GB2256082A - Infrared radiation source incorporating a raman medium for isotope separation - Google Patents

Abstract

A tuneable infrared radiation source has a laser cavity 10, 13, a Raman medium positioned within the cavity, and a subcavity 16, 13 through which a Raman shifted beam (Stokes or anti-Stokes) can be emitted. The cavity has minimum output coupling and minimum losses so that a laser pulse is fully developed before Raman threshold is achieved, and the output Raman beam is emitted through the subcavity. The Raman shifted wavelength can be tuned using grating 10. A germanium etalon separates the subcavity, which may contain para-hydrogen or deuterium, from the main cavity which may contain a CO2 laser. The laser cavity may accommodate a focal beam waist (56, Fig 5). The laser may operate between a mirror (60, Fig 6) and the surface of a lens (68, Fig 6) with the Raman subcavity operating between inner surfaces of lenses (64, 68, Fig 6). The source produces 16 mu m radiation for Uranium Hexafluoride isotope separation. A series of sources may be fired parallel to each other through supercooled UF6 gas (L1-L5, Fig 7), and a beamsplitter (76, Fig 7) may be used. <IMAGE>

Description

A SIMPLE 16 Fm HIGH EFFICIENCY TUNABLE LASER SOURCE FOR ISOTOPE SEPARATION The invention of a laser system for producing high energy tunable 16 gm radiation at very high efficiency and repetition rate, with minimum capital and maintenance costs is described. The system is an ideal source for a commercially efficient Uranium Hexafluoride (UF6) isotope separation plant.

There are two methods for separating Uranium isotopes with lasers: The Uranium Atomic Vapour Laser Isotope Separation (AVLIS) method and the Uranium Hexafluoride (UF6) Molecular Laser Isotope Separation (MLIS) method. It had generally been accepted in the past that for a commercial separation plant the MLIS process utilizing Uranium Hexafluoride (UF6) is the most preferable method for the enrichment of Uranium since the AVLIS process has inherent severe problems in its commercial realization due to the corrosive nature of the metal vapour, the low working pressure, compatibility with the rest of the fuel cycle from a fluoride to a metallic product, expensive and inefficient laser systems etc.It had also been stated often in the past that the molecular route method (MLIS), which has not yet been developed to the prototype stage would finally overtake the AVLIS method (e.g. Dr Herman Roser US Dept. of energy, 1983). In addition the handling of a familiar gas (UF6) makes the processing much easier to operate. Thus for a commercial separation plant the Uranium liexafluoride (UF6) is the most attractive gas for the enrichment of Uranium. It is well known that in the molecular route (MLIS) method half the cost would come from the capital and maintenance expenditure of the laser systems (e.g.

1982 Los Alamos Science report on isotope separation). Thus an efficient high repetition rate, low cost laser system is of paramount importance in the commercial separation of UF6 isotopes.

The basic step in the MLIS method is the production of narrow band tunable laser radiation in the 16 Fm region and more specifically near the absorption band of 235up at a centre frequency of 628.32 cm'l. Many systems have been developed in the past few years which produce narrow band tunable radiation at these frequencies. However the only system which can produce sufficient energies for commercial applications is stimulated Raman scattering in para-Hydrogen pumped by a high pressure C02 laser.Due to the very low Raman gain in para-Hydrogen at 16 Em it was necessary to develop large multi-pass Raman cells (more than 6 m long and up to 25 passes of the C02 beam) in order to achieve threshold and conversion of the C02 10 pm radiation into Raman wavelengths at 16 Em. The problems associated with multi-pass Raman cells are numerous and the systems are cumbersome,con- plicated, difficult to operate and costly to construct and maintain.

It is well known that a 16 pm system suitable for commercial UF6 dissociation experiments has to meet the following requirements: (1) Tunability between 620-640 cm'l, (2) Pulse energies in the region of 10-150 mJ.

(3) Pulse duration less than 40 ns.

(4) Repetition rates greater than 5X103 Hz.

(5) Low capital and maintenance costs.

A very simple system for achieving threshold and converting the C02 10 Fm into 16 Fm radiation using Raman scattering in para-Hydrogen, and which easily satisfies all the above conditions, is described below.

The basic concept of the invention involves the positioning of the Raman active medium with small gain coefficient within the cavity of a laser with niininrum output coupling andminnimum losses. A subcavity collinear with the main laser beam is constructed which forms a Raman oscillator and allows the Raman shifted beam to emerge. In the early days of Raman scattering with lasers, the introduction into the laser cavity of cells containing Raman active media spoilt the operation of the laser system and produced disastrious results.

