JPRS ID: 10709 USSR REPORT PHYSICS AND MATHEMATICS

• Document Type: CREST
• Collection: JPRS
• Document Number (FOIA) /ESDN (CREST): CIA-RDP82-00850R000500090006-1
• Release Decision: RIF
• Original Classification: U
• Document Page Count: 122
• Document Creation Date: November 1, 2016
• Sequence Number: 6
• Content Type: REPORTS

APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500090006-1
FOR OFFICIAL USE ONLY
JPRS L/~107.09
4 Au~gust 1982
I~SSR Re ort ~
p
PHYSICS AND MATHEMATICS
C.F~O U O 6/~8 2~)~
Fg~$ FUREIGN BROADCAST INFORMATION SERVICE
FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500090006-1

APPROVED FOR RELEASE: 2007/02/49: CIA-RDP82-00850R040500090046-1
NOTE
JPRS publications contain information primarily from foreign
newspapers, periodicals and books, but also from news agency
transmissions and broadcasts. Materials from foreign-language
sources are translated; those from English-language sources
are transcribed or reprinted, with the original phrasing and
other characteristics retained.
Headlines, editorial reports, and material enclosed in brackets
are supplied by JPRS. Processing indicators such as [Text]
or [Excerpt) in the first line of each item, or following the
last line of a brief, indicate how the original information was
processed. Where no processing indicator is given, the infor-
mation was summarized or extracted.
Unfamiliar names rendered phonetically or transliterated are
enclc?sed in parentheses. Words or names preceded by a ques-
tion mark and enclosed in parentheses were not clear in the
original but have been supplied as appropria*e in context.
Other unattributed parenthetical notes within the body of an
item originate with the source. Times within items are as
given by source.
The contents of this publication in no way represent the poli-
cies, views or at.titudes of the U.S. Government.
COPYRIGHT LAWS AND REGULATIONS GOVERNING OWNERSHIP OF
MATERIALS REPRODUCED HEREIN REQUIRE THAT DISSEMINATIOh
OF THIS PUBLICATION BE RESTRICTED FOR OFFICIAL USE 0~]LY.
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500090006-1

APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500090006-1
JPRS L/10709
4 August 1982
USSR REPORT
PHYSICS AND MATHEMATICS
_ (FOUO 6/82)
CONTENTS
LASERS AND MASERS
Kinetics of Chemical Processes in Plasma of Mol~cular ~
Gas Discharge Lasers 1
Abstracts of Pape~a on Quantum Electronics 9
Chemical-Gasdyr~amic Cd2 Laser Using CO + 0+ M
Recombination Reaction Products ...........................o... 18
OPTOELECTRONICS
Theory of Dynamic Image Selection Ef'_`ect in P~~otorefractive
Media 29
Using Optical Data Processing 38
PLASMA PHYSICS
Nonequilibrium Low-Temperature Plasma Kinetics 90
~ Dynamics and Radiation of Open (Vacuum~ Plasma~ynamic
Discharges of 'Plasroa Focus' Type: Survey 95
- a- [III - USSR - 21.H S&T FOUO]
~ FOR OFF[CIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500090006-1

APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500090006-1
F'OR OFF'1CIAL USE ONLY
LASERS AND MASERS
UDC 537.525
KINETICS OF CHEMICAL PROCESSES IN PLASMA 4F MOLECULAR GAS-DISCHARGE LASERS
Moscow KHIMIYA VYSOKIKH ENERG?Y in Russian Vol 16, No 3, May-Jun 82 (manu-
script received 5 May 81) pp 267-272
[Article by V. I. Volchenok, V. N. ~Comarov, V. N. Ochkin and N. I. Sobolev,
Scientific Research Physicochemical Institute imeni L. Ya. Karpov]
[Text] Molecular plasma systems are classified within the
framework of simplified three-component plasmochemical kin-
etics from the viewpoint of their operating ,life. The class-
ification is made according to the ratio of the rates of re-
versible and irreversible (homogeneous and heterogeneous)
interactions. It is shown that the active media of molecular
lasers (C02, CO and N20) investigated earlier can be described
by this scheme. The role of catalysts in extending the life-
time of the systems is determined.
A low-temperature molecular plasma is now used to implement a wide range of
processes. All the devices in which these processes are realized can be di-
vided from the viewpoint of application into two main grov.ps: 1) devices that
utilize the plasma as an active medium of a chemical reactor, in which the in-
itial materials are reprocessed; it is desirable that these s~evices guarantee
high rates and degree of transformation of the initial materials to the required
reaction products; and 2) devices that utilize the plasma as an object with re-
tained properties, for example, in light sources or active media of electric-
discharge lasers. In this case, unlike the first case, it is necessary that
the concentration of initial (working) molecules be maintained at a level that
exceeds some minimum concentration required for operation of the devices for as
long a time as possible. We will consider the second case in the given paper.
The most detailed investigations of chemical processes in a molecular plasma
have now apparently been conducted for active media of continuous infrafred
electric-discharge lasers (see, for example, survey [ll)� It is feasible in
this regard to attempt to classify the tnain types of these systems from the
viewpoint of the chemical kinetics of the gas-phase and heterogeneous processes
occurring in them.
If laser transitions occur o~ m~lecules A present in the initial working mix-
ture of gases, then the dependence of their concentration on time (t) can be
represented in the general case as
1
FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500090006-1

APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500090006-1
FOR OFFICIAL USE ONLY
[~l (c)=[-~l (r-ii)I(~)� c~)
The form oi function f(t) can be found from solving a system of nonlinear dif-
ferential equations of the type of Pauli equations that describe the total
kinetics of homogeneous and heteroqeneous chemical processes with participation
of all particles formed in the discharge at different energy states. However,
solution or ar:alysis of this system in qeneral form is impossible. Therefore,
the prablem can be simplified for qualitative analysis and classification of the
main cases. We ~rill coneider a plasma in which chemically different components A,
B and C undergo mutual conversion and inter~ct with the surface.
Let us also suggest that the presence of particles of other kinds in the plas-
ma, including charqed and excited particles, is takE~n into accou~t by effective
rate constants of reactions A, r~ and C and therefore the kinetics is formally
described by first-order equations.
. .c~ 'P'~~
f - ~ ~I f I
~1 ~ b
~
` ~
L` t t~ ~
C~%
.f ~
c d ` ~
s0 ~x x
x
f, ~ N
\ M
~
?
Z O ~/0 ~~Ci.
- Figure 1. First Case of Classification of S stem (a-c) (see text):
a--rp = 0, kA = 0; b--rA = 0, kA = 0, kA > kA; c--rA = 0, kA = 0.
The dependence (d) of t~,~ concentration af N20, N2, 02 and NO mole-
cules on the time the mixture is in the discharge, the :nitial N20-He
~ mixture (1:4) at pressure of 0.532 kPa, current density of 31.8 A/m2
and mass apectrometric measurements; ~--N20, O--N2, X--02 and +--NO
If molecules A further break down in the plasma into particles B and C:
~x[li] I,i[(:](a+~~-=1)~ then
~,1~ --:_r,.,[~~l+r~�"[~;)-+-r;,:"l~=]-~~~~~�~~~ ~r~ c2a)
_ 2
F'OR OFF7CIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500090006-1

APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-04850R000500090006-1
� FOR OFFICIAL USE ONLY
~13~ .~-li~~[I>~+ak�(AJ--k~"[lij-k�"'~li]-f-r� (2b)
((;j =-k,:[~1+~'~~[~1l-h~"fC]-h~;"'[CJ+r~:, (2c)
where kp, is the rate constant of breakdown of A into B and C, kg and k~ are
the rate constants of breakdown of B and C into subsequent fragments, k~ and
kA are the rate constants of reversible processes B-? A and C-? A, k~, kB and
k~ are the rate constants of irreversible heteroqeneous binding processes of
molecules A, B and C and rp, rg and r~ are the rates of appearance of mole-
cules A, B and C from external sources (separation of molecules from the sur-
face of the discharge elements, from special getters, from a ballast tank and
so on) .
By combining equations (2a)-(2c), let us write
lca"~ak�IE1~-~~~~'re)
[A]=-(k~+h,,") [Al+ + (3)
kn-~-knn-}.,kDw
1cc"~~k,+~A~-~C~~'rc)
+ ~~+k~~+k~W r,,
and let us consider typical limiting cases, having first noted the obvious
circumstance that the caaes when [A](t*` t) ~[A], where ~A]* is the density
of the working molecules similar to the optimum with regard to laser param-
eters and t* is the desirable operating time of the system, are of practical
interest among them.
1. The working molecules A break down in a plasma at considerable rate, while
the reversible reactions are hardly effective: ~;~[A]~ k$";B] and k~"[C] .
Solution of equatior.s~(2) and (3) yields
[A]=(~~]oc.1~(-k~t) {exp(-k~"'t) ~1+ [AJo(!c +~.~W) X (4)
X (exp (k,~t+k~"'t) -1) 1 ~ ,
~
where [Aj o- [~11 (t=0) . Equation (4) can be represented in the form [A] _
= [ J acp (t) (t ) , where 'I~ ~t) =exp (-k,t) describes the gas-phase process of di-
rect breakdown of m~lecules A in the plasma. In the sim~lest case of the ab-
sence of external sources (rp, = 0), function ~p(t)=exj~(-k�"~t) describes the ab-
sorption of the working molecules by surfaces. If rp, # 0, then separation
- into two functional dependencies for homogeneous and heterogeneous processes
xs unjustif ied, ~'(t) (the expression in brackets) becomes dependent on the
rates of the gas-phase processes and with long times [A] ~~=-r�,, (~�~+k,,`''
The relationship between kA, k~ and rp determines the process that limits the
presence of molecules A in the system. ~ao limiting cases are possible in
this case.
3
FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500090006-1

APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500090006-1
FOR OF~ICIAL USF. ONLY
. 1-a. The gas-phase process of breakdown of working molecules is faster than
heterogeneous processes. Some examples are presented in Fiqure la, b. If
rA = 0, then on the time scale corresponding to Figure la, const = 1 and
elimination of A is determined by gas-phase breakdown f=�. If rA # 0(Fig-
ure lb), then ~r' is an increasing function and F>~, reaching an asymptotic
value of [A]~,/[A]p. The situation correspondi~ to Figure la is typical, for
example, for an N20 laser [2]. The corr~sponding data of mass spectrometer
measurements are ~resented on Figuxe ld, showing how the densities of N20,
N2, 02 and NO molecules (in % with respect to N20 in the initial mixture)
depend on the time that the discharge (with current density of 31.8 A/m2
and pressure of 95.32 kPa) acts on the initial N20-He mixture (1.4). The
_ working N2 molecules are transformed (with fulfillment of the nitrogen and
oxygen balance in the gaseous phase) to 02 and N2 with formation of NO at the
intermediate stage. If the active medium af an N20 laser is connected to the
b allast gas tank containing N20, then this corresponds to the situation shown
in Figure lb.
1-b. Irreversible heterogeneous binding processes of working molecules ~'e
faster than the gas-phase process of their breakdown, f= ~y, and const = 1.
An example is presented in Figure lc. This situation is typical, let us say,
for C02 or N20 lasers cooled by liquid nitrogen, where the rate of elimination
of working molecules is determined by diffusion with subsequent concentration
on the wa1Zs and kA - D/R2 (D is the diffusion ccefficient and R is the radius
of the tube).
2. Irreversible heterogeneous processes occur with high rate for transforma-
tion products of working molecules withoui. a~fecting the wor~Cing molecules
themselves. Thus, lc�`�~1~�"; k.,N>t 1 in front of the nozzlewas broken, an equi-
~.i Cj ~�S libriian f low of 0+ 02 + Ar entered the mixing
zone, where it was mixed wiL�h the other reagent
Fig. 1 (CO + He mixture) discharged from the cavity
of valve 4 through injector 3. The configuration
of the zone in which flow intermixing and chemical
, /
reactions occurred in the system was determined
by the prof ile of the lateral surface of the cylin- u~,~?~~~.\ /
drical injector and the geometry of the subsonic
part of the nozzle (Fig. 2). The configuration CO~liei
of the mixing zone and the positions of the injector ~
orifices determined the durationg f floy intermixing o�o�*nr-f T
and the length of time that the as sta ed in this ~
zone.
Shaped nozzle 19 with profile described in Ref. 6 Fig. 2
was used in the experiments. Nozzle parameters:
19
FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500090006-1

APPROVED FOR RELEASE: 2407/02109: CIA-RDP82-00854R000500090006-1
COR OFFICIAL USE ONLY
1~eiRht of critical cross section h* = 1.0 mm, vertex half-angle a~ 42�, degree
of expansion of the flow A/A*= 23. It is shown in Ref. 5 that the time during
which the gas discharge from the nozzle is steady-state is -0.2 s. The dura-
tion of existence of steady-state conditions in the "hot" plug behind the
shock wave in the 02 + Ar mixture is considerably shorter: To= 0.15-0.30 ms.
Therefore triggering of the electromagnetic valve was synchronized with the
arrival of the shock wave at the end face of the shock tube in such a way
that the flow injected from the channel cavity was formed with a lead of
200-300 us before the instant when the diaphragm in front of the nozzle burst.
The physicochemical processea in our research were diagnosed on the basis
of registration of intensities of spontaneous (a = 4.3 um) and s~imulated
10.6 um) infrared radiation in the working cross section of the gas flow,
and also spontaneous recombination emission accompanying the reactions
0+ 0-? 02 + hv (Schumann-Runge bar.d. a= 401.0 nm) and 0+ CO-~ C02 + hv (continuum
280-700 a= 488 nm). A diagram of the optical channels of the system is
shown on Fig. 1. The course of the recombination reaction CO +0+ M was monitored
by measuring the concentration of oxygen atoms in the hot plug (see Fig. 1, 2)
and in the supersonic flow of reaction products in the cross section situated
at a distance of x= 50 mm from the plane of the critical cross section of
the nozzle.
The concentrations of oxygen atoms in the hot plug were determined from measure-
ments of intensity Iyol of radiation of 02 molecules on wavelength 71= 401.0 nm.
Radiation from the pre-nozzle region 2 was coupled to photomultiplier 22
(FEU-39A) through f iber-optics guide 20, quartz lens 18 and monochromator
17. Luminescence intensity Iyol is related to the concentration of oxygen
atoms in the hot plug by the expression [Ref. 7]
leoi = C~o~l~aoi ~~~o~ ~O1"-,
where Iyol(B) is the magnitude of the si~nal registered on the oscilloscope
screen corresponding to radiation on wavelength a= 401.0 nm, Cyol is the cali-
bration factor of the registration system in V�~?�sr/(W�cm 3), I'yol(Tb) in
W�sr-1�cm3)/(~�mole2) is the value of function I'J~(To) that describes the
distribution of radiation intensity in the Schumann-Runge band at 401.0 nm,
To is the temperature in kelvins in the hot plug after equilibri~ dissociation
of 02 molecules is established in the mixture of initial composition 0.402+
0.6Ar, and [0] is the density in mole/cm3 of oxygen atoms in the hot plug.
The intensity of visible radiation in the supersonic flow of reaction products
(at a distance of 50 mm from the plane of the critical cross section of the
�ozzle) is determined mainly by the cross section in the chemiluminescent
reaction 0+ CO-~ C02+ hv, and hence depends on the concentration of oxygen
atoms. In this paper measurements were made of the radiation intensity in
the spectral range of a= 488 � 7.5 ~ by using fiber-optics guide 9, inter-
ference filter 8 and photomultiplier 7(FEU-39A). According to Ref. 8, the
intensity of radiation in the given spectral interval is related the concen-
tration of oxygen atoms by expression
I~e~ = C~e~r~ee~T) ~nI ICO~. ~ ~2)
20
FOR OFF7CIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500090006-1

APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500090006-1
FOR OFFI('lAl, l1SE ONLY
Calibration factors Cyui. C488 were experimentally determined. The values
of T4ae~T) at T= 293 K are given in Ref. 8. To extrapolat~: function Taea(T)
to other values of T we used the temperature dependence of the rate constant
of chemiliuninescent reaction CO + 0+ C02 + hv on temperature [ sic ][Ref . 9] .
At Tb = 3500-3850 R in the hot plug, molecules of 02 (B3Eu, ~v' = C,1) make the
principal contribution to radiation (a = 401.0 � 7.0 nm, Sch~mmann-Runge band) of
p2(B3E-) ; 02(X3E-)) [Ref. 7]. In accordance with the Fran=k-Condon principle
this correspondsgto excitation of vibrational levels v" > 18 of the lawer elec-
tron state of 02(X3Eg) [Ref. 10]. The equilibrium popu5ation of vibrational
levels v" > 18 of molecules 02(X3Eg) does not exceed 10- of the number of
� molecules of 02(X3Eg, v" = 0). Analogously in the course af chemilu!ainescent
reaction C0+0-> C02 + hv (a = 488.0 � 7.0 nm) we ~et excitation of vibrational
levels of the ground electron state of COZ(lE ) that are situated ~48 kcal/mole
higher than the ground vibrational level v1=vg = v3= 0 of the COZ(lE+g) molecule
[Ref. 9]. These levels are also practically unpopulated under conditions
of supersonic flow with translational temperature T= 150-300 K. Because of
low population, the given levels of the 02 and COz molecules make no contri-
bution to radiation transfer, and therefore in recording recombination radia-
tion of 0+ 0-~ 02 + hv and 0+ CO-? C02 + h~~ in the given system there is no need
to account for reabsorption of radiation. Thi~ enables us to use relations
(1) and (2) that are valid for conditions of an optically thin layer.
The systematic error in measuring oxygen atbm concentration in the
hot plug is due to the error in determination of the quantity I'yol(T) given
in Ref. 7, and also to the error in measurements of the ca.libration factor
Cypl. The overall error of ineasurements of the quantity E~ in the hot plug
did not exceed 20~.
In the case of optical measurements in the supersonic f lot~~ (x = 50 mm) , an
error associated with the uncertainty of values of transl~.tional temperature
is added to the above~mentioned sources of errors. Deviat.ion of the actual
values of T in the flow from those calculated in the apprc~ximation of uniform
steady-state discharge through the nozzle is selated to Lhe possible increase
in T due to V-T relaxation of excited reaction products ir~ the flow. Without
consideration of this factor, the calculated values of temperature T in the
expression are understated, and consequently the measured values of concentration
of oxygen atoms are overstat~ed. Therefore the systemat:ic error in deter-
mination of in the flow was from +30 to -70%.
Infrared diagnosis of the state of the nonequilibrium flow of reaction products
in cross section x= 50 mm was based on measurements of intensity of spontaneous
radiation I4,3 of C02 molecules in the 4.3 um band, as we?1 as measurements
of th: small-si~nal gain (or absorption) Ko of the probinf; C02 laser radiation
(a= 10.~ um). The vibrational temFeratures of asymmetric (T3) and collective
(T2) modes of the C02 molecules that are reaction products were determined
from the measured values of I4.3 and Ko. The method of determining vibrational
temperatures from characteristics of spontaneous and stimulated emission is
given in detail in Ref. 11. In the given experiments, a beam from an LG-23
C02 electric-discharge laser (a = 10.6 um) (see Fig. 1, 5) after reflection
from flat mirror 6 passed through BaF2 windows 10 located in cross section
21
FOR OFF(CIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500090006-1

APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500094006-1
FOR OFFICIAL USE ONLY
x= ~i0 rtun and dispersion filter 15, and was incident on photoresistor 14 (FSG- .
'l'1A). From the same cross section of the flow by means of spherical mirrors
11 and 16 and narrow-band filter 12 (J~= 4.41� 0.05 um) spontar.eous radiation
was measured in the band of the CQ2 molecules with center 7~= 4.3 um (photo-
resistor 13).
Calibration of the system for registration of infrared radiation is described
in Ref. 11, according to whic.h the accuracy of ineasurements of vibrational
temperatures T3 and TZ is determined mainly by the random error and amounts
to 10~.
The parameters of the gas behind the reflected shock wave in the mixture of
initial composition 0.402+ 0.6Ar wpre: po= 15 � 2.0 atm, To= 3800 - 4820 K.
Results of calculation of the relative cuncentration
~r.i . of oxygen atoms in the hot plug (assuming establishment
1 � ~ of equilibrium dissociation of 0 molecules) are shown
o~~- ~ on T'ig. 3. Characteristic time of establishing condi-
~ ~ - ~ ~
~ - ~ tions of equilibriwn dissociation under the conditions
~ r~ j~�? of the experiment was -12 us. After establishment of
1 ~ ' '
~t ~i equilibrium dissociation of 02 molecules, the temperature
o e~-- in the hot plug according to theoretical calculation
~ _I dropped to Tb= 3450-3850 K. The experimentally measured
�,JO ~ concentrations of oxygen in the hot plug are shown on
r ! '
~ s.~, ~ i, K Fig . 3. The relative concentrations of oxygen atoms
" ' under the conditions of the given experiments were
Fig. 3 = 0.06-0.15 depending on the equilibrium temperature
behind the reflected shock wave Tb. The measured values
of ~~j average 15X less than the calculated values.
On the basis of results of Ref. 12, estimates were made of the duration of
the process of intermixing flows as a jet of CO + He is injected in a direction
close to the normal with respect to the direction of the flow of 0+ 02+ Ar
in the cross section where the Mach number of this f low is M~ 0.2. These
estimates showed that intermixing of flows is completed at distance Tm ~2.5 mm
from the injection point. The time of intermixing of flaws (T~ ~ 5 us) is
0.7T2, where T2 is the time interval from the instant of in~ection of gas
CO + He into the mixing zone to the instxnt of arrival of the mixeC flow in
the critical cross section of the nazzle. Comparison of the characteristic
time of the recombination reaction CO + 0+ M('cchem~ and mixing time Tm shows
that these quantities are of the same order cf zaagnitude.
For efficient chemical pumping (TChem~Trel ~1) under conditions of a mixing
GDL, conditions must be favorable for the chemical reaction. This requires
organization of rapid mixing of flows (Tm = Tchem~ to praduce a reaction mixture
that is uniform with respect to camponent makeup.
The composition of the mixture in the mixing zone was calculated ir~ the approxi-
mation of a model of instantaneous mixing of flows under initial conditions
- in the hot plug of po= 15 atm, To = 4820 K. The relative content of principal
components in the mixing zone was: 0.070 + 0.2102+ 0.14C0 + 0.37Ar + 0.21He.
22
FOR ORFICIAL USE GNLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500090006-1

APPROVED FOR RELEASE: 2007/02/49: CIA-RDP82-00850R440500090006-1
FOR OFFICIAL USE OR':.Y
The kinetic scheme of processes that take piace in the mixing zone is described
by the followi~g principal reactions:
0 CO -E-1~1-~ Cl)2 -}-111, (3)
Q -{-114 0_ At, ~4)
p pl =}-111-~ 0, -}-11I,
0 03 2U~,
CO 0_-~- CO.~ U,
where M= 0, 02, C0, COZ, Ar, He.
Calculation of the relative contribution Y made by sid~ reactions (4)-(7)
to the consumption of atomic oxygen showed that y comprises about 20% of the
overall consumption of atomic oxygen. The calculation used rate constants
given in Ref. 13, 14. Thus it has bPen demonstrated that recombination reaction
C0+ 0+ M under the conditions of the xperiment r~akes the pri.ncipal contribution
to consumption of atomic oxygen fortned in the ho': plug.
As a conse.quence of the strong dependence of t?~a rate of three-particle recom-
bination on particle density in the flow k- p~, the chemical reactions in
the flow are practically frozen after p~ssicig a single diameter downstream
from the critical crosG section of the nozzle. Therefore, comparison of rE:la-
tive concentrations of oxygen atoms in the hot plug and in the supersonic
flow (x = 50 mm) gives information on the completeness of chemical processe:a
with participation of atomic oxygen. It has been experimentally shown that
the relative concentration of oxygen atoms in the superson.ic flow is
= 0.005-0.008, which is considerably less than the concentration of .*_hese
atomG in the mixing zone ~p =0~07 at Tb= 4820 K). This result shows the
almost total completion of chemical processes wit:~ partici.pation of atomic
oxygen in the mixing zone.
~ I Fig. 4 shows the way that radiat.ion intensity in
- ~''~'i the 4.3 um band of C02 molecules as recorded in
- a supersonic flow (x = 50 mm) of reaction products
' depends on relative concentratic~n of oxygen atoms
in the hot plug. The quantit.y Iy.3 increases
j linearl with increasing in t:he range from 0.08
~ Y
n to 0.13. The increase in relative concentration
in the given experiments was achieved by increas-
Fig. 4 ing the initial temperature behi.nd the reflected
shock wave in the mixture of 0.402+ 0.6Ar. This
led correspondingly to an increase in translational tempei�ature T in the super-
sonic flow; however, as shown in Ref. 15, changing T has J.ittle effect on the
intensity of radiation in the /.3 u*.n band of C02 molecule:~.
ProbinR the flow of reaction products with e2ectric-discharge C02 laser radia-
tion in cross section x= 50 man revealed absorption of laser radiation
1b.6 um) with absorption coefficient Ko= 0.0-0.03 m 1. Vibrational temperatures
that describe the populations of levels of asymmetric (T3) and collective (T~;
modes of the C02 mclecules that are reaction products werP determined from ~he
23
FOR OFFiCiAL USE ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500090006-1

APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-04850R000500090006-1
FOR OFFICIAL USE ONLY
measured values of Ko and the intensity of spontaneous radiation in the 4.3 um
band of co2 molecules.
Vibrational temperatures were calculated for experimental conditions with
~ maximum (in the given experiments) content of atomic oxygen in the mixing
zone ~b - 0.07 (at To = 4820 K). The amount of energy stored in a unit volwne
of gas during recombination reaction CO + 0+ M is proportional to the number
of C02 molecules that are formed, and consequently to the initial concentration
of atomic oxygen Thus the conditions of calculation correspond to the
maximum contribution of energy from recombination reaction CO+O+M
to the total amount of energy of the gas, and consequently to the vibrational
degrees of freedom of C02 molecules.
The state of the gas in the cross section of the flow (x = 50 mm) to which
the calculated values of vibrational temperatures T3 and T2 are referred cor-
responds to : 0.06C02 + 0. 2302 + 0.09 C0+ 0. 39Ar + 0.23He, T= 190 K, p= 2. 5� 10-Z
atm. The values of the vibrational temperatures of C02 molecules were:
T3 = 2080 K, T2 = 1100 K.
The experimentally obtained high values of T3 evidence appreciable excitation
of vibrational levels of C02 as a result of reactions that occur in the mixing
zone. We will show that excitation of the asymmetric mode of C02 molecules
is due to the recombination reaction CO + 0+ M, and that the contribution of
other chemical processes is minor. Consider the kinetic scheme of chemical
processe s with participation of hydrogen-containing impurities H2(H20). Ac-
cording to the technical specifications of the degree of purity for the gases
used in the experiments (02, C0, Ar, He) the content of impurities H2(H20)
does not exceed 6�10-5 mole fraction. Additional purification of 02 and CO
was done by fractional distillation of the gases. Argon was dried in a trap ,
with liquid nitrogen (T= 77 K). With low content of impurities H2(H20), the
kinetic scheme of the reactions in. the system CO + 02+ H2(H20) in addition
to reactions (3)-(7) includes the following processes:
p H~ ~ OH -f- H, ~8)
kg
H OZ ~ OH 0,
~~~o
_ p I1~0 20IT, ~10)
1[.~ 11~pt~~11.: ! 01(, ~1~~
. u
"'a (12)
CO-}-OII~~CO.~~~- li,
O-~- II M i~--? OIi `i, ~1'~~
n14 ~1~~~
0,~ I I-~- nt iI02 ni.
24
FOR OFF[CIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500090006-1

APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500094006-1
FOR OFFI('IA1, l1SF ONI.Y
The values of reaction constants (8)-(14) are given in Ref. 13, ~4. Under
the conditions of the experiment we have the ratio of conc~=ntra~ions [O],
Io2l � IH2)(IH2o])�
Analysis of kinetic reaction scheme (8)-(14; can be conveniently divided into
two cases: a) hydrogen H2 is the predominant impurity in the initial mixture;
b) the principal hydrogen-containing impurity is from H20 ~nolecules.
In case a), calculation in the approximation of rapid establishment of quasi-
cteady concentrations of atoms H and r~dicals OH (d[H]/dt=:d[OHJ/dt = 0) with
consideration of conservatior. of the number of atoms H in a unit volume of
the reacting mixture written as
[ll] + [Olt] + 2UT~] 3l Il~la, (95)
leads to the rela*_ion
� (oll~ _ ~ ke ~ (1G)
~ ~ 11�~n ~-o ~e ~~1'?/r_n 1;~ Irblk D.
With variation of the parameters in the mixing zone T~ = 1660-1820 K, = 0.04-
0.07, the value of Z ranges from 0.16 to 0.18. Thus the content of OH radicals
is practically constant in the formulated experiments (assuming the same ini-
tial content of H2 impurities in the investigated mixtures).
Consideration of the other mechanism of formation of OH radicals with partici-
pation of water va~or impurities (case b) leads to the relation [OH] = 2[H20]
since chemical equilibrium in the main reaction (10) of this mechanism is
strongly shifted to the right. Analysis of both mechnisms of formation of
OH radicals shows that the OH content in the mixing zone remains ~onstant
with an appreciable change in conditions in the hot plug: To= 4000-4820 K,
F~ = 0.07-0.13.
If the vibrationally excited molecules of C02(v3) were formed in reaction
(12), the rate of formation of C02 molecules would be determined by the re-
lation
d ~co~ (~~)1 k~z (~~I [DII] _ lcoz (~~)1- ~~~a ~v~)1~ ~171
rrc T~
where T1 is the characteristic time of relaxation of molecules of C02(v3)
in the reaction zone, [C02(v3)] is the equilibrium content of C02(v3) in the
reaction zone at T~ = 1820 K, p~ = 15 atm. Under conditions of dynamic equiltb-
rium between processes of chemical pumping and deactivaticn of levels of
CO?(v3) (at d[C02(v3)]/dt= 0) we have the condition
[COL(v,)] _ [Cl)~(v,)l + Ir,~t',[C01 [Olll. (18)
The time of establishing this mode is ~0.02 us. From an examination of the
quantities appearing in the right member of equation (18) we see that the
values of T1, k12, [C02(v3)] vary weakly over a range of T~ = 1660-1820 K,
E,~ = 0.07-0.13. The concentration of OH radica?G. as shown by analysis of
kinetic reaction scheme (8)-(14) with participation of impurities of H2 (or
li~U) a~td CO molecules, remained almost constant in all experiments.
25
FOR OFFIC[AL USE ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500090006-1

APPROVED FOR RELEASE: 2007/42/09: CIA-RDP82-00850R000500090006-1
FOR OFFICIAL USE ONLY
Expression (18) shows that the content of vibrationally excited molecules
of C02(v3) should be unchanged in the mixing zone in the supersonic flaw.
The experiments showed linear dependence of I4.3(~~j) (see Fig. 4) at ~g =
0.07-0.13. Since Iy.3~ [C02(v3)J, comparison of the experimental plot of
I4.3(~~) and expression (18) demonstrates the validity of the assumption of
the insignificant contribution of reactions (8)-(14) with participation of
H2(H20) to formation of C02(v3).
The concentration of molecules of C02(v3) under conditions of quasi-steadiness
of reacti.on C0+ 0+ M; C02 + M and processes of vibrational relaxation satisfies
the rel.ation
~ Equation (19) shows that I4.3- [C02(v3)]~ ~g (or ~~j). Thus the assumption
of chemical pumping of C02(v3) during recombination reaction CO+O+M-~C02+M
is in agreement with the experimentally obtained plot of I4.3(~~)�
In the given analysis of the dependence I4.3(~~), we used results of infrared
diagnosis of vibrationally excited molecules of C02(v3); therefore the question
of the possible contribution of reaction (12) to formation of C02 molecuies
in the ground vibrational state requires separate consideration. When the
value of k12 from [Ref.] 14 is used, it is shown that the maximum content
of COz molecules formed in reaction (12) does not exceed 0.3X of the total
concentration of particles in the reaction zone. Thus the given analysis
of the kinetics of chemical processes in the system of reagents using equations
(3)-(14) and analysis of the experimental dependence I4.3(~~) indicate that
the main reaction leading to formation of C02 molecules is recombination reac-
tion C0+ 0+ M.
The values of T3= 2080 K found in the experiments correspond to a much greater
population of levels of the asymmetric mode, of C02 molecules that are the
reaction products of recombination (by a factor of 1.6) than when the levels
of this mode are populated in the mixing GDL based on inert mixtures of
C02+ N2+ He at initial temperature in the hot plug of To = 3850 K. The con-
siderable population (T2 = 1100 K) of levels of the collective mode of C02 mole-
cules is the reason for absorption of probing radiation (1~= 10.6 um), despite
the high values of vibrational temperature of the asyrmnetric mode T3. Reali-
zation of conditions that favor relaxation of levels of the collective mode
of C02�molecules that are reaction products should lead to population inversion
in the system of vibrational levels of thsse molecules.
Experiments were done with displacement of the probing region downstream to
the flow cross section situated at distance x= 95 m~ from the plane of the
critical cross section of the nozzle. In this case, amplification of the
C02 probe laser radiation was observed in the stream of C02 molecules. The
I-- - I
measured gains were =1.3 m~l (Fig. 5) . ~ - . ~ ,
In this research, experiments were done with ~ o._ _
partial replacement of the helium in the CO + He ~ I Fig. S
mixture by carbon dioxide injected from the ~.5
valve cavity. These experiments enable esti- 0 3 6~cot
mation of the influence that C02 molecules
26
FOR OFF'iCIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500090006-1

APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500090006-1
FOR OFFICIAL USE ONLY
formed in the initial stage of mixing of streams of 0+ 02 + Ar and C0+ He,
cind possibly having lost a considerable part of the vibrati_onal excitation,
have on the gain in the expanding flow. The action of thi:; mechanism was
simulated by introducing "cool" C02 molecules into the CO + He mixture.
The measured values of gain Kp=1.3 m 1 showed little chanf;e as ~~p2 ranged
from 0 to 0.03 (see Fig. 5) (~~p2 is the relative concentration of C02 mole-
cules introduced into the CO + He mixture as calculated in t:he approximation
of the instantaneous mixing model for streams of 0+ 02 + Ar and C0+ C02 + ci~) .
A further increase in concentration in a range of ~C02= 0.()3-0.08 led to a
reduction in gain from 1.3 to 0.6 m-1. Since the rate of r.eaction 0+ C02-~
CO + 02 is small, introducing "cool" molecules of C02 into ~~he mixture injected
from the valve cavity h3s little effect on chemical equilil~rium in the system.
The observed reduction in Ko with increasing ~~p2 is thus ~iue to the partici-
pation of C02 molecules introduced into the CO + He mixture in relaxational
V-T and V-V' processes.
Thus in this research we have realized chemical pumping of levels of the asym-
metric mode of C02 molecules populated in the course of re~ombination reaction
CO + 0+ M. Population inve~sion was achieved in the system of levels (00�1-
10�0) of C02 molecules formed in the c:ourse of this reacti~n under mixing
GDL conditions.
REFERENCES
1. Biryukov, A. S., TRUDY FIZICHESKOGO TNSTITUTA IMENI P. N. LEBEDEVA,
Vol 83, Moscow, Nauka, 1975.
2. Carrington, T., Garvin, D., in "Vozbuzhdennyye chastitsy v khimicheskoy
kinetike" [Excited Particles in Chemical Kinetics], Moscow, M:ir, 1973.
3. Koopmann,8. K., Saunders,A.R., J. QUANT. SPECTR. RAD. TRANS., Vol 10,
1970, p 403.
4. Anderson, J. D., PHYS. FLUIDS, Vol 13, 1970, p 1983.
S. Kroshko, V. N., Soloukhin, R. I., Fomin, N. A., in: "Gazovyye lazery"
[Gas Lasersj, Moscow, Nauka, 1977.
6. Ktalkherman, M. G., Mal'kov, V. M. et al., FIZIKA GORE,NIYA I VZRYVA,
Vol 15, 1979, p 6.
7. Myers, B. F., Burtle, E. R., J. CHEM. PHYS., Vol 48, T~o 9, 1968, p 3935.
8. Pravilov, A. P., ZHURNAL FIZICHESKOY KHIMII, Vol 52, rio 8, 1978, p 1863.
9. Clyne, M. A., Th�ush, B. A., PROC. ROY. SOC. LONDON, Col A269, 1962, p 404.
10. Gilmoure, F., J. QUANT. SPECTR. R.AD. TRANS., Vol 5, 1965, p 369.
11. Kudryavtsev, N. N., Novikov, S. S., Svetlichnyy, I. B., KVANTOVAYA
ELEKTRONIKA, Vol 6, No 4, 1979, p 690.
27
FOR OFF(CIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500090006-1

APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500090006-1
FOR OFFICIAI. USE ONI.Y
12. Kompaniyets, V. Z., Ovsyannikov, A. A., Polak, L. S., "Khimicheskiye
reaktsii v turbulentnykh potokakh gaZa i plazmy" [Chemical Reactions
in Turbulent Flows of Gas and Plasma], Moscow, Nauka, 1979.
13. Kondrat'yev, V. N., "Konstanty skorostey gazofaznykh reaktsiy" [Rate
Constants of Gas-Phase Reactions], Moscow, Nauka, 1974.
14. Baulch, D. L., Drysdale, D. D., Horue, 0. G., "Evaluated Rinetic Data
for High-Temperature Reactions", Vol 1, 2, Butterworths, 1972.
15. Kudryavtsev, N. N., Novikov, S. S., in: "Materialy vtoroy Vsesoyuznoy
konferentsii po metodam aerofizicheskikh issledovaniy" [Materials of the
Second All-Union Conference on Methods of Aerophysical Research], Novo-
sibirsk, 1979.
COPYRIGHT: Izdatel'stvo "Nauka", "Fizika gorenlya i vzryva", 1982
6610
CSO: 1862/184
28
FOR OFF[CIAL USE ONLY Y
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500090006-1

APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500090006-1
MOR OFMI('IAI. USE ONLY
OPTOELECTRONICS
THEORY OF DYNAMIC IMAGE SELECTION EFFECT IN PHOTOREFRACTIV.~ MEDIA
Leningrad FIZIKA TVERDOGO TELA in Russian Vol 24, No 2, Fe~ 82 (manuscript
received 16 Jul 81) pp 337-343
[Article by V. V. Bryksin, L. I. Korovin, M. P. Petrov and A. V. Khomenko,
Physicotechnical Institute imeni A. F. Ioffe, USSR Academy of Sciences, Lenin-
grad]
[Text] A model of a photorefractive medium is proposed
that l.eads to the phenomenon of dynamic selection of images
--flashing of images in the recording light after it is
switched off in the geometry of a PRIZ optical electro-
modulator. The results of the theory for diffraction ef-
ficiency of the medium as a function of the time of exposure
of the recording beam and the frequency of its spatial modu-
lation agree qualitatively with experiments done on Bi12Si020
crystals.
~ I.?ve4ti~arton of pr~c~~sps of ortical data +-ecording in pb.otorefractive crys-
tals has led to the detection of an unusual efYect suppression of statiu?~ary
parts of images being recorded, and isolation of the nonstationary parts [Ref.
1]. This phenomenon, which has been termed the effect of dynamic selection
of images, may be treated under certain conditions as time: differentiation
of an unsteady pattern presented to the crystal, whereas t.he response of light-
sensitive media is usually proportional to the integra*.ed action of an optical
input signal. Experimentally, the effect of dynamic selection of images has
been observed in image recording on a thin (=0.5 mm) B12Si�020 single crystal
plate cut in plane [111] or [110] when an electric field af 1-2�104 V/cm is
applied to this plate by transparent electrodes applied ta the faces of the
. plate parallel to the given planes. If an image in blue or green light is
presented under these conditions to the given structure (which is called a
PRIZ light raodulator), with readout at the same time in polarized red light,
the readout light will show mainly the nonstationary parts of the recording
ima~e. The principal characteristics of this experiment are readout light
intensity of =10-5-10-4 W, rate of change in input information =1-0.1 s,
de~ree of suppression of stationary part of the image 10-20 dB.
Experiments done with the PRIZ modulator show that image r.ecording arises
due to excitation of photocarriers by the recording light, the charge
29
FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500090006-1

APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R400500090006-1
FOR OFFICIAL USE ONLY
redistribution being proportional to the distribution of intensity of the
recording light. The resultant charge distribution causes electric fields
to appear in the crystal, giving rise to spatial modulation of birefringence.
The polarized readout light undergoes a change in the state of polarization
due to the transverse electro-optic effect. It has been experimentally estab-.
lished that with slow changes of input signal intensity J(x, y, t), the phase
difference ~~(x, y, t) between the ordinary and extraordinary readout beams
at the output of the device is proportional to dJ/dt [Ref. 2]. Although at
present the relation between the e3.ectro-optical characteristics of the crystal
and the condition of image recording has been fairly well established both
experimentally and theoretically [Ref. 3], and we also have some idea of the
nature of charge distribution under steady-state conditions [Ref. 2], nonethe-
less the dynamic behavior of the system leading up to the effect of dynamic
selection of images has not been studied, and there are not even any reliable
qualitative ideas about the nature of this effect.
This paper is the first to present a theoretical analysis of a model picture
that allows us to establish the ma~or causes leading to tfie effect of dynamic
selection of images. ,
1. Qualitative Picture of the Phenomenon
When the transverse electro-optic effect is used, the phase difference ~~(x,
y, t) is proportional to the integral characteristic I(x, y, t) of the trans-
verse field El that is set up in the specimen [Ref. 4]
do
l(x, U. = f~~ lt. Y, z. dz. ' ~1~
u
where the z-axis is opposite to the directior~ of the external field Eo applied
to the specimen, and do is the thickness of the specimen. An important charac-
teristic of the photorefractive medium is diffraction efficiency r1� In the
case where e� < 1, it can be considered with fair accuracy that r1~ (~~)2, and
consequently r~ is proportional to the square of the amplitude of modulation
of the quantity I(x, y, t).
9
do
~ T
' ~ ~ 2 1
~ ~
~ ~o
~ ~ ~
~ ~ ~
~ f!r ~ Ey a z
Fig. 1. Schetaatic picture of the change in direction of
transverse fields Ey in time. Solid arrows in regions 3
and 4 indicate the direction of Eq under lighting, and the
broken arrows indicate the direction of Ey after lighting
is switched off. The direction of the light beam is shown
to the left of the y-axis. It is assumed that light is
completely absorbed at distance a-1.
30
FOR OFF[CIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500090006-1

APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500090006-1
FOR OF'FICIAL USE 01LY
The qualitative picture of dependence of I on time t, which is what determines
the effect of dynamic selection of images, can be explained in the following
way. To simplify the discussion, let us asstune that there are illuminated
and unilluminated sections of the plate surface. In an illtaninated section
(region 1), after the light is switched on, photoelectrons and ionization
donors appear (see Fig. 1). Charge separation begins in an external field
Eo. If diffusion processes are disregarded, the space charge will extend
over the illuminated regions along the z-axis (region 2). Assuming that dif-
fusion takes place slowly compared with motion in the external field, the
entire process of establishment of the steady state in the system can be broken
down into two stages: establishment of a quasisteady state in regions 1 and
2, and spreading of the charge in direction y on the second stage (into regions
3 and 4). On the first stage the negative charge is balanced that consists
of the electrons in the band and those captured by traps in regions 1 and 2.
This balancing under conditions of electroneutrality of the specimen leads
to charge separation. In the process of charge separation, region 1 is posi-
tively charged with respect to region 1; the transverse field is directed
from 1 toward 3(solid arrows on Fig. 1). In region 2, which is negatively .
charged, Ey takes the opposite direction. Conditions in an experiment are
usually such that ad~> 1(a is the coefficient of absorption), and therefore
region 1 is smaller than region 2. !~s a result, the shielding action of the
metal electrodes (shielding length of che order of a period of spatial modula-
tion Zo) reduces the total field E� more strongly in region 1 than in region 2.
Thus on the first stage the main co?:tribution to I(t) is from the field between
regions 2 and 4. This stage is characterized by an increase in image bright-
ness.
On the second stage, the negative charge of electrons in the band and on traps
is equalized by diffusion along y. In such equalization the contribution
to I(t) from the negative charge gradually disappears. The quantity I(t)
begins to decrease in absolute value, vanishes, and after sufficiently long
times reaches a steady-state positive value that is determined by the contri-
bution to field Ey from nonuniformly distributed positively charged donors.
As we can see from the following, image storage takes place on this stage.
After lighting is switched off, the system arrives at the initial equilibrium
state, and free electrons in region 1 recombine on donors. Recombination
will continue due to electrons from region 2 that enter region 1 through the
external circuit. And if diffusion in the transverse direc[ion is the slowest
process, no negative charge will remain in regions 1 and 2, whereas a positive
charge will remain in region 1 since some of the electrons are distributed
in regions 3 and 4. On this stage, a transverse field arises between regions
2 and 4(broken arrows on Fig. 1) so that the directions of the field coincide
between regionc 2, 4 and 1, 3. This leads to a further increase in I(t) as
compared with the steady-state value. Diffusion from regions 3, 4 into 1, 2
begins to reduce the field Ey, and the electrons entering region 1 recombine
on donors. As a result this leads to total disappearance of the local charge
and the internal field.
31
FOR OFFICIAL liSE ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500090006-1

APPROVED FOR RELEASE: 2007/42/09: CIA-RDP82-00850R000500090006-1
FOR OFF'iCIAL USE ONLY
2. Description of Model and Principal Equations
Let us consider a dielectric (photoconductor) with two systems of discrete
levels: donor levels with concentration nd and trap levels with concentration
nt. Of all possible transitions between levels and the conduction band, in
future we will consider transitions from donors in the conduction band under
the action of light and recombinations from the conduction band on donors,
and also capture of band electrons by traps and thermal e,jection of electrons
from trap levels into the band. With consideration of this, the system of
equatior,s studied below for determining electron concentration n in the band,
the concentration of ionized donors n,~ and charged traps n_ can be written as
ed (n n-- ?++)Idt - e�Eodn/d:. pkTd=rei ~J'~ l~)
dn+ld~ =8e ~to - t) F (z. Rd - n+/ ~d, (3)
d n_~d t- nn t f Jt - n_; c!, (4 )
Here e is the modulus of the electronic charge, u is carrier mobili[y in the
conduction band, gF(z, y) is the rate of generation of electrons from donor
levels under conditions of nonuniform illumination. 6(ta- t) is the Heaviside
function (step) that switches off the light at time to; light is switched
on at time t= 0. It is assumed that the light is modulated along the y-axis.
Processes of recombination and capture in equations (3) and (4) that are non-
linear in nature are approximated by linear processes: recombination on donors
is described by term n+/zd, and capture on traps is described by term nnt/9t
(rd is the effective time of recombination on donors, At is the constant of
capture on traps). This approach disregards depletion of donor levels and
the nonlinear term that limits filling of traps. These approximations are
valid when conditions r~,�nd, n_ �nt are met (n_ is the concentration of
charged traps), which corresponds to illumination levels that are not too
high. After switching off the recording light, the system can be returned
to the initial state only if the electrons are moving in a direction perpen-
dicular to the external electric field ~o. In this connection, equation (2)
takes consideration of electron diffusion along the y-axis. Accounting for
diffusfon processes along the external field should not cause any qualitative
change in the results. The right-hand part ef equation (2) is the divergence
of current density in which the term n7~ is omitted, i. e. the influence of
internal fields on electron motion is disregarded. Here ~ is the potential
of the internal fields, i. e. the field~ that arise as a result of charge
redistribution after discounting field Eo. The distribution of electric fields
is determined from Poisson's equation
_ (4~e~ t) (n - n- - n+).
where e is the permittivity of the crystal. Let us note that equations (2)-
(4) are a linearized version of the general system of nonlinear equations
that describe the way that internal fields in the crystal depend on time and
coordinates in the presence of spatially modulated ill~mmination and under
conditions of current passing through the specimen [Ref. 5).
32
FOR OFFICIAL liSE ONL~'
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500090006-1

APPROVED FOR RELEASE: 2007102109: CIA-RDP82-00850R000500090006-1
FOR OFFICIAL USE (3NLY
For further analysis it is convenient to convert to dimensionless variables
IV - n/nd, N* = nt/n~, C= az. a!/' ~
~ = �EoaSd, L = PkTastd~e. ~ = n!'~dl8t~ ~ ~s~
w = Td~4j, S = t/td, ~ _ ~a 'c~~~lEtnd).
in which equations (2)-(5) take the form
J (IV 1V- - N+)Id~ - ~JN~dC bd"N~d~2,
JN+Idt =89 (to - 21 '~dF ~C, E1 - N+~
dN_/di = QN - ~N_~ To = ~o~'~d~
[(v21vC=) (d`I~Es)1 ~ = N N-' N*'
It is assumed below that lighting intensity is modulated by a cosine law,
and damps out exponentially into the plate, so that F(~, is equal to
F~ (C, EI = exp (-C1 I1 b cos (x. E)1, x= 2a~(alo)� (a)
3. Integral Characteristic of Transverse Field
In virtue of the linearity of equations (6) and the selected form of light
modulation, al~ unknown quantities (concentrations of electrons, ionized donors
and charged traps, and also the internal field potential) are representable
as A+ B cos K~, where A and B do not depend on Since a contribution pro-
portional to A does not lead to a transverse field, we consider below only
the contribution proportional to B, i. e. the first (and in this sense the
only) Fourier component of the internal field. Therefore only the first har-
monics of the concentrations and potential are considered below (without chan-
ging the symbols).
In the internal transverse electro-optic effect, the diffYaction efficiency n
of the light-sensitive medium is determined by modulation of the phase dif-
ference Therefore we investigate below the first harn!onic of the dimen-
sionless characteristic of the transverse field (see formula (1))
e
I = x ~ 'I' (C, dC~ d = aao� ~9~
0
The solution of system of equations (7) is given in the Appendix. I(T) takes
the f orm
1t~I~bBTd~~ - S ~~I O (TO - S~ - ~W ~S~ - W ~T T0~ ~ ~t - `0.'~~ \10~
where W is the sum of the exponentials and of the oscillating part
W(t) (tI C~P~) exP t-Poi) C~Po) eZP I-PoT) C~ exP (-S) I~1)
(analogously for W~T- Tp~~. S(1) (see II, 6Z corresponds to the steady-state
value of I, i. e. when To-~�� and T-~~, and W is the sum of the oscillating
terms -
4G m x=b i~tlk) (exp ~-P~T) eXP ~-�PkT) ll (!2)
tiZ (t) d~ Re {~1- tak) (x` -I- a~) L 1(P~) + 1 lNk) J1'
k-1
J ~P) = P ~1 - P) -I- Qw (w - P)~5~~ llk ~ 2ak~d~ ~ (13)
G = ~i - exp (-d)] tl~ (xd~2).
33
F'OR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500090006-1

APPROVED FOR RELEASE: 2007/02109: CIA-RDP82-00850R400500090006-1
FOR OFFICIAL USE ONLY
The period of oscillations and damping of exponentials in (12) is determined
by quantities pk
p~ - 2-~ {0 u~ ic'b t~8k d( W-}- x~- ~ 8kj- - 4~ (s L+ i k)}� (!4)
The nature of the oscillations of I(T) is analogous to that described in
Ref. 5. Coefficients C and C1 are equal to
~(a) = a tt2Gld) - x ~t - exp (-a)111-' ca). (!5)
Cz = ~t - 9~-1 {[G ctl~ (qd/2) - x9 - exP ~-d~~~ ~4 x=a/~~ ~4= - s=]-i
I1-i- x'a1HI S (1I}, (!6)
where q= q(-1) is given in the Appendix (II, 1). Curves of the time dependence
of I calculated by formulas (10)-(16) are given on Fig. 2 and 3.
/�~0i x0
1 r .
~ I
,t ~ ~
J l \
0 _ _1 ~
' 6 s t~,o�~ as .
~ ~ ~ ~
,l _ - ~
I ~ -
-3 l ' ` 6i.f0~~
~ I
I .
I
Fig. 2. Time dependence of integral _p.s ~
characteristic of transverse field
at different values of the external
field Eo. Values--S: 1--10-4, 2--
S�10-'', 3--~; w= 10'S, S2= 10-2, aZo =
100, d= 10, d= S� 10-3, bg Td ~ 10-~. Fig. 3. Time dependence of integral
Curve 3 increases from 0 to the extre- characteristic of transverse field
mum value in time T~52-1 at the se- at different times of switching off
lected values of parameters in time the li~hting. Values of T~: 1--
T~ 100 0.4�10 , 2--1.5�10~, 3--3.~�10'.
~olid curve corresponds to To+
w= 10-5, St= 10-2, aZ~= 10, d-~~,
d= 10-4, S=-10' , bg Td a 10-~
4. Discussion of Results and Comparison With Experiment
As can be seen from Fig. 2 and 3, the proposed model leads to a time dependence
I(T) that in general features coincides with that found in section 1 from
qualitative considerations. If conditions are met that preclude oscillations
of the quantity I, i. e. for the interval of fairly weak and f airly strong
fields (Fig. 2), then the time dependence of I2 behaves as follows. I2 passes
through a maximum, vanishes, and then settles to a steady value. The vanishing
of the I2 should show up in the experiment as a single flicker of the image.
A rather signif icant property of this model is the appearance of a max3mwn in
34
FOR OFF7CIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500090006-1

APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500090006-1
N'OR OFMI('IA1, l1SE ONI.Y
the response of the medium to switching off illumination, e. one of the
manifestations of dynamic selection of images. The maximun of the intensity
of the flash in the readout light after switching off the iighting depends
strongly on the switch-off time, and for small switch-off times practically
disappears. For example the ratio of the maximum of I2 for. curve 3 to the
maximum for curve 1(shawn by broken lines on Fig. 3) is 3Q.
Let us now turn to analysis of the dependence of I2 on the frequency of spatial
modulation of light K/2~ (let us recall that I2 determines the diffraction
efficiency r1 of the medium). Since the function I(K) is p:irametrically depen-
dent on time, it is convenient to study the dependence on at small T, where
I(T) depends linearly on time. If we assume that w�S~, d
~
L ~ p~ ~3 . Pi ~ u
' � c
3: j(x) f a a~i f ~
L~ ~ . L: Pl 3~iW Ps c:.
u
Fig. 6.5. Optical schemes of
� spatial filter formation: L--
L' ~ lens; 3--~mirror; P1--plane of
1. � a -
: 3~ ~ 3 assignment of function f(x) ;
. ~ . = P2--plane of registration of
holographic filter
where Uo(u) = Uq(u)~exp i�(u) is the field distribution in the reference beam;
F(u) _ ~F(u)~e~e1u~ is the spectrwn of the sought signal. We can get the neces-
sary value of F*(u) by varying the value of ~Uo(u)~ and ~(u). By combining
the resultant hologram with a positive that has transmission 1/N(u), we con-
struct the filter
T (u) = IUo (u)I~ + I~''(~)I~ + ~F +VoF'
N (uj . N (s) N N ~
If (U~(u~=C and ~(u)=6u~ then
= A (u) H ~ (u) e''~ yle+~a~ .
N (u)
where q ~u~=1Vo (~)I~ + IF (++)la ~ f~ ~u~= CF' (u) ~
~ N (a) N (u)
C and b are constant coefficients. The third term of this expression corre-
sponds to the transfer function of the sought f ilter. The filter is set in
plane P2 of the coherent optical system (see Fig. 6.2), the signal to be ana-
lyzed being formed at the input of this system (plane P1). In observation
plane P3 the field distribution g(x) is equal to the Fouri.er transform of
the product of spectrum F1(u) of input signal fl(x) and the filter T(u)/N(u),
i. e.
g(x)'=2a F~ ~u) 4(u) e~r.s ~..2n S Fi ~u~ N~ (u) e'~~`-0j"~?u -
-r ~M
Y
+2a ~ Fi ~u) Hi ~~t) e~~~-�~~`du. (6. 6)
The first term of expression (6.6) characterizes distribution localized near
the optical axis of the system, and in the given analysis is of no interest.
The second term characterizes the filtered signal, the field distribution
being localized near a point that is displaced relative to the optical axis
by an amount x= b. It will have the maximum value when the spectra F1(u) and
47
FOR OFFICIAI. USE ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500090006-1

APPROVED FOR RELEASE: 2007/42/09: CIA-RDP82-00850R000500090006-1
FOR OFFICIAL USE ONLY
F(u) coincide. The field distribution characterized by the third term will
be shifted relative to the axis by an amount x=-b. With appropriate selection
of the values of b, all three patterns are separated in space relative to
one another. In the simplest case of steady-state white noise N(u) = const
ve observe the function of cross-correlation and the function of convolution
of ~he useful input signal with the signal recorded on the filter simultane-
ously in plane P3:
-
f~~E)f' ~x-I-b-~-E)~tE=~f~~x-~b) ~E I' ~x-I-b)]; ' . (6.7)
-w
f f~(E)f ~x-b-E)dE=lf~~x-b) ~E f (x-b)J. ~ (6.8)
If fl(x) contains a series of signals that are separated in space, then the
maximwn value is observed for the signal that is determined by the autocorre-
lation part of expression (6.7). In this way the problem of recognition and
detection of the sought ob~ect is solved. When several filters are recorded
on the same light-sensitive medium, several signals can be recognized simul-
taneously without using the operation of inechanical scanning or change of
f iiters [Ref. 87].
Note should be taken of some peculiarities inherent in the given method.
Translational displacement of the input signal leads to the same displacement
of the signal of convolution or correlation in the observation plane. A change
in scale of the signal without altering overall intensity leads to signal
attenuation. In an analogous way, filtration is sensitive to rotations of
the initial signal. The permissible displacements and rotations of filters
depend to a considerable extent on the nature of the sought signal [Ref. 190].
Most often, the problem of optimum filter placement is resolved by changing
the scale and angular orientation to maximize the signal at the output of
the system. Unfortunately, such an operation consumes considerable time.
The given method has been used to recognize a known shape or part of a text
against a background of a complicated transparency [Ref. 187]. It has been
suggested that the method can be used with appropriate updating for high-
speed recognition (reading) of printed symbols [Ref. 124, 131J.
An interesting possibility has opened up with the use of matched optical fil-
tration for correcting the direction of flight of an aircraft [Ref. 191].
In this technique, the signal coming from the radar screen of an aircraft
flying at a certain altitude (radar map of the region) is compared with the
signal recorded on a hologram filter (marker inside the given territory).
The same method has been used for recognition of markers in spaceflight [Ref.
142]. An advantage of the method is simplicity of recognition. It is only
required that the ob~ect to be recognized fall within the field of view of
a telescope carried on the spacecraft.
The input optical image can be made in the form of an ordinary transparency
that provides amplitude modulation of the radiation passing through it. An
optoelectronic holographic device can be used in the same way witti deformation
of a ref lecting surface, which enables phase modulation of the incident light.
48
FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500090006-1

APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500090006-1
FOR OFFICIAL USE ONLY
The experiment used photographs taken on missions of the Gemini-4 and Gemini-5
spacecraft. The setup of the experiment was simi~ar to that diagrammed on
Fig. 6.2. The light source was a helium-neon laser.
Landmarks were recognized in eight regions characterized by different dimen-
sions and terrain from photographs of the northeastern part of the Arabian
Desert. In every case, recognition was successful. The signal/noise ratio
was in a range of 100-250. It was established that high contrast of objects
is not a mandatory requirement for reliable recognition.
Recognition of individual valleys and even sections of desert was completely
satisfactory.
Recognition was possible even with photographs of poor quality. It was suc-
cessful even when drawings and maps were used for making the spatial filters.
It was established that recognition of an object is possible even with cloud
cover of 90X of its surface area. This effect can be attributed to the fact
that the intensity of luminescence of points in the plane of recognition de-
pends nonlinearly on the area of the given object.
The dimensions of the optical image and the spatial filter~ may differ by 15%
without destroying reliability of recognition. There are methods that can
ensure recognition even if the dimensions of the c or resp ond in g images differ
by a factor of ten. The same methods can be used for dete:rmining such naviga-
tional parameters as altitude and slant range. Spatial fi.lters have been
used for recognizing stellar objects, which are ideally suited for this purpose.
In this case, the signal/noise ratio reached 400.
To guarantee a predetermined error, the focal length must be increased with
increasing distance of the vehicle above the surface of the planet. For the
described system, the error did not exceed 10-``. A source of error is the
limited resolution of the vidicon; the magnitude depends directly on the dimen-
sions of the field of view of the telescope, and can be re~duced by a correspond-
ing increase in focal length. In the case of flight at an altitude of 160 km,
the focal length can be taken as equal to 15.2 cm. Then r.he angle of view
of the telescope is 6�, and the magnitude of the error in measuring the position
of a landmark does not exceed 15 m. At higher altitudes, the same precision
can be achieved by increasing the focal length.
Methods of optical filtration and holography are used to :~olve the problem
, of posterior correction of photographic images, which in principle is analogous
to construction of an ideal linear system [Ref. 166]. To do this, it is neces-
sary to know the scattering function of the forming system in the presence
of distortions. It has been possible in practical cases to improve the images
of objects such that each point has undergone aberrations that are the same
in the statistical sense. This means that the distortions should not disrupt
the spatial invariance of the forming system. From the expression character-
izing the structure of the image on the photographic plate as recorded in
the spatial region,
8' ~x) _ ~ ~x) ~K ~h ~x)I' (6. 9) .
49
FOR OFFIC[AL USE ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500090006-1

APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500090006-1
FOR OFF'ICIAL USE ONLY
and the corresponding expression for the frequency region
O ~u)= ! (u) L (u), ~ � I6:10)
where
~ L (u)= ~ {~h (x)r),
we see that to isolate the sought image it is necessary to have a matched
filter with transmission function T= L*/~L~2, since GT = I. Here I(x) is the
distribution of intensity of the undistorted image; ~h(x)~z is the scattering
function of the shaping system. The component L* of the filter is made by
recording the spectrum of a point source (positive) obtained in a system with
aberrations on a Fourier hologram. Component.~L~2 is formed by recording
the pawer spectrum of this image on a photographic plate placed in the focal
plane of the corresponding lens. By installing filter T in plane P2 (see
Fig. 6.2), and the positive to be corrected in plane P2, we get an improved
image in one of the diffraction orders ir~ plane P3. It should be emphasized
that the giv~~n method corrects images only within the limits of those spatial
frequencies that have passed through the system. Thus at the output one ob-
serves a signal of the form
8~ ~x)= / (z) ~E ~xH'~ ~ (6.1 I )
where ~hl(x)~2 is the scattering function (intensity-response function) of
the corrected system. The difference between a system with transfer function
L1(u) _~{Ihl(X)IZ} and a system with transfer function L(u) _~{~h(x)~2} con-
sists in the fact that in the latter system the spatial frequencies are shifted
in phase and have lower amplitude than those of system L1(u). The existence
of a finite number of zeros in function L(u) has no significant influence
on the effect of filtration.
If the photographic image 'nas been obtained in the presence of a steady-state
turbulent medium and the exposure time is much greater than the time of fluc-
tuations, the photographic emulsion fixes the average intensity distribution
[Ref. 166]: ~
~8 ~x)) ~ ~ ~x) ~E ~ I~+i ~xH'~�
The scattering function of the system, assuming quasimonochromatic illumination,
is written as
~ ~x1~' ) _ ~ Y ~x) ) Ihe ~z)~',
where is the averaged function of mutual intensity in the plane of the
entrance pupil of the forming system as determined by the point source in
the object plane; ho(x) is the response function of the forming system. In
the frequency region we have
(o(u))=!(u) (I'(u))IH(u)I'
50
' FOR OFF'ICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500090006-1

APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500090006-1
FOR OFFICIAL USE ONLY
or, assuming that all spatial frequencies characterizing the ob~ect are re-
corded,
(~(u)) =~(u) (r(u))�
It is clear from this that to realize the operation of matched filtration
we must have a filter of the form
. T(u)=c/(r(u)>.
In the general case, its realization necessitates the use of hologra~ns; how-
ever, for most cases, such as inhomogeneous steady-state media, Y(x) is a
real and even function. Consequently, conventional photogr~phy can be used
for recording .
The effectiveress of this method has been experimentally studied [Ref. 162].
A turbulent steady-state medium was modeled by heating air in the region
situaced between the object and the shaping lens. The image of the object,
and also an image of a point source located in the center of the ob,ject plane
were fixed on photographic emulsion. The matched filter was made by photo-
graphing the spatial spectrum of the image of the point source and processing
the photographic emulsion with contrast coefficient equal to unity. The image
was corrected in the system diagrammed in Fig. 6.2. The e~xperiments showed
that the quality of the photographic images is considerab].y imgroved if the
blurred scattering function is gaussian.
The idea of posterior correction of images was further developed in Ref. 178.
Let the blurred image of an object be defi:~ed by expressic~n (6.11). If signal
g(x) passes through a system with response function hl*(x), where
. r
Ih,(x~'~E hi(x)= ~ ~~(E)I'h'~x-~-E)dE=a(x), ~s. iz~
_m
then we get a deblurred image characterized by function I(.x). Condition (6.12)
for the frequency region is written as LH1* = 1.
Realization of the proposed method is optically accomplished by two techniques.
In the first case, the function g(x) is recorded on a Fourier hologram. As
a result, we get a transparency with transmission function
'(rc)=1-I-~(il'-}-O-~-Q'=1-~~Q~'-F-IL-~-I'L'. (6.13)
Placin~ this hologram in the original position and illuminating it with wave
H1, we isolate the field distribution characterizAd by the last term of equa-
tion (6. 13) : I*L*H1 = I*.
posi ive observation
Fig. 6.6. Diagram of image ~holo am PZane
deblurring using a pcsitive
T
h, ~x.rl ~
51
FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500090006-1

APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500090006-1
FOR ONNI('IA1. USE ONI.Y
As a result, an improved image is reconstructed (Fig. 6.6). Wave Hl~ulaced
formed by illuminating the (positive of) image of the function hl(x) p
at the location of the point source.
In the second case, a Fourier hologram of the image is recorded with an extended
reference source hl(x). The transmission function of the holograms is
T (u) - (O tl ~a -I-~H~)' =~~P IH~P 1 LH~ /'L'H,.
If LH* = 1, the corrected image is formed in a beam of light characterized
by the term ILH* = 1.
For this purpose, the resultant hologram is placed in the original position
and illuminated by a wave from a point source situated in the location of
the extended source hl(x) (Fig. 6.7).
� , observation The given technique has been used to
holo ram Plane reconstruct lensless Fourier holograms,
~ which were produced by using an extended
object in the form of a frosted plate
. ~ , and a set of point sources (a two-
dimensional grating of 9000 point
diaphragms [Ref. 179]). Reconstruction
Fig. 6.7. Diagram of reconstructing was done by the arrangement diagrammed
deblurred image from a hologram in Fig. 6.7 using the same reference
sources. Since condition (6.12) was
satisfied by using exten~ied reference sources of the indicated type, an image
of good quality was observed. The experiments showed that the setup is quite
critical to exact placement of the hl(x) transparency. With a mismatch of
-8 um, the reconstructed image was completely suppressed by noises [Ref. 719].
It should be emphasized that the effectiveness of the proposed techniques
f or image deblurring depends in large measure on the accuracy of ineasuring
the scattering function of the system. Only in the case where ~hl(x)IZ is
known and does not vary from experime~nt to experiment (steady-state system)
can a matched filter be formed or the function hl(x) be selected. In principle,
the scattering function is calculated on the basis of certain assumptions
relative to the statistical characteristics of introduced distortions. How-
ever, such an approach has so far had little success in practice [Ref. 143].
The methods considered above have applied. to classical methods of optical
filtration. The technique of holography has only appreciably simplified the
formation of optimum filters. The principles lying at the basis of holography
have led to the development of new methods of optical filtration that had
not existed before. Let us consider a relatively simple method of compensating
phase distortions based on the property of holograms to reconstruct both a
real image and a complex-con~ugate virtual image [Ref. 21, 148, 153].
An inhomogeneous optically transparent mediinn is described by the spatial
~rariation of the index of refraction n(z). For inhomogeneities with geometric
dimensions much greater than a wavelength, the scalar ea:sation of propagation
52
FOR OFFICIAL LJSE ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500090006-1

APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R004500094006-1
FOR OFFIC'IAL USE ONLY
takes the form
0'u k'n' (z) u= 0, (G. I 4)
where k= 2~r/J~, u is complex field amplitude. For waves propagating at a small
angle to the z-axis, the solution of equation (6.14) is written as
u=f ~x)e re=, (6. 15)
Here f(z) is a slowly varying function, R=const. The fan~ilies of rays charac-
terized by solution (6.15) and by the complex-conjugate solution have the
same optical path, but propagate in opposite directions. Therefore if a wave
front that has been scattered and distorted by the inhomogeneous medium is
f ixed on the hologram, and then during reconstruction a cample~-con3ugate
wave front is passed through a system that introduces the same distortions,
the phase aberrations of the wave front are compensated.
This property is extensively used for eliminating lens abe:rrations [Ref. 182J,
and for suppressing noise in interference studies [Ref. 132J.
Investigations have been made of the effectiveness of us'_ng this method
to compensate atmospheric distortions [Ref. 109].
An optical diagram of the facility is laser
shown on Fig. 6.8. The reference 0 ~ MN
beam was formed by objective lens
0 and collimator C, and directed n~, , B
by mirror M2 to hologram T'. The =
� working beam bassed through beam , f
splitters B1 and B2, through a ~
glass plate with thickness of 2 cm
with inhomogeneous air pockets,
and was incident on target T. A `
p flat mirror was used as the target. . M
- Part of the radiation ref lected by MJ BJ ~ 71, = T~ .
the target passed a second time
through the distorting medium and Fig. 6.8. Diagram of experimental
beam splitter B3 and was incident setup for studying t:he method of com-
on hologram I'. In this way, the pensating atmospher:Cc distortions:
actual conditions of observation 0--objective lens; (:--collimator; M--
of objects through the atmosphere mirror; B--beam spl:itter; D--inhomo-
were simulated for. ill~unination geneous glass plate; T--target; I'--
by a coherent light wave scattered hologram; F--plane of interference
by inhomo~;eneities. On the recon- pattern ot~servation
struction stage, the hologram was
il.luminated by the wave reflected by mirror M1. The complex-conjugate image
of the distorting mediucn wa~ projected in plate D. As a~-esult, phase aberra-
tions were compensated, and an image of the target illumiiiated by the scattered
light wave was observed. To bring the directions of propagation of the working
and reconstructing beams into register, a cyclic interferometer was used (the
Hari-Haran interferometer [Ref. 127]). Alignment accuracy was checked by the
interference pattern in plane F.
53
FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500090006-1

APPROVED FOR RELEASE: 2407/02/09: CIA-RDP82-00850R000500490006-1
FOR OFFICIAL USE ONLY
An advantage of this method is that it is not necessary to precalculate or
measure the distortions that are introduced compensation for them is rea-
lized automatically. Among its disadvantages are the necessity of bringing the
directions of propagation of the reference and reconstructing beams into exact
registration, and also the need for good collimation. The slightest deviation
from these conditions negates the effect of interference suppression. Let
us also point out that practical implementation of the given scheme necessitates
fast-acting light-sensitive media. The entire procedure of recording and recon-
struction must be carried out in a shorter time than the characteristic period
of oscillation of fluctuations. ~
To distinguish an optical signal against a background of steady-state noise,
the technique of subtraction of optical signals has been suggested [Ref. 133,
105J. The operation of subtraction is accomplished by means of interfercnce
between the reconstructed signal recorded on the hologram and the signal arriv-
ing at the input of the system.
Let there be a hologram with transmission function
~(x)=~-~-YTu'Ix)~-YTuoulx) le'~Ra.~-n.~�+,er~R.x-.c~�l~ (6.16
on which signal u(x) is recorded. Here T is exposure time; y is the coefficient
of contrast; uo is the field of the reference wave. By reconstructing the
holo$ram with wave uoeikax~ and simultaneously illuminating it with wave
uei~`X~, we observe in one of the diffraction orders distributions of the field
of form
(Cu'(x)-~-Y7'u'(x) u'(X)~e~.~x~ _YTuou~x~~~lrt~j? FR~. ~t~,17)
Considering
y= -CJTu',
we find from (6.17) that complete suppression of the signal in the given dif-
fraction order requires satisfaction of the condition
uo ~o u (x)
u' (x) a (x) ao ~
and since usually uo� [u(x)]~aX, it is sufficient that uo/u~(x) = uo/u(x).
Thus if there are two optical signals with certain differences in a small ~e-
gion of space, we can distinguish their difference, e. g. the image of an c~b-
ject or part of it included in this region.
This method was used for analyzing ordinary photographic images in Ref. 105.
Holograms of different types were taken from a positive image (Fresnel, Fourier
and focused-image holograms). After processing, they were placed in the origi-
nat position in the corresponding optical system, and were then exposed with
~i reference wave and with a wave that had been diffracted by a transparency
installed at the input of the system. Appearing in the plane of observation
was an image of the part of the transpareiicy recorded on the hologram.
54
FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500090006-1

APPROVED FOR RELEASE: 2007/02/49: CIA-RDP82-40850R040500094006-1
FOR OFFICIAL USE ONLY
.Iust as the preceding methods, this technique requires exact placement of the
filter hologram in the proper position, and location of the intensity in the
linear region of the characteristic curve of the light-sensitive media.
One of the effective ways to deal with random interference is signal averaging.
In this case, the repeated addition of the regular useful signal intensifies
it, and at the same time the random sign-alternating random signal is attenuated
and suppressed.
It has been suggested that the averaging procedure be done by recording a defi-
nite number of uncorrela~ed reali~ations of the wave front and adding them
on the reconstruction phase [Ref. 34]. To do this, the holograms are trans-
illuminated sequentially so that the k-th wave front corresponding to the real
image is the reconstructing wave for the (k + 1)-th hologr~m. As a result of
such a process and isolation of the component with the sun�nary phase, we form
a field distribution
~
t ~p~~x~
u.K = �~e ,
where uo is the field distribution in the plane of the hol.ogram in the absence
of a distorting mediinn; ~k(x) is the phase distortion distribution function.
Using a calculator, we extract the n-th root and get the value of the field
averaged with respect to n realizations
n
~ A ~ 4K~X~
k~~
~u~=uoe ~ .
Tiie corresponding average value of intensity in the sought image under con-
dition of isotropism of the random medium is determined by the correlation func-
tion of the entrance pupil of the system Kp(x), the corre"~ation function of
field distribution without distortions Kuo(x) and the corr.elation function
of distortions K~(x):
(1 ~x)) CKn~ Kp ~x) K~ Ix)� .(6. l 8
For pure phase distortions distributed by normal law,
' K~~> ~ (6.19)
K, = e , .
n
where QZ is the dispersion of phase distribution function �(x);
Kvc.~~�~(.X) ~F ~P* (-x)�
Expressions (6.1.8) and (6.19) imply that for sufficiently large n the influence
of aberration can be considerably reduced. It has been shown by ntunerical
calculation that distortions are practically totally suppressed by averaging
100 times [Ref. 37]. The averaging procedure can be modified if individual
realizatioT?s of the wave front are accumulated on the stage of formation of
the holo,c-,ram. Technically, this is done by sequentially recording n exposures
55
FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500090006-1

APPROVED FOR RELEASE: 2047102109: CIA-RDP82-00850R400504090046-1
FOR OFFIC'IA1, l1SF. ONI.Y
on ~hotoKraphic emulsion or on a photoelectronic medium. Upon reconstruction
of such a complex hologram, an image appears in the plane of observation with
intensity determined by the average over the set of realizations of the wave
front. For sufficiently large n, the fluctuations of intensity introduced
by the inhamogeneous medium become insignificant [Ref. 27].
The given method of interference suppression has a serious disadvantage. This
is due to the fact that it is necessary to record a certain nianber of statisti-
- cally independent realizations of wave fronts. Practical implementation of
such recording is complicated when observing moving ob3ects through inhomogene-
ous media.
This flaw is eliminated by using a method proposed in Ref. 138. The hologram
recording arrangement is shown in Fig. 6.9. The object is illwninated by a
coherent light source (not shown on the
f diagram), and may be either two- or three-
1 ' dimensional [Ref. 179]. Next to the object
1 11 is a small reflector that forms a reference
wave. If the reflector and object are
sufficiently close together, the aberrations
~11 of the working and reference waves will
i"''� J)) be approximately equal. Distortions are
compensa te d on t he recons truc t ion s tage
~ as a result of interference of the rays
~ in the plane of the photographic plate.
Fig. 6.9. Diagram of holography
through a distorting medium: 1-- Let us evaluate the resolution of the re-
object; 2--reference ob~ect; 3-- constructed image. Let ~o be phase distor-
distorting medium; 4--photographic tions in the reference beam, and let ~(x)
plate be the phase distortions of the wave re-
flected by some point of the object. In
reconstruction, the field distribution in the light beam that constructs the
real image takes the form
uuo exP ~l If - ~o)~~ (G. 'l0 )
where u and uo are the field distributions on the hologram produced by a point
of the ob~ect and by the point reference source. We can see from expression
(6.20) that compensation of phase distortions takes place with satisf action
of the equality exp i� =exp i~o, or according to the Rayleigh criterion, the
condition ~ - ~o ~t/2.
This circumstance imposes a restriction on the dimensions of observable objects
for a certain strength of perturbations.
Ref. 134 gives a detailed analysis of imaging of two-dimensional ob~ects based
on this method. Let us limit ourselves to consideration of inedia such that
the logarithm of amplitude and phase fluctuations of the wave front incident
on the entrance pupil can be considered a locally steady process describable
by a Gauss statistic. Atmospheric fluctuations conform to such conditions
[Ref. 126].
56
FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500090006-1

APPROVED FOR RELEASE: 2007/02/49: CIA-RDP82-40850R040500094006-1
FOR OFFI('IAL USE UNI.Y
The long-exposure case is analyzed, where intensity on the hologram is averaged
over all values of the random function that characterizes fluctuations of the
atmosphere, and also the short-exposure case, where inhomogeneity of the atmos-
phere leads to shifting (blurring) of the diffraction pattern in the plane
of the hologram, and intensity averaging takes place over the field of the
set of reconstructed images.
It is found for a long exposure that the influence of the inhomogeneous mediwn
boils down to a reduction in brightness of the reconstructed image with in-
creasing distance between the object and the reference sot.rce. As long as
the brightness of the reconstructed image exceeds the noi:;e threshold of the
detector, the sought object is observed with good resoluti.on.
In the second case, there is a loss of resolution over the~ entire field of
the image. If the distance between object and reference source is small, the
reconstructed pattern contains the sought image with superimposed noise that
depends only on the amplitude of variations of the inciderit wave front.
Experimental results on determining the possibilities of t:he method are given
in Ref. 139. Lensless Fourier holograms were formed in ttie presence of atmos-
pheric interference. A Q-switched ruby laser was used as the emissi~n source.
Beam divergence was =200 rad. An ob~ect was imaged that c:onsisted of a set
of reflecting plates fastened on a board. A corner reflector was mounted on
the edge of the board to form the reference background. 7'he scattered radiation
was collected by a newtonian telescope system with mirror diameter of 1.2 m
and focused on a photographic emulsion. The size of the tiolograms was =7.5 ~n
and the distance between the object and the reception part: was 12 km. Distance
between the axis of propagation of the beam and ground tei~rain ranged from
35 to 800 m. Quality of the reconstructed images depended considerably on
the nature of atmospheric perturbations. In particular, t~nder conditions of
good visibility a hologram was obtained for mirror object:~ situated at a dis-
tance of 100 and 12 cm from the corner reflector. Resolution of the recon-
structed images was =0.5', which was approximately 1/5 of the diffraction limit
of the telescope. Imaging was also done on more complicat:ed objects. The
reconstructed images were of poor quality. One of the mo::t important factors
leading to image distortion was the presence of a large r~inge of reference
wave intensity fluctuations. Because of this, some regioris of the hQlogram
had long exposure, while in other regions the exposure wa:~ below the limits
of sensitivity of the photographic emulsion. Thus only a part of the working
aperture of the hologram participated in image formation. It is proposed that
an alternative solution to this problem would be to use electronic detectors
such as orthicons or multilayered photographic emulsions.
Let us take up the advantages and disadvantages of the given holographic
methods.
The method of optical matched filtration enables differenriation of the sought
signal against a noise background. To do this, it is necE~ssary to know the
parameters of the signal itself or of the interference. Depending on the prob-
lem, the coordinates of an object are sought or its image is reconstructed.
The method has be~n most extensively used for posterior processing of optical
57
FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500090006-1

APPROVED FOR RELEASE: 2047102109: CIA-RDP82-00850R400504090046-1
FOR OFFICIAL USE ONLY
imu~;es. These may be compound transparencies with a set of patterns, images
taken from radar screens, photographs of objects in the presence of atmospheric
distortions or aberrations of forming systems. On the current stage of devel-
opment, the technique of making holographic filters is complicated, and the
operation of installing them is time-consuming. This situation impedes the
practical introduction of the method for high-speed signal processing. Fur-
thermore, it cannot be used to detect unknown targets or to suppress unknown
interference.
The technique of subtracting optical signals has the same disadvantages. The
method of compensation when a reconstructed complex-con~ugate wave front passes
through an inhomogeneous medium has been successfully used for suppressing
phase interference. Knowledge of interference parameters is not mandatory
for its realization. It has been put to extensive use in eliminating lens
aberrations and diffraction distortions.
A peculiarity of this method is that on the stage of hologram reconstruction
the wave front passes through an inhomogeneity that causes distortion of the
sought signal during recording. Thus the speed of the system must be at least
the same as the characteristic time of fluctuations of parameters of the inhomo-
geneity.
From this standpoint, a more promising method is averaging of optical signals
by fixing statistically independent realizations of wave fronts on different
detectors or with accumulation of holograms on the same light-sensitive meditan.
When an object that sets up a reference perturbation is close to the sought
target, the process of suppressing distortions is relatively simple. It is
sufficient to fix one realization of the wave front and to use the technique
of optical Fourier transfonaation on the reconstruction stage. Imaging with
long and short exposures is possible during recording. This fact is of particu-
lar importance for observing moving ob~ects.
These properties determine the tremendous outlook of this method for detecting
and recognizing ob3ects through the atmosphere.
The most important advantage of optical systems over electronic analogs is
*_heir capability for handling enormous amounts of information over a short
time interval at the limit, over the time of light propagation in the system.
Operating in terms of communication theory, we can state that ogtical systems
have a large value of the product of input signal area multiplied by the square
of the maximwn linear frequency included in the signal.
For example, consider the linear operation of filtration (see Fig. 6.2). Let
the input signal have area S and limiting resolution R, and let the filter
have area S' and resolution R. For every element at the input, the optical
system performs S'R2 operations of multiplication, and S'R2 operations of sub-
traction. These operations are done SR2 times simultaneously. On the whole,
SS'R'' operations are needed to get the necessary signal at the output. If
we asswne that S= 104 mm2, S'= 4 mm2 and R= 100 lines/mm (values that are
easily attainable at present), then the numher of operations is 4�1012. Let
58
FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500090006-1

APPROVED FOR RELEASE: 2407/02/09: CIA-RDP82-00850R000500490006-1
FOR OFFICIAL USE ONLY
us assume that it is necessary to process a signal in this way in 1/30 s(the
usual scanning time in television systems) by an electronic system. Its pass-
band must be at least 12�1012Hz, and this is for the case of processing only
a single signal. In optical systems, several filters can be placed simultane-
ously in the filtration plane, i. e. a large number of two-dimensional signals
can be processed at the same time. Thus with a data input-output time of 1/30 s
in optical systems, we get an operator speed much greater than in electronics.
At the same time, optical data processing systems have a number of peculiarities
that prevent extensive application in place of electronic systems. To handle
radio signals, an electrical signal must be preconverted to optical form, and
so far we have not developed any adequately convenient and rapid methods for
doing such operations. Light-sensitive media that are used for recording op-
tical signals require considerable time for processing or have low resolution.
Nonetheless, even considering these specifics, the f ield of application of
optical data processing methods is quite extensive. Optical systems are used
for high-resolution spectral analysis, correlation analysis of functions and
fields in space and time, distinguishing signals on a background of noise and
interference, the operation of matched filtration and for handling other jobs
in radio astronomy, engineering, medicine, geophysics and acoustics.
6.2. Using Holography Without a Reference Beam
Methods of holography in which a coherent reference beam is used have found
wide application in various kinds of research [Ref. 33, 26, 71, 65, 127].
However, the conditions that ensure the necessary requiretr.ents impose limita-
tions on the class of solvable problems. Among these is rigidity of the geam-
etry of recording holograms, coherence of sources of radiation and high resolu-
tion of light-sensitive materials. These difficulties can be eliminated by
using the technique of holography without a reference beau~ [Ref. 2, 4].
Image reconstruction is realized if a wave scattered by part of the ob~ect
is used as a reference wave [Ref. 80, 82, 175].
Having a known part of the signal in form
�o (x) _ .4o Ix) eXP I~~Po ~x)~~
we define the amplitude transmission of the hologram t(x) as
t (x) ~-ao ~X) e'r.t f+, (x) e'r~tX1l~ ^ ~rro (x)~'
. -E-u; (x)no (X)-~-u?(x) uo(x)~
where ul(x) = A1(x)ei~~X~ is the unknown wave perturbation.
The deficient information on the object is contained in tr.e last two terms
of the given expression. To recons'truct it, the spectral component correspond-
ing to one of these te~-:as is isolated and divided by the signal uo(x)[uo*(x)].
Tl~is is done by forming an additional hologram and transparency with amplitude
transmissions tl(x) and t2(x) respectively:
59
FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500090006-1

APPROVED FOR RELEASE: 2407/02/09: CIA-RDP82-00850R000500490006-1
FOR OFFICIAL USE ONI.Y
~l~X~=~~ro,~+u~~r)~'=1-~-1~(x)I'-~--u.(.x)e r�,~-~uo(x) e~��`~
. . ~ ) ~ ~
, t x= ab (x).uo (x) luo (X) I~ '
The coef f icient v o= 2 n a/ a determines the phase distribution of the f ield
in the plane of the hologram, which sets up a reference wave with wavelength
a that propagates at angle a to the optical axis. Transparency t2(x) is made
by recording signal uo(x) on a photographic plate and processing to coeffi-
cient of contrast Y= 2. Ai~er adding the three transparencies t(x), tl(x)
and t2(x), they are illtnninated with a Flane monochromatic wave. In the direc-
tion of light beam propagation ~a, we get the field distribution in the form
_~o~r '
u ~X) = e {u~ ~x) ~u~( x) 2 uo (x) + u, ~x) + r+~~~
o (X)i~x ~x)~ .
I o (
The last two terms of this expression give the sought image and its conjugate.
The spurious effect of their superposition is eliminated by special selection
of the geametry of recording hologram t(x). To do this, the part of the object
that scatters wave uo(x) is situated to the side or at a different depth from
the object of observation ul(x). Then the reconstructed images are localized
in different regions of space [Ref. 67].
The procedure of reconstructing a hologram without a reference beam is simpli-
f ied if it is illuminated by a wave that coincides in localization and phase
with a wave scattered by part of the object [Ref. 50]. Let there be a hologram
on which an ob3ect is fixed that consists of a discrete set of N points. Its
transmission at some fixed point is written as
~y N �
At e- IRrr 1~ elRrv ~
T ~s ~ ~P
where Ag, Ap are the complex amplitudes of the sources that are points of the
ob~ect; rg, rp are distances from points of the ob~ect to a point on the holo-
gram; k= Zn/~; g, p= 1, 2,..., N are the subscripts of points of the ob3ect.
On the reconstruction stage, the hologram is illuminated by a spherical wave
of form e~~" emitted by a point source. If it coincides in localization
- rp
with one of the points of the object, the following set of images is reproduced:
an image corresponding to the initial object (p = n):
N
i 1~~~~ ~1 s e i..r~ ' '(6. ZO)
2 r~ ji~ ~s
its conjugate image (g = n)
N �
2 ~A`~' e C~A,Or
~ ~o
60
FOR OFFlC1AL USE ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500090006-1

APPROVED FOR RELEASE: 2407/02/09: CIA-RDP82-00850R000500490006-1
FOR OFFICIAL USE ONLY
and N- 1 more pairs of images (g, p# N)
n
~ A~Ap e=~R~~~--ro~~ As e i~r~ +
2 r�ro ~ !s
+ AnAr e~(' A~'s) N~p ej ~P .
r�r~ ~ ~v
With an increase in the nwnber of point sources coincidin~; with points of the
object, there is an increase in the number of reconstructed images. In this
connection, images (6.20) that correspond to the object are superimposed with
respect to localization and phase,whereas such superposition does not occur
for the other images. Thus when a hologram is reconstructed by a wave from
some part of the ob~ect, the useful image stands out in iritensity, and all
other images form a comparatively faint background. Unfortunately, it should
be noted that reconstruction of the missing part of the ot>ject necessitates
availability of considerable a priori information about it. For example in
experiments done by the authors on reconstructing an objec:t consisting of four
points, it was necessary to illuminate the hologram with a wave formed by three
points of the ob~ect.
Reconstruction of unknown information about the object is realized by using
Fourier holograms [Ref. 161]. To do this, the wave scattered by the object
is focused in the plane of the light-sensitive medium. If ta(x) is the trans-
mission function of the unknown part of the object, and th(x) is the trans-
mission function of the known part, then the recorded intNnsity distribution
is written as
/ (v) {t, (x)} -f- ~ {t.tX)?I'=1Ta (v) +T? (y)l'.
where } denotes the operation of Fourier transformation.
When such a hologram is reconstructed by a wave of forai T~(v), we get the field
distribution
U(v)=Ti(v)~Ta~y)-I-Te~'~)J ~Ta~'~)~'Ts~v)~�
Let us note that Tb(v) is the spatial spectrum of perturbation tb(x). Thus
for reconstruction we can use a wave diffracted from the corresponding part
of the ob~ect. By optical realization of inverse Fourier transformation, we
~et the following field distribution in the plane of observation:
u(x)=~to(x)-~-1~ (x)1 ~E (x) ~E [ta(-x)-}-tb(--x)],
where ~ denotes convolution.
When the condition
t~(x) ~ [ta(-x)-}-ts(-x)] ^:a(x) (G.'ll )
is met, we get the reconstructed image ta(x)+ tb(x). The condition (6.21)
is realized if the field of perturbation from the ob~ect has random phase
61
FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500090006-1

APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500090006-1
FOR OFFICIAL USE ONLY
distribution in space. The other case of image reconstruction is realized
if distribution tb(x) has sufficiently large amplitude, i. e. an object is
holographed w~.th an exceptionally bright point [Ref. 4, 5, 6].
One variety of holography without a reference beam is holography with a local
reference wave [Ref. 112J, which has found wide application in interference
studies. Part of the radiation scattered by the ob~ect is used on the stage
of forming the hologram. ,This radiation is passed through a spatial filter
to form a reference wave with~a simple wave front spherical or planar [Ref.
110, 111]. The procedure of reconstructing the image from su~h a hologram
is no different fram that of reconstructing conventional holograms with side
(oblique) reference beam.
It is well known that holography of a focused image is insensitive to the shape
of the reconstruction wave front [Ref. 173]. Therefore, by using par.t of the
radiation not diffracted by the object, we can get an image with rather good
resolution [Ref. 49, 104].
Extensive possibilities are opened up by using three-dimensional light-sensitive
media for making holograms without a reference beam. Three-dimensional holo-
grams can be treated as a set of two-dimensional holograms. In reconstruction,
each of them will form the above-mentioned images, with the useful images being
added in phase and intensified. In this case, we can get by with fewer "read-
ing points." And in fact reconstruction of a two-dimensional object from a
three-dimensional hologram is realized by using a single point source [Ref.
5, 7].
If during recording the entire holographic scene is focused in a given voliane
of the light-sensitive medium, then in reconstructing one of the rays emanating
from the object, we reconstruct the entire system of rays that,produce its
image. A three-dimensional focused image of the ob~ect can be observed behind
the hologram in the region that is free of the direct beam [Ref. 7, 45].
An interesting possibility opens up for studying various processes associated
with change in phase of the light wave when diffractograms are formed by dif-
fracting elements and reconstructed as holograms without a reference beam
[Ref. 2].
s -
Consider the optical arrangement diagra~ned ~
f
on Fig. 6.10. A ray from coherent light 1 ~
source 1(Fig. 6.10a) is shaped by lenses j
2 into a parallel beam that passes through T ~
the investigated ob~ect 3 and diffaction
grating 4 and is incident on plate 5. The s a
intensity pattern formed by diffraction ~ '
of the plane wave by the grating, and the i
pattern distorted by the ob~ect are sequen- ~T
tially fixed on the photographic emulsion. 6~ ~ �
On the reconstruction stage (see Fig.6.lOb)
the diffractogram 5 is illuminated by a Fig. 6.10. Diagram of record-
parallel monochromatic light beam passing ing with diffraction grating:
through lens 6. Several diffraction orders a--recording; b--reconstruction
62
FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500090006-1

APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500090006-1
FOR OFFICIAL USE ONLY
with intensity that decreases with increasing distance from the optical axis
are focused in the focal plane of lens 6. The reconstructed image is observed
in oiie of the orders by supplemental lens 8. Slit disphrag~u 7 isolates the
corresponding diffraction order and suppresses the spurious background.
The operation of the arrangement is described as follows. if /jx)= ~ a(X-r rtx~)
is intensity distribution in the plane of the photographic plate as determined
by a grating with period xo, and f(x) is.a function that characterizes the
intensity redistribution introduced by the ob~ect, then for. case f(x) < x the
distribution of blackening on the photographic emulsion takes the form I(x) +
f (x) = I (x) + I' (x) f (x) .
The transmission of the photographic plate is written as
- / (x) - (x) ~ =C / (x) - ! (x) f' (x).
When such a hologram is reconstructed in the focal plane of lens 6, the ampli-
tude distribution that depends on x will be
� -
-
f 1(x) I' (x) e~.~.~x = a(v - nvo) ~E F(y)� ~ F tro - nv, ,
~ n
A
where F(v) is the Fourier transform of function f'(x).
We can see from the resultant expression that in reconstruction of holograms
by Fourier transformation, there ia a Fourier spectrwn of the first deriyative
of function f(x) in each interference order.
Let us note that information with predetermined field intensity in the plane
of the diffraction element can be transmitted by such a method. Let us aseume
that the transmission function of the grating takes the form
�
t ~xj = ~ a (x - nz,).
n~~M
If a plane monochromatic wave is normally incident on the grating, then at
a distance z= 2xon/~, n= 1, 2, 3,... away from the grating, the field is equal
except for a constant coefficient to the field distribution immediately behind
the grating:
�
ul (z; x)= ~ a
(x- nx~).
.
Using the relation
.
, �
~ ~ a(x-nx.) o ~ a(ti-n~~)=v~ ~e(~--nvo).
A~~M
. ~-~w A~~r .
we write the recorded intensity dietribution as
~ , _ _ M
/1(t, x)=~ % ~.s(ro-nva~ ~ ~a(v- nvo) ,
. ~ e..
63
FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500090006-1

APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500090006-1
. FOR OFF[CIAL USE ONLY
where ~F denotes the operation c,f sutocorrelation.
In filtration plane 7 we have
(v) ~ a(~v- nv.) ~E ~ a (v - nv~) =~o " e (ro - na~�)�
- � � � n..-.
Isolatiag the first diffraction order vod(v - va), we get the follawing field
distribution in observation plane 9:
~ lU"~ (~)I =Ce''~. _ (6. 2l)
`When the ob~ect is inserted immediately in front of the grating, the ~ield
distribution behind it takes the form
_ f '
u, (x)= j (x) ~ e (x- nx~).
~
The corresponding intensity distribution fixed on the~diffractogram is
.
/,(z~ x)=~ vo� ~ F(v-KVO~ ~f' ~ (v-it~r~)~
�
where i ~ F (v)-m { j (x)}� . .
If the spatial spectrum of the investigated ob~ect fits into the gap between
diffraction orders of the grating spectrum, then a sigaal can be isolated
in the f iltration plane characterized bq autocorrelation function F~ F*.
Then for the filtered first order, we get
U*(a)='ao[F (ti-'%) ~E F' (v-tiaJ=
=~o {~f (p) ~E F� (v}] a (v-v,)}.
This implies that the field in plane (9) takes the form
~'lu3(~)I =1f (xK'er.~..
Fig. 6.lla [photo not reproduced] shows a diffractogram and reconstructed
image (Fig. b.llb [photo not reproduced]) of a candle flame taken by the method
described above. The reconstructed picture ia a shadawgram of an inhomogeneity.
An OKR-11 laser source was used in the experiment. The ob~ect was placed
immediately in front of the grating, and the aiffractogram was recorded in
the plane where a clear image of the line atructure of the grating was observed.
The diffraction grating was made by winding metal wire on a frame with pitch
of 0.5 mm and diameter of 0.4 m~.
One of the possible elements that introducea the phase modulation necessary
for forming diffractograms ia space itsel~.
As an example, consider an arrangement for recording the diffractogram of
a phase o~ject (Fig. 6.12). A parallel monochromatic light beam illuminatea
64
FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500090006-1

APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500090006-1
FUR O~FIC'IAL USE ONLY
th.e ob3ect placed in plane xo. The diffraction
pattern is fixed on the photographic plate in
plane xl.
~ ~
" Let us write the transmission function of an
ob~ect that has phase characteristic ~(xo),
_ z-__ t~xo)=e-~n=.~_
- Fig. 6.12. Diagram of re- Let us ~enote the field immedia.tely behind the
cording phase objects: 1-- ob~ect by uo(xo). The field distribution in
ob~ect; 2--uhotographic plane xl is defined as
plate
u~ ~x~)= u.Ix~) ~E ho ~X~)�
The spatial spectrum of che field ul(x) is
~ (x,))=C, (d~ {u,(xl)} exp (!v'%4b)]. ' (fi. `l3)
Here b= k/2z, Ci are constant coefficients, i= 1, 2,..., n. ~ince for real
objects there is an interval of spatial frequencies (=vo, vo) in which most
of the energy of signal ui(xl) is concentrated, we can select the condition
of observation
vo .vos~
4b ~ 4a 1~ (6. 24)
Performing the operation of i.nverse Fourier transformation in expression
(6.23) and taking (6.24) into consideration, we have
a ui ~x~) u~ ~x,~ - ! d~ao ~x)
4b dx~
Substi.tuting the value of the fi~ld ao(xo), we get
a~ (x~)=C~e ~KX~~ j ~ 46 (X~)~~- 46 ~x~)} .
~
The corresponding 3.ntensity distribution is written as
I (x~)-rc, (x,) ui ~x,)=Cz {1- 26 ~�ix,)~' 18b~(~"(x,){'-~
16b~ (~P~ ~x~)~~}-C~ f ~ - 2b ~x~~~ .
Thus when condition (6.24) is met, the recorded diffraction pattern contains
information on the second derivative of Che phase distribution function on
the object.
Decoding (reconstruction) of such a diffractogram is done in the optical sys-
tem presented on Fig. 6.13. A parallel monochromatic light beam shaped by
collimator 2 illuminates diffractogram 3 located in the rear focal plane of
65
FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500090006-1

APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500090006-1
FOR OFFICIAL USE ONLY
2 . ~ A~ ~
Fig. 6.13. Diagram of phase ~
object visualization: 1-- j
laser; 2--collimator; JIi, T
J72--focusing lenses; ~--fil~er; .
4--visualization plane f t f f
_ lens JI1. Lengen theanlane ofeconjugationtand foci atThe fieldedistributioneL
~ is located i p
in the f ilter plane takes the form
4~ { 1- (x)) =C (,a (v) z~'~ {~p{x)}J.
If a transparency is placed here that suppresses the zero order of radiation
component d(v) and has a transmission function inversely proportional tc the
square of the distance from the optical axis in the remaining region [t~= 1/v2],
then as a result of the operation of secondary Fourier transformation realized
by lens JI2, a field is reconstructed in the plane of observation that is de-
termined by real function ~(x). By observing the condition of linear recording
of intensity ~(x) on the photographic emulsion and then bleaching it, we get
a transparency with transtnission
~1 ~x~_e~9K~)
C~eff icient S describes the relation between phase shift and intensity of the
wave incident �e mon chromatichwave,uasfield isereconstructedrthat reproduces
nated by a plan
the wave front scattered by the investigated object.
In order to solve some problems, it is sufficient t~ have information on the
object in the form of an autocorrelation function of scattering (transmission)
of the investigated object. In this case, holography without a reference
= beam is a simple means of isolating optical signals against a noise background.
1 2 ' ~ ~ ~
Fig. 6.14. Diagram of holography y �
through an inhomogeneous medium:
1--ob3ect; 2--inhomogeneous me-
dium; 3--focusing lens; 4--pho-
tographic plate ~ �
FiR. 6.14 shows one of the arrangements of recording holograms that is used
for observing objects through an inhomogeneous medium. Object 1 is illim?inated
by monochromatic coherent radiation. The wave front distorted by inhomogeneity
2 is f~cused blanenof3theTlensoloSubsequentcreconstructiongispaccomplished
in the focal p
by the method of optical Fourier transformation.
The hologram in this case is in fact a difracto~ram that contains information
on the observed object.
The process of recording and reconstrucbengheudistribution ofitheaf ield on>
is described as follows. Let ul(x, y)
65
FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500090006-1

APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500090006-1
FOR OFFICIAL USE ONLY
the object; u2(x, y) the field distribution on the entrance pupil of the
lens. Field ul(x, y) is related to field U(u, v) in the focal plane of the
len~ by Fourier transformation
[I(u~ v)=Ci~ {u.i(x� y)1,
where } denotes Fourier transformation with respect to spatial frequencies
u= kx/f ; v= k~ /f; f is focal length; C is a constant coeff`icient.
Intensity I(u, v)= C2U(u, v)U*(u, v) is recorded on a phot:ographic plate in
- the rear focal plane.
In reconstruction, inverse Fourier transformation is realized with respect
to the same spatial frequencies as in recording. We have the following field
distribution in the observation plane:
u ~x+ y) _ { t (u~ '~)l = ~ ~x~ 9) ~ uz ( - x, - y).
If there is no intiomogeneous medium on the stage of hologram formation, the
response function of free space Ho(x, y) is determined by the relation
H. (x~ ~)=C exp [12s ~'r'-}- y')~ ' (6. 25)
where z is the distance from the object to the entrance pupil of the lens. We
have u2(x, y) = ul(x, y) ~ ho(x, y). Frrnn this and expression (6.25) we get
rc(x~ ?J)=ul(x, ~l) ~E ui(-x. --y). ,
since ho(x, y)*ho(-x, --y)=~(�Y, y).
Thus the field distribution in the plane of observation is determined by the
autocorrelation function of the field distribution on the ob~ect.
In the presence of ~ distorting medium, field u2(x, y) is written as integral
~
t1: (xs; ys) = J J ui ~xt~ 9i~ hi ~X~~ ys, xi. 9i) dx~ ~yl�
w �
For an isotropic medium, the response function hi(x2, y2, xl, yl) depends
only on coordinate difference x2- xl; y2-yl. In this case, field u2(x2, y2)
is written as
� (x, ~Jl = u1(x~ ~E ui ( - x. - y~ ~E hi (x~ y) ~F h': ( - x~ - y)�
and the reconstructed pattern takes the form
u~ (x� y:) � uo ~Xs, 9:) ~ h~ ~x~, y:), .
If the spatial spectra of object and inhomogeneity differ from one another,
then the distortions introduced will be insignificant. In the special case
67
~ FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500090006-1

APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500090006-1
FOR OFFICIAL USE ONLY
of white noise hi(x, y)~ hi(-x. -y)-~d(x, y), the autocorrelation function
of the field distribution function on the object is observed ~ust as in the
absence of interference.
In the experiment, the object was a transparency of the letter "H" with spec-
trum in the region of low spatial frequencies. The ob~ect was illuminated
by a plane wave. The light source was a helium-neon laser operating in the
multimode state. For effective resolution of low spatial frequencies, the
focus of the lens was taken as ~1.5 m, and the hologram was recorded on MIKRAT-
300 photographic emulsion. A frosted plate was used as the inhomogeneous
medium. The f rosted screen was rotated during exposure to simulate fluctua- .
tions of the elements of inhomogeneity. The same optical system was used
on the reconstruction stage as in recording. In this case the frosted screen
was removed.
Fig. 6.15, 6.16 show photographs [not~reproduced] of holograms and reconstructed
images for ob3ects taken through stationary and rotating frosted screens.
For comparison, Fig. 6.17 [photo not reproduced] shows a hologram and recon-
structed image of the same object produced in the absence of an inhomogeneous
medium. We can see from the given photographs that distortions introduced
by the medium completely change the nature of the spatial spectrum recorded
on the hologram. However, the general form of the reconstructed outline deter-
mined by the autocorrelation function of field distribution on the ob~ect
is the same.
When the above described method is used for recognizing ob,jects against a
noise background, the problem arises of~uniqueness of the signal representation
at the output of the optical system in the form of an autocorrelation function
[Ref. 2, 172].
It is known that a single autocorrelation function may correspond to a broad
set of signals with the same duration constraint [Ref. 39, 141]. However,
only one signal out of this whole set is formed by a system with limited trans-
mission spectrum. Any optical system is just such a system. Therefore the
autocorrelation pattern will correspond to a unique object that can be recorded
by the given optical device.
Regarding the method of holography, it should be noCed that it is one of the
possible versions of solution of a more general phase problem, i. e. the prob-
lem of finding the signal phase from its amplitude [Ref. 39, 2].
The necessity for mandatory introduction of a reference beam has usually been
justif~ied by the fact that in registering a signal
u(x)= A(x)el~l~
on a square-law detector, we get a distribution equal to
rt (x) a* (x) = A' (x),
in which there is no function that characterizes phase. However, even in
the first research on holography it was noted that the scattered wave is formed
68
FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500090006-1

APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500090006-1
FOR OFFICIAL USE ONLY
in accordance with the Huygens-Fresnel principle, i. e. ttie amplitude at each
observation point is determined by the sum of amplitudes taken with their
phase from each point of the object. Therefore the problem has been reduced
to methods of recording both information components in explicit form. The
same goal is pursued by the so-called methods of holography without reference
beam.
The general formulation of the problem is correct if sign~?1 u(x) is defined
as a signal of arbitrary nature. However, we are dealing with electromagnetic
radiation, and are treating the object as some boundary condition that influ-
ences the nature of distribution of the scattered field. In the far zone
that realizes Fourier transformation of a spatially limite~d function, we have ~
a distribution that is described by analytical functions. Thanks to this
circumstance, an additional equation arises that relates L-he amplitude and
phase of the sought distribution and is defined by the thEeory of functions
of a complex variable. And this means that a single-beam hologram (diffracto-
gram) carries complete information on the amplitude and ptiase of th~ light
wave scattered from the ob~ect.
6.3. Reconstruction of Object Image From Modulus of Autoc:orrelation Function
The problem of finding the image of an ob~ect from the modulus of the auto-
correlation function consists in the following. Knowing the modulus of the
autocorrelation function, it is necessary to find its phase, and then by
using the resultant amplitude-phase information to determine the function
corresponding to this autocorrelation function.
This problem arose for the first time in connection with iinding the intensity
distribution of an incoherent light source.
It is known that the image of a remote incoherent light saurce can be recon-
structed if the second-order optical correlation function is known [Ref. 159,
160J. The correlation function is usually determined from interferometric
measurements, for example by using a Michelson stellar inCerferometer. The
phase of the correlation function is related to the positi.on of the inter-
ference bands, and its modulus is related to the ltuninosit:y function.
Since technical difficulties preclude sufficient accuracy in determining the
disp2acement and position of the bands, experimenters frec~uently run up against
the problem of reconstructing the unknown function from it:s modulus.
An analogous problem arises in x-ray crystallography, where the phase of the
wave signal is usually unknown, in the theory of particle scattering when
determining the scattering cross section, and in other prc~blems [Ref. 89, 136].
Some scientists have discussed the loss of phase informati.on from the stand-
point of the limitations stemming from the analytical properties of the corre-
lation function [Ref. 39, 122, 123, 141, 165, 195]. Superposition of the
reference light beam with the field carrying information on the source of
radiation has been suggested [Ref. 157]. This method is analogous in nature
to the holographic method of recording amplitude-phase information [Ref. 137J.
69
FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500090006-1

APPROVED FOR RELEASE: 2007/02/49: CIA-RDP82-00850R440500090006-1
FOR OFFICIAL USE ONLY
A~:utl~er method of finding the phase was suggested in Ref. 156. In this case,
exponential filters are used for analytical continuation of the correlation
function into the complex region. If the modulus of the complex correlation
function is known outisde the real axis, then its phase is determined from
the modulus by integrating Cauchy-Riemann differential equations.
The last two methods of finding phase in�ormation have been experimentally
studied (Ref. 149]. The correlation functi~n T(x) in the far zone of dif-
fraction, which is related to intensity distribution of the light source by
the formula
.
I'(x)=C ~ S(u)exp(-lk,x~t/R)r1rt, Z~i)
, , .
was measured by a prism interferometer (Fig. 6.18). Here R is the distance
from the source to the plane of registration, x is the displacement of the
interfering rays, C is a constant coefficient, ko� 2n/ao~ ~o is mean wavelength.
~frosted screen
, motor ~ ' be m splitter
� 'll
~ ~i 1 x
1
source ~ ~ ~
. . . 1 ' .
" slit
~ . II
11
, demodulator .
chart .
recorder filter photo-
multiplier
. .
~ ~ amplif ier
Fig. 6.18. Diagram of facility for making correlation mea-
surements
The interference pattern is described by the relation
~I' (xj~ cos [vx a rg I'.(x) coi~st,
where v is the average spatial fraquency.
A scanning photoelectric detector was used in recording the interference pat-
tern. The x-coordinate is proportional to scanning time t, i. e. the signal
at the detector output is
_ IP (~t)I cos [v(~ a rg I' I~t)] -~-const.
70
FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500090006-1

APPROVED FOR RELEASE: 2007/42/09: CIA-RDP82-00850R000500090006-1
~OR OFF'ICIAL USE ONLY
The signal from the detector output went to an amplifier and filter that did
not pass the constant component. A chart recorder was used for registration
of the interference pattern. The recorded pattern contained information on
the modulus and phase of T(x). A helium-neon laser and rotating frosted glass
were used to simulate an incoherent source.
The first method of finding phase information is analogous to the use of a
reference beam in holography.
Using the formula of inverse Fourier transformation, we write
~ - .
A (u) _ ~ S (n') S (u' + a) rtu~ _ (krlnc~R) f (r(x)I' exp (lxukalR) dx.
-s -w
( 6. L7,
Let us assume that the unknown source S(u) has limited dia~ensions a< u< b,
and G(u) is a reference source situated at distance b- a from source S(u).
Then we have
o (u) =S (a) Cla (u - u,),
where C1 is a constant corresponding to the intensity of the reference source.
Now expression (6.27) takes the form
(k~/2nCll) ~ IrIX~I'exp(lxuk~/R)dx-~-C~d(u)-}-
. .
M ~ � .
~ S (v' ) S (u' u~ du' C1S (u uo) CiS
j.th - u). . (6. 28)
The third and fourth terms of this expression correspond t.o the real and imagi-
nary (mirror) images of the source. The uniqueness of the reconstruction
is determined from the a priori information on the positian of the reference
source relative to the ob~ect.
In principle, there is no naed for the reference source to be situated at
distance b- a from the object, but in this case, two measurements are required:
one with the reference source and one without it. The int:erferometric device
shown on Fig. 6.18 was used t~ form the interference bands. The measurement
source was modified as shown in Fig. 6.19.
Fig. 6.20 shows the reconstructed image obtained by Fourier transformation
of function ~1'(x)I2 in accordance with expression (6.28). The ob~ect was
a 95 um slit.
Reconstructed images of more complicated objects are shown on Fig. 6.21.
~ Let us consider the method of determining phase by exponential filters. Fol-
lowing Ref. 156, we get the analytical continuation of the correlation function
by replacing the real variable with the complex variable z= x+ iy, where x
and y are real quantities. We have
~
I'(x-}-ly)-C~ S(u)exP(kouy//1)exp(-ikoir.r;/1)rlu. (1;.:'!?)
71
FOR OFF[CIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500090006-1

APPROVED FOR RELEASE: 2007/02109: CIA-RDP82-00850R400500090006-1
F'OR OFF7CIAL USE ONLY
~moving screen
~transparency , ~
1 splitter
laser~
beam / ~
~ .
?
eference '
point reconstructed image of slit
source
' , ~ Fig. 6.20. Reconstructed image of 95 Wn
~ � slit. The central peak corresponds to
' laser ' the first two terms of equation (6.28).
beam The image is situated symmetrically rela-
~ tive to the central peak
Fig. 6.19. Diagram of formation
of source image
slit sawtooth disk '
aperture
, test
pattern
~ , . . . . .
a b c d
Fig. 6.21. Reconstructed images: a--95 11m slit; b--sawtooth
radiation source 95 ~tm wide; c--iwo disks with diameterP of
58 and 116 1?m; d--test pattern of three lines of 19 ~tm size
spaced 38 11m apart
If T(x + iy) is treated as a function of the real variable x, then (6.29) has
the same form as formula (6.26) when S(u) is replaced by S(u) exp (kouy/R).
The new source can bE constructed by placing the filter close to source S(u),
the filter transmission varying accordirig to an exponential law relative to y
In cases where it is physically possible to locate such a filter 3lrectly
behind source S(u), measurements of 1'(x + iy) with and without the filter will
give information on analytical continuation of the correlation function into
the complex plane. Such measurementa enable us to determine the derivative
2~T'(x+ iy) ~ /dy.
It is known that correlation function I'(z) is regular in the lower half-plane
� of complex pl~ane z. Therefore the modulus and argument of I'(z) are related
by Cauchy-Riemman relations:
.a Ir ~ ax ~s~)I _ ~r ~x.~..1J)I `~r ( a+ ~~r) ; ~r. 3n?
alr(a+l~N =_Ir(x-~-jy)) ~~a.r !y)
tr '
72
FOR OFF'ICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500090006-1

APPROVED FOR RELEASE: 2007102109: CIA-RDP82-00850R000500090006-1
FOR OFFICIAL USE ONLY
where I'(x-{-ly)=~I'(~-{-ly)~exp[1~(x-}-ly)~. Inteqrating expression (6.30), we have
~
1 d ~ I' (x' + ~Y)I ?
~(x-~-1?)-I (y)=-~I~(r' +ld)I dy ttx
0
where f(y) is some function of y. From this we get
� s
~ = a~ ~ in ~r - f- iy)(~x'. . ~ (s. 3l )
Thus in principle we have a procedure for determining the phase of the corre-
lation function from measurement of its modulus. If we need to know T(x+ iy)
only in the neighborhood of y= 0, the exponential function can be approximated
by a straight line, i. e. the exponential filter can be replaced by a linear
filter.
a
a b c b '
Fig. 6.22. Shapes of sources � � ~
used for method of reconstruc-
tion with exponential filters: Fig. 6.23. Machine recurding of interfero-
a--initial source; b--source
modified by exponential func- 8rams: a--typical inteY�fero ram fixed on
tion with damping duration of chart recorder; b--graph of ~T(x)I produced
1232 um; c--source modified by by using the given interferogram. The source
exponential function with damp- Was the transparency of Fig. 6.22a
ing duration of 616 ~nn I~~x~~
The experiment used the interferometer shown fsitnald
8
on Fig. 6.18. The function of source S(u) a
was simulated by a slit mask with height cor- _
responding to the selected function. The '
exponential modification was realized by phase 5~r
changing the shape of the mask ;see Fig. b_
6.22) . Fig. 6.23 shows interferograms re-
corded on a strip chart, and a graph of the
modulus of the autocorrelation function calcu- Fig. 6.24. Correlation
lated by computer. Fig. 6.24 shows the same function obtained by fil-
autocorrelation function obtained after fil- tration: a--modulus of I'(x);
tration, and the phase calculated by equation b--phase of T(x)
(6.3). The source was reconstructed by calcu-
lating the Fourier transform of function I'(z) and the phase obtained frrnn
equation (6.31).
73
FOR OFFICiAL USE ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500090006-1

APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500090006-1
FOR OFFIC[AL USE ONLY
~ ~ ' A
1 '
. i
b '
a
� � sawtoo~H source [sic]
c . ~ d .
~ � slit s0urce [sic]
e f :
. step~source ~ ~
Fig. 6.25,. Reconstructed images of sources: a--398 um _
slit reconstructed from measurement of function T(x); b-
reconstructed from function ~T(x)~ with (broken liiie) and
without (solid line) exponential filter; c--sawtooth source
398 um wide reconstructed from measurement of function T(x);
d--reconstructed from measurement of ~T(x)~ with (broken
line) and without (solid line) exponential filter; e--source
in the form of a double step 616 um wide reconstructed fram
measurement of function T(x); f--reconstructed from measure-
ment of function ~T(x)) with (broken line) and without (solid
line) exponential filter
The reconstructed images for three shapes of sources are shown on Fig. 6.25.
The resultant experimental data effectively demonstrate the feasibility of
determining the shape of a source and the phase of the correlation function
from its modulus.
It should be noted that the given methods of finding the phase of an optical
signal from its modulus are not the only solutions of the phase problem. As
has already been pointed out (see chapter 1), all optical signal.s belong to
the class of signals with finite spectrum. Therefore (see section 1.4), the
phase can be represented as
~ - ~ P ` lo~ctx-}-~arg s-si
s._s s-s~
'P ~ ~ - n ~ . ~
~
where A(z) is the amplitude of the optical signal analytically continued into
the complex plane, and zi ia the value of zeros in the complex plane. By
74
FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500090006-1

APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500090006-1
. FOR OFFICIAL USE ONLY
~ ~
u ~ ~ ~
~ � . ~ .T
~
~ ^ M ~
W ~-~1 ~ w r--~
a ^ w ~T O~ 1~
N N1 V1
u ~ ~..r .~i
3 3 w ~ ~ y
Cl Gl N 'd 'b 'd
.
.
~ p o a~~c a$ a~~
o oo a~i ~ a~i ~ +i ~a
u a~ ~ b b b
~ d ~ ~ u~~-�i u~
~ ~ � ~ � 6 � a~ d .~c d ~ a~ d
U ~a ~$rl e~0 O 0 0 ~
~o~ ~o~~ u � a~i ~ a�, y � a~i~
'n~ 'nd~c~'o ~ a~i ~ ~ a~i ~ a~? �
a~-a a oac~ oa oa~
~ ^
b t `k6
N N a ~ 4 ~ F
~o ^ N ^ ~ k ~ I v i~ 't
v ~Q
ra ~ N N ~ I v~ b? ti~ o
~ ~ I N N ~ l . : + ~
h
H yl w ~ pp N ~ p N a ~ , Y ~ a v
O ~
{.1 e + ~ A � k , ~ ~ e ~ i?
v k
V ~ C.t 1~ t a~. I u ~ u~'~
P4 ~o 4 4' u v~~.
: II "~k + -IK ~
Y ~ y.~r r0~ 1~ I~ k
il ~e
a e - ~ "
e ~ ?
~ ~
a
~
u~
N d a ~ a~
r~l N N ~
- a.~ ~ Ul ~ri N
~ Gl u'O vl,~ vl,~ W~ ~ 3
~n ri a~i u'~' n M. a ~n e~t ^
~ ~ u v ~ ui u~'i ~ ~O I'~'.' ~e I ~O ~ ~
a�~i ,-~i ~ ~ n 4 a a~i ~ u.' a~i d
H tn r-1 tb u^' I IC I K ~U n ~
Gl I ~ ~^n ',a~ ~ N l11 W ~ I ~ ~ I ~ ~
a ~ tl N~p, 9 Y~ rl P4
~ G! O ~o A
tn t~0 m O
.C � N
p~ G) U ~A
rl rl i-i O
aa~N
Y+ ,a o a?
~ u m N
75
FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500090006-1

APPROVED FOR RELEASE: 2007/02/49: CIA-RDP82-00850R040500090046-1
FOR OFFICIAL USE ONLY
imposing an additional condition on the optical signals, namely the Fermat
principle d~(z) = 0, we get
w
z_ t p ~�g ~A (s~)~ dz. (6. 32)
rt ~ s-:'
Thus, optical signals that are formed by passive optical systems (systems
without sources) contain zeros only on the real axis. Therefore the phase
of the optical signal is uniquely determined by its modulus.
The aiffraction pattern obtained in coherent light contains complete amplitude-
phase information on the optical signal, and in this sense we can speak of
holography without a reference beam (one-beam holography).
Let us note that phase reconstruction from such a hologram is possible not only
analytically in accordance wtth expression (6.32), but also in a coherent
optical system that realizes Hilbert transformation. The method proposed in
Ref. 140 can be used to get a transparency with transmission that varies ac-
cording to a logarithmic law.
In conclusion, the methods of solving the phase problem are su~narized �~n
Table 6.1.
6.4. Phase Problem in Radio Astronomy
The phase problem as initially formulated in optics has analogous treatment
in other regions o� the spectrum, and particularly in the x-ray and radio
bands. The problem is closely related to use of the luminosity function first
introduced by Michelson in optical interferometry, and subsequently generalized
by Wolf, who saw it as a"complex function of mutual coherence." Let us look
more closely at the formulation of the problem and some possible ways of solv-
ing it as applied to radio astronomical research.
It is known that one of the ma~or problems of radio astronomy is determin$tion
of the distribution of radio brightness over a source of space radiation.
The distribution of radio brightness B is related via inverse Fourier trans-
formation to the complex luminosity function U of the lobe structure of the
radio interferometer as measured on different bases, which in the one-dimen-
sional case is represented as
U u e~�~Edu~, " ~ (6.33)
B(E)- f ( - .
where uX = 2nDx/a is spatial frequency; DX is the base of the interferometer
with respect to coordinate x; a is working frequency. The complex luminosity
function can be represented as
U ~u.~~= ~V ~us~~ e ~~r`~ � (6. 34~
In principle, experiment zan give us the modulus and phase of luminosity:
76
FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500090006-1

APPROVED FOR RELEASE: 2007/42/09: CIA-RDP82-00850R000500090006-1
FOR OFFICIAL USE ONLY
I~ Iu~II = P~'= - P~~� ;
Ps~: pels �
~ (p,~)=2net/t.~
where Pmax' Pmin are the maximum and minimwn values of the main lobe of the
interferometer relative to the zero value; et is the time of delay of the
maximum of the envelope relative to the calculated interference maximwn; to
is the time interval between ad3acent maxima, i. e. the period of the inter-
ferogram [Ref. 68].
While it is easy to measure the modulus (see for example Ref. 24), there are
a number of difficulties in determining the phase. Nonetheless, knowledge
of the phase of the luminosity function is considered necessary for determining
the distribution of radio brightness.
Thus the formulated problem reduces to determination of the phase of the com-
plex luminosity function followed by substitution in (6.33) and (6.34).
The key to solving the phase problem is use oc the analytical properties of
the investigated functions, enabling reciprocally single-valued determination
of their real and imaginary parts on the basis of Hilbert transforms.
In application to our problem, we note that functions U(uX)
In U (u,~) -1 n U (u,~) - (uX) (6. 35)
are analytical in the lower half-plane J_ of the crnnplex Flane. Expression
(6.35) gives the possibility of determining phase ~(uX), which in general
form is written a3 [Ref. 181]
' ~f~lt.~)=~P~~u~)~'~Pe(~l.r)~ (~;.36)
~ ~ in U~u~ ) (l~. 37)
where ~,~(u,~)=--P , ~ du'X;
n ~j-ux
, u~ - n~i (6. 3~)
~e~u.~)= arg . .
. u~ - ~si
The first term ~(uX) is the value of the minimum phase that in the absence
of zeros in J_ uniquely relates the phase and modulus U(u~;) through Hilbert
transformation.
The second term ~(uX) is the Blaschke factor that evaluates the contribution
of zeros of uXi in J_. Summation extends to all zeros of the function U(uX)
in J_.
Thus phase ~(uX) is determined from knowledge of the moduZus of function
U(uX) and the position of its zeros in J_. We should take note of the impor-
tance of this statement, as until recently it had been assumed according to
Rayleigh that the amplitude (modulus) and position of interference lobes
(phase) are mutually independent quantities, and that both meas:urements (space
and time) have to be made separa*ely to reconstruct the spectra [Ref. 24].
77
FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500090006-1

APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500090006-1
FOR OF'FICIAL USE ONLY
Tha problem of the influence that the Blaschlce factor has on reconstruction
was considered by Wolf [Ref. 195], who concluded that in the case of consider-
ation of the function of time coherence U(T) the differences in the curves
of the reconstructed spectra in the "miniphase" solution ~witY~ consideration
of ~ alone) and in the complete solution -~b) do not exceed a few percent
with respect to major parameters.
Examples can be given of spectra that do not have zeros in J_ [Ref. 122].
Furthermore, for the two extreme cases that are of interest for radio astronomy,
namely ideal blackbody radiation and monochromatic emission, the phase problem
has b~en considered and a solution has been found in the research.
Thus we can make a general recammendation that the simple and elegant mini-
phase solution in form (6.35) be used to get information on the phase in the
' first approximation, and that the question of sufficiency of the resultant
solution that is associated with the presence and influence of zeros be de-
termined separately in each individual case.
In accordance with this, we arrive at the followirig "skeleton" of a solution
for the problem of reconstructing radio brighr.ness from experimental recordings
of W:
W ~t)=1 U ~u~)~ } ~ ~u.~) U (a,~): B ~E)~ (G. 3J)
where the symbols H and ~ denote the Hilbert and Fourier transforms respectively.
As has been noted, determination of the modulus U from W presents no diffi-
culties, and therefore, writing out (6.33) we get
~ i(~~+rs+~,~E1~u (6.4U)
. B (E~ = r U ~ur~ ~ .
b
where ~X ~ . .
In Uf s{) du, In U(a;~ d~~ (6. 41)
'P~=11m r , x~-J .
~y~ r
~
is the principal value of the Cauchy integral..
As we can see, analytical Fourier transform (6.40) differs from conventional
Four ier transform (6.33) in the change of limits of integration and the inclu-
sion of Hilbert transformation as an internal operation.
Denoting the analytical Fourier operator by symbol we can write (6.40) as
~ . � .
B (E) ~ U (u,~). (6. 42)
Let us note that expressions (6.40)-(6.42) are the fundamental solution of
the problem for a symmetric interferometer without consideration of issues
of practical realization relating to the influence of the inhomogeneous medium,
smoothing effects of antennas, inequality of interferometer axms, etc. The
78
FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500090006-1

APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500090006-1
FOR OFFICIAL USE ONLY
solution has been obtained for asymtnetric inhomogeneous distributions B(~).
for sy~etric [distributions] relative to the center of gravity ~o of the
source, expression (6.40) is simplified:
B~E~ - r~\uJl C0, V~~uF. l6. 43)
~ '
A solution of form (6.43) can be extended co time frequencies as well, thus
giving time spectra of the radiation sources.
In this case we have
Q (~(t), (6. 44)
where operator being extended to time frequencies, has the above-mentioned
meaning, and T is time delay introduced into the arms of the interferometer.
Radio astronomical interferometry experimentally realizes both kinds of co-
herence: spatial (SC; and temporal (TC). The former is realized due to a
change in base of the interferometer, and the latter is due to rotation of
the earth. In this sense, we have an analog of the Michelson and Young op-
~ical interferometers [Ref. 18]. Thus we find that the camplex luminosity
function reflects the concepts of SC and TC and crnnpletely determines the
characteristics of radiation from cosmic sources arriving at the earth, which
in the final analysis gives information about the morphology of tr?e sources,
and about the nature of processes occurring in them. This is reflected in
the following transcription of the solution found in the most general form:
E ~E~ U (u,~~ . (G. 45)
where E(~, w) denotes the distribution of radiation energy of the source with
respect to spatial coordinates ~ and temporal frequencies w.
Assuming in (6.45) T= const, i. e. singling out the spatial frequencies, we
get the spatial spectrum of the source ("static one-dimensional portrait"
of the source) that reveals its morphological features.
Setting w= const in (6.45), the interferometer is contracted to a point, which
is equivalent to recording on an isolated antenna, i. e. by isolating the
time frequencies we get the time spectrum of the source. This gives infor-
mation on the dynamics of processes that take place in the source and on the
_ na[ure of its radiation, i. e. we get a"dynamic portrait" of the source.
It should be noted that in the general case, spatial and temporal coherence
are not independent, conforming to the two-wave equations [Ref. 18]:
1 I 6~U (u j~ ro)
p~~~U(uX, Y)= cs r ~ ~ (6. 4G)
1.
where oi~2 is the laplacian with respect to coordinates of the end points
of the base in predetermined direction x; c is the velocity of light in vacuum.
79
FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500090006-1

APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500090006-1
FOR OFFICIAL [JSE ONLY
In essence, solution (6.46) is the well-known Van Zittert-Zernike theor~m
which in the most general form is repzesented by expression (6.45). Thus
even from (6.46) we get in case U(uX, Tp) the spatial spectrum corresponding
to SC, and in case U(uo, T) the tesporal spectrum of the source correspond-
ing to TC, if uo, To are constant parameters.
For spectrally pure processes, the so-called "reduction formula" is applicable
[Ref. 24]: ~
U(W~, t)=U (rcx~ '~o) U~u.~, Y- to),
where the first factor in the right-hand member denotes spatial coherence,
and the second denotes temporaJ. coherence.
Let us note that the resultant solution of the phase problem is applicable
not only in interferometry where the informer is the function of coherence
as the correlation of emission at two points of the wave front, but also in
the case of signal registration at one point on an isolated antenna (here
the meaning of "point" is arbitrary).
In the latter case the signal can be considered as the resultant of elementary
interferometers with different bases fitting in the aperture of the antenna.
In principle, any point of the wave front contains information about all de-
tails of the emitting ob~ect.
]
A certain analog in classical holography is the well-known fact of recanstruc-
tion of the wave image of an ob~ect from a fragment of a hologram.
Aperture synthesis, which has come into extensive use in radio astronomy,
can be considered as a spatial radio hologram, and with appropriate recording
and processing, e. g. for the radiation of cosmic sources "space masers"
(spectral atomic and molecular lines of H, OH, NH3, H20 and so on) we can
reconstruct a"portrait" of the emitting regions.
Correlation processing of signals recorded by elements of aperture synthesis
can be done both with respect to high frequency and with respect to the enve~
lope. In the latter case we get the analog of the so-called "intensity~ inter-
ferometer", where the informer is the square of the moduius of the luminosity
function. Even in this case, after elementary conversion, the phase of the
complex luminosity function is reconstructen from the known modulus, i. e.
there are no losses of information on phases of high-frequency signals.
6.5. Systems Approach to the Optical Cavity Problem
,q The use of systems theory gives new possibilities for analyzing and synthesizing
open cavities [Ref. 1, 14].
The essence of the systems approach to the optical cavity problem consists
in the following. Some distribution of the electromagnetic field is taken
as the input signal simulating the open cavity, and the field distribution
in the same plane produced by the input action as a result of its repeated
80
FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500090006-1

APPROVED FOR RELEASE: 2007/02/49: CIA-RDP82-00850R040500090046-1
FOR OFF(CIAL USE ONLY
propa~ation from one Rrain to another is taken as the output signal. Such
~;ystem is completely described by the space-frequency characteristic. Be-
sides, the inverse Fourier transform of the frequency response gives the re-
sponse of the system to the input signal described by a delta function, or
else gives the Green's function of the optical cavity problem. Knowledge
of the Green's function enables us to find the modes of the open cavity, since
it is the sum of the eigenfuncti.ons (modes) of the given problem.
In most solutions of electrodynamic problems, finding the Green's functiuns
presents serious difficulties. The simplicity of the elem~nts that comprise
the optical cavity (mirrors or generators of frequency-modulated waveforms,
space as a linear phase filter [Ref. 14]), enable us to construct a model
of the optical cavity as a linear system with feedback [Ref. 14], to determine
the space-frequency characteristic of such a system, and by carrying out its
Fourier transf~rmation to get a pulse response, i. e. the Green's function
of the optical cavity problem.
The general space-frequency characteristic of the open cavity is found in
the following way. Let H(u) be the space-frequency characteristic of the
system that describes one circuit of the cavity by an electromagnetic pertur-
bation.
AEter each circuit of the cavity, the electromagnetic perturbation of a certain
plane is again incident on this plane, in virtue of which the cavity is a
. system with feedback, and its general space-frequency characteristic has the
f o rm
o ( ) p�j
H u =~-H(~)'
where pd~ (u)= for ~u~ < uo is the pupil function that determines the band
{ 0, f or ~u~ ~ u,
of transmitted space frequencies. Here use is made of the property of the
pupil function
W ithin the framework of idealization that uses the Huygens-Fresnel principle:
p ~u~ (6. 47 )
ye(u)= .
1--r~rzexpl ( ~ ~ a~-(~i+~x)x''
[-\4 (2a-~~) 4 (2a-~z)
where rl, r2 are the mirror reflectivities, a= k/2L; k= 2n/a is the wave ninn-
ber; L is the distance between mirrors; S= k/2f; f is the focal length of
thc mirror.
Inverse transformation of the Fourier characteristic gives the pulse response
of the system
m
ho~x~-~2~ Ho~u~~~~Xdu.
J
Let us reFresent this integral in the form of the convolut.ion of Fourier trans-
,
' furms of the cofactors P�~(u) and H~u~ of function Ho (u)
81
FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500090006-1

APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500090006-1
FOR OFF'ICIAL USE ONLY
h~~C)=~ (Pr~~u)) ~ ~ ~I - H(a), ' ~G. 48)
that defines the Green's function of the cavity problem. (Here -~E dnotes con-
volution of the two functions.) The first function in convolution (6.48)
is sin uox/nx, and accounts for finiteness of the mirror dimensions. The
- second function defines the solution of Che problem of modes with infinite
~ mirrors, and is expressed in form EeiumX, where
u 4(di +~i) (24 - PJ('Za - Ps) xs+2(2a-~J(2a-~i)~`lrcnt-In ~irz).
- 4a - + ps) 4a - (Pi + ~s)
Thus the general solution is written as integral
~
/~e (.r) -r e~�mX sin u~ (x - E) ~E
fi~~l ~ x - E
m ~
As an example we consider a cavity with identical infinite mirrors (R1= R2= R)
and we find the set of eigenfunctions that correspond to subscript m= 0. We
restrict ourselves to the approximation of real mirrors, i. e. we set r1= r2= 1.
We have
uo=1x y2v-v~~ v=~l~=L/R~
?s,=.�s+
implying ~?o(x)=e'~~~=e .
Under these conditions, in the case of infinite plane-parallel mirrors (s1= R2=
L/2R = 0) the optical cavity is a spatially invariant system and its general
space-frequency characteristic will be in accordance with (6.47)
f~~ ~u~ _ i ~ .
1 - exp ! (-Lu~/k)
The pulse response of such~a system will be equal to
~
ho ~x~ _ ! ~ e~o.~~u. � (6. 49)
2n J . L
. 1 = eap (-i - u~l ~ ,
k /
The integrand has first-order poles at points
u,~ _ � j~2nmk/L,
and a second-order pole at point u= 0. The residue at point u= 0 is zero.
tr ~ Thus the ulse response of the cavity
The residue at point u~ is t~/l k~,~. p
will take the form of the sum
ho ~x~ - ~ cus (a,~x)
la~,/k
82
FOR OFFICIA~. USE ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500090006-1

APPROVED FOR RELEASE: 2407/02/09: CIA-RDP82-00850R000500490006-1
FOR OFFICIAL USE ONLY
l~c,r :i planar cavity we get a solution in the form of cosines with argtnnent that
includes the distance between mirrors. The latter circtnnstance enables us
to find the solution of a cavity with a mirror that has periodically spaced
apertures.
Let us consider a cavity in which one of the plane-parallel mirrors has a
periodic structure that consists of alternating reflecting and transparent
strips of identical width xo/2.
The space frequency characteristic of the system that describes the process
of propagation between two planes separated by distance z takes the form
[Ref . 33]
H,(u)=exp(I2n V 1-( 2a (G. i(1)
t ~
Let us limit ourselves to consideration of the case of two spatial harmonics
(m = 0, 1). Then the output signal of the system will have a spectrum defined
by the expression
F.~~ ~u)= F~a (u) ~ H(m2n/x,) a(a+ m2n~xo)=
m_o,~
= FR N -f- F.: ~2n~X�~ H ~2n~xa~'
(Russian subscripts sbix = output; sx = input]
In the case of optical systems, the function FHbpt(u) is equal to the amplitude
of the field in the exit pupil. In the given case it is equal to the ampli-
tude of the field in plane z. The intensity of this field is
Fws (u) F.w: (u) _ ~ F,a ~F~ 12n1Xo)I1-~"
2 ~Fsa ~0~~ ~Fa ~2n~xo~~ Cos - ~P ~2n~xo)]~
where �(u) is the phase of frequency response (6.50).
With consideration of expression (6.50), the intensity in plane z is represented
in the following form:
~ F.~ (u) F~: (u);- IFa (~H' (Fa (2n/xn)I' -f-
2 (Fa ~Fa ~2n/xo)~ cos [2rcz/~ - 2ni/l?1~ 1= (a/x�)'] .
This implies that the intensity of bands formed in plane z is maximum if the
distance to this plane is
zN= ~
MY , M=1,2,...
l - 1 /x
If period xo �a, we get the relation
z� = 2Mx~a.
~ The resultvnt relations coincide with the known Rayleigh relations [Ref. 74].
It was RaylPigh who called attention to this effect of lattice reproduction
83
FOR OFFICiAL USE ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500090006-1

APPROVED FOR RELEASE: 2007102109: CIA-RDP82-00850R000500090006-1
FOR OFFICIAL USE ONLY
In coherent light. Subsequently, the effect was experimentally verified and
a nuc~ber of papers were devoted to it, e. g. Ref. 38, 125. Space-frequency
analysis of an open optical cavity gave the capability of determining the
form of the natural oscillations of the resonant cavity periodic modes
a phenamenon similar in nature to the lattice reproduction described by Ray-
leigh.
In the general case for a cavity with mirrors having radii of curvature R1
and R2, we get the condition of existence of periodic modes from the expres-
sion that defines the Green's function:
r~L -~41 CRi + Rs 1 J Nx�~
1.
where N= 1, 2, 3...
Fresnel Sandwich as Optical Cavity
The Fresnel sandwich is a device in which the distribution of the field of
electromagnetic radiation repeats itself.
Finding the conditions of self-reproduction of the field is the problem of
~ mode determination.
Consequently, if we make the reflectivity of one of the mirrors vary like
a Fresnel function (which is reproduced in the sandwith), then under certain
conditions the distribution of the field
reflected from this mirror will reproduce
i I f~p~ itself after each pass through the cavity,
~ and the Fresnel sandwich can be used as an
;I f optical cavity. According to Ref. 58, a
Fresnel sandwich made up of three Fresnel
~ zone plates (Fig. 6.26 realizes Fourier trans-
formation at certain parameters so that its
~I ~one plates~ ~I action is described by the following equation:
_ ~ j
F(u)=zo(x)i[f (X)Zo(~) ~E Zo(x)1}(u). (6.51)
Fig. 6.26. Fresnel sandwich
where f(x) is the initial function to be
transformed by the Fresnel sandwich. This function may be the amplitude dis-
tribution of the electromagnetic field in plane z= 0, i. e. in direct nroximity
to the first Fresnel zone plate (see Fig. 6.26):
Zo (x) = rto ~~2exp f lu'a.~c') is the Fresnel f unction.
Here we are considering one-dimensional distribution of the electromagnetic
field f(x). The results of such consideration are easily extended to the
case of real two-dimensional distribution f(x, y).
Transformation of the image f(x) via the Fresnel sandwich leads to the Fourier
transform of the initial function. For this purpose it is necessary that the
84
FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500090006-1

APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500090006-1
FOR OFFiCIAL USE ONLY
zone pl ate in plane z= Z 1 be reduced with respect to zone plates in planes
z= 0 and z= Z1 + Zz by a factor of Z1/Z1 + Z2, and that the parameter of the
Fres~el function Zo*(x) satiGfy equality u2= 1/2. The Fourier image of the
initial function is formed in the plane directly behind tr.e third zone plate,
i. e. in plane z=(Z1 + Z2) .
Now let the surfaces of the plane-parallel mirrors that camprise the open
optical cavity have reflectivity that varies as the transa:ission function
of the Fresnel zone plate, and let the corresponding zone plate be placed
between such mirrors in plane z1= Z1. The action of such a cavity, like the
Fresnel sandwich, is described by equation (6.51), i. e. ~.t certain parameters
such a cavity realizes Fourier transformation of the electromagnetic field
distribution after one pass from one mirror to another.
The first multiplication f(x)ZQ*(x) in the right-hand part. of equation (6.51)
takes place upon reflection of the electromagnetic wave from the first mirror
in plane z= 0. The operation of convolution is also reali.zed just as in the
Fresnel sandwich as a result of light transmission through the zone plate
- located in plane z= Z1. The last multiplication is realized upon reflection
of light from the second mirror located in plane z=(Z1 + Z~). The Fourier
imaKe of function f(x) is formed in plane z=(Z1+ ZZ) in direct proximity
to the surface of the second mirror.
In virtue of symmetry, its eigenfunctions will be defined by the equation
0
~~~r~X)= f ~R (EI e-'E"dE~, (G. 5'')
where ~k are eigenvalues corresponding to eigenfunction ~~(x), and a is the
transverse dimension of the mirror.
Equation (6.52) for the modes of the described cavity coiticides with the equa-
tion that defines the eigenfunctions of a confocal cavity.. These functions
are proportional to spheroidal wave functions, and they h~ive a number of prop-
erties that are useful for a plications. For example ~(x} has the maximum
possible energy on interval ~x~~ a, which dictates minimum possible diffrac-
tion losses in the optical cavity. In addition to this pr.operty of the eigen-
functions of the Fresnel sandwich, the feasibility of using it as the optical
cavity in a laser is due to the capability of coupling elE~ctromagnetic energy
out through apertures in the zone-plate mirrors, the infliience on output beam
parameters (reduction of divergence as a consequence of ari increase in spot
diameter on the mirror), and more complete utilization of the active substance
of the laser due to diffraction by the zone structures of such a cavity.
6.6. Holographic and Integrated Optics
A certain number of optical devices are used to transform the amplitudes,
phases and polarization characteristics of light beams: mirrors, prisms,
lenses, birefrin~ent and phase plates, polarizers, various kinds of gyrators
and diffraction gratings. Optical components of this type are usually cumber-
some and expensive, and their capabilities for handling light beams are limited.
85
FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500090006-1

APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500094006-1
FOR OFFICIAL USE ONLY
This has given rise to the problem of developing compact optical equivalents
of known optical components and devices that have new properties. These prob-
lems have been studied in a number of papers [Re�. 22, 36, 61, 62, 113, 114,
145j.
Holographic methods enable us to produce holograms that are analogs of con-
ventional optical components and that in addition have a number of interesting
peculiarities (Ref. 62]. The capability of superimposed holograms to act
independently allows the development of complex cambinations of such camponents
and even creation of new devices that have no analogs in classical optics.
The new f ield of application of holographic methods of producing and prolifer-
ating both elementary and complex optical components is developing rapidly.
Hologram-Mirrors. The simplest optical component is a flat mirror (total
internal reflection prism) an optical device that changes the direction
of a light beam. Its principal characteristic is reflectivity; multilayered
dielectric coatings are used to increase this parameter. Such mirrors are
essentially artificial holograms adapted for operation in monochromatic light.
With appropriate production they can alter the nature of polarization of re-
flected light beams in a predetermined way. Extensive use is made of multi-
layered coatings that do not change the nature of polarization of light that
is incident on them at a certain angle. Sometimes such hologram-mirrors can
be used as light filters and polarizers. Such components are characterized
by less irregular scattering of light than conventional optical mirrors.
Hologram-Lenses and Holographic Ob,jectives. The diffraction analog of a lens
- is the~zone plate. A zone plate forms an image of an extended object in the
same manner as a conventional lens, and its resolution at the focus is equal
to that of an ideal lens of the same aperture. The main disadvantage of the
zone plate (both phase and amplitude plates) is the presence of a series of
f ocal lengths corresponding to real and imaginary foci that depend on the
wavelength of light. In contrast to an ideal lens, the zone plate is not
tautochromic. In imaging ob~ects that are a finite distance away or off the
optical axis, the light reaches the end zones and the center at different
times, leading to dephasing of waves in the focal plane.
An important advantage of zone plates is their two-dimensionality and their
capability for focusing infrared, ultraviolet and soft x-rays. This property
is used in making compact light equipment for space research [Ref. 113, 11].
Zone plates are made most precisely and simply by a holographic method. Zone
plates can be recorded on both sides of a thick photographic emulsion by
Yu. N. Denisyuk's technique [Ref. 36] with subsequent reconstruction by white
light with selection of one wavelength. Such zone plates give only the real
image. When a zone plate is recorded in the light of three lasers operating
on different wavelengths, it can be use.d to get a color image.
Holographic recording of zone plates in the off-axis arrangements gives plates
with spatial separation of the principal real and imaginary foci from each
other and from the zero diffraction order. Such a recording scheme almost
completely eliminates background and considerably improves image quality.
86
FOR OFFICIAI. USE ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500090006-1

