ProoreJ~ in Nuclear £neroy, Vol. I, pp. 27 to 39. Pergamon Press 1977. Printed in G r u t Britain
T H E S E P A R A T I O N N O Z Z L E PROCESS F O R E N R I C H M E N T O F URANIUM-235 E. W. BECKER
Institut ffir Kernverfahrenstechnik, Karlsruhe Nuclear Research Center,* Federal Republic of Germany (Receit'ed I0 September 1976)
1. INTRODUCTION The separation nozzle process for the enrichment of U 2as has been developed at the Karlsruhe Nuclear Research Center as an alternative to the gaseous diffusion and centrifuge processes l-s. Isotope separation is achieved by the same basic mechanism as in the centrifuge method. However, the serious mechanical problems of highly stressed rotating machines are avoided, since, in the nozzle method, the centrifugal separating forces are generated by the deflection of a high speed jet of uranium hexafluoride and a light auxiliary gas. Since 1970, the German company STEAG, has been involved in the technological development and commercial implementation of the nozzle process 6. In 1975, the Brazilian company, NUCLEBRAS, and the German company, I N T E R A T O M , joined the effort. The primary objective of the common activity is the construction of a separation nozzle demonstration plant with an annual capacity of about 200 t SWU. This paper covers the basic features and the most important steps in the technological development of the new process.
end of the deflection the flow is split into a lighter and a heavier fraction by means of a skimmer. In Fig. 2 the elementary isotope separation effect ~A
nl
(1 - n,)
=
n, (1 -- hi)
1,
(nl, n, = mole fractions of U zas in the light and heavy fractions of UF6) is plotted versus the UFe-cut flow rate o f UF6 in the light fraction vu = flow rate o f UF6 in the feed gas Within the range covered the measured separation effect rises continuously, as the cut becomes smaller. For a cut of 5 % an elementary isotope separation effect of 4.2 % is reached. Besides the generation of high flow velocity of UF6 the light auxiliary gas makes another important contribution, which can be explained by the lower curve in Fig. 2. The lower curve represents the theoretical upper limit of the so-called equilibrium separation effect ~* given by Am
In v U
m "UU -- ] '
2. BASIC FEATURES OF THE SEPARATION NOZZLE METHOD The basic features of the separation nozzle method can be understood from Fig. 1, which shows a cross-section of the nozzle system used in the commercial implementation of the process. A jet of gaseous uranium hexafluoride mixed with hydrogen expands along a curved fixed wall. The hydrogen, which is present in a large molar excess, increases the flow velocity of the UF6 and, hence, adds to the centrifugal force determining the separation. At the
(Am/m = relative difference in masses of the isotopic molecules.) The fact that the 6,t values measured considerably exceed those calculated for equilibrium conditions can be explained by the fact that the light auxiliary gas. causes the two isotopes to approach their equilibrium distributions at slightly different rates.
* Equilibrium separation characterizes the condition in which the separating pressure diffusion is balanced by the remixing ordinary diffusion caused by the concentration gradients. For a given cut the isotope separation effect cannot exceed a specific maximum value, even for infinitely high flow velocities, since the UF6 is concentrated in an increasingly narrower layer as the flow velocity increases ~.