This effect can be attributed to the fact that with short wavelength lasers and media with high gain coefficients, the threshold was easily achievable and Raman scattering was initiated long before the laser pulse had fully developed. It subsequently depleted the laser light which had been generated within the cavity up to that instant, thus preventing the generation of more laser light through stimulated emission in the laser active medium. In the case of long wavelengths and Raman active media with small gain coefficients, the situation would be entirely different. The low Raman gain enables the laser pulse to develop fully until the population inversion in the laser active medium has been depleted, before achieving Raman threshold.

The principle on which the present Invention works is as follows.

A laser is constructed with minimum or no output coupling, min- imum losses and with as high stability as possible. A subcavity at the Raman wavelength collinear with the laser is constructed with an outlet for the Raman beam. When the laser is fired-the laser pulse develops and it is trapped within the two reflectors travelling back and forth. In doing so it traverses the Raman active medium as many times as possible. It will continue to do so until Raman threshold is reached and it starts being converted into the Raman wavelength, the latter being emitted through the output mirror of the Raman sub cavity. Specific cavity and subcavity, designs allow for repetitive refocusing of both the laser and Raman beams enabling threshold to easily be achieved even with very small systems.

Based on the concepts and principles described above the present invention for converting the C02 10 Fm radiation into 16 pm radiation by stimulated Raman scattering is described below with reference to figures 1- 7 : Figure 1 shows an arrangement embodying the concepts and principles of the invention and depicting most of its practical aspects.

Figure 2 shows the rotational Raman linewidth of the S(1) Hydrogen level.

Figure 3 shows the calculated rotational Raman gain coefficient of para-Hydrogen and Deuterium.

Figure 4a shows a laser cavity comprising a focal beam waist and figure 4b indicates the definitions of the parameters in figure 4a.

Figure 5 shows the system incorporating a common focal beam waist for both the laser cavity and the Raman sub cavity Figure 6 shows the development of the system in figure 5 using suitable hard coatings to accommodate the common focal beam waist in a straight line.

Figure 7a illustrates an application of the system to the Molecular laser isotope separation process where the UF6 supercooled molecular gas is repetitively irradiated along the direction of its expansion, each system being fired at a repetition rate corresponding to the expansion velocity of the gas. Figure 7b illustrates how an extension of the system in figure 7a can be applied to more than one expansion chambers.

Figure 1 shows a schematic arrangement of the system. A CO, laser is constructed with no output coupling. Oscillation takes place between the 100% mirror 13 and the grating 10, the latter also selecting the wavelength of oscillation.

A thin Germanium etalon 12 is placed within the cavity at the Brewster angle and thus it does not affect at all the laser oscillation which takes place with its plane of polarization parallel to the plane of the paper, the active laser medium C02 being 11. This thin Ge etalon, however, can reflect 98.6% of the light at the polarization perpendicular to the plane of the paper and thus a subcavity 13-12-14 is constructed, 14 being a partially reflecting mirror at the Raman wavelength. Since the Raman scattering in para-Hydrogen occurs at a polarization perpendicular to that of the pumping beam, the cavity 13-12-14 serves as the Raman oscillator. A Raman cell 15 is constructed around 14, 12 and 13 and filled with para-Hydrogen at the optimum temperature and pressure. The shaded section 17 represents the cross-section of the C02 10 pm laser beam and the dotted section 18 represents the 16 pm Raman shifted beam. The output Raman shifted beam emerges at 16.

The Brewster angle for Germanium is 75.980. At this angle the C02 light (polarized parallel to the plane of the paper in Fig. 1) is completely transmitted without any losses whati so ever. A thin Ge etalon at this angle has a reflectivity of 98.6 at 16 pm for light polarized perpendicularly to the plane of the paper. It is well known that at the required centre wavelength of the 235bF6- absorption band at 628.32 cm-l there is a small absorption in the Ge. However this is so small, less than 2% per mm, that it hardly affects the reflectivity of the Ge etalon.Thus for, say, a 1 mm thick Ge etalon its reflectivity would still be greater than 97% since according to Fabry-Perot interferometer theory, small absorption has no more than a minute effect on the reflectivity of an etalon. The mirror 14 can be made of any material to suit the required output mirror reflectivity of the Raman oscillator. Note that it is possible to construct the entire system with no coatings at all on any of the reflecting or transmitting surfaces.

Due to the fact that the Brewster angle of the Germanium is so large 75.980, the intensity of the beam falling on it, is very small. For a 1 cm diameter cross section laser beam, the major diameter of the ellipse on the Germanium is greater than 4.1 cm. The area covered by the beam on the Germanium is 4.1 times greater than the cross section of the beam.

Thus the intensity falling on the Ge etalon is 4.1 times smaller (l/cos75.980) than the intensity falling on any element placed perpendicularly in the beam. In addition the mirror 13 can be made of copper or any other high damage threshold material. The versatility of the system is also unique. The intensity of the C02 laser beam in the cavity 10-13 can be easily controlled by increasing the length 1213 to eliminate completely any danger of damage on the Ge etalon or on the grating 10. This does not affect the Raman process at all since the interaction length 12-13 of the two beams in the para-Hydrogen cell is also increased, keeping the overall Raman gain per pass the same. Thus there is no damage problem at all in employing the Ge etalon within the cavity of the C02 laser.