APPROVED FOR RELEASE: 2007/02/49: CIA-RDP82-40850R040500094006-1
FOR OFFICIAL USE ONLY
In addition to the classical zone plate, a hologram-lens can be made that
is formed by two point sources producing diverging spherical waves. High-
quality objective lenses have been recorded on holograms an this arrangement,
The resultant holograms are the analogs of the objective :_enses. The use
of a hologram-lens for recording an image to produce focused-image holograms
enables simultaneous production of both the focused image and a reference
beam in the form of undiffracted light.
Computers can be used to improve the diffraction efficiency of hologram-lenses
made on photographic emulsions. In this case, the calcul:ited relief is applied
to a large surface, then reproduced by precision optics. The diffraction
efficiency of such hologram-lenses (kinoforms) reaches 90;e [Ref. 145].
, The strong chromatic aberration of the hologram-lens can l~e used in spectral
instruments. The hologram-lens can serve as a dispersing component, replacing
expensive concave diffraction gratings.
An important advantage of diffraction optics is the capability of producing
complex optical components. One of these is the holugraphic image multiplier
on which the Fourier transform of a set of point sources is recorded. Such
a device produces multiplied images with resolution of 6 Enn on a S x 5 cm field.
Holographic Diffraction Gratings. These are made in sizes up to 150 mm, band
frequency 3000 lines/mm, resolution of the order of 1~i. The holographic
method produces gratings that give forward diffraction of a wave of given
shape, enabling correction of aberrations of optical systems of spectral in-
struments.
Holograms as independent optical elements produce a real image of an object
without using additional lenses and ob~ectives. Holographic methods that
permit recording of several holograms on a single photographic plate open
up new possibilities for making light and compact optical devices capable
of performing the functions of several components simultaneously [Ref. 22].
One of the important jobs on the road to development of holographic optics
is to work out a method of getting holograms that have high diffraction ef-
ficiency, i. e. low losses [Ref. 107, 108]. Obviously these must be phase
holograms since amplitude holograms inevitably absorb part of the energy passing
through them.
The development, standardization and classification of artificial optical
components enables us to make optical devices in accordance with principles
analogous to those in electronic systems.
Optical effects in thin-film waveguides can be used for optical and optoelec-
tronic data processing. Thin-film dielectric waveguides can be used as a
basis for miniature laser devices, optical modulators, filters, parametric
generators and other components for communication systems with large informa-
tion capacity, high-speed computers and for optical data processing systems.
The production of thin-film optical components on a flat dielectric backing
opens up the serious possibility of integrated optical circuits [Ref. 145].
87
FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500090006-1