* The Nuclear Research Center is operated by Gesellschaft flu" Kernforschung m.b.H., D-75 Karlsruhe, Postfach 3640. 27
28
E. W. BF~ER
concentration profiles, and it decreases to a limiting value when equilibrium conditions are established. The angular position o f the maximum isotope separation effect depends on the diffusion coefficient. Therefore, the desirable coincidence between the maximum of the isotope separation effect and the position of the skimmer can be obtained by proper choice of the inlet pressure of the nozzle system. 3. OPTIMUM OPERATING CONDITIONS Fig. 1. Cross section of the separation nozzle system used in the commercial implementation of the process. 5I
J
Nz/OFs-MIXTURE,
3 ~/pL:8
=2""~ ~
,.,.,
/
0
EQUILIBRIUMSEPARATION
0;2
o'.L
0.6
UFs-CUT Fig. 2. The upper curve indicates the experimentally determined relation between the separation effect ea and the UF6-cut vu (mixture of UFe/H2 with 1.6 mole ~e UF6; expansion ratio 8). The lower curve gives the upper limit for equilibrium separation calculated for infinite flow velocity. Figure 3 shows the results of a theoretical analysis of uranium isotope separation in a curved flow of a UFe/H2 mixture ~. The separation effect increases to a maximum value during the evolution of the 2
i
'/ /
l°.t.,°, s,,,,,,0. i..1
•
In order to achieve the enrichment required for reactor operation, the elementary uranium isotope separation effect supplied by the separating elements must be multiplied in a separation cascade. The gas expanded in the individual separation nozzle stages is recompressed before being fed to the separation nozzle systems of the next stages. The geometric parameters and the operating conditions of the separation nozzle systems, of course, must be set so that the separative work produced by the whole plant is as inexpensive as possible. A study of the cost structure of the method shows that, at the present state of development, the price per separative work unit is determined mainly by power costs and the capital cost fraction, which is proportional to the power consumption. Accordingly, the optimization can be carried out in a first approximation by finding the minimum of the specific ideal isothermal compression energy*: 2RT Inn
E, ffi
N o ~A VU(1 -- VU)
where R = gas constant, T ----absolute temperature, = ----expansion ratio, No --- UF6 mole fraction. Fig. 4 shows an example of the dependence of the compression energy on the expansion ratio and on the U F s mole fraction for the separation nozzle system shown in Fig. 1. The minimum value of the compression energy accordingly will be reached at an expansion ratio of 2.1 and a U F s mole fraction of approximately 4 ~o. The results described in Fig. 4 apply to a constant uranium cut v, ----25 % which has turned out to be the economic optimum in a series of experiments. The corresponding cascade diagram is shown in Fig. 5. Fine adjustment of the uranium cut is possible in the individual separation stages by means of
°o( eo' I~'u'~80' 2Lo' 300" ANGLE OF OEFLECTION Fig. 3. Elementary effect of isotope separation, e,(, versus angle of deflection, calculated for a H2/UFe-mixture with 1.5 mole % UFe and a UFe-cut vu = 1/3.
* The specific ideal isothermal compression energy is the ratio between the ideal isothermal compression energy to recompress the gas expanded in the separation nozzle and the separative work produced. See also (1).
The separation nozzle process for enrichment of Uranium-235
E
i
taa
I'-
//
2 3 t S 0 O 10 EXPAHSIO RATI It O lifoMOLE % Fig. 4. Dependence of the specific ideal isothermal compression energy on the expansion ratio (left side) and on the UFe mole fraction (fight side). Mixture of UFe/Hi; UFe-cut = 0.25.
~
29
4. PRODUCTION OF COMMERCIAL SEPARATION ELEMENTS Since the optimum operating pressure of the separation elements is inversely proportional to the characteristic dimensions of the nozzles and a high operating pressure is favorable for economical operation, the characteristic dimensions of commercial separation elements are chosen to be as small as possible. To operate separation nozzles of the design shown in Fig. 1 with inlet pressures between 0.25 and 0.5 bar, the diameter of the deflection groove must be about 0.2 mm. Several methods of inexpensive mass production of separation elements of such small characteristic dimensions have been worked out in cooperation with industry. Figure 6 illustrates the design of a commercial separation element manufactured by the MESSERSCHMITT - B O L K O W - BLOHM [company of
IBll,q~IlOllSTAB[ LIGHT FRACTION
U6HI'FRACTION HEAVYFRACTION
S FEED GAS
.... " C;.;,J:
;
HEAVY FRACTION
Fig. 5. Cascade flowsheet for a UFo-cut of 0.25.
throttle valves installed in the suction lines of the heavy fractions 1. The optimum inlet pressure Po of the separation nozzle method is inversely proportional to the characteristic dimensions of the nozzle system. The optimum Reynolds number relating to the smallest nozzle diameter is of the order of 100. The operating conditions planned for commercial application and the performance data they produce are summarized in Table 1 : UFe/H, mixture with 4,2 mole ~o UF6 Expansion ratio ,r 2.1 UFe cut vv 0.25 Separation effect eA 1.48 x 10 -2 Table 1. Optimum operating conditions and performance data of the separation nozzle system shown in Fig. 1.