Recapitulating the above, the operation of the device shown in Fig. 1 is as follows. The C02 laser beam (shaded area) is trapped between two reflectors, the grating 10 and the 100% mirror 13. It is polarized parallel to the plane of the paper and its wavelength is tuned with the grating 10.

It will continue to oscillate back and forth until Raman scattering sets in the para-Hydrogen at the perpendicular polarization (dotted area) and conversion takes place. The Raman oscillator is formed by 13-12-14 with the output of the 16 pm beam emerging at 16. A simple system, like the one shown in Fig. 1 can easily produce between 100-200 mJ tunable 16 pm radiation at high repetition rates.

The advantages of the system are enormous: (a) It is very easy to construct.

(b) It requires the minimum number of components which are all standard.

(c) There is the best possible mode matching and overlapping of the C02 laser beam and the Raman scattered beam.

(d) Very good mode output at 16 pm (628.32 cm~l).

(e) No coatings are required on any of the components thus greatly increasing the damage threshold.

(f) The high pressure C02 laser is easily tunable even at frequencies which lie in between the C02 emission lines since the losses of the cavity (two 100% reflectors) are extremely small and the radiation is contained for along time until oscillation at the required frequency is ach ieved.

(g) The intensity falling on the Ge etalon is very small due to the very large Brewster angle (4.1 times less intens ity than for normal incidence).

(h) The damage threshold of the elements in the cavity 10-13 can easily be controlled by increasing the length 12-13 without affecting at all the Raman threshold.

(i) The mirror 13 can be made of copper or#any other high damage threshold material.

(j) It converts the entire C02 beam energy into a 16 pm beam with no losses and therefore highly efficient.

(k) It emits radiation only at 16 pm and therefore no extra apparatus for the separation of the C02 beam is needed.

(1) Tunability of the 16 pm output is achieved by simply turning the grating 10.

(m) It requires the minimum of maintenance.

In short the system satisfies all the conditions (1) to (5) above (page 2) for a commercial Uranium Hexafluoride (UF6) isotope separation plant.

It must be pointed out, however, that the system can generally be used for the production of tunable infrared rad iation using different Raman active media and different pump lasers. It is also important to notice that the mirror 14 can be substituted with a grating blazed at the Raman wavelength with the mirror 13 being made partially reflecting at this wavelength i.e. forming the output mirror of the subcavity 14-12-13 in Figure 1. In practice however such a modification of the system would be applicable only in a limited number of occasions under favourable circumstances. One drawback would be that since it would not be possible to make the mirror 13 out of copper, the number of high damage threshold materials suitable for the conditions required for making it would be very limited.Also great care must be taken in applying this modification with regard to the efficiency and tunability of the Raman shifted wavelength in any particular medium. In particular, although the Raman gain is independent of the pump laser bandwidth and the bandwidth of the Stokes signal generated from noise is identical to that of the pump, any attempt to narrow the Stokes bandwidthwith- out also narrowing the pump bandwidth lowers the gain. Thus, in general this modification would only be applicable on a limited number of occasions.

The most favourable conditions for the efficient operation of the system are those for which the Raman gain coefficient is maximum and the losses are minimum. The gain co efficient 9, 9R(Cm/Watt) for Raman scattering is given by

where Vp and Vs are the pump and Stokes frequencies respectively, AVs is the spontaneous Stokes linewidth N is thenum- ber of molecules per cc,

is the differential scattering cross section at sne pump wavelengtn ana ils Is tne ret- ractive index at the Stokes wavelength; T is the temperature of the Raman medium and Vy is the frequency shift of the Stokes transition. The theoretical expression (1) has been found to fit very closely the experimenti results. The differential scattering cross sections at two different pump wavelengths are related by

From equation (2) the differential Raman scattering cross section can be calculated for any wavelength provided its value is known at a particular wavelength.

Figure 2 shows the variation of the rotational Raman linewidth of the S(1) Hydrogen level as a function of pressure at four different temperatures. The differential scattering cross section at 10.2 pm has been calculated using equation (2) to be

cm/sr. mol . Using equation (1) and the results of Figure 2 the variation of the rotational Raman gain coefficient at 10.2 pm, gR(lo.2) with temperature has been plotted in Figure 3 at two different pressures of the para-Hydrogen gas. Care has been taken in the calculations to take account of the fact that as the temperature increases there are less and less molecules in the S(O) rotational level of para-Hydrogen for the 354.33 cm#1 Raman transition.The graphs shown in figure 3 are in very good agreement with the experimental results. It is evident that the optimum operational conditions for para-Hydrogen are for pressures of around 1/2 atm. and temperatures of around 60 K.

