

• Document Type: CREST
• Collection: General CIA Records
• Document Number (FOIA) /ESDN (CREST): CIA-RDP81-00120R000100060029-0
• Release Decision: RIFPUB
• Original Classification: K
• Document Page Count: 8
• Document Creation Date: December 20, 2016
• Document Release Date: November 6, 2002
• Sequence Number: 29
• Publication Date: January 1, 1962
• Content Type: CONT

PRn('FFnTIVr'c nF TTTR TRP
Approved For Release 2007/09/21 : CIA-RDP81-0012OR000100060029-0
Some Potentialities of Optical Masers*
B. M. OLIVERf, FELLOW, IRE
Summary-This paper, originally presented at the 1961 WESCON
Convention, is intended as an introduction to the principles and pos-
sible applications of the optical maser. Very little prior knowledge on
the part of the reader is assumed. The intent is to acquaint him with
these exciting new devices with the hope that he, in turn, will dis-
cover applications not foreseen by the author. The methods of gen-
erating coherent radiation, of focussing it, and of collimating it into
tight beams are described. The use of lasers for communication is
explored, and certain medical and other applications are suggested.
INTRODUCTION
N OT LONG AFTER the development of the mi-
crowave maser by Townes, Bloembergen and
others, it became apparent that the same princi-
pies could be applied to the generation of radiation of
far higher frequency. In a historic paper in the Physi-
cal Review' Schawlow and Townes predicted the possi-
bility of extending maser principles into the optical re-
gion, described general types of structures which would
be required, and speculated upon the performance possi-
ble. Very few developments in recent years have excited
the imaginations of so many scientists and engineers as
has, the optical maser, or "laser." The reason, of course,
is that this device offers for the first time a means for
producing and amplifying coherent light and therefore
opens up the opticalspectrum for exploitation by all of
the techniques currently used in the radio spectrum.
In July, 1960, Maiman [1] of Hughes Laboratories
announced the first successful production of coherent
light on a pulse basis using optically pumped ruby.
Early in 1961 Jawan, Bennett and Herriott of Bell
Laboratories announced the successful CW operation
of a gaseous optical maser [2]. In addition, there has
been a great deal of activity in this field at many other
laboratories, and the effort is increasing. -
The basic principles of the optical maser are the same
as those of the microwave maser. Fig. 1 illustrates in
schematic form the essential ingredients. First of all,
there must be it resonant cavity. In an optical maser this
is formed by two precisely oriented mirrors, one of
which is slightly transparent. It can be shown that
resonant modes will exist between these mirrors at fre-
quencies for which the spacing is an integral number of
half wavelengths. In this space between the mirrors is
placed an active medium which may be either a gas or a
crystal doped by certain atones (such as chromium in
the case of ruby). The medium must possess two atomic
states separated in energy by an amount corresponding
to the frequency desired, and it must be possible to over-
* Received by the IRE, September 25, 1961; revised manuscript
received, November 20, 1961.
t Hewlett-Packard Co., Palo Alto, Calif.
i A. L. Schawlow and C. 11. Townes, "Infrared and optical mas-
ers," Phys. Rev., vol. 112, pp. 19.10-1949; December 15, 1958.
populate the upper of these states with respect to the
lower. This is done by "pumping" the atoms front a
ground state to a higher energy state either electrically
or optically. From this higher energy state the atoms
usually decay nonradiatively to the upper of the two
energy states involved in the desired transition. From
this upper state some atoms will decay spontaneously to
the lower state and emit light just as occurs in any neon
sign. The light caused by such spontaneous emission is
incoherent and is radiated in all directions at random.
However, in the presence of the resonant cavity some of
this spontaneous emission will excite one of the resonant
modes of the cavity, and the field associated with the
resonance will induce emission in the medium. This in-
duced emission is phase coherent with the field which
induces it and as a result, if the interaction is strong
enough, a coherent electromagnetic wave will build up
corresponding to one of the modes of the resonant cav-
ity. Some of this energy will leak through the partially
transparent mirror forming one end of the cavity and
emerge as it sharply defined beam of coherent light. The
significant thing is that this beam of light is it plane co-
herent electromagnetic wave just as would be produced
by a radio transmitter, but of vastly higher frequency.
There are two aspects to wave coherence: spatial and
temporal. A wave is spatially coherent if there exist sur-
faces over which the wave amplitude as a function of
time is highly correlated. If the coherence is complete,
the correlation will be unity, and the voltage at the one
point will he proportional to the voltage at other points
oil the surface. As an example of spatial coherence, con-
sider the amplitude at any two points on an equiphase
front in the light from it distant star. The amplitude at
two points is the same function of time, and similarly at
a given time, the amplitude along different rays is the
same function of distance.