Fig. 6. Schematic representation of a commercial separation element tube manufactured by the company MES$ER$CH MITT-BOLKOW-BLOHM, Munich. Munich. An extruded aluminium tube, which is shown schematically in the lower part of the figure, serves as a basic unit of the element. The tube has 10 slit-shaped nozzle systems on its periphery. As shown in detail in the upper part of Fig. 6, the nozzle systems proper consist of the
30
E. W. Bv.crdEg
deflection groove machined in the tube and aluminium strips, whose edges are designed as the skimmer and the nozzle wall. The aluminium strips are fixed in the dovetail grooves of the tube and held in place by balls forced in between the strips. The feed gas for the nozzle systems is admitted into S channels at one end of the tube, the heavy fraction is withdrawn from the rest of the channels at the opposite end of the tube. The light fraction is pumped off from the space around the tube. Up to now the company, MESSERSCHMIDTBOLKOW-BLOHM, has produced separation elements with a total separative work capacity of several thousand SWU/a. Life and performance tests on these separation elements have shown that their separative work capacity did not change over a test period of approximately 4 years and that uniform separation characteristics are maintained*. Besides the mechanical method of fabrication, the SIEMENS company of Munich introduced the photoetching technique to produce commercial separation elements. This method is outlined in Fig. 7. The separation nozzle structure and the ducts for feed gas and heavy fraction are photoetched into a metal foil. Separation nozzle chips are obtained by stacking such foils. These assemblies are integrated into separation element tubes, one half of the tube being supplied with feed gas and the other half being used for the removal of the heavy fraction. The light fraction is withdrawn to the outside of the tube. Figure 8 shows a separation nozzle structure produced by the etching technique. Considering that the radius of curvature of the deflection wall is only 1/10 mm, the picture demonstrates the accuracy of the fabrication method. The SIEMENS technique is particularly suited for the commercial production of separation nozzle configurations even more complicated than the version shown in Fig. 1, which will probably have to be utilized in the course of a long-term development of the process a. An automatic production line for manufacturing photoetched separation elements has been set up by SIEMENS. At present its capacity is about 2500 m of slit length per annum. 5. SEPARATION STAGES A large number of commercial separation elements have been tested in two prototype separation stages of the kind required for the construction of a commercial scale separation nozzle plant. Figure 9 shows the cross section of the so-called large stage containing about 80 separation element tubes of 2 m length s . The stage consists of the tank holding the
V
:lion
'ion
v
. . . . . .
Fig. 9. Cross-section of the large prototype separation stage. separation elements, the crosspiece for gas distribution, the two-stage gas cooler, the two-stage centrifugal compressor, and the motor rigidly coupled to the compressor. Because of the high gas flow the stage has been tested in closed-circuit operation. The heavy fraction is recycled by the upper closedcircuit loop and mixed with the light fraction coming from the separation elements. Figure 10 shows a photograph of this stage, which was installed in July 1970. The separation experiments carried out were successful from the outset. No failures occurred in the major components of the stage, although they were severely stressed in more than 400 starts and by testing the limits of maximum load and stable operation. On the basis of the operating experience gained with the first prototype stage, an advanced version of a separation stage has been developed by the STEAG Company6. Figure 11 shows the insertion of the separation elements into the so-called small stage*. The separation elements are arranged as a compact unit to allow easy installation. All major components of the small stage were designed with a view to the requirements of mass production. The * The gas flow of the small stage is about one-third of the gas flow of the large stage.
The separation nozzle process for enrichment of Uranium-235
31
Fig. 7. The assembly of a commercial separation element manufactured by means of photo-etching (SIEMENS company, Munich).