The peak of the Raman gain coefficient at these low temperatures partly reflects the fact that as the temperature decreases more and more molecules are in the para-Hydrogen rotational level S(O). It is also fortunate that under these conditions the stimulated Brillouin scattering gain coefficient for Hydrogen and Deuterium is very small, more than two orders of magnitude below the Raman gain coefficient.

One problem which arises in the 16 pm laser systems for the MLIS isotope separation process based on stimulated Raman scattering in para-Hydrogen (at 354.33 cm~l) is that the required shifted wavelength at 628.32 cm-l corresponds to a pump C02 frequency at 982.65 cam'1. This frequency falls in the middle of the R(30) (982.09625 cm#1Y and the R(32) (at 983.253 cm'l C02 laser lines. Subsequently a high pressure C02 laser is needed and a number of engineering problems arise as well as the fact that the capital and maintenance costs increase.What is often overlooked is the possibility of producing stimulated Raman scattering at the S(2) rotational level of Deuterium. If the Raman frequency shift of the S(2) rotational level of Deuterium at 414.71 cm-l isused for producing the desired 628.32 cm'l wavelength, the required C02 wavelength corresponds to the very strong C02 line P(24) of the 001-020 transition at 1043.163316 cm#1 and lies within its 3 GHz gain linewidth at atmospheric pressure. Moreover at room temperature 0.388 of the Deuterium molecules are in the S(2) rotational level and thus at atmospheric pressure there is a high number of molecules per cc taking part in this rotational Raman transition.Assuming that the variation of the spontaneous Raman linewidth of the S(2) rda ational level of Deuterium varies in the same way as for Hydrogen (Figure 2), the stimulated Raman gain coefficient has been calculated to be 1.05 X 10-5 Cm/Mw at room temperature and atmospheric pressure. In fact the optimum stimulated Raman gain coefficient of the S(2) line of Deuterium is at 200 0K and atmospheric pressure and has been calculated to be 1.5 X10-) Cm/MW . This is even more handy as the Brillouin scattering at lower temperatures is even smaller.Thus the Raman gain coefficient for the S(2) level of Deuterium at 414.71 cm'l is of the same order of magnitude as that for the S(O) level of para-Hydrogen at 354.33 cm'l, and is only slightly smaller, for producing 16 pm radiation. In addition, the frequency shift required for producing the precise wavelength for isotope separation corresponds toaverystrong line of the C02 laser P(24) of the 001-020 transition and lies within its 3 GHz bandwidth at atmospheric pressure. On the basis of the present invention, it is thus possible to produce a 16 pm laser system for the MBIS process based on the S(2) level Raman frequency shift in D2, in which both the C02 and D2 will be at atmospheric pressure and room temperature.

The population distributions of the rotational levels of the hydrogen (H2) and Deuterium (D2) molecules together with their rotational Raman frequency shifts are shown in thetable below (R.L. Byer, IEEE J. of Quantum Electronics, Vol.

12, 11, p. 732, 1976) ROTATIONAL RAMAN FREQUENT SHIFTS AND POPULATION DISTRIBUTION Temperature 25 OK 50 do 75 K 1004K 2000K 3006K S(0) 354. 33 cm'l 0 992 0.788 0.535 0.388 0.193 0.133 S(1) 586.85 cm-l 0.008 0.211 0.462 0.603 0.719 0.667 5(0) 179.04 cm-l 0.952 0.769 0.608 0.449 0.267 0.182 sCi) 297.52 cm-l 0.047 0.208 0.291 0.313 0.260 0.205 S(2) 414.71 cm#1 - 0.023 0.097 0.188 0.367 0.388 It is evident from the table above that at room temperature most of the Deuterium molecules (D2) are in the 5(2) rotational level.This rotational state of Deuterium produces a Raman frequency shift at 414.71 cm-l and this corresponds to precisely the difference between the P(24) line of the C02 laser at 1043.163316 cm-l and the required frequency for iso- tope separation at 628.32 cm~l. In fact it lies within the 3 GHz gain bandwidth of the C02 laser at atmospheric pressure.As pointed out before the stimulated Raman gain coefficient for this transition (S(2)) has been calculated to be 1.5x10-5 cm/Mw and this is of the same order of magnitude as for the S(O) transition of para-Hydrogen at 354.33 cmll In general the physics of the ortho- and para- states in the two cases of the H2 and D2 molecules are also very similar and can be found in most advanced text-books of Quantum Mecb- anics. In addition, the fact that the P(24) is a very strong line of the 001-020 transition of the C02 laser and the precise frequency required lies within its 3 GHz bandwidth enhances the efficiency of the system as well as greatly improving all its engineering and commercial aspects. Simple C02 laser units for producing this radiation at atmospheric pressures are readily available in the market and they operate at well over 10 kHz repetition rate. Thus using the present invention with Deuterium-it is possible to construct the entire laser system needed for the commercial separation of Uranium isotopes using the MLIS process in a simple packed bench unit.