APPROVED FOR RELEASE: 2407/02/09: CIA-RDP82-00854R000500090006-1
FOR OFFI('IAL l1SE ONLY
InteKrated optics sprang up as a result of studying the principal properties
of optical waveguides on a dielectric backing, development of the method of
input and extraction of light through a waveguide surface based on the optical
tunnel effect, and accomplishment of laser frequency conversion in a nonlinear
thin-film waveguide.
- One of the problems of integrated optics is creation of passive optical compo-
nents (lenses, prisms, diffraction gratings) in film. Just as prisms and
lenses are formed in conventional optics by appropriately shaping the surfaces
of transparent dielectrics, so components that act as prisms or lenses can
be formed in thin films by appropriately shaping the boundaries of a region
of change in the index of refraction. For practical purposes, this can be
done by a local change in thickness of the film. In another technique, prisms
and lenses are formed by introducing a suitably shaped layer with high index
of refraction into a waveguide film or substrate. Film prisms can be used
to analyze the frequency spectrum of a waveguide light beam in any mode, or
for spatial separation of the light of different modes on the same frequency.
Thin-film lenses for a surface wave are formed by shaping a curved boundary
of the region with changed f ilm thickness.
An ~.mportant part in making prisms and lenses in thin-film optics is played
by dispersion of the index of refraction, which is readily altered over a
wide range by selecting film thickness. It should be noted that we can have
refraction on the boundary of two regions without dispersion, which is important
for making achromatic prisms and lenses.
In addition to prisms and lenses, diffraction gratings can be made in two-
dimensional optics by applying closely spaced depressions on film surfaces,
or by applying dielectric strips with low index of refraction. The constant
of propagation of the surface wave in such a structure undergoes periodic
variations, leading to diffraction effects similar to light scattering on
a standing acoustic wave in a volimmetrir medium. These phenomena can be used
in a number of thin-film devices of the spectral f ilter and mode selector
- type. Material with high saborption applied instead of a dielectric on the
film surface causes rapid attenuation of the optical surface wave. This effect
can be used to create thin-f ilm equivalents of amplitude masks, spatial filters,
gratings and lenses of the Fresnel zone plate type.
Dielectric films with thickness of the order of a wavelength of light or less
are important in integrated optics. Critical film thicknesses corresponding
to law-order surface waves lie in this range. At the minimiua effective thick-
ness of the waveguide, the maximum gradient of the effective index of i�efrac-
tion is reached. The optical tunnel effect based on a prism is used for
coupling the radiation into and out of thin-film waveguides. It is a compli-
cated job to get the optimum profiles of the film and gap that ensure total
input of a given li~ht beam into a predetermined mode of a thin-film waveguide.
In addition to t~xnnel input of radiation, it is highly effective to excite
surface waves by a phase diffraction grating applied on the surface of the
waveguide film. The input devices use both planar and volwnetric diffraction
gratings that are applied directly on the wavetuide film. Radiation can be
coupled in and ou*_ through a t~~pering,edge of waveguide film.
88
FOR OFF7CIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500090006-1