Fig. 8. Separation nozzle structure produced by the photoetching technique (SIEMENS company, Munich).
32
E . W . BECKER
Fig. 10. View of the large prototype separation stage.
The separation nozzle process for en~chment of Uranium-235
33
1,00 3OO
""'-.'-..". "'..-L-..-:
"" ..
~., 2O0
100
2OO
~ 6OO HOURS OF OPERATION . - - - - - - -
8OO
1000
Fig. 12. Separation capacity vs time of the small separation stage.
typical test operation of the small stage is recorded in Fig. 12; the separation capacity is shown to be constant within the limits of error. One target of test operation has been to work out the most economical methods of assembling the components of a commercial separation nozzle plant. Up to now, no 'clean conditions' have been applied. Instead, dust particles present in the process loops have been collected in a filter during test runs prior to insertion of the separation elements. 6. CASCADE DF-~IGNAND UF6-RECYCLING Figure 13 shows a schematic representation of an industrial separation nozzle cascade with two types of separation stages. The use of two types of stages with a 1:3 ratio of flows makes the performance approach that of a corresponding ideal cascade with an efficiency of approximately 90 %. A total of about 500 stages must be connected in series in order to UFr RECYCLEFACILITIES \
~
PRODUCT
p~H_V~0C,E N ' ~ m m ¢ i
LIC-*HTI=I~CTI~ i'~. RECYCLED Of: TOP STJ~GIES[.J~ H~/UFi-MIXTURE
'
r uP1
t i I
I I
Lt
p
3.Owl% U-235
I I I
t
p NORHAL FEED
i
i
I
oEs
J I I f
~ f I i
LOlwt% U-23S
Fig. 13. Schematic representation of a separation nozzle cascade with UF6 recycle facilities.
produce enriched uranium containing 3 % U T M and to strip to some 0.3 % the U "3s content in the waste. A separation nozzle cascade produces a net upward transport of the light auxiliary gas, which is of the order of the stage throughput. To prevent enrichment of light gas in the cascade, the upward transport has to be extracted from the top and recycled to the bottom of a section, as is illustrated in Fig. 13. For this purpose, the light fractions of the top stages are processed in the so-called UF6recycle facilities shown in Fig. 14. Here, the UF~content of the upward flow is stripped off with high efficiency and fed back to the top stage feed flows. The flowsheet (Fig. 14) of such a facility includes a special separation nozzle stage backed by a low temperature freeze-out heat exchanger system 9. The separation nozzle stage results in some 80 to 90% of the UF6-content being fed back to the top stage in continuous operation. The remaining ~OF6 is frozen out at temperatures between --30°C and -120°C in a group of counter-current heat exchangers which are run in cyclic operation. As shown in Fig. 15, the heat exchangers include additional passages for the coolant flow by which the temperature profiles can be controlled properly. The UF6 desublimes in relatively large passages equipped with serrated fins of the multi-entry type. A prototype of such a freeze-out system was built by the LINDE company of Munich. Experiments performed on UF6/H2 mixtures under process conditions have demonstrated that even with the highest temperature differences achievable between the gas stream and the heat exchanger wall no undesired desublimation of UF~ in the gas phase has been detectable. This made it possible to run the heat exchanger with a steep temperature gradient as a function of time, as is shown in Fig. 16. By proper shifting of the temperature profile along the heat exchanger, the UF6 is deposited in a layer of
34
E . W . BECKER
SEPARATIONOZZLE N UGHTFR~TIOFROM H
I I
LOW TEMPERATURE FRFFZE-OLfT SYSTEM
I
PUREHYDROGEN
12"/,U"JF¢1
HIPAVYF"RAI:TIO, i/,'/,U[ TOTOPSTAGES
UF$- RECYI:,,(:(~ LE OFs-BOFFER-NEATSUPPLY
(00"el Fig. 14. Flowsheet of a UF6 recycle facility.
{'OLD SUPPLY
1-120'C)
mhU,,e,
H /o.3%
COOLANT FLOW
PURE
- 00 . . . . . . . . . . . . .