It must also be pointed out that the present invention can be used for up-conversion by generating anti-Stokes Raman scattering. In Figure 1, mirror 14 can be made to be high ly reflecting at the Stokes wavelength and transparent at the anti-Stokes wavelength. The interaction of the pump pulse with a high intensity Stokes beam within the Raman medium will then generate a high intensity anti-Stokes beam emerging from 14. The versatility of the system enables also the anti-Stokes beam to be tuned at a required wavelength simply by tuning the grating 10. Any other parasitic wavelength which could arise can be filtered out in the section 14-12 of the system in Figure 1. The conical divergence generally associated with anti-Stokes Raman scattering is very small and the beam, with proper focusing, can propagate several metres collimated.It is thus possible to easily use such an up-converted anti-Stokes beam in most practical applications.

In all applications of the invention the ratio of the lengths of the pump laser cavity and the Raman subcavity can easily be adjusted for optimum mode matching and timing.

The construction of very small systems with output energies below 10 mJ at 16 pm is of paramount importance in the application of the invention to the commercial separation of the Uranium Hexafluoride isotopes. It has been experimentally established that the selectivity of the desired isotope 235UF6 at low pumping powers is excellent even for a strai#t- forward passage of the 16 pm beam through the supercooled molecular gas. A series of such small, highly efficient 16 pm laser systems, fired with a time delay between them equivalent to the rate of expansion of the UF6 gas ( 104 cm/sec) enables the supercooled gas to be repetitively irradiated along its direction of expansion rendering the MLIS process an outstanding selectivity in a single highly selective step (see later).In order to construct very small systems operating with output energies below 10 mJ the cavity of the laser shown in Figure 1 will have to be modified in such a way that the intensities of the two beams in the position of the Raman medium attain very high values, sufficient to reach Raman threshold. This means that a cavity must be constructed with a focal beam waist within the Raman medium. The system then becomes even more versatile as it can operate at any required level of output energy or power.

Figure 4a shows how a laser cavity can be constructed to accommodate a focal beamwaist. The light energy is contained between mirrors 40 and 46 with the electromagnetic mode being contained within the cavity with the aid of lens 44, all beamwaists, radii of curvature and Rayleigh lengths being calculated according to the [ ABCD3 law. To construct such a cavity let us start from the desired beam waist WO at the plane mirror 40. This beam waist can be selected by aperture 41 placed in front of the mirror, the active medium 42 placed in the position shown. In practice, the physical size of the selecting aperture 41 should be about 30% bigger than the calculated spot size in order to accommodate the entire cross section of the electromagnetic mode and avoid any diffraction effects.The radius of curvature of the beam Ro and the beamwaist at the position of the lens can be calculated from ordinary Gaussian beam theory. For a lens with focal length f the radius of curvature of the beam R1 after the lens, the position of the focal beamwaist e1 from the lens and the focal beamwaist W1 can subsequently be calculated again from the Gaussian beam propagation theory. A mirror 46 placed after the focal beamwaist 45 with a radius of curvature of the Gaussian beam at the position 46will return the beam back on itself to lens 44 resulting in a cavity operation completely equivalent to the cavity depicted in Figure 4b.The transverse electromagnetic mode can easily be defined by apertures 41,43 and 45 or any other apertures placed at suitable positions within the cavity, their diameters calculated at those particular positions according to ordinary Gaussian beam theory. The laser will operate between mirrors 40 and 46 with the beam parameters being depicted in Figure 4b. In many ways the cavity 40-46 is more stable than a straightforward two mirror cavity since the parameters of the electromagnetic mode inside the cavity (beam curvatures,beam waists etc) are precisely defined at their particular values by lens 44, the mirror 46 and the apertures. For any particular transverse mode size the mirror 46 can be placed at any position after the aperture 45, provided its radius of curvature matches the beam radius at that particular position.

Since the mirror 46 is opaque and can be made of very high damage threshold material such as copper, it can be placed nearer the aperture 45, especially for low energy systems.

Figure 4b depicts the laser beam within the cavity and defines the various parameters of the electromagnetic mode of the confined beam. The parameters of the design of a resonator such as the one shown in Figure 4a have to satisfy the self- consistency condition, that is we require that a stab- le eigenmode of the resonator exists which reproduces itself after one round trip (see below). Other cavity designs with focal beam waists within the oscillating cavity are possible but the cavity 40-46 in Figure 4a is the simplest one which is directly applicable to the present invention. In fact it can be shown on the basis of standard Gaussian beam theory that any optical resonator which is formed by replacing any two phase fronts of a propagating Gaussian beam by reflectors is stable.The stability condition of the resonator shown in Figure 4a can be deduced by choosing an arbitrary reference plane in the resonator and requiring that after applying the [ ABCD ] law through the various elements of the cavity the beam parameters reproduce themselves after one round trip, starting and ending at the chosen reference plane.