A wave exhibits time coherence to the degree that
there is correlation between the amplitude of the wave
Approved For Release 2007/09/21 : CIA-RDP81-0012OR000100060029-0
 Approved For Release 2007/09/21: CIA-RDP81-0012OR000100060029-0
at a given point at one time and at some later time. A
single frequency represents the extreme of time coher-
ence; so does any combination of single frequencies,
harmonically related. If the line components of the
spectrum are broadened due to random modulation, the
time coherence is lessened and in the extreme, when the
spectrum of the wave consists of a smooth distribution
of frequencies, as is true of black body radiation, for ex-
ample, time coherence virtually disappears. Spectral
purity (i.e., the degree to which the spectrum ap-
proaches a line spectrum) is thus a measure of time co-
herence.
The following analogy may be helpful in visualizing
the difference between coherent and incoherent radia-
tion. Incoherent radiation may be thought of as the
three dimensional analog of the pattern of waves on the
surface of a swimming pool in its usual state just after
the swimmers have left. Waves of all different wave-
lengths are racing every which way at random, and
there is little correlation between the time functions
representing the amplitudes at two widely separated
points. If the pool were surrounded by quiet water at
the same level and the walls suddenly removed, the agi-
tation would spread out in all directions in a way which
resembles the radiation of incoherent light. By contrast,
if the surface of an otherwise quiet swimming pool were
set in motion by the up and down oscillation of a float
extending clear across one end, a series of plane waves
would be produced. These would exhibit high spatial
coherence, because the wave amplitudes as a function of
time at different points would be highly correlated. And
if the motion of the float were periodic, the waves would
exhibit time coherence as well. If the walls were again
removed the wave pattern would propagate out in a
beam normal to the float, and if the wavelength were
short compared with the length of the float, the beam
would exhibit very litt,'e spread with distance.
Ruby lasers operate on a pulse basis and give quite
high peak power output. Characteristically, pulses on
the order of 10 kw output power and pulse durations of
the order of 1 cosec are obtained. The total energy per
pulse is therefore on the order of 10 joules. Although the
spectral line is clearly narrowed by the maser action,
the time coherence is still relatively poor in the ruby
laser. The ruby rods tend to oscillate in several modes
simultaneously and to produce a series of short spikes
instead of a single pulse. The line width of each mode is
on the order of 10 Mc as a result of the spike amplitude
modulation. The spatial coherence also is far from per-
fect and is thought to be limited by optical imperfec-
tions in the ruby crystals themselves. Hopefully, these
are defects which will be eliminated with further re-
search, and in our discussion to follow we will assume
ideal coherent operation to be possible. The CW gaseous
masers operate at relatively low power levels-on the
order of 20 milliwatts-and exhibit excellent coherence.
Line widths of approximately 1 cps have been reported,
February .
to believe that the power output of the gaseous laser
can be increased, and these devices look very promising
for communication applications.
COHERENT OPTICS
The availability of coherent light greatly increases
the scope of things that can be done with optical sys-
terns. In particular, the spreading of beams of light be-
comes limited only by diffraction. In an ordinary search-
light the beam spread is principally due to the finite size
of the source of light. A point of the source lying on the
optical axis produces a beam parallel to the optical axis,
while various points of the source lying off the optical
axis produce beams at various angles with respect to the
axis. The totality of all these beams thus spreads at a
rate determined by the greatest extension of the source
and by the focal length )f the objective. By contrast,
plane waves radiated by an optical maser spread in the
same fashion as would the beam from an antenna having
the same size measured in wavelengths. Having coher-
ent light is tantamount to having a point source.
When the light from a star is imaged by a small tele-
scope objective under ideal seeing conditions, a diffrac-
tion pattern called an Airy disk is formed. It consists of
a central patch of light surrounded by a series of rings,
the intensity as a function of radius, r, being given by
2
[2J(r) 1
\Xfr)
J
L
(1)
where
d=diameter of the objective,
f =focal length of the objective,
X = wavelength,
and Jl is the first order Bessel function. This same dif-
fraction pattern is formed when the light from an opti-
caI maser is brought to focus with an ideal lens, assum-
ing the beam illuminates the lens uniformly as shown
in Fig. 2. In general, the diffraction pattern is the (two:
dimensional) Fourier transform of the aperture illumi-
nation. In this case the intensity at the center of the
spot is
A
Io = P
X2f2
(2)
where A =area of objective, and P=power in light
beam.' Taking a value of 10 kw for the power output of
the ruby laser at a wavelength of 0.7 ?, we find the power
density at the center of the image to be about 1018 watts
per square meter. This is a power density far in excess of
anything normally obtained in the laboratory. As a com-
parison, the power density at the surface of the sun is
less than 108 watts per square meter. Thus, the ruby
laser is theoretically capable of producing a power
and the spatial coherence seems to agree with what 8 This relation can be derived directly from the radio transmission
expression, (5), by setting D=f, Ar=A, PT-P and I0=PR/Art.