H2
PROCESS GAS Hz/UF 6
100,
i~~ ~[[t B.)OF6~GhLAYER BUILDUP .
1
Fig. 15. Compact freeze out heat exchanger with an additional passage for the coolant flow. constant thickness. In this way, a high freeze-out capacity is achieved with a minimum pressure drop. A UF+-content of the light gas well below 1 ppm is easily achieved. Accordingly, only negligible losses of separative work are associated with recycling of the light gas from the top to the bottom of the cascade sections. 7. STABILIZING THE OPERATING CONDITIONS ALONG THE CASCADE In the separation nozzle method, a high degree of stability is reached for the pressure distribution along the cascade, because local differences in the nozzle inlet pressure give rise to major local changes of the relatively pronounced upward transport mentioned in the previous section of the light additional gas in the sense of a stabilization. However, inherent
o WARM LENGTH OFEXCHANGER PASSAGE COLD END END Fig. 16. Time dependence of the temperature profile (A) along the heat exchanger of a UF6 recycle facility and resulting UF6 layer (B). stability of the UF6 concentration along the cascade can be achieved only by proper choice of the characteristics and operating points of the plant components. A ten stage pilot plant equipped with Roots compressors was built to work on these problems 1° (Fig. 17) and served for experimental studies both of the steady state and the dynamic control behavior of separation nozzle cascades 11. The UF6 distribution was found to assume an inherently stable state t, nder all operating conditions studied in the plant
The separation nozzle process for enrichment of Uranium-235
Fig. 17. View of the 10-stage separation nozzle pilot plant.
35
37
The separation nozzle process for enrichment of Uranium-235 (Fig. I8A). Without any active control measures the full anticipated multiplication of the elementary effect of uranium isotope separation was found
(Fig. 18S). A theoretical analysis of the stability behavior clearly indicates that the stabilization o f the UF~ concentration is determined primarily by the influence o f the UFe mole fraction on the UF~-
I,:i:ii~
1.'/2
r
L~'--
~
:
.... --
i l'J EssmMm0 I I
l.0k
cut. The separation stage must be made to react to an increase in concentration by raising the UF6-cut so that any excess UF6 is transported in the direction of the UF~ recycling system, which acts as a UF6 buffer ~2. Figure 19 shows that the separation elements are characterized by this very pronounced stabilizing
" o.2s EXPA~O-M
-:-7 -~/-/-V-
-
-
1.e/2.1I/2.,
I
2
/i
1
/
I
I
I
! ~
iI
5 "/,
6
o~ B.) UFG-CU'T01r SEPARATIVESTAGE
0.30
0.25
020
~ ;
2
"3 /~ mL~ P,~Tm OFuF6
Fig. 20. Compression ratio of a centrifugal compressor (A) and resulting UFe-cut of a separative stage (B) as a function of the UFe mole fraction.
! //"
j
0.~
t
lsl
Fig. 18. Upper diagram: UF6 mole fraction of the stages of the pilot plant, showing nearly uniform UF6 content along the cascade. Lower diagram: Experimental enrichment factor of Uranium 235 relative to the heavy fraction of stage !.
i
OF STAGE COHPR[SSOR
/
100 1 2 3 & S 6 7 0 9 10 STAGEIIgMBE0
0.35
property: the UF~-¢ut rises monotonically with increasing UF~ mole fraction. On the other hand Fig. 19 reveals that the UF6-¢ut decreases with increasing expansion ratio, i.e. with increasing compression ratio of the stage compressor. Since the Roots compressors used in the pilot plant indicate only a slight dependence on the gas composition of the compression ratio, the use of this type of compr~-~sor leaves the stabilizing property of the separation elements nearly unaffected. Therefore inherent stability of the UF~ distribution is observed in the pilot plant under all operating conditions of practical interest. The situation is somewhat more complex with the centrifugal compressors used in commercial separation stages. Figure 20 shows that in this case a higher UF6 mole fraction entails a considerably
3 ~ FICTIONOFUF6
5 "/.