To be more precise let us design the cavity in Figure 4a by calculating the various beam parameters according to Gaussian beam focusing. The parameters of a Gaussian beam propagating in the z-direction whose spot size (beam waist) is WO and its Rayleigh range is z0 are defined by

where A is the wavelength of the radiation and R its radius of curvature.A Gaussian beam defined by relations (3) is transformed by a lens with focal length f to a Gaussian beam with new parameters obtained from

where Ro is the radius of curvature of the beam immediately before the lens, R1 is the radius of' curvature of the beam immediately after the lens, We is the beam size at the lens, e1 is the distance of the new beam waist from the lens, and W1 , Z1 are the new spot size (beam waist) and Rayleigh range respectively. By defining the beam waist Mg which is selected by aperture 41 in Figure 4a, we can successively calculate its radius of curvature Ro and its beam size at the lens 44 using equations (3).Subsequently, using equations (4) we can calculate the focal beam waist W1 and its distance e1 from the lens 44, where the aperture 45 is placed.

Using the new beam parameters and equations (3) we can calculate at any required distance from the focal beam waist the radius of the beam Rm. A mirror 46 with radius RRm is constructed and placed at this position.

The mode stability criterion for the cavity of Figure 4a can be obtained by applying the [ABCD] law through the vari- ous elements of the cavity and requiring that all beam parameters reproduce themselves exactly after one complete round trip of the cavity. The resulting equations describing the oscillation of the transverse radius of the ray r, are

where s designates the unit cell of the periodic structure (one round trip) and the ray matrix elements are

where f is the focal length of the lens 44, fm=R2mis the focal length of the mirror 46, d is the distance 40-44 and x is the distance 44-46 as shown in Figure 4b. Straightforward algebraic calculations show that AD - BC = 1 , as expected for a stable cavity configuration with unimodular optical elements.Solution of equations (5) in a standard way results in the stability criterion for a confined beam in the cavity of Figure 4a

The mode stability criterion condition of inequality (7) must be satisfied for any cavity design such as the one shown in Figure 4a. In designing such a cavity an easy way to see straight away whether its calculated parameters form a stable configuration is to check whether the two constituent parts of the cavity shown in Figure 4b, the plano-concave cavity between mirror 40 and Ro and the concave-concave cavity between R1 and Rm#, both satisfy the stability criterion for a two-mirror cavity on their own : for a stable resonator either the centre of curvature of one mirror or that mirror itself, but not both, must cut the axis between the other mirror and its centre of curvature.Note that for x = 2 Rm, f = m R2m and d = Rm the cavity becomes symmetrical concentric and inequality (7) gives Ibi 1 the cavity becoming on the verge of instability as expected. Also for x = 2f , Rm= 2f and d= f the cavity be comes confocal and inequality (7) gives |b|= O , the cavity being again on the verge of instability as expected. For f ~ co and Rm # oo the cavity is plane parallel and again inequality (7) gives b 1 1 the cavity becoming on thever- ge of instability as expected. Note that because the beam waist at the lens 44 can be made as large as necessary there is no difficulty with regard to damage problems in placing hard antireflection coatings on the lens surface, such as for example on a Ge lens, at low laser energy levels below 30 mJ.Other laser cavity designs incorporating a focal beam waist are possible. For example, if a very tight focal beam waist is required at 45, the lens can be constructed with two different radii of curvature with one of the surfaces producing a very tight focal beam waist at 45, the parameters of the electromagnetic mode being calculated using eqs.

(3) and (4) through the various optical surfaces of the cavity, as described earlier.

Let us give a practical example in designing a cavity like the one shown in Figure 4a. For a cavity with d.=0.7 m and a lens with focal length f = 0.12 m , choosing the beam waist at the mirror 40 to be Wo=1,8 mm we obtain using equations (3) the beam parameters in the d-section of the cavity to be z0 = 0.9787 m We , V#=2.2lX1O3 m , Ro = 2. 07 m Subsequently, using equations (4) we calculate the beam parameters in the x-section of the cavity to be R1 = 0.1274 m , z1 = 0.0109m, #1=0.1265m , WI=0.19X103 in We see that the focal beam waist W1 at 45 is ten times smaller than WO, resulting in the intensity at 45 being at least two orders of magnitude higher than anywhere else in the cavity. For a mirror placed at 0.05 m from W1 we obtain Rm= 0.0524 m , x =#+ gl, 0.05 = 0.1765 m, R1 + Rm = 0.1798 m We see that R1 + Rm > x as expected for a stable cavity in the x-section.The spot size (beam waist) at the mirror is Wm=O.9Xl0-3 m the intensity of the beam already becoming very small and can be sustained by any copper mirror.