APPROVED FOR RELEASE: 2007/02/49: CIA-RDP82-40850R040500094006-1
FUR OFFI('IAL USE ONLY
The waveguide film may consist of dielectric layers of optically nonlinear
material. The properties of nonlinear optical waveguides can be used to excite
radiation on combination frequencies of the initial waves, and also to get
the second harmonic. Thin-film optical waveguides can be used for spatial
scanning of surface waves. Of greatest promise here is an acoustic method--
interaction of light with elastic waves propagating in the film. The capa-
bility for getting high concentrations of optical and acoustic energy in thin
films allows creation of efficient thin-film spatial modulators. The acoustic
method of converting surface wave modes is not unique. It is possible to
convert optical modes in films with anisotropic and gyrotr~~pic backings and
boundary interfaces.
At the present time, a number of active components have been suggested and
realized in thin-film waveguides. An investigatior.. has been made of stimu-
lated emission of gelatin f ilms of different thicknesses and under different
pumping conditions. In addition to the capability for single-frequency lasing
in a film 14 1na thick, the possibility has been noted of multifrequency lasing
with strong pwnping and in thicker films. The multifrequency nature of the
radiation is associated with stimulated emission of different modes of the
gelatin film. A change in the period of spatial modulation of the index of
refraction or gain gives the capability of tuning laser wavelength with dis-
tributed feedback. For these purposes we can use the property of an organic
dye solution to change gain and index of refraction upon absorption of intense
optical radiation. Creation of a thin-film ring laser in which strong feedback
is set up by the simplest method enables expansion of the class of thin-film
lasers by using media with moderate gain. The thin-film ring laser can be
quite simply joined to a flat film on a backing, and is used for coupling
radiation into the film.
Electroluminescent and semiconductor lasers can be used as thin-fiLn active
elements, and have a fine outlook in integrated optics in view of their small
size, high efficiency and other advantages over lasers of different types.
In recent years, integrated optics has reached a l~:vel of development where
its capab ilities revealed in laboratory studies are findin.g ever wider appli-
cation for optical and optoelectronic data pxocessing.
COPYRIGHT: Izdatel'stvo "Mashinostroyeniye". 1976
6610
CSO: 1862/163
89
FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500090006-1

APPROVED FOR RELEASE: 2007102109: CIA-RDP82-00850R000500090006-1
FOR OFF7CIAL USE ONLY
pI,ASMA PflYSICS ~
UDC 533.9.01
NONEQUILIBRIUM LO'W-TEMPERATURE PI.~?SMA KINETICS
Moscow KINETIKA NERAVNOVESNOY NIZKOTEI~ERATURNOY PLAZP'n in Russian 1982 (signed
to press 10 Dec 81) pp 2-5, 373-375
[Annotation, preface and table of .:ontents from book "Nonequilibri~ Low-
Temperature Plasma Kinetics", by Leon Mikhaylovich Biberman, Vladimir
Sergeyevich Vorob'yev and Igor' Tevfikovich Yakubov, Institute of Low Tempera-
ture Physics, USSR Academy of Sciences, Izdatel'stvo "Nauka", 2350 copies,
376 pages]
[Text] This book is the first to cover a wide class of problems in nonequi-
librium low-temperature plasma kinetics. Data are given on collisional and
radiative elementary processes. Radiative transfer of excitation is con-
sidered. Criteria of equilibrium under a variety of experimental conditions
are discussed. Nonequilibrium distributions of atoms by levels are described
as well as electron energy distributions. Methods are outlined for calculating
coefficients of ionization and recambination. Unsteady processes are discussed.
Some questions of kinetics in a molecular plasma are conaidered.
The book is intended for scientiats and engineers working in the fihlsical
plasma physics and plasma chemistry, electric discharge in gases, p y
electronics, and also for graduate students and upperclassmen majoring in
physical and technical fields. Figures 128, tables 44, references 490.
Preface
The first plasma research was done in connection ~;ith atudying electric dis-
charge in gases. Physiciats had been concentrating on partly ionized plasma
with kinetics deteranined by various collisional and radiative processes. This
trend of resean~h ~a~hee~i~ inedevelopingsgasrdischargenlightpsourcesl recti~s
that were urge �
fiers, inverters.
Since the early fifties, interest in plasma physics has taken a sharp upturn.
This applies primarily to the investigation of completely ionized plasma with
its various collective pheno~ena, instabilities, interesting and at times
unexpected effects associated with propagation of electromagnetic waves in
such a plasma and the action of external electric and magneyic f~hedvariethe
plasma. Interest in hot plasma has been stimulated not onl roblems that have
and novelty of physical processes and effects, but also by p
arisen in connection with controlled nuclear fusion.
90
FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500090006-1

APPROVED FOR RELEASE: 2007/42/09: CIA-RDP82-00850R000500090006-1
FOR OFF[CIAL USE ONLY
The appearance of new technical fields in the early sixties, such as gas-
discharge lasers, magnetohydrodynamic generators, thermoemission converters,
plasma chemistry, plasma engines, various methods of plasma technology and
the like, rekindled interest in weakly ionized low-temperature plasma. This
applies primarily to nonequilibriimm plasma that is distinguished by extra-
ordinary diversity of states and properties. It is in a weakly ionized non-
equilibrium plasma that population inversion of excited states of atoms and
ions is realized, high electrical conductivity of low-enthalpy gas flaws is
attained that is necessary for installations of direct energy conversion,
and selective excitation of individual states of atoms is realized that ensures
eff iciency of plasma-chemical reactions. Specific oscillations and instabili-
ties that originate in nonequilibrium low-temperature piasma are of consider-
able physical interest.
Low-temperature plasma kinetics is determined by a combination of a large
number of elementary proceases, among which we mention inelastic collisions
of electrons with excited and unexcited atoms, inelastic collisions of atoms
and ions, processes of associative ionization and dissociative recombination
and many others that are absent or of little significance in a hot plasma.
- Processes of radiative excitation transfer play an appreciable part. As a
result, the kinetics of low-temperature plasma is in some sense also more
complicated and varied than the kinetics of a campletely i.onized plasma,
largely due to the presence of atoms and molecules with their nwnerous excited
states.
Despite the considerable advances that have been mRde in recent decades, low-
temperature plasma kinetics has not yet bean duly reflecte.d in monographs ~
devoted to plasma physics. The widely known series "Probl.ems of Plasma Theory"
' edited by M. A. Leontovich, and books by L. A. Artsimovicri and R. Z. Sagdeyev
"Plasma Physics for Physicists", I. Shksrovskiy, T. Johnston and :~1. Bachinskiy
"Kinetics of Plasma Particles", N. Kroll and A. Trayvelpis "Fundamentals of
Plasma Physics" and others have dealt primarily with other problems. In this
book the suthors have attempted to fill this gap. Ma3or e:mphasis has been
given to rinetics of ionization, excitation, recombinatior~, energy distribution
among plasma components and individual degrees of freedom of these components,
energy exchange between plasma and ambient medium. The p].asma is treated
as an interrelated system of electrons, ions and atoms in different energy
states. Interaction among particles in the presence of er:ternal perturbations
gives rise to compromised nonequilibrium states. A more ~letailed exposition
of some problems can be found in other publications. The state of a plasma
in magnetic fields is described in the series "Problems of Plasma Theory".
The'book by A. V. Yeletekiy, L. A. Palkina and B. M. Smirr~ov "Transport Phe-
nomena in Weakly Ionized Plasma" contains a fairly c o mp lete exposition of
the physics of transport phenomena in law-temperature plasma. Instability
problems are covered in a series of surveys by Ye. P. Vel~.khov et al., and
also in the book by A. V. Nedospasov and V. D. Khait "Oacillations and Insta-
b ilities of Law-Temperature Plasma". There is a series of monographs dealing
with problems of kinetics that arise in connection with development ef specific
applied devices: MfID generators,.lasers and the like. In our monograp:~ we
cover problems that are in somc: measure common to all these applied studies.
Nonequilibrium plasma theory is presented simultaneously with numerous experi-
mental results.
91
' FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500090006-1

APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500090006-1
FOR OFF7CIAL USE ONLY
C1?apter 1 outlines the region of parameters corresponding to low-temperature
plasma, giving the major concepts of plasma physics necessary for further
exposition.
Chapter 2 is devoted to elementary collisional and radiative processes. It
is a reference chapter that provides the minimum required information on proba-
bilities of elastic and inelastic collisions and radiative transitions. Con-
siderable attention is given to approximate and sewiempirical relations for
probabilities of various elementary processes.
Radiative excitation transfer plays a rather important part in nonequilibrium
plasma kinetics. This question is treated in Chapter 3.
The authors have deemed it necessary to set forth the criteria of arisal of
various kinds of nonequilibria individually in the fourth chapter. Research
results given in this chapter are presented on the basis of physical consider-
ations, and can be used without studying the subsequent chapters where they
are more rigorously substantiated.
Chapter 5 examines the distribution of atoms with respect to excited states.
This topic is of interest not only from the standpoint of the optical proper-
ties of plasma and its diagnosis. L'xcited atoms play a quite important part
in the kinetics of ionization and recombination, being as it were the rungs
of a ladder over which the electron passes from the bound to the free state
and back.
Chapter 6 outlines the kinetics of ionization and recombination. Various
mechanisms of ionization and recomb ination are discussed, methods are given
for calculating coefficients of ionization and recombination, and also reference
data. Nonequilibrium ionization in a plasma is cor..sidered.
Chapter 7 is devoted to the velocity distribution of electrons, and to the
energy balance of an electron gas. Considerable attention is given to the
influence of inelastic collisiona on energy distribution of electrons, and
to the relation between nonequilibrium electron energy distribution and the
distribution of atoms by excited states.
_ In chapter 8 we consider some questions of an unsteady nonequilibriiuo plasma;
primarily the rather important criteria of quasistea.~iness, problems of ioniza-
tion relaxation, distribution of atoms by excited states and the electron
energy distribution function. Also discussed are some problems of low-tem-
perature plasma instability that are closely related to the kinetics preser~ed
above.
Molecular plasma kinetics is much more complicated, and at the present time
has been less developed than the kinetics of atomic plasma. Some questions
of molecular plasma kinetics are discussed in chapter 9. With a view mainly
to gas lasers, we briefly consider the kinetics of population of vibrational
states of molecules. This section of chapter 9 was wri.tten by A. A.
Likal'ter, who has the authors' sincere p,ratitude.
Abbreviated notation is used in tables: e. g. 6.37-2 means 6.37�10-2
92
FOR OFF'ICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500090006-1

APPROVED FOR RELEASE: 2007/02/49: CIA-RDP82-00850R440500090006-1
F~R OFFICIAL USE ONLY
Contents page
Preface 3
Chapter 1
Low-Temperature Plasma. General Information 6
1.1. Quasineutrality. Debye shielding 6
1.2. Ideal plasma 8
1.3. Equilibrium plasma 9
1.4. Local thermodynamic equilibriwn. Elementary processes 11
1.5. Particulars of transport phenomena 13
1.6. Nonequilibrium law-temperature and high-temperature plasmas 16
Chapter 2
Elementary Processes in Low-Temperature Plasma 19
2.1. Elastic collisions 19
2.2. Inelastic collisions of electrons with atoms, ions and molecules 22
2.3. Inelastic collisions with heavy particles 44
2.4. Elementary radiative processes 59
2.5. Average energy transferred to atom in collisions 66
Chapter 3
R~..iiative Excitation Transfer 74
3.1. Major peculiarities of radiative excitation transfer 74
3.2. Equation of radiative excitation transfer
3.3. Approximate method of effective lifetime 80
3.4. Radiative transfer of excitation in inhomogeneous medium 84
3.5. Limits of applicability of the theory 89
Chapter 4
Criterion of Arisal of Nonequilibrium States 91
4.1. Criterion of electron temperature separation 92
4.2. Criterion of equilibrium ionization and equilibrium distribution of
atoms by levels 96
4,3, Criterion of violation of maxwellian distribution 109
Chapter S
Kinetics of population of excited states 114
5.1. Qualitative pattern of population distribution in nonequilibrium
plasma 114
5.2. System of kinetic balance equations for populations of excited
states 118
5.3. Numerical methods of solving kinetic system of equations for
populations 122
5.4. Diffusion approximation 125
5.5. Discr.ete methods and mudified diffusion approximation 132
5.6. Comparison of analytically found populations with data of computer
calculations and experiments 141 .
5.7. Influence of atom-atom collisions on population distribution 145
5.8. Accounting for sources of excited atoms in system of balance
equ2 ~:ions 147
5.9. Peculiarities of shock-radiative kinetics in rarefied plasma 153
5.10. Some applications of the theory 157
Chapter 6
Kinetics of Ionization and Recombination 164
6.1. Elementary kinetics of ionization and recombination 16~
93
FOR OFFICIAL USE ONtiY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500090006-1

APPROVED FOR RELEASE: 2407/42/09: CIA-RDP82-40850R000500490006-1
FOR OFFICIAL USE ONLY
6.2. Principal equations of ionization and recombination kinetics and
results of numerical solution 174
6.3. Coefficients of shock-radiative recombination in diffusion and
modified diffusion approximations 183
6.4. Electron concentration in nonequilibriimm steady-state conditions 205
Chapter 7
Electron Energy Distribution and Electron Energy Balance 209
7.1. Kinetic equation and electron energy balance ' 210
7.2. Inelastic collisions. Their effect on electron energy balance,
frequency of excitation and ionization 222
7.3. Self-consistent electron energy distributions and distributions of
atoms by excited states 239
7.4. Electron energy distribution in strong electric field 245
Chapter 8 '
Unsteady Nonequilibrium Plasm~ 255 ~
8.1. Criteria of quasisteadiness 255
8.2. Ionization relaxation 263
8.3. Radiation of unsteady plasma 2~3
8.4. Relaxation of distribution function 276
8.5. Instabilities of nonequilibrium plasma in external electric field 282
Chapter 9
Some Problems of Molecular Plasma Kinetics 293
9.1. Electron energy balance 293
9.2. Electron energy distribution function 299
9.3. Distribution of molecules by vibrational levels 304
9.4. Electron-ion recombination in molecular gases 315
9.5. Some problems of kinetics of atomic~molecular plasma 324
Appendices 330
References 352
COPYRIGHT: Izdatel'stvo "Nauka", 1982
6610
� CSO: 1862/166
94
FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500090006-1

APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00854R004500090006-1
FOR OFFICIAL USE ONLY
UDC 533.92
DYNAMICS AND RADIATION OF OPEN (VA~UUM) PLASMA-DYNAMIC DISCHARGES OF 'PLASMA
FOCUS' TYPE: SURVEY .
Moscow TEPLOFIZIKA VYSOKIKH TEMPERATUR in Russian Vol 20, No 2, Mar-Apr 82
(manuscript received 25 Feb 80) pp 359-375
[Article by A. S. Kamrukov, N. P. Kozlov and Yu. S. Protasov, Moscow Higher
Technical Academy imeni N. E. Bauman]
[Text] The paper gives the results of an experimental study
and an aly si s of the dynamics, space-time structure and
emission propertie,s in the range of energies of quanta of
2-350 eV for open (vacuum) intense plasma-dynamic discharges
of the "plasma focus" type. A method is described for com-
prehensive investigation of the radiation properties of
intense emitting discharges in the spectral region from
the visible to the extreme ultraviolet (XUV). It is shown
that thz emission spectrum of a plasma focus is sharply
different from planckian, and is due mainly to the recom-
bination continuum of typical groups of ions that determine
the properties of plasma at a given temperature and density,
70-90~6 of the emitted energy belonging to the XUV region
~ of the spectrum.
The authors consider possibilities of effective control
of the emission spectrum over a wide range of energies of
quanta (influence of macrostructure of hypersonic flow,
chemical composition, etc.) and attainment of high bright-
nesses in the XUV. Brightness temperatures exceeding 5 eV
have been reached for the first time in the region beyond
~ the helium ionization potential.
Recent years have seen a considerable upsurge of interest in development of
powerful sources of radiation with high brightness temperature in the extreme
ultraviolet region. This is due to expansion of the sphere of scientific
and applied problems whose solution in large measure depends on development
of such sources. In the first rank among such problems are promising develop-
ments in design of powerful lasers for the visible and near ultraviolet region
for which the absorption bands of working media are in the XW region (lasers
that use allowed electron transitions. of molecules [Ref. 1], excimer
~ 95
FOR OFF(CIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500090006-1

APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00854R004500090006-1
~ FOR OF~'ICIAL USE ONLY
photodisaociation lasers [Ref. 2]), and also research on the pos~ibilities
of developing new types of coherent radiators (photoionization lasers [Ref.
3]). Furthermore, such sour~es can be used for other purposes as well, e. g.
= for studying photochemical reactions, processes of interaction of radiant
high-density fluxes with condensed media, etc.
The o utlo o k for developing high-intensity plasma sources in the XW region
is determined in large measure by the capabilities of getting sufficiently
large volumes of dense (Ne = 1017-1019 cta 3) plasma with temperature Te =
2-10 eV, for which a considerable part of the radiation may be in the extreme
ultraviolet and associated with photorecombination of electrons to the ground
states of double, triple and higher-multiple ions. High selectivity of the
emission spectrum of such a plasma, which is important for example for exci-
tation of active laser media, can be attained at fairly small optical thickness
of the plasma, where recombination maxima show up in the emission spectrum
and have appreciable intensity.
Attracting the most attention among methods of getting a dense radiating plasma
is an electric-discharge method with plasma containment by the magnetic self-
field of the discharge current. However, in developing XUV radiation sources
based on intense self-compressed discharges of the z-pinch type, fundamental
difficulties arise that are associated with shielding of short-wave radiation
by layers of cool plasmar expanding at thermal velocity (vacuum discharges
in mstal vapor [Ref. 5]), or with arisal of an ionization wave that stabilizes
the brightness temperature of the radiating surface on a level of 2-3 eV (dis-
charges in inert gases [Ref. 6]). In contrast to the mentioned sources of
radiation, conditions may be reallzed in the intense plasma-dynamic discharges
of a magnetoplasma crnnpressor of the "plasma focus" type [Ref. 7] under which
~hielding of short-wave radiation of the hot plasma is excluded in principle
[Ref. 8]. As is known [Ref. 7], a characteristic feature of self-compressed
discharges of magnetoplasma compressors is the high average mass velocities
of the radiating plasma (v~~- 50 km/s), considerably exceeding its thern~al
~ velocities (Mach number in the flow M~ 5-10 [Ref. 9]). Thus even in a quasi-
steady state of flow, an optically dense shielding layer is nct formed due .
to satisfaction of the condition v~~�v~~ vtherm (vl is the velocity component
perpendicular to the axis of flow).
Of interest in connection with this is a detailed study of the space-time
structure and peculiarittes of spectral distribution of emission of open
plasma-dynamic discharges for the purpose of determining capabilities for ef-
fective control of the emission spectrum over a wide range of quantum energies
hv - 5-100 eV, and attainment of high brightness characteristics in the extreme
ultraviolet.
1. Experimental Conditions and Research Method. Studies of dynamic and radia-
tion properties of open (vacuum) intense plasma-dynamic discharges were done
on a facility consisting of a steel discharge chamber 0.8 m in diameter and
3 m long pwuped out by a high-capacity vacuum system to a pressure of 10-3 Pa,
a capacitive energy storage device, a charging module with initiation unit,
and also measurement and diagnostic instrumentation.
96
FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500090006-1

APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500090006-1
FOR OFFI('IAL USE ONLY
10
_
~
~
8 I .
!1 ~ 1
7 = 4
. 3 Z 5
9 I
6~ ,
Fig. 1. Diagram of experiment: 1-=vacuwn section; 2--
magnetoplasma compressor; 3--quartz (glass) calorimeter;
- 4--DFS-29 diffraction spectrometer; 5--differential high-
vacuum pumping units; 6--open ionization chamber; 7--vacuum
standard based on KRIS; 8--bolometric section; 9--photo-
electric sensor with M-foils; 10--scintillation spectrometer;
11--Xr-Co thermocouples
The plasma sources in the described experiments were magnetoplasma compressors
- of erosion. type comprising a coaxial system of cylindrical or specially shaped
electrodes separated by a dielectric sleeve. The plasma-forming substances
were products of erosion of the electrodes and/or products of ablation of
the separating sleeve. The chemical composition of the electric-discharge
, plasma was controlled by appropriate choice of the construction materials
of the magnetoplasma compressor. For example, to get a discharge in metal
vapor, heat-resistant dielectrics wer~ used as the material for the separating
sleeve (boron carbonitride BNC, zirconium dioxide Zr02 or alundum A1203),
and the electrodes of the magnetoplasma compressor w.ere ma.de of the appropriate
metal (Cu, Cd, A1, Mo). The main yield of inetal in the discharge came from
the central electrode (cathode); the partial fraction of erosion of the outer
electrode was less than lOX, and did not significantly change the chemical
composition of the plasma. When a separating sleeve of ablatable material
was used (fluorocarbon plastic (C2Fy)n, polyformaldehyde (CH2O)n, cesium iodide
CsI, etc.) with electrodes of erosion-resistant metals (Mo, W, Cu), the plasma
composition was determined mainly by the products of dissociation and ablation
of the d~ie~ectric.
Magnetoplasma compressors were used in the experiments with geometry of the
electrode system similar to that described in Ref. 10; diameters of the outer
electrode were varied over a range of 30-8~ m~n. The magnstoplasma compressor
models were installed on the dielectric end flanges of the discharge chamber,
and the investigated discharges were observed and diagnosed through optical
side windows in the chamber. All expe~riments were done.with negative polarity
of the central electrode in the first half-period of the discharge current.
97
FOR OFFICIA?. USE ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500090006-1

APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500090006-1
FOR OFFIC'IAI. USE ONI.Y
The electric energy accumulator was a bank of low-inductance IlrIIri-5-150 capaci-
tors connected in parallel. Maximum capacitance of the bank in the described
experiments was C= 750 uF, maximwn charging voltage was Uo = 5 l~V.
The electrotechnical parameters of the discharge were measured by a standard
technique, using calibrated Rogowski loops and r.ompensated RC voltage dividers.
The discharge was periodic in nature with strong damping. As a ruie, a current
pulse contained two or three half-periods with duration varying over a range �
of 17-20 us for different models. Maxim~ current was reached in 6-8 us of
the discharge, and in operation on a fluorocarbon plasma at Co= 750 uF was
~420-460 kA (at a rise time dI/dt up to 0.8�1011 A/s). Current damping~factor
was -7.7�104 s-1. The total inductance and impedance of the discharge circuit
together with the magnetoplasma compressor as calculated from the electrical
engineerin~ equation for the current in the RLC tank were ~40 nH and
-(3-5)�10' S~ respectively.
The energy input to the discharge was determined in Each specific experiment
from oscillograms of the current and vol*_age, and at Co= 750 uF was ~75-85X
of the energy stored in the capacitor bank, about 90-95X of the invested energy
being introduced during the first half-period of the discharge. Typical values
of peak electric powers were ~0.8 GW.
The space-time development of open plasma-dynamic discharges was studied by
SFR-2M high-speed cameras operating in single-frame and slit-scan modes.
The density of flows and the time-integrated radiation output in individual
spectral intervals of the optical band were determined by light-filtered photo-
cells calibrated by the EV-45 reference source. Recording of the time-scanned
radiation spectrum was done with an ISP-30 quartz spectrograph equipped with
disk slit chronograph. The space-time distribution of brightness temperature
Tbr of the radiation in the visible region was studied by methods of mono-
chromic photometric comparison of densities of blackening of negatives of
the slit scans of the discharge and the EV-45 reference source,~and the spec-
tral distribution of Tbr was determined by the method of photoelectric compari-
son of radiation intensities.
The brightness temperatures of discharges in the XUV range were measured by
- double open ionization chambers, the integrated energy output of short-wave
radiation was determined from heating of a quartz calorimeter and by b~lometric
methods, the relative distribution of radiation over the spectrum was studied
by a photoemission-scintillation spectrum, the XUV spectrum was recorded by
the DFS-29 vacuum spectrograph. Fig. 1 shows a diagram ~f an experimental
setup for studying the emission characteristics of plasma-dynawtc discharges
in the extreme ultraviolet.
2. Dynamics of Open Self-Compressed Plasma-Dynamic Dischirges. Fig. 2[photos
not reproduced] shows typical photochronograms that illt~~trate the dynamics
of development of an open plasma-dynamic discharge. The discharge was ignited
by a high-voltage (20-50 kV) ignition pulse to the third auxiliary electrode
installed in the dielectric sleeve in the middle of the discharge gap in the
central electrodes of the magnetoplasma compressor.
98
FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500090006-1

APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-04850R000500090006-1
FOR OFFICIAL USE ONLY
The plasma of the ignition discharge ini~iates a powerful streamer discharge
over the surface of the dielectric, leading to intense vaporization and ioni-
zation of the working substance. Streamer formation as a~~ontracted current
channel (Fig. 2) causes amplitude inhomogen.eity of the plasma flow on the
initial stage of the discharge (T = 2-4 us).
The plasma in the interelectrode gap gives rise to volumet~ric electric current
whose radial component jr upon interaction with the azimut:~al magnetic self-
field B~ sets up a longitudinal component of ampere force FZ = jrB~ that ac-
celerates the plasma. The process of plasma acceleration .is accompanied by
electromagnetic cumulation of flow along the axis of the system and formation
of a compressed zone the plasma focus beyond the tip of the magnetoplasma
_ compressor (Fig. 2a). The mechanism of electromagnetic cwnulation reduces
to a two-stage process that shows up as Hall compression of the flow toward
the central electrode (cathode) in the interelectrode gap, and collapsing
of the plasma due to the pinch effect in the entrained streams beyond the
tip of the magnetoplasma compressor [Ref. 7].
The region of compression is localized in space, and is a macroscopically
stable formation, i. e. a fairly clear outline of a quiescent ~et is observed
over nearly the entire first half-period of the discharge. Maximum compression
occurs 1-2 us after attainment of the maximum current in cross sections
1.5-3 cm away from the tip of the magnetoplasma compressor. The transverse
size of the flow in this case is a few mm, length of the compression zone
is 7-15 cm, divergence of the jet is 12-16�.
The plasma flow in the compression zone is contained by the magnetic field
of the discharge. Estimates of the ratio cf magnetic pressure pM and gas-
dynamic pressure pr
f,,: ~i,,/!=/2 10-'/=Jrcd'
..p~ ~ N(l+z)l.'l
( ,V~ ~ N; 1 Ir7'
\ /
~ show that for typical plasma parameters (see below) the concentration of
heavy particles N~ 1019 cm 3, avera~e chsrge of ions z=Ne/ENi - 2, temperature
kT ~ 5 eV, diameter of the plasma column d~ 0.6.cm, magnetic forces predominate
over gasdynamic forces even at currents greater than ~40 kA.
An important characteristic of the magnetoplasma compressor as a device for �
producing dense high-velocity plasma flows is the coefficient of utilization
of the working substance r~,s, which is the ratio of the amount of material
flowing through the zone of the plasma focus to the total consumption of mass
over the discharge pulse. In its physical sense, this parameter characterizes
tt~e macro.scapic structure of the plasma flow discharged from the magnetoplasma
compressor. At high values of n~,s ~(G.9-0.95) (Fig. 3a, c[photos not repro-
duced]) the flow is a dense magnetohydrodynamically formed plasma ~et with
small angular divergence. A reduction of r~.s (Fig. 3b [photo not reproduced])
results in a considerable part of the erosion mass not flowing through the
focus zone, and being discharged from tne magnetoplasma compressor as a super-
soriic jet that is practically not contained by the magnetic field of the dis-
charge and expands freely into the vacuum. Capabilities for controlling r~,s
99
FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500090006-1

APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500090006-1
FOR OFFICIAL USE ONLY
and consequently the macrostructure of the flow in discharge systems based
on magnetoplas~na compressors of erosion type involve appropriate choice of
energy conditions, the geometry of the discharge gap, thermophysical properties
and the method of introducing the plasma-forming substances into the discharge.
As will be demonstrated below, the macroscopic structure of the flow has a
considersble effect on the radiation properties and spectral distribution
of emission energy of an open plasma-dynamic discharge.
The microscopic structure of the plasma jet escaping from the magnetoplasma
compressor is revealed on photochronograms of the discharge taken with high
space-time resolution. Fig. 2c [photo not reproduced] shows a slit scan of
a discharge across a fluorocarbon plasma obtained with orientation of the
slit along the axis of the f low. The formed magnetoplasma compressor flow
is characterized by a dis~rete structure observed on the photochronograms
in the form of alternating dark ~^d light bands. The characteristic dimension
of inhomogeneities in the direct_~n of the axis of flow and the recurrence
rate were determined by photometry of the negatives of the scans, and are
0.2-0.5 cm and 5-10 MHz respectively. It is known that the interrupted flow
structure observed in a number of papers [Ref. 11-17] in the investigation
of pulsed accelerators of gas-discharge and erosion plasma is due to the exis-
tence of individuai microplasmoids [Ref. 17]. Investigation of the nature
of erosion destruction of the surface of electrodes and dielectric inserts
of magnetoplasma compressors and r.omp~rison with the results obtained on other
erosion plasma systems [Ref. 17] shows that the formation of individual micro-
plasmoids is due to the discrete nature of arrival of the mass of plasma-
forming material in the discharge, whi~h is caused by space-time inhomogeneity
of energy release on the surface of the electrodes and dielectric inserts.
v, km/s If the inhomogeneities are detached plas-
~ moids, the velocity of their propagation
s~ ~ 2 (which can be determined from slit scans
that are x-t diagrams) will be the velocity
of the flow. The results of processing
' ~ slit scans obtained at different initial
1 voltages Uo of the discharge are shown
~ on Fig. 4. Typical of the initial dis-
iu ~0 ` av t.i~ T~ us charge stage is considerable time inhomo-
geneity of velocities. With increasing
Fig. 4. Velocity of plasma in Uo, inhomogeneity increases rap~idly. The
free ~at at distance L= 10 cm average mass velocities of the plesma
from the tip of the magnetoplasma flow are ~40-50 km/s. The resultant dis-
compressor (C = 750 uF, C2Fy tribution of velocities agrees with the
plasma) at Uo: 1--3 kV; 2--4 kV; results of ineasurements of the dynamic
3~-5 kV characteristics of the discharge by using
double electrostatic probes [Ref. 18].
The transverse structure of the flaw in the compression zone was studied by
slit scans obtained with the slit of the streak camera normal to the dis-
charge axis. Fig. 2d [photo not reproduced] shows a typical photochronogram
corresponding to the cross section at distance L= 2.5 cm from the tip of the
magnetoplasma compressor. It can be seen that tbe plasma column preserves
100
FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500090006-1

APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500090006-1
FOR OFF[CIAL USE ONLY
ics dimensions and position in space practically throughouL- the first half-
period of the discharge, i. e. force instabilities of the type of pinches,
kinks and the like that are typical of intense self-compressed discharges
are absent [Ref. 4, 19]. At the same time, the initial stage of the discharge
shows azimuthal inhomogeneity due to streamer formation and subsequent propa-
gation of the ionization front over the surface of the dielectric leads to
development of a helical instability with characteristic frequencies of oscil-
lations that vary over a range of 200-300 kHz. Four micro:3econds after supply
of the initiating pulse the corkscrew instability becomes :small-scale. A
peculiarity of the structure of the plasma flow in the comUression zone is
also the presence of transverse small-scale high-frequency oscillations
("noises") with characteristic frequencies of ~10 MHz. Let us note that
small-scale structures of this kind are also observed in ti?e region of cam-
pression of a gas-discharge magnetoplasma compressor [Ref. 20].
3. Space-Time and Spectral Characteristics of open Plasma-Dynamic Discharges
in the Quantum Energy Region of hv ~ 2-350 eV. Ref. 7, 21, 22 devoted to the
investigation of radiation properties of a dense plasma focus give data that
reflect the integrated emission characteristics of open plasma-chemical dis-
charges, i. e. that characterize the spectral distributions of energy densities
and radiated power averaged over the entire emitting surface of the discharge.
At the same time, knowledge of the local parameters of the radiation, their
space-time and spectral distributions, is necessary for a more complete idea
of the potential capabilities of the discharge as a source of radiation, as
well as to determtne the mechanism of rzdiation and the nature of the emission
spectrum of the plasma.
The space-time distribution of brightness temperature (spectral brightness
of radiation) of an open plasma-dynamic discharge was studied by the method
of photometric comparison of the blackening densities of photographic film
y~ith time scanning of luminescence of a plasma focus and a reference light
source on the SFR-[M high-speed camera. Photometry was done on f ilms of the
slit scans of the plasma focus taken with orientation of the camera slit both
along and across the discharge axis. The measurements were made in the blue
region of the spectrum on an effective wavelength of aef = 430 � 20 nm isolated
by a set of light filters (SS4 + SZS-22 + ZhS-11). An EV-45 pulsed light source
was used as the reference. In photographin� the reference source, a nine-step
neutral optical wedge [Ref. 23, 24] was installed on the focal arc of the
SFR-2M; this enabled coverage of the entire optical range of photographic den-
sities of the film anticipated in photographing the plasma focus. The position
of the rotating mirror of the camera was synchronized with the flash of the
EV-45 source in such a way as to expose the photographic film behind the opti-
cal step wedge in the time period when the brightness of the reference standard
was constant.
Identity of exposure conditions in photographing the marks of the reference
standard and the investigated effect does not require the use of reciprocity
law in determining temperature [Ref. 23-26], thus eliminating possible con-
comitant errors. The films of each series of experiments and the photographs
of the reference sources were developed concurrently to eliminate the effect
of photochemical processing on measurement results.
101
FOR OFFiCIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500090006-1

APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500090006-1
~ FOR OFFICIAL USE ONLY
The use of a source with brightness temperature close to t1:3t being measured
(see below) as the comparison standard reduces the error of the method. The
error in relative measurement of energy brightness does not exceed �5X (error
of the method), and the error in absolute measurement of energy brightness
(error of the method and refer~nce) does not exceed *_15X for the visible region
[Ref. 23]. The corresponding accuracy of brightness temperature measurement
is �5-10%.
Tbr, eV
. _
b ^ ~
a '1 ~
~ ~
2
4 us
~
' ~ j
~ y._
~ ~
i ~ 1
. � Zoo ~ni: aou S00 IP, Ka
~ p ~ i ~ ' _._.1_-~.-L-,._ ~
s ro ~,s ~n 2;~ au t, us Z 4 s uo, kv
Fig. 5. Results of photographic measurements (C2F4 plasma,
C o= 750 uF) in the center of a plasma f ocus (L = 25 mm) :
a--typical densitometer plot of the discharge (Wo= 9.4 kJ);
b--time dependence of brightness temperature of the discharge:
1--Uo~ 5 kV; 2--2 kV; c--dependence of maximum brightness
temperature on the electrotechnical parameters of the discharge
Fig. 5a shows a typical densitometer plot of the discharge obtained by micro-
photometry along the time axis of a slit scan of the central zone of the plasma
focus in the cross section at a distance L= 25 mm from the tip of the ma~neto-
plasma compressor (slit oriented across the flow), and Fig. Sb shows the re-
sults of ineasurement processing. At stored energy Wo= 9.4 kJ the maximum
~ brightness temperature in the zone of the plasma focus is ~3 eV (~35,000 K).
Maximum temperature is reached on the 6-8th.microsecond of the discharge,
and corresponds in brightness to the instant of maximum compression of the
f law. Dependence of the maximinn brightness temperature on electrotechnical
parameters of the discharge is shown in Fig. Sc. With increasing discharge
energy the brightness temperature of the zone of maximum compression of the
flow shows a tendency to stabilize. Comparison of this resul.t with the power-
law dependence of spectral brightness Bv av averaged over the lateral surface
(spectral density of radiation energy E~~, Bv, av"'Ev~ U3-4 [Ref. lOJ shows
an increase in effective radiating surface of the discharge with increasing
energy input.
The space-time distribution of brightness temperature of an open plasma-
dynamic discharge is shown in Fig. 6a, b. The region of plasma compression
with radiation brightness temperature greater than 20,000 K has average dimen-
sions of ~50 mm in the axial direction, and ~6 m4n in the radial direction.
Maximum brightness temperatures are reached on the axis of the flow at dis-
tances of 2~-30 mm from the tip of the magnetoplasma compressor.
102
FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500090006-1

APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00854R004500090006-1
FOR OFFICIAL USE ONLY
Tbr, eV
a b Tbr, eV
T = 8 us
J J -
- 6p0/
L=ZSMM
y ~ utl 2
iw , io
~ ' ~ / a0
~6
- ---1---' -1. y-.L_-1-~J
~ 40 d0 IZU L,MM Z~ 10 ~ 1~ Zd %,MM
Fig. 6. Distribution of brightness temperature a.long the
discharge axis at different times (a) and in different cross
sections of the jet at the instant of maximum current (b)
(C2Fy plasma, Wo = 9.4 kJ, Ua = 5 kV)
The spectral dependence of hrightness of radiation and brightness temperature
of th~ plasma in the zone of maximum flow compression was determined by the
method of photoelectric comparison of the emission intensity of the plasma
focus and fhe EV-45 reference source in the visible and negr-ultraviolet re-
gions of the spectrum [Ref. 27]. Measurements were made both on individual
wavelengths by using an FEU-39A photomultiplier crossed with a DMR-4 double
quartz monochromator as the radiation receiver, and over a fairly broad spec-
tral range ~hv ~ 1 eV by using photocells with optical filters: F1 photocell
with ZhS-3 filter, F1 with UFS-2 filter, and F7 with UFS-1. Both the photocells
and photomultiplier were equipped with cathode followers to preserve the neces-
sary time resolution (~T - 1 us).
JOU 650 SSf!
I
9 8 15U' I ~~i
~ ~ t ~
~ n _ 1
~ d~- -tt~' - - � 4 ~ � T ~ . ~
~ VS
y ~ J
- -~~t- F=75 F=275 2
o ~p i i
ll ~ 12 ~
~
to S8-2
Fig. 7. Diagram of optical measurements: 1--magnetoplasma
compressor; 2--EV-45; 3, 7, 10--diaphragms; 4, 5--achromatic
ob~ective lenses; 6- mirror; 8--DNIIt-4 monochromator; 9--
FEU-39A; 11--photocell; 12--aligrunent laser
A diagram of the optical measurements is shown on Fig. 7. An optical system
made up of quartz achromatic ob~ectives 4 and 5 pro3ects a triply magnified
ima~e of capillary 2(or of the central zone of the plasma focus) on diaphragm
7 or through rotating aluminized mirror 6 onto diaphragm 10.
Diaphragms 7 and 10 isolate the central part of the capillary and determine
- the size of the area on the surface of the plasma focus from which light is
incident on the photocell. Diaphragm 3 limits the solid angle in which raci~~-
tion is registered.
103
FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500090006-1

APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500090006-1
FOR OFFI('IAI. USE ONLY
Measurements on each wavelength or in a separate spectral interval were made
in two stages. On the first stage, a reference source was used to calibrate
the optical channel, the axis of the capillary being aligned by a reference
laser beam with coordinates corresponding to the coordinates of the zone of
maximum compression of the plasma flow. On the second stage, after setting
up the magnetoplasma compressor model and evacuating the dischar e chamber,
measuremznts were made. To reduce statistical scatter (QX = E(x - x) /n(n - 1))
the measurenents on each wavelength were repeated at least 5-6 times. The ad-
vantage of this method of ineasuring the spectral brightness of radiation is
that the ratio of the signals from the reference and the investigated sources
depends only on their brightness temperatures, while the influence of geometry
and transmission of the optical sys[em is completely eliminated. The absolute
accuracy of the given method of ineasuring brightness temperature is determined
mainly by the accuracy of calibrating the reference source, and is equal to
�7 in the visible region of the spectrum and �lOX in the ultraviolet [Ref. 27].
B~, W�cm'1�sr'1(c~n 1)-1
I / . r
~
! I S~ ~ ~ 3
1~,~:.~~ ~ _ '
9 ~ l1 ' 1 1 4
I z
!
S
C , , , ~
p Z 4 6 8 10 12 hv, eV
- ' ~
500 30U Z00 150 120 100 nm
Fig. 8. Spectral brightness of central zone of plasma focus
(L = 25 cm, C2Fy plasma, Wo= 9.4 kJ, Uo=S eV): 1--measure-
ments by photomultiplier; 2--measurements by photocells; 3--
photographic measurements; 4--theoretical emission spectrum;
5--ideal blackbody isotherm with T= 4.5 eV
Results of optical measurements are shown on Fig. 8. Photoelectric measurements
made by photocells and by a photomultiplier crossed with a monochromator give
similar values of spectral brightnesses, and agree well (~IOX) with the results
of photometric measurements of the brightness temperature on eff~ctive wave=
len~th of aef = 430 nm (hveg= 2.9 eV). The results show that as quantum energy
increases, the brightness temperature of plasma focus emis~ion drops off.
In the visible region, the brightness temperature is 3-3.5 eV, and in the
near ultraviolet ~2.3 eV. The reduction in brightness temperature is due
to spectral dependence of the coefficient of continuous absorption of the
plasma. To confirm this, and also to determine the nature of the observed
104
FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500090006-1

APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500090006-1
FOR OFFI('IAL USE ONLY
radiation spectrum, the authors calculated the coefficients of continuous .
abs~rption and the emission spectrum of a fluorocarbon plasma.
In the general case, calculation of the coefficient of absurption is an ex-
tremely difficult problem since it requires accounting for different kinds
of transitions, including free-bound and bound-bound transitions, whose calcu-
lation requires knowledge of exact wave functions of atom.s and ions. How~ever,
under conditions of high densities and temperatures (and it is just such con-
ditions that are of interest in developing high-intensity Light sources, and
that are realized in high-current radiating discharges), due to the strong
interaction of plasma particles with each other and with radiation it is suf-
ficient to consider only continuous emission either in the hydrogen-like ap-
proximation of atoms and ions [Ref. 4, 28), or with consideration of the indi-
vidual structure of ternis, by introducing additional correcting functions
or coefficients [Ref. 29].
Coefficients of decelerating and photoionization absorption were calculated
by the approximate theory of Ref. 29 based on the quantum defect method [Ref.
30, 31]. In this case, the photoionization crass sections for upper excited
states were determined by integrating the Burgess-Seaton equation [Ref. 32-
35], while calculation of the coefficients of decelerating absorption utilized
matrix elements obtained by extrapolating the matrix elements of bound-free
transitions with consideration of the specifics of free-free transitions to
the field of multiply charged ions [Ref. 36, 37].
The overall coefficient of bound-free absorption by z-tuple ions and free-
free absorption in the field of z�l-ions takes the form
~~.+t v-/,+~\v -+1 v"N:, ~1~
r..�. L' 7' eap ( 1~ ~ ( )
\
where C=16n2c-3/3~ch = const, UZ and UZ+1 is the statistical sum of the initial
and final ion, IZ is the potential in eV of the z-tuple ion, ~v is the displace-
ment of the photoienization potential due to interaction of particles in the
plasma in eV, NZ is the concentratior. of the z-tuple ion in cm 3, ~Z(v) is
a function that accounts for the specific structure of terms of the z-ion
(i. e. the deviation from the hydrogen-like approximation).
Formulas for calculating ~Z(v) are given in Ref. 29, 32. For most elements
the function ~Z(v) is weakly dependent on plasma temperature (or completely
independent) and has a value of the order of unity. Values of the function
F(v) for atoms and ions of carbon and fluorine are given in Ref. 38. Formula
(1) is valid for frequencies less than the cutoff frequency vg equal to the
maximinn threshold frequency of the photoelectric effect for excited states
that are accounted for integrally. For frequencies v~ vg the coefficient
of continu,~us absorption is calculated by the formula
. ~
x~.,.~.. x ~~'s, 7') ~ ~v) / 1 + xn ~ T ~2)
~ (~'.w) ~ I 1_I
where vn~l is the coefficient of photoionization absorption from a single
level with quantum nwnbers n and Z.
105
FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500090006-1

APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500090006-1
FOR OFF7CIAL USE ONLY
llsing the Saha formula, the ccefficient of continuous absorption can be written
as
cxp(v!T)
~3)
. x~.=;~ ~~�10''iN~~~:+i~:+t)Z c~p T- ~:ly) Ty~y~ �
\
For a plasma of complex composition that contains ions of various multiplicity,
the absorption coefficient takes the form
Xv.7' L X(v. ~~)R t"-~
A,'
~
(']Ci1~V~T~ -1 l~V:-1^A~s-1
- _ ~,'r."r ~~I - ~ c.~? r ~ (v) N,~=n-,
~ T
k,:
where k is the index of the chemical elem~nt, and swmnation is don~ with re-
spect to k and z. Quantities NZ, vg and ~(v) depend on k and z, while ~v and
AI are assumed to depend only on z.
The coeff icient of absorption with consideration of stimulated emission is
determined by the experession
x~r~'=x~r~ I-cxp(-v / T) J. ~5)
By using formulas for the coefficient of deceierating and photoionization
absorption, We can evaluate the concentration of the electronic component
in the zone of magnetohydrodynamic compression of the flow from the experi-
mentally measured density of continuous radiation and plasma temperature.
~ The magnitude of the f actor for deviation from the hydrogen-like approximation
~(v), as already pointed out above, differs insignif icantly from unityl; esti-
mates of the factor exp[(~vZ-1- ~IZ-1)/ T] ahow that over a wide range of
plasm~ parameters in the compression zone its value changes only slightly,
being ~1.1-1.25. Therefore in a first approximation these cofactors can be
disregarded in the formula f or the coeff~cient of absorption. Then the ex-
pression for the coeff icient of overall absorption with consideration of stimu-
lated emission can be represented as
, Y '1,45�10~-"N.= , ~ v ~ / N=.~ ykZ~
r.~' ~ ~ ` N~ !
k,:
v usp(v/T)-i N:.k
~ ( 1 ~ ~ ~-N' 1 ~~Z = z~
~ 7' / (~'I~~)'~ ~
h.:
where z is the effective atomic number of the ions. The quantity 'z was deter-
mined from calc;ilating the composition of the fluorocarbon plasma [Ref. 39].
1For ions of fluorine and carbon, ~(v) at v= 3 eV varies over a range
of 0.8-1.0 [Ref. 39].
106
' " FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500090006-1

APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00854R000540090006-1
F'OR OFFICIAL USE ONLY

~
~ ,
xY, a--- - --1--- xy,
~M-~ i ~ ~M-1
N~ Z�IOrs
~ - -
/0~ ~ - ! I
,b a
_ I 6 !0 ~a ~
r ' I - Ol - -
t0" - i - ~�~0'a
' ~ q01 - - _ _ - -
~o-s . : . _ 10,~
~ :
~ ~
~
' ' 0,00?
- 0,6 1 2 4 6 70 ZO
U 2 4 L' 8 T, eV N~ � 10"'~cn+-~
Tbr, eV H~=2� r0'~cM-~
5 ~ ~ is~ ~0's
~
4 1,2�101~
~0~~ Na � i0-;ecM�~
J d
40
~ 6 10~~ ~0
~I I
J� 10 IO I I
1 ~~Id I ~
~ ~--~--L- ' - l-
~ --~-----L--- ' ' - ~ ~ ~ 6 B ~T, eV
0 2 4 6 8 !0 T, eV "
Fig. 9. Coefficient of continuous absorption of a fluoro-
carbon plasma as a function of temperature (a) and electron
concentration (b); c--brightness temperature of a plane
layer of fluorocarbon plasma with thi.ckness of d= 0.5 cm;
d--region of plasma parameters that ensure a brightness
temperax.ure of the layer equal to 3 eV at thickness 8= 0.5 cm
Results of calculations of the coefficients of continuous absorption for quan-
_ tum energy hv = 3 eV in the temperature range of 1-10 eV and electron concen-
tration of 1018-2�1019 cm 3 are shown on Fig. 9a and b.
The spectral density of radiation intensity (or brightness) for a plane homo-
geneous plasma layer in the direction along the normal to its surface is
1;, =1~,i,(~ ~'~I~(- Y,�'l)l, (i)
where
'1,~.:~ ~ n,t~sSv:~
It,.,, ~,~~~(~�/7')- I ~~~p(v/7')- I' ~W~~cm2 �sr�cm-1) J
107
FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500090006-1

APPROVED FOR RELEASE: 2447/02/09: CIA-RDP82-44850R444544494446-1
FOR OFFICIAL USE ONLY
is tt~e equilibrium intensity of radiation (v and T are expressed in eV), Z is
the thickness of the layer. For cylindrical plasma formations, the equi~alent
- thickness of plane layer is determined by the formula Z= nD/4.
Fig. 9c shows the results of calculation of the brightness temperature of
emission of a plane layer of fluorocarbon plasma of thickness Z= 0.5 cm cor-
responding to the average thickness of the discharge in the compression zone,
where v= 3 eV at different plasma temperatures and electron concentrations.
The temperatures of the plasma focus experimentally observed on these frequen-
cies are ~3 eV and at the anticipated plasma temperatures in the crnnpression
~one (T < 10 eV) can be realized only at electron concentrations of greater
than 1019 cm 3. The region of plasma paraaneters Ne and Te that ensure a bright-
nes temperature of radiation greater than 3 eV at the given thickness of the
plasma formation (Z = Q.5 cm) and on rhe given frequency (v = 3 eV) lies above
the curve shown on Fig. 9d. Experimental da.ta on the plasma temperature in
the compression of a magnetoplasma compressor operating on a fluorocarbon
plasma are given in Ref. 7, 40. The electron temperature determined by the
method of self-inversion of spectral lines [Ref. 40] is Te = 3.5-5 eV, and
it is noted that with increasing energy input to the discharge this tempera-
ture is stabilized at a level of 4.5-5 eV due to intense radiation cooling
of the plasma. Measurement of ion temperature by Doppler broadening of lines
(Ref. 40] and by registration of Mach ref lection on a thin plate [Ref. 9]
as well as estimation of relaxation times have shown that maxwellization of
the plasma occurs in the zone of the focus along with equalization of the
electron-ion temperature. Existence of local thermodynamic equilibrium in
the vicinity of the plasma focus was established in Ref. 40.
Taking the plasma temperature in the zone of the focus as Te = 3.5-519V, w3
get (Fig. 9c) the eiectron concentration in the plasma: (1.2-2)�10 cm .
Let us note that this is a lower estimate since consideration was not taken
of the radial distribution of electron density and temperature in the focus
zone.
The emission spectrum of a fluorocarbon plasma in the quantum energy range
of ~v = 0-60 eV was calculated at typical parameters of the compression zone
Te = 4.5 eV and Ne = 1.5�1019 cm 3. The plasma composition was detFrmined from
solution of a system of Saha equations for sequential stages of ionization
together with equations of quasineutrality and constancy of composition [Ref.
39J. lteduction of ionization energy in the plasma was accounted for in accor-
dance with the Debye-Huckel theory [Ref. 42]. The values of statistical sums
and ionization energies necessary for the calculation were taken from Ref. 42,
and the missing data were calculated by the approximate method of Ref. 43.
Displacement of the photoionization threshold was accounted for by the ap-
~ proximate Ingliss-Teller theory [Ref. 44, 45]
eV.
tiii:~�to "(.;+~)~~".v,.
The photoionization cross sections for individually considered levels were
calculated by the Kramers formula [Ref. 28)
n `a
~ (1Q)
7,;1� 111-is -
nn.='
-1~ \
ZUO
FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500090006-1

APPROVED FOR RELEASE: 2447/02/09: CIA-RDP82-44850R444544494446-1
FOR OFFICIAL USE ONLY
where n~ is the principal quantwn number of the level, vn is the minimum energy
of a quantum capable of ionizing an atom or ion that is in the n-excited state.
The coefficient of absorption ir~ determined by the expression
(ti)
xr.n ~ 7.:.n ~I:.n ~Y).
1....~
Here ay~n is the reJ.ative c~ncentration of ions of multiplicity z on level
with number n determined in the case of boltzmannian distribution by the formula
1:--/,,,r \ 1 (I'l)
. ~e:.~~ -oxp ( I
. U, ('1', ,\I: ) \
where gn~Z, In,z are the statistical weight and ionization energy ~f a z-tuple
ion from level n, UZ(T, DIZ) is the statistical sum of the ion.
xy, cna-~ a
Fig. 10. Coefficients of con-
x~ _ , tinuous absorption (a), theo-
~ CIII ~ i'~ 1~ retical (b) and measured (by
ionization chambers (c) and
~ ~ . IF~ photoemission-scintillation
)0"~ w~`r~ . CII
'\Fj~~. ~ � spectrometer (d)) emission spec-
,
~ ~ tra of f luorocarbon plasma
10-t ~ ~'1'.
B pX = 80 . 8
15 ~ ~ 4191. 4, 44
i ' T-aSeV b
!0 ~Tbr = 2 . 4 eV \ - .
Z,~Z 15 Ne \
i s ~RvropM e Ar , He
6 ~
u ~
~ 1J5 .
3 N ~ 27
~ ~ ~ ~ ~ t
~ F11,pstP~G112ptP~F112p~~P CI/!Zs= ~ti FIIf2p~ ~`S~
ai 4 - Ne ~
f- -
2 fe~
i~ Ha l
0 _.~-1 . L- ' ' -
10 ?.0 JO 40 SO 60 h1; eV
~.rel. unit5
~ J d .
~ ~
, ~
.
- 1 ~
' 1--- ~ ~ -
0 )U '1.0 3U 4U SO ~ 60 1~y eV
109
FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500090006-1