?.=...,t~i he expected theoretically. There is every reason Again it assumes a uniformly illuminated aperture.
Approved For Release 2007/09/21 : CIA-RDP81-0012OR000100060029-0
 aser
sing
-ases
sys-
be-
arch-
size
i the
axis,
Stical
o the
at a
3urce
trast,
zn the
tying
oher-
tele-
ffrac-
sts of
rings,
=3 n by
(1)
_te dif-
1 opti-
sssum-
shown
- (two.
-dlumi-
of the
(2)
light
put of
? power
watts
ccess of
a com-
- sun is
ruby
power
smission
Pe/AR.
J'6.
Oliver: Some Potentialities of Otitical Masers
Approved For Release 2007/09/21 : CIA-RDP81-0012OR000100060029-0
Fig. 2-Focusing of coherent light.
"EYEPIECE"
Fig. 3-Optical "antenna" with power gain of (f2/fl)2.
density one hundred million times that of the surface of
the sun! This high-power density is accompanied by a
correspondingly high-electrical field strength given by
E = 4./rtlo
where
71 =120ir=impedance of free space.
For the above case we find
(3)
E=V/120irX1016
= 2 X 109 ;volts per meter
= 2 million volts per millimeter.
At such fields it should be possible to produce many ef-
fects heretofore unobservable. Such possibilities as the
alteration of the construction of molecules, the disrup-
tion of chemical bonds in small regions inside homo-
geneous substances, etc., suggest themselves.
If, after having been brought to a focus by an initial
lens, the light from the diffraction pattern is allowed to
propagate further, it can illuminate a much larger lens
which can in turn recollimate the light as shown in Fig.
3. Such an artlangement will be recognized as the simple
astronomical telescope used in reverse. Just as the re-
solving power of a telescope is increased in proportion
to the diameter of its objective, so the beam spread of
the emerging beam from this "optical antenna" is in-
versely proportional to the diameter of its objective. In
fact, the width of the major lobe from the peak to the
first null is found from (1) by setting J1= 0, and is the
usual formula for the resolving power of a telescope
rd r -ird
f - = - 0=3.8317??
A
0=1.22-?
d
(4)
As a result of the decreased beans spread, distant points
will be illuminated more intensely, and the antenna will
exhibit a power gain equal to the square of the normal
magnification of the device as a telescope. Used as a
searchlight, a laser followed by a telescope can produce
a remarkably small spot of light at great distances. For
example, a 12-inch diameter telescope on the earth
02
ba
10
1.2
0.05
EXAMPLE: j A ? 0.T}e
ll 2a- 1211
.05% LOSS IF:
b ? a ? 20 MILES
Fig. 4-Transmission between apertures.
would produce a central spot of light only 8800 feet in
diameter on the moon. The illuminated patch, of course,
corresponds to the figure of confusion of the same ob-
jective when used as a telescope.
The extremely small beam spread possible at optical
frequencies with coherent light suggests that it should
be possible to transmit power over considerable dis-
tances with relatively little loss. This turns out to be the
case. For a given maximum size of antenna (consider-
ing both the receiving and transmitting antenna) the
least power loss will occur if the beam intensity as a
function of radius off axis is properly shaped. This shape
is approximately Gaussian and is, in fact, the distribu-
tion which arises naturally when the resonator of the
laser consists of two confocal concave mirrors as de-
scribed by Fox and Li [3]. The beam from a laser em-
ploying such a resonator one meter in length is only
about a millimeter in diameter: a fine thread of light.
How+.ver, its radial intensity distribution will be pre-
served after passing through an antenna of the type
shown in Fig. 3, and so it is simple to create a beans of
large cross section having this Gaussian distribution. If
the diameter of the transmitting aperture and receiv-
ing aperture are both 2a and if the distance between
them is b, as shown in Fig. 4, then, with such a beam,
the power loss will be as given by the table and example
in that figure. A loss of one twentieth of one per cent of
the power in a twenty mile hop is less than occurs with
the average transmission line. However, this low loss
can only be achieved if scattering and refraction in the
medium are absent, and hence can be realized only in
free space or if a controlled atmosphere is provided in a
pipe connecting the two apertures. The possibility of
transmitting power by optical means from earth to a
satellite or from one space vehicle to another is a very
real one.
Approved For Release 2007/09/21 : CIA-RDP81-0012OR000100060029-0 .,,_,.-.