The mirror 46, however, can be constructed and placed at any position away from W1 to allow for any desirable intensity level to be falling on it. For this particular example substituting the above values in expression (7), we obtain for the stability condition that |b| Ib 8 0.8 < 1, , from which we see that the cavity is stable. For all designs of the cavity shown in Figure 4a the validity of the stability con dition in inequality (7) must always be checked. The cavity can be made longer or shorter for intensity control and the tightness of the focal beam waist can be selected at will by using different focal length lenses or adjusting the other beam parameters. The construction and alignment of the cavity shown in Figure 4a is a simple procedure and it is no more difficult than an ordinary two-mirror cavity.

Let us now apply the principle of the above cavity design to the operation of the 16 pm system of the present invention. Figure 5 shows the complete system incorporating the Raman subcavity, with both the C02 laser cavity 50-57 and the Raman subcavity 57-53-58 having a common focal beam waist at 56. The shaded region shows the cross-section of the C02 10 pm laser beam whilst the dotted region shows the cross-section of the 16 pm Raman shifted beam. Because the wavelengths of the pump beam (- 10.4 pm) and the Raman shifted beam (- 16 pm) are very close, the beam parameters of the two cavities are very nearly the same.At these long wavelengths in the infrared the transverse mode can very easily be controlled with the apertures, and the alignment conditions are not stringent at all. Care must be taken to make the main controlling mode aperture at the focal beam waist 56 large enough to avoid any diffraction problems.The ratio of the lengths of the two cavities can also be adjusted for optimum mode matching and timing by making, if necessary, a slight curvature on the output mirror 58. The lens 54 can be made of any infrared material which is transparent to both wavelengths. Special Germanium with no absorption at these wavelengths and with hard antireflection coatings is ideal for such thin lenses. Any other suitable material can be used. The beamsplitter can be made of athin Germanium etalon at the Brewster angle as previously described or any other suitable dichroic device or dispersive element. Note that because we are interested in low energy systems there is no damage problem at all for these components which are readily available with excellent quality in the market. The mirror 57 can be made of copper or any other high damage threshold material and can be placed at any position after the focal beam waist 56 provided its curvature is constructed to match the Gaussian beam curvature at the position at which it is placed, according to thecav ity design. Because the length of the cavity is adjustable as well as all other parameters and the focal beam waists occur only in the para-Hydrogen gas, this system can also be designed to produce high energy outputs without any damage problems at all, as previously. The Raman cell 59 is constructed around 57, 53 and 58 and filled with para-Hydrogen gas at the optimum temperature and pressure.

The incorporation of the Raman subcavity and the matching of the electromagnetic modes of the two cavities is very easy to construct. From eqs. (3) we see that the necessary condition for the radii of curvature of the two beams to be the same at the same distance from the beam waists, is that the Rayleigh ranges of the two beams, the laser beam and the Raman shifted beam, are the same.- Denoting all the Raman beam parameters by adding a subscript R to the corresponding laser beam parameter, for the best possible matching of the two beams it is best to start from their minimumbeam waists at 56 and require that z1 = Z1R . In this way the radii of curvature of the two beams are the same and therefore their wavefronts are in phase along the whole of the interacting length.This is very appropriate for the interaction process. In applying eqs. (3) and (4) to the laser example given previously we obtain the following parameters for the Raman beam Z1 = Z1R = 0.019 m , W1R= 0.235X 10 m giving RmR - 0.0524 m , 0.325 in and R1R = 0.1274 m , WeR= 2.74X 10-3 m , RoR= 2.066 m the latter requiring the spot size (beam waist) at the output Raman mirror to be WOR 2.23X10 3 m (note that the focal lengths of the lens at the two wavelengths are virt ually the same and no other adjustment in this respect may be necessary).The radii of curvature of the two beams and their phase fronts are identical along the whole of theinteraction length, easily facilitating the interaction process. We see from the above calculations that there is only a very small difference in the cross-sections- the two beams. The Raman beam cross-section is slightly bigger than the laser beam cross-section (taking the spot sizes at the 1/ field points). In actual system-operation however this is an advantage. In any laser cavity containing a high gain medium the beam cross-section is not exactly Gaussian but flattens out beyond the 1/e field points due to the high gain of the active medium. In a laser with no cavity losses this effect will be more pronounced.Thus in practice the converted Raman beam will be in perfect match with the laser beam, in size as well as in all other parameters along the whole of their interaction length. The defining mode apertures 55 and 56 must therefore be adjusted to suit the Raman beam mode so that they are wide enough to avoid any diffraction effects occuring. In general, also, a meniscus lens would be the most preferable lens design to be used within an operating cavity.