Fig. 19. UF~-cut of the seimration elements as a function
of the UF6 mole fraction for different expansion ratios.
higher compression ratio, which enhances the effective expansion ratio at the nozzle and in this way reduces the UF~-cut. The marked stabilizing effect of the separation element is considerably diminished in this way by the destabilizing effect of the compressor characteristics. However, if the operating point has been chosen properly and use is made of the gas dynamic properties of the backpressure valve of the heavy fraction, it is possible, to achieve inherent stability of the UF~ distribution also with the compressors used in commerci,! application.
38
E.W.
8. ASPECTS OF BASIC RESEARCH Besides the development work on process technology, extensive studies on the physics of the separation nozzle method have been performed at Karlsruhe. These investigations relate to theoretical studies on gas dynamics and diffusion problems, flow measurements in enlarged separation nozzle models, and laboratory scale separation experiments with UFe. As an example, some activities dealing with interacting gas jets will be mentioned in more detail: In the separation nozzle system shown in Fig. 1, streamline curvature is generated by the deflection of a gas jet at a fixed wall. The steep velocity gradients existing within the flow region near the wall result in relatively high friction losses and adversely affect the centrifugal force determining the separation. The generation of curved flow lines and the simultaneous avoidance of steep velocity gradients is possible by a mutual deflection of gas jets 2.3.]3.~4.js. A separation nozzle system based on this concept is shown in Fig. 21. In this case,
FILAJ:i'IOll
.,
HEAVY FRACTION
,i,i.~ L
~
FEED
...............................
~
.' "
•
-
""
~ . ,
: ~ , . , .. ,, ~.~ . ~
.---.'..
""
FIt~IION
FEED
--'.':.'- "~ , ' ,' 4" -.': " : - - " ' - .
....'-.
.
~
-
', ;'..'.!.'... ,,-' . ' t .~'.~. . . . . ',,,,,,"
.'.-'-:~-.-..
NEAVY FRACTIOH
"-
A~
~'~
Fig. 21. Separation nozzle system of the opposing jet type. The arrows indicate the local flow directions for the heavy component of a He/SFo mixture.
streamline curvature is produced by two opposed jets. Figure 21 shows the results of flow measurements carried out by means of free molecular probes using a mixture of He and SF~. The local flow directions are indicated for the heavy component. It is apparent that a large angle of deflection and a uniform curvature of the flow lines are achieved. Separation experiments with UF6/H2 mixtures indicate that, under special conditions, the so-called opposing jet
BE~CII~R
scheme may be economically attractive for the separation of uranium isotopes ~s. It is of particular interest with regard to better exploitation of the intermediate enhancement of isotope separation by the light auxiliary gas s. Besides the work dealing with the interaction of jets under gas dynamic conditions, some work has been done at our Institute on crossed beams under molecular flow conditions 16-1s. This activity, which started from experiments on cluster beam generation withcurved nozzles 19, is mainly devoted to the mass separation of clusters. Experiments with xenon 17 demonstrated that appreciable isotope separation can occur. Molecular beam processes, however, have the inherent weakness of extremely high pumping requirements. Therefore, such concepts are hardly suitable for commercial uranium enrichment, even if comparatively high separation effects would be obtained.
9. ECONOMICS To obtain reliable cost data on the separation nozzle process, a number of industries were commissioned to develop design and fabrication methods for the most important plant components under the requirements of mass production. These orders, among other items, concerned the separation elements, compressors, cooling system, piping, tanks, and valves. On the basis of the costs determined in this way, detailed optimization studies were performed on the design and operation of industrial scale separation nozzle plants 6. These studies as well as the development work performed to date on the separation nozzle technology have provided a safe basis for analysing the economics of the process. Cost evaluations corresponding to the present development status result in specific investment costs which qualify the separation nozzle process as an economically attractive technique. At the present state of development, the specific power consumption of the separation nozzle process corresponds approximately to that of the existing U.S. gaseous diffusion plants 2°. However, there is no doubt that a further significant reduction of the specific power consumption of the separation nozzle process is to be expected. In conclusion, it can be stated that the erection of the separation nozzle demonstration plant can be recognized as the implementation of an enrichment process which combines a reliable and comparatively simple technology with a high potential for further improvement.