 Approved For Release 2007/09/21 : CIA-RDP81-0012OR000100060029-0
.138
February
COMMUNICATION BY COHERENT LIGHT creased. It is important to remember that with fixed
antenna sizes any improvement in transmission with
PROCEEDINGS OF THE IRE
The use of coherent light for communication. purposes increased frequency comes about from this cause alone.
is an obvious application, and in this section we will con- There are practical limits to the directivity of beams
sider the potentialities of lasers in this service. The fac- which can be achieved even in the optical region. In or-
tors which must be considered are the bandwidths af-
forded der for the theoretical beam spread to be attained, an
by the new type of channel, the transmission objective mirror must be well within a quarter wave-
loss, the inherent noise, and the cost. Let us look at length of the true figure over its entire surface. While
these in turn. the surfaces of a lens are less critical, there are more of
In the red end of the spectrum the frequency of light them. As the size is increased this accuracy requirement
is approximately 4X101' cps. A 1 per cent band in this becomes increasingly difficult to meet. The 200-inch
portion of the spectrum has a width of 4 million i\'Ic, telescope, for example, is far from accurate enough to
enough for a billion simultaneous telephone conversa realize its s full resolving power.
tions. Thus, it should be possible in theory to transmit Even the resolving
were perfect, the turbu-
all of the conversations going on anywhere in the world lence of the atmosphere would limit the usable beam
simultaneously over a single thread of light one milli- sharpness. It is very seldom that seeing conditions are
meter in diameter. Apparently we have bandwidth to d enough to permit the resolution of points sepa-
burn. This is especially true when one considers the fact goorated by less than one half second of arc, so that beams
that because of the extremely directional beams that sharper than this would not be suitable for communicat-
can be produced, many simultaneous channels can exist ing from the earth's surface to space. Along the earth's
in the same frequency band without mutual interfer the situation is even worse because of the
ence. The problem is not one of lack of spectrum space, surface greater length of high density airpath, so that unless a
but of how to make use of it: of how to modulate the controlled atmosphere or vacuum is provided, beams of
optical maser'or its output. so as to fill up the spectrum. angular spread of at least a few seconds must be used.
Work in this direction is going on at the present time, If we now further assume that atmopsheric refraction
and beams from lasers have been modulated at frequen- effects are absent, as in'an evacuated pipe or free space,
cies up to X band, so that channels 10,000 Mc wide have there still remains the problem of pointing an exceed-
been achieved but at the cost of large modulating ngly 1 sharp beam in the right direction. Vibration of
powers. There seems little doubt that ways will be mounting structures, bending of supports by thermal
found to produce extremely wide-band modulation and expansion
and, in the case of free bodies, changes in the
thus utilize the bandwidth capabilities of optical Chan- moments of inertia due to motions of the parts will alt
eels, but even on a less ambitious bandwidth basis; introduce pointing errors difficult to reduce below one
these channels are still attractive, as we shall sec.
nd of arc
.
seco
The transmission loss in an optical channel is small In addition to turbulence troubles in the medium, a
because of the very directional beams easily achieved, great deal of power can be lost over a long path through
as illustrated by the example of power transmission. molecular scattering and scattering due to suspended
The basic formula of transmission loss in the optical particles. Again the use of vacuum or perhaps a filtered
PR ATAR
PT
where
PT = transmitted power,
PR = received power,
AT= area of the transmitting antenna,
A R = area of the receiving antenna,
X = the wavelength,
D = the distance between the antennas.
(5)
This formula assumes uniform illumination of the trans-
mitter aperture, which indeed gives the least power loss
on axis. The X2 in the denominator shows that the
power received is proportional to the square of fre-
quency and this reflects the increasing concentration of
energy by the transmitting antenna as frequency is in-
3 Fqs. (1), (2) and (5) may all be derived from Huyghens' prin-
ciple. Eq. (5) also follows from the fact that the aperture of an iso-
tropic antenna is a2/47r, as may be proved thermodynamically.
tt
dis:
11101
114.
acl;
col,.
ste,
crc:
cie-
helium atmosphere would reduce this loss. Further, any
terrestrial path must be bent to conform to accessible
routes and to the earth's surface. This can be done
through mirrors or prisms or lenses disposed along the
path at frequent intervals. It is difficult to make such a
beam bender with less than 1 per cent loss, so that an
attenuation on the order of one or more nepers per 100
bends is to be expected.
Quantum mechanical analyses [51 have shown that
even an ideal amplifier has a noise power spectral den-
sity referred to the input which is given by
0(v) =
+ hv, (6)
exp - 1
(1)
where T is the absolute temperature of the source (re-
sistance for circuits, weighted average of the field of
view for antennas). The first term in_this equation rep-
resents thermal noise and -t low frequencies (for which
hr