APPR~VED F~R RELEASE: 2007/02/09: CIA-RDP82-00850R000500090006-1
F'OR OFFICIAL USE ONLY
The results of calculation of coefficients of continuous absorption and the
resultant radiation spectrum of a fluorocarbon plasma are shawn on Fig. 10a.
The partial fraction of ions of different multiplicity in the overall coef-
ficient of continuous absorption varies depending on the spectral range being
considered. In the infrared, visible and near-ultraviolet regions of the
spectr~un the maximum contribution is from singly charged ions of fluorine
and carbon (-60 and -36X respectively). In the far ultraviolet, an appreciab?e
role is played by double ions. The radiation spectrum of a fluorocarbon piasma
with characteristic dimension of the plasma formation Z= 0.5 cm is considerably
different from planckian at a temperature equal to that of the plasma, which
is due to optical transparency of the plasma in the region of quantum energies
2. 7< v< 31 eV (optical thickness TV = KuZ ~ 1~ .
The theoretical radiation spectr~n is compared with the results of optical
measurements in the visible and near-ultraviolet region on Fig. 8. Values
~ of spectral brightness of the plasma focus uieasured both on individual wave-
lengths and in wide spectral intervals agree fairly well (about 20X) with
calculated radiative characteristics. The reduction in brightness temperature
of emission of the focus with transition from the visible to the near-ultra-
violet region is due to reduction in the optical thickness of the plasma.
In the region of quantum energies v- T, continuous radiation of the plas~~a
is due to decelerating and photorecombination mechanisms. The contribution
of decelerating processes for the given plasma parameters in the resultant
radiation spectrum is shown on Fig. lOb. Even in the near-ultraviolet region
of the spectrum the recombination mechanisms predominate, and the plasma radi-
ation spectrum is determined mainly by photorecombination of electrons to
the upper excited states of singly ionized ions.
In the quantum energy region of 5< v< 10 eV, the distribution of radiation
is measured by a photographic method using the DFS-29 parallel-incidence vacuum
spectrograph. To eliminate the influence of self-sensitivity of the photo-
graphic material to hard photons, the XUV spectra of plasma focus radiation
were registed on type ML-2 photographic film sen~itized with sodium salicylate,
which has a constant quantum yield (in a range of 10-lSy) in the spectral
region of 340-40 nm [Ref. 46, 47]. The sensitivity of the spectrograph-
photomaterial syst.em in relative units was determined by a pulsed continuous-
spectrum source made in exact accordance with the recommendations of Ref.
48 and calibrated with respect to brightness in the extreme ultraviolet. The
reference source is based on a pulsed electric discharge through a capillary
in vacutm?. Inthis case, the source of the continuous spectrum in the XW
is a Lyman continuum produced by radiation of a high-pressure plasma formed
from the material of the capil~ary walls (polyformaldehyde) when an intense
discharge current pulse passes through the capillary. Upon satisfaction of
conditions with respect to current (Im X= 10 kA, TPulse"'S us) and capillary
geometry (diameter d~ = 3.5 mm, length = 30 mm), according to the results
of Ref. 48 the central zone of the capillary 2.5 mm in diameter is homogeneous
(with accuracy of 10%) as a light source that emits in the XW region of the
spectrum up to a= 120 man as an ideal blackbody with temperature T= 37,000 t
3,000 K.2
2Using the EV-39 standard, we predetermined the brightness temperature of
[he capillary in the region a= 250 � 20 nm: T~= 37,000 � 10% K, which agrees
well with data of Ref. 48.
110
FOR OFFICIAL USE ON!.Y
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500090006-1

APPROVED FOR RELEASE: 2007/02/49: CIA-RDP82-00850R040500090006-1
FOR OFFICIAL USE ONLY
In determining the relative spectral distribution of plasma focus emission
in ehe XUV from the known (as a result of calibrating the reference source)
sensitivity of the optical system of registration, deviation from reciprocity
law was disregarded since the difference in the duration of the emission pulse
of the light sources being compaxed is small. For example, the calibrate3
source had an effective duration of continuous radiation of ~5 �s, while the
investigated discharges had T~ 15-20 �s. It is known ~Ref. 49-51] that in
this exposure ti~e range the iso-opaque line of the photographic material
has a constant section, and deviations from reciprocity law are practically
unobserved.
Absolute calibration of spectral distribution was done witti respect to the
emission brightness of the central zone of the plasma f~cua in the region
~v = 4.35-5.0 eV registered by photoelectric methods. Fig. 11 [photo not re-
producedJ shows typical XUV spectrograms of the plasma focus for various ex-
perimental condi~ions,3 and Fig. 8 shows the distribution of emission brightness
in the near vacuwn ultraviolet obtained by spectral measurements for a plasma-
dynamic discharge on a fluorocarbon plasma. Tt~e spectra of the plasma focus
in the XW region consist of intense cw emission and radiation lines that
belong mainl~ to single, double or triple ions of the mate�rial of the dielec-
tric washer of the magnetoplasma compressor; lines of elements of the electrode
materials are weakly represented in the resultant spectra. When processing
the densitometric plots of the spectra, consideration was taken of only the
continuous radiation of the plasma since the contribution to the total emitted
energy by lines in region a> 120 nm is small. The experimental spectrum in
! the quantum energy range ~v = 5-10 eV obtained by the photographic method cor-
~ relates well with the calculated spectrum; the brightness temperature of the
central zone of the plasma focus in region a= 150 nm is ~27,000 K. The drop
in intensity of radiation in the short wave region is due to the reduction
in~optical density of the plasma, which is also confirmed by the increase
, in the relative fraction of line emission in the overall discharge emission
spectrum.
The spectral density of radiation intensity of the plasma focus in the extreme
ultraviolet region was measured by open double ionization chambers filled
with spectrally pure inert gases [Ref. 22]. The region of spectral sensitivity
of the ionization chamber is determined on the long-wave boundary by the ioni-
zation potential, and on the short-wave boundary by a fa11-off in the photo-
ionization cross section of the filler gas. By using inert gases, the spectral
region from 12 to 65 eV was broken down into four partly overlapping inter-
vals: 12.1-22 eV (Xe), 15.8-28 eV (Ar), 21.6-40 eV (Ne) and 24.6-65 eV (He).
Results of ineasurements of the spectral brightness of the central zone of
the plasma focus in the extreme ultraviolet are shown on Fig. lOc. The ex-
perimental conditions are similar to those used for optical measurements in
the visible and near-ultraviolet regions of the spectrum (Fig. 8).
To compare experimental results with the calculated emission spectrum of a
fluorocarbon plasma, the authors calculated the average values of brightness
3The emission spectra of discharges in the XW region were taken by
S. G. Shashkovskiy and A. G. Opekan.
111
FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500090006-1

APPROVED FOR RELEASE: 2007/02109: CIA-RDP82-00850R040500090006-1
FOR OFFI('IA1, l1SE ONI.Y
- in spectral intervals corresponding to regions of sensitivity of the chamber
with different fillers
t
= f 11,. d~�.
~'=-v, J
ValueS of brightness averaged over the calculated spectrum are shown on Fig. lOb.
Comparison of the calculated emission spectrum of the fluorocarbon plasma (Fig.
lOb) with that measured by ionization r.hambers (Fig. lOc) shows fairly good
qualitative agreement between calculation and experiment.
In contrast to measurements in the visible and near-ultraviolet regions, the
experimental values of spectral brightnesses in the extreme ultraviolet were
lower than the calculated levels. This is due to the difference in spatial
resolution of the ionization chamber used (~x ~ 6 mm) from that of the arrange-
ment for optical measurements in the visible and near-ultraviolet regions
(~x = 1.5 mm), resulting in averaging of brightness in the extreme ultraviolet
over a larger area of the radiating surface of :::e plasma focus.
Measurements by ionization chambers showed an increase in the brightness tem-
perature in the short-wave region of the spectrum, which may be evidence of
the presence of recombination maxima in the emission spectrum of the plasma
focus. However, the width of regions of spectral sensitivity of the chamber
does not permit registration of the position of individual emission maxima,
or determination of the specif ic form of spectral distribhoto missions$ion
energy in the extreme ultraviolet. For this purpose, a p
scintillation spectrometer was used to measure the relative distribution of
radiation of the plasma focus in the range of quantum energies from 10 to
350 eV. The method used was based on registration and analysis of the spec-
trum of photoelectrons knocked out of the photocathode by direct radiation
of the plasma [Ref. 52, 53]. The method of ineasurements and theectralidistri-
facility are described in detail in Ref. 54. The form of the sp
bution of radiation obtained by the photoemission-scintillation spectrometer
is shown on Fig. lOd. The nature of the measured spectrum, the amplitude
of maxima and their shape (with consideration of the spectral resolution of
the instrument ~v"'Sto~the~distributionrof emissiontbrightnesstmeasur dbbyaioon
continuum, and also
nization chambers.
A maximum of focus emission in the region of 30-35 eV is due mainly to photo-
recombination to the ground state 3P and to the lower excited state 1D of
- the first fluorine ion FII; a second maximwn in the reg~ou �of20oweremixed
due to recombination to the ground level 2P� and to a g p
terms ''P, 2D, 2S of the carbon ion CII. Recombination maxima of the second
ions CIII and in the region v~ 45 and 60 eV did not show up in the measured
spectrum which may be due to the appreciable optical thickness of the plasma
in this range of quantum energies. The increase in optical density of
the plasma in the short-wave region of the spectrum is also associated with
the increase in brightness temperature registered by the ionization chambers
in the absorption bands of neon (Tbr - 3.9 eV) and helium (Tb= ~ 4.2 eV). A
112
FOR OFFiCIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500090006-1

APPROVED FOR RELEASE: 2047/02/09: CIA-RDP82-00850R000504090006-1
FOR OFFICIAL USE ONLY
distinguishing feature of the emission spectrum of the plasma focus is the
presence of a hard radiation component in the region of 100-350 eV ("super-
planckian" for Te ~ 5 eV) registered in measurements by spectrometer [Ref. 54]
and bismuth bolometers [Ref. 7]. The nature of this radiatior~ is possibly asso-
ciated with microinhomogeneities that are typical of the initial unsteady
stage of discharge in the magnetoplasma compressor.
Thus these studies show that radiation of the plasma focu~ is due mainly to
the recombination continuum of characteristic groups of icns that determine
the properties of the plasma, the percentage of XUV emissien in the overall
spectrum of the plasma-dynamic discharge being ~70-90X as evaluated from the
results of integrated measurements by a quartz calorimetei~ and bolometers
[Ref. 21, 55]. By appropriately altering the chemical anc ionizational compo-
sitions and parameters of the plasma in the zot.~= of M~ID compression, we csn
realize fairly effective control of the radiation spectrum of a plasma focus,
which enables creation of selective sources of radiation in the far ultra-
- violet.
The spectral distribution of plasma-dynamic discharge radiation energy may
be profoundly influenced by the macroscopic structure of rhe plasma flow ema-
nating f rom the magnetoplasma compressor. This flow structure depends on
the configuration of the discharge gap, the energy conditions of operation
and the conditions of introducing the plasma-forming substance into the dis-
charge. The structural influence shows up in varying degrees of shielding
of the short-wave radiation of the plasma focus by periphe~ral layers of cool
erosion products that are not captured by the magnetic field of the discharge
current. Depending on the chemical cor~position and particle density in the
peripheral annular layer rhat surrounds the hot zone of the discharge, the
shielding may be in the nature of either filtration of short-wave radiation
in individual spectral intervals that as a rule are fa~rly narrow, or in lines,
or else an abrupt reduction (by a factor of several ten:o) in the magnitude of
emission intensity of the plasma focus, beginning at son.e cutoff wavelength,
with a simultaneous increase in the luminous output of the discharge in the
visible and near-ultraviolet regions of the ~,~ectrum. The latter is due to
rapid heating, ionization and radiation of the products of erosion as a result
of their broad-band absorption of high-energy XUJ radiation of the plasma
focus. Since the radiating surface of the peripheral layE~r is larger than
the surface of the focus, the balance of the radiation and the absorption
in it sets in at a lower temperature, i. e. conditions are realized for trans-
formation of short-wave radiation to emission in the visible and near-UV bands.
The XUV spectrograms of the plasma focus shown on Fig. 11 [photo not reproduced]
that were obtained under different experimental conditions, illustrate the
- influence that macrostructure of the plasma-dynamic discharge has on spectral
distribution of rad:ation energy in the short-wave region. Microphotometric
analysis of spectra shows that discharges on fluorocarbon and polyformaldehyde
plasmas (Fig. llb and c[photos not reproduced]) characterized by approximately
the same coefficient of utilization of material (r~,s ~ 0.94) have similar
emission characteristics in the XW region, but differ consider.ably with re-
spect to the nature of spectral distribution of radiation from discharge in
CsI vapor (Fig. lld), in which the coefficient of utilization of working
113
~ FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500090006-1

APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500090006-1
FOR OFFICIAL USE ONLY
substance for the given magnetoplasma compressor model and the selected energy
conditions are considerably lower than in discharges on (CH2O)n and (C2F4)n
plasmas (Fig. llb, c).
According to spectroscopic measurements in the region of quantum energies
~v = 4-8 eV, the CsI discharge is characterized by higher (by ~4-5 times)
energy yield of radiation than discharges on fluorocarbon and polyformaldehyde
plasmas; however, in the regic.n of v> 8 eV (a < 150 nm) the intensity of continu-
ous emission of the cesium iodide plasma is insufficient for exposure of thz
photographic emulsion during a single discharge puls~. These data agree well
with the results of photoelectric measurements of the tntegrated energy yield
of radiation of open discharges in the near ultraviolet region (~v = 4.35-5 eV)
[Ref. 10] and also with measurements in the range of ~v = 12-45 eV made by
ionization chambers, according to which the spectral brightness of the CSI
discharge decreases in the indicated region by 30-50 times compared with the
- discharge on fluoroplastic [Ref. 22].
Thus Lhe assurance of high coefficients cf utilization of working substance
is a necessary condition for eliminating effects of shieldizg short-wave radia-
tion and attainment of high brightness temperatures of the plasma focus in
the XUV region of the spectrum. When this condition is met, the brightness
temperature of the discharge is limited from above only by the true temperature
of the plasma in ~he zone of the focus, which can be increased by increasing
the degree of magnetohydrodynamic compression of the plasma flow as the density
of electromagnetic energy in the interelectrode gap and the zone of entrained
currents of the magnetoplasma compressor is increased. Thus, as a result
of optimizing the geometry of the discharge gap in the magnetoplasma compressor,
the energy conditions of operation and the conditions of introducing the plasma-
forming substance in the discharge, it has been possible for the first time
to reach brightness temperatures of radiation exceeding 60,000 K[Ref. 56]
in the spectral region beyond the helium ionization potential (~v = 24.6-65 eV).
Let .,s also note that such flux densities in the far XUV region have so far
been reached only on facilities of synchrotron type [Ref. 41]. Fig. lla shows
the emission spectrum of an optimized plasma-dynamic discharge. The steep
drop in the intensity of continuous radiation on the short-wave boundary
(a< 90 tun) recorded on the spectrogram is apparently due to an abrupt increase
in the coefficient of reflection of the alinn~.num coating of the diffraction
grating of the spectrograph [Ref. 46].
Let us point out one more possibility for controlling the spectral properties
of open plasma-dynamic discharges, which consists in producing plasma layers
of a given chemical composition surrounding the zone of A4ID compression (in
the general case, a chemical composition differing from that of the radiating
plasma of the focus), and with given density, and hence having certain prede-
termined optical characteristics. These layers may act as plasma filters
that have "transparency windows" in individual (defined for the specific prob-
lem) spectral inr_ervals, or totally blocking the hard radiation of the plasma
focus beginning at some quantum energy that is undesirable or even harmful
for the selected application. Technically, the realization of such layers
may be accomplished by using ablating separative sleeves in the magneto;~lasma
compressor with chemical composition profiled along the radius in the aone
of ti~e deflagration current layer.
114
FOR OFF[CIAL USE ONl.Y
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500090006-1

APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500090006-1
FOR OFFICIAL USE ONLY
The authors thank P. A. Ovchinnikov and A. G. Opekan for assisting with the
research.
REFERENCES

~ 1. Borovich, B. L., "Feasibility of Making Optically Pumped Gas Lasers Using
Allowed Electronic Transitions of Molecules", ZHURNAL ERSPERIMENTAL'NOY
I TEORETICHESKOY FIZIKI, Vol 61, 1971, p 2293.
2. Mikheyev, L. D., "Gas Lasers With Wide-Band Optical Pumping", KVANTOVAYA
ELEKTRONIKA, Vol 5, 1978, p 1189.
3. Rozanov, V. B., "Feasibility of Producing Inverse Medium by Photoionizing
Inner-She1l.Eiectrons in Atoms", PIS'MA V ZHURNAL EKSPERIMENTAL'NOY I
TEORETICHESKOY FIZIKI, `Jol 12, 1970, p 486.
4. Aleksandrov, A. F., Rukhadze, A. A., "Fizika sil'notochnykh
elektrorazryadnykh istochnikov sveta" [Physics of Intense Electric-Discharge
Light Sources], Moscow, Atomizdat, 1976.
5. Pukhov, A. M., "Exploding Wire Vapor Shielding of Pulse Discharge XW
Radiation", ZHURNAL PRIKLADNOY SPEKTROSKOPII, Vol 22, 1975, p 922.
6. Zvorykin, V. D., Klementov, A. D., Rozanov, V. B., "Characteristics of
Intense Discharge in Neon at High Pressure", KVANTOVAYA ELEKTRONIKA, Vol 4,
1973, p 43.
~ 7. Kozlov, N. P., Leskov, L. V., Protasov, Yu. S., Khvesyuk, V. I., "Experi-
mental Investigation of Plasma Focus in Erosion Plasma Accelerators: I",
ZHURNAL TEKHNICHESKOY FIZIKI, Vol 43, 1973, p 730; "II: Energy Character-
istics of Dense Plasma Focus", ZHURNAL TEKHNICHESKOY FIZIKI, Vol 44, 1974,
p 2519.
8. Kamrukov, A. S., Kashnikov, G. N., Kozlov, N. P. et al., "Feasibility
of Making High-Brightness Sources in the Far Ultraviolet Based on Hyper-
sonic Flows of Dense Plasma", PIS'MA V ZHURNAL TEKHNICHESKOY FIZIKI, Vol 2,
1976, p 447.
9. Kozlov, N. P., Leskov, L. V., Protasov, Yu. S. et al., "Measurement of
Mach Number in Plasma Jets", TEPLOFIZIKA VYSOKIKH TEMPERATUR, Vol 12,
1974, p 697.
10. Kamrukov, A. S., Kozlov, N. P., Malashchenko, V. A., Frotasov, Yu. S.,
"Experimental Investigation of Plasma Focus in Erosion Plasma Accelerators:
III. Radiation Properties of Dense Plasma Focus", ZHURNAL TEKHNICHESKOY
FIZIKI, Vol 47, 1977, p 1673.
11. Kvartskhava, I. F., Meladze, R. F., Suladze, K. V., "Experiments on Elec-
trod~~namic Plasma Acceleration", ZHURNAL TEKHNICHESKOY FIZIKI, Vol 30,
. 196~, p 289.
115
FOR OF~'ICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500090006-1

APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00854R004500090006-1
FOR OFFICIAL USE ONLY
12. Skvortsov, Yu. V., Komel'kov, V. S., Tserevitinov, S. S., "Structure of
Magn~tic Fields in Plasma Jet With Self-Currents", ZHURNAL TEKHNICHESROY
FIZIKI, Vol 34, 1964, p 965.
13. Grigor'yev, V. N., Plasmoid Structure in Electrodynamic Accelerator",
~ZHURNAL PRIKLADNOY ,MEKHANIRI I TEKHNICHESROY FIZIKI, Vol 2, 1965, p 35;
"'Pinches' Observed in Rail Plasma Accelerators , ZHURNAL PRIKLADNOY
MEKHANIKI I TEKHNICHESKOY FIZIKI, Vol 4, 1965, p 146.
14. Osadin, V. A., "Energy Release in Intense Plasma Discharge", ZHURNAL
TEKHNICHESKOY FIZIKI, Vol 35, 1965, p 1230; "Problem of Clot Formation
in Pulsed Plasma Accelerators", ZHURNAL TEKHNZCHESKOY FIZIKI, Vol 35,
1965, p 1327.
15. Derevshchikov, V. A., "Measuring Velocities of Plasma Components by High-
Speed Photography in ~:onochromatic Light", ZHURNAL TEKHNICHESKOY FIZIKI,
Vol 37, 1967, p 315.
16. Rushaylo, A. M., Flow Inhomogeneities in Pulsed Electromagnetic Accel-
erator", ZHURNAL PRIKLADNOY 1~KHANIKI I TERHNICHESKOY FIZIKI, Vol 4,
1970, p 165.
17. Min'ko, L. Ya., "Polucheniye i issledovaniye impul'snykh plazmennykh
potokov" [Producing and Studying Pulsed Plasma Flows], Minsk, Nauka i
tekhnika, 1970.
18. Gubarev, V. Ya., Kozlov, N. P., Leskov, L. V. et al., "Experimental Deter-
mination of Velocity Characteristics of Pulsed Erosion Accelerator" in:
"Plazmennyye uskoriteli" [Plasma Accelerators], Moscow, Mashinostroyeniye,
1973.
19. Aleksandrov, A. F., Zosimov, V. V., Rukhadze, A. A. et al., "Study of
Equilibrium and Stability of Self-Compressed Discharges in Optically Dense
Plasma", ZHURNAL EKSPERIMENTAL'NOY I TEORETICHESKOY FIZIKI, Vol 64, 1973,
p 1568.
20. Kovrov, P. Ye., Morozov, A. I., Structure of Compression Region in Magneto-
plasma Compressor", ZHURNAL TERHNICHESKOY FIZIKI, Vol 46, 1976, p 2508.
21. Zvorykin V. D., Kashnikov, G. N., Klementov, A. D. et al., "Radia+tion
of Plasma Focus of Magnetoplasma Compressor in Visible and Ultraviolet
Regions of the Spectrum", KVANTOVAYA ELEKTRONIKA, Vol 2, 1975, p 2416.
22. Zvorykin, V. D., Kamrukov, A. S., Klementov, A. D. et al., "Investigation
of Plasma Focus Emission in the XW Region by Ionization Chambers",
KVANTOVAYA ELEKTRONIKA, Vol 4, 1977, p 290�
23. Zatsepin, Yu. A., Popov, Ye. G., Tsikulin, M. A., "Brightness of Shock
Wave Front in Some Gases", ZHURNAL EKSPERIMENTAL'NOY I TEORETICHESKOY
FIZIKI, Vol 54, 1968, p 112.
24. Tsikulin, M. A., Popov, Ye. G., "Izluchatel'nyye svoystva udarnykh voln
v gazakh" [Radiative Properties of Shock Waves in Gases], Moscow, Nauka,
1977.
116
FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500090006-1

APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-04850R000500090006-1
FOR OFFICIAL USE ONLY
25. Model', I. Sh., "Measuring High Temperatures in Intense Shock Waves in
Gases", ZflURNAL IICSPERIMENTAL'NOY I TEORETICHESROY FIZIKI, Vol 32, 1957,
p 714.
26. Dubovik, A. S., "Fotograficheskaya registratsiya bystroprotekayushchikh
protsessov" [Photographic Recording of Aigh-Speed Processes], Moscow,
Nauka, 1967.
27. Basov, N. G., Borovich, B. L., Zuyev, V. S. et al., "Intense Discharge
in Gases: I. Experimental Study of Luminosity and Energy Characteristics
of Powerful Discharge in Air", ZHURNAL TEKIiN1CHESKOY FIZIRI, Vol 40, 1970,
p 516.
28. Zel'dovidh, Ya. B., Rayzer, Yu. P., "Fizika udarnykh voln i
vysokotemperaturnykh gidrodinamicheskikh yavleniy" [Physics of Shock Waves
and High-Temperature Hydrodynamic Phenomena], Moscow, Nauka, 1966.
29. Biberman, L. M., Norman, G. E., "Continuous Spectra of Atomic Gases and
Plasma", USPEKHI FIZICHESKIKH NAUK, Vol 91, 1967, p 193.
30. Seaton, M. J., "Method of Quantum Defect", MON. NOT. ROY. ASTRON. SOC.,
Vol 118, 1958, p 504.
31. Norman, G. E., "Substantiating the Quantum Defect Method", OPTIKA I
SPEKTROSKOPIYA, Vol 12, 1962, p 333.
32. Burgess, A., Seaton, M. J., "Cross Sections for Photoionization From
Valence Electron States", REV. MOD. PHYS., Vol 30, 1958, p 992.
33. Biberman, L. M., Norn?a.:, G. E., "Plasma Recombination Emission and Brems-
strahlung", J. QUANT. SPECTROSC. RAD. TRANSFER, Vol 3, 1963, p 221.
34. Biberman, L. M., Norman, G. E., "Calculating Photoionization Absorption",
OPTIKA I SPEKTROSKOPIYA, Vol 8, 1960, p 433.
35. Biberman, L. M., Norman, G. E., U1'yanov, K. N., "Photoionization of
Excited Many-Electron Atoms and Ions", ASTRONOMICHESKIY ZHURNAL, Vol 39,
1962, p 107.
36. Norman, G. E., "Free-Free Electron Transitions in Ion Field", OPTIKA I
SPEKTROSKOPIYA, Vol 14, 1963, p 521.
37. 13orman, G. E., "Photoionization Cross Section: oF Lower Excited States,
and Oscillator Strengths of Some Lines of Carbon a*:d Nitrogen Atoms",
OPTIKA I SPIICTROSKOPIYA, Vol 14, No 5, 1963, p 593.
38. Bondar', V. A., Kiselevskiy, L. I., Truk,~an, Ye. P., "Using Pulse Discharge
With Restricted Channel Cross Section in :::-^erimental Spectroscopy",
ZHURNAL PRIKLADNOY SPEKTROSKOPII, Vol 9, 1968, r 73?.; Vol 9, 1968, p 928.
39. Ablekov, V. K., Kashnikov, G. N., Kozlov, N. P. et ~1., "Raschet sostava
plotnykh mno4okomponentnykh plazm" [Calculating Composirion of Dense
117
FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500090006-1

APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00854R000540090006-1
FOR OFFICIAL USE ONLY
Multicamponent Plasmas] in: "Teplofizicheskiye svoystva nizkotemperaturnoy
plazmy" [Thermophysical Properties of Low-Temperature Plasma], Moscow,
Nauka, 1976.
40. Kozlov, N. P., Malashchenko, V. A., Protasov, Yu. S., "Interferometry
of Dense Plasma Flows", ZHURNAL PRIRLADNOY SPIICTROSROPII, Vol 22, 1975,
p 210.
41. "Sinlchrotronnoye izlucheniye" [Synchrotron Radiation], TRUDY FIZICHESKOGO
INSTITUTA IMEriI P. N. LEBEDEVA, Mascow, Nauka, Vol 80, 1980.
42. Darwin, H. W., Felenbok, P., "Data for Plasmas in LTE", Paris, 1965.
43. Traving, G., ~achec, B., Holveger, fl., "Zur Berechnung von zugtand S~men",
Abhandl. Aamburger Sterrnaarte, 1966.
44. Ingliss, D. R., Teller, E., "Ionic Depression of Series Limits in One-
Electron Spectra", ASTROPHYS. J., Vol 90, 1939, p 439.
45. Grim, G., "Spektroskopiya plazmy" [Plasma Spectroscopy], Moscow, Atamizdat,
1969.
46. Zaydel', A. N., Shreyder, Ye. Ya., "Spektroskopiya vakuumnogo ul'tra-
fioleta" [Extreme Ultraviolet Spectroscopy], Moscow, Nauka, 1967.
47. Samson, J. A. R., "Techniques of Vacuum Ultraviolet Spectro~tcopy", New
York, London, 1967.
48. Podmoshenskiy, I. V., Pukhov, A. M., Yakovleva, A. V., "Pulse Continuous-
Spectrum Sources Calibrated With Respect to Brightness in the Extreme
Ultraviolet", ZHURNAL PRIKLADNOY SPEKTROSKOPII, Vol 16, 1972, p 415.
49. Kartuzhanskiy, A. L., "Violation of Photochemical Reciprocity Law for
Photographic Layers", USPEKHI FIZICHESKIKH NAUK, Vol 51, 1953, p 161.
50. Ryabtsev, A. N., Sukhodrev, N. K., "Investigation of Deviations From
Reciprocity Law in Photographic Layers in the Extreme Ultraviolet Region
of the Spectrtan", ZHURNAL NAUCHNOY I PRIKLADNOY FOTOGRAFTI I
KINEMATOGRAFII, Vol 15, 1970, p 167.
51. Zatsepin, Yu. A., "Investi~ation of Deviations From Reciprocity Law at
Exposure Times of 10-5-10 s" ZHURNAL NAUCHNOY I PRIKLADNOY FOTOGRAFII
I KINEMATOGRAFII, Vol 5, 1960, p 60.
52. Vekhov, A. A., Nikolayev, F. A., Rozanov, V. B., "Photoemission-Scintillation
Spectrometer for Studying Plasma Emission in the Extreme Ultraviolet",
PRIBORY I TEKHNIKA EKSPERIMENTA, Vol 6, 1974, p 198.
53. Vekhov, A. A., Nikolayev, F. A., Rozanov, V. B., "Investigating Radiation
of Intense Pulsed Discharges in Metal Vapor in the Vacuum Ultraviolet",
PIS'MA V ZHURNAL EKSPERIMENTAL'NOY I TEORETICHESKOY FIZIKI, Vot 17, 1973,
p 750.
118
FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500090006-1

APPR~VED F~R RELEASE: 2007/02/09: CIA-RDP82-00850R000500090006-1
FOR OFFICIAL USE ONLY
54. Vekhov, A. A., Ramrukov, A. S., Kozlov, N. P. et al., "Emission Spectrwn
of Plasma Focus in 10-350 eV Region", ZHURNAL TEKEINICHESKOY FIZIKI, Vol 3,
1977, p 661.
55. Kozlov, N. P., Protasov, Yu. S., "Radiation Properties of Dense Plasma
Focua", TEPLOFIZIRA VYSOKIKH TF�MPERATUR, Vol 10, 1972, p 1319.
56. Kamrukov, A. S., Kozlov, N. P., Protasov, Yu. S., "Emis~ion Spectrum of
Plasma Focus in Quantum Energy Region of 0.64-350 eV", AOKLADY AKADEMII
NAUK SSSR, Vol 237, 1977, p 1334.
~ COPYRIGHT: Izdatel'stvo "Nauka", "Teplofizika vysokikh temperatur", 1982
6610
CSO: 1862/183 - ~ -
119
FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500090006-1