The separation nozzle process for enrichment of Uranium-235 REFERENCF_~ 1. Becket E. W., Bier K., Bier, W., Sch0tte R. and Seidel D. (1967) Anoew. Chemie, Intern. Edit. On English) 6, 506. This paper includes a list of previous pubfications on the nozzle process. 2. Becket E. W., Bier W., Biey P., Ehrfeld U., Ehrfeld W. and Eisenbei$ (1973). Atomwirtschaft 18, 524. 3. Becker E. W., Berkhahn, W., Biey P., Ehrfeld U., Ehrfeld W. and Knapp U. (1975) International Conference on Uranium Isotope Separation, London. 4. Becker E. W., Bier W., Fritz W., Happe P., Piesch D., Schubert K., Schfitte R. and SeidelD. (1975)International Conference on Uranium Isotope Separation, London. 5. Becker E. W., Bier W., Ehrfeld W., Schubert K., Schfitte R. and Seidel D. Nuclear Energy Maturity, Proc. o f the Paris Conference 1975. Progress in Nuclear Energy Series 1976, p. 172. Pergamon Press, Oxford and New York. 6. Geppert H., Schuhmann P., Sieber U., Stermann H., V61cker H. and Weinhold G. (1975) International Conference on Uranium Isotope Separation, London. 7. Becker E. W., Bier W., Ehrfeld W. and EisenbeiD G. (1971) Z. Naturforsch. 26a, 1377. 8. Becker E. W., Bier W., Ehrfeld W., Eisenbeil3 G., Frey G., Geppert H., Happe P., Heeschen G., Lficke R., Plesch D., Schubert K., Schfitte R., Scidel D., Sieber U., V61cker H. and Weis F. (1971) 4th United Nations International Conference on the Peaceful Uses o f Atomic Energy, Geneva, A/Conf. 49/P, paper 383 (in English).
39
9. Fritsch H. J. and Schfitte R., KFK-lkricht 1437, Gesellschaft ~ Kernforschano, Karlsruhe (1971). J. Schmid, Schfitte R. to be published. 10. Becket E. W., Frey G., SchOtte R. and Seidel D., KFK-Bericht 854, Gesellschaft ff~ Kernforschu~, Karlsruhe (1968). 11, Fritz W., Hoch P., Linder G., Schifer R. and Sch(ltte R. (1973) Chemic Ino. Techn. 45, 590. 12. Sch(itte R. (1974) KFK-Bericht 1986, Geseiischaft ff~ Kernforsctumg, Karlsruhe. 13. Becket E. W. (1961) Verfahren zum Treanen yon gasoder dampffdrmigen Stoffen, insbesondere Isotopen. German Patent 1 096 875. 14. Becker E. W., Bier W., Ehrfeld W., Schubert IC, SchOtte R. and Scidel D. (1974) KFK-Bericht 2067, Gesellschaft ~ KernforschanO, Karlsr~e. 15. Ehrfeld W. and Knapp U. (1975) KFK-Berieht 2138, Gesellschaft ff~ ICernforschuno, Karlsruhe. 16. Gspann J. and Vollmar H. (1972) Proc. RGD 8, Stanford. 17. Gspann J. and Vollmar H. (1973) Proc. 4th Intern. Conf. on Molecular Beams, Cannes. 18. Gspann J. (1974)Proc. RGD 9, G6ttingen. 19. Becker E. W., Gspann J. and K6rting K. (1972) Z. Natarforschg. 27a, 1410. 20. Becket E. W., Bier W., Ehrfeld, W., Schubert, K., Schfitte R. and Seidei D. (1975) ANS Meetino, San Francisco, Nov. 1975; KFK-Bericht 2235, Geseilschaft fdr Kernforschuno, Karlsruhe.