Let us briefly apply the Raman threshold conditions to the present invention. In a Raman oscillator the gain should be sufficient to compensate for the round trip losses, the oscillation condition becoming (for a cavity of length LR)

or using eq.(l) we find the threshold intensity (Ie)th to be

For operation in a para-Hydrogen gas at 60 0K and 0.5 atm, we obtain using Loschmits number (2.69XlOl9inOl/cm3atS.T.P) that the number of hydrogen molecules per cc is NH 6Xl019 moVcm3 , and since from the table above only 0.7 of the molecules are in the para-EIydrogen state N = 4.2X1019 inOVcm3.

At this temperature and pressure (7d,#)p= 2.61X10 36 cm/sr.mol as calculated earlier. On considering a Raman cavity to be at least as long as the Rayleigh range 2z1R and with a cavity reflectivity R= 0.8 , we obtain from eq.(8) above (for the example previously cited z1=l.O9 cm) the threshold intensity for Raman oscillation to be (I#)th = 2.3 X109 W/cm2 .

This is the intensity needed for only the length of the Rayleigh range of the focal beam waist. In reality when the whole length of the interacting beams is taken into account the laser threshold intensity needed at the focal beam waist is well below 1 GW/cm2 . If in the example previously cited (W1R= 0.019 cm) we consider that the laser energy emitted and staying within the cavity is 20 mJ and that this is emitted from the laser active medium within 50 X10 9 sec, for a cavity of about 80 cm long the pulse is folded at least 20 times from beginning to end. The intensity at the focal beam waist becomes 7.5 X109 W/cm2 . We can consider that the Raman active medium is continuously subjected to this laser intensity at the common focal beam waist.We see that even at C02 laser energies as small as 20 mJ the laser intensity at the focal beam waist is more than an order of magnitude higher than the pumping intensity needed for the Raman cavity to oscillate. Note that the design of the system enables the diameter of the beam to be made as large as desirable and the focal beam waist as small as required.

Control of the interaction length and Raman mirror reflectivity further enhance the versatility for the achievement of Raman threshold and operation of the system.

Let us look at the threshold condition from another more simplistic approach, that of a trapped beam traveling up and down the cavity and being refocussed through the focal beam waist (W1R= 0.019 cm, 2z1R= 2.178 cm). If we consider a 10 mJ pulse of duration 50 X10 9 sec folded at least 20 times within the cavity, its intensity at the focal beam waist is 3.5 X109 W/cm2 , resulting in X1O#11X3.5Xl09X2.l78 = 0.38 . A gain of must be reached for stimulated Raman scattering to initiate.

Therefore the pulse must traverse the cavity only 20 times in order to achieve threshold. This is very easily achievable. We see from the above numerical analysis that using cavity designs for the present invention such as the one shown in Figure 4a, threshold for Raman scattering can be very easily achieved even for small laser pumping energies below 10 mJ.

The operation of the system is very easy to comprehend.

The laser beam which is contained within the electromagnetic mode of the cavity 50-57 in Figure 5 has no means of escaping since the cavity has no output coupling. As it travels up and down the cavity it is repetitively refocussed at the focal beam waist 56 within the para-Hydrogen gas. It will continue to do so until it is converted into the Raman shifted wavelength which is emitted through the output mirror 58 of the Raman subcavity 57-53-58. The common focal beam waist is produced by lens 54 and mirror 57. The transverse modes of the laser cavity and the Raman subcavity are very easily controllable with apertures placed at 51, 55 and 56. The C02 laser active medium is placed at 52 and a Raman cell 59 is constructed around the Raman subcavity to contain the Raman gas.Note that threshold is very easily attainable even for very small pumping energies below 10 mJ since the Raman scattered light which is contained within the Raman subcavity 57-53-58 is also repetitively refocused at the same focal beam waist 56. The intensities reached at the focal beam waist 56 even at such small energies below 10 mJ, considerably exceed any intensities which can be reached in the enormous and cumbersome multi-pass Raman cells which use up to 1 J in order to attain threshold. A great advantage of the system is that the intensity levels which can be reached as well as the interactibnlength in the Raman medium are very easily controllable by the cavity design. In addition, the number of refocusing times is very large and the interaction length in the Raman active medium enormous.The flexibility of the system is also unique : The cavity can be constructed to attain threshold at any power level and the output energy at 16 pm can completely be controlled at any required level for very small or for very large systems by the cavity design.

The system may further be developed into an even more simple and straightforward arrangement by using the fact that with the latest technology extremely hard coatings in the infrared are possible with very high damage threshold intensities. This development uses the fact that a meniscus lens, made of a material with high refractive index such as Germanium ( = 4.0032), is used as the focusing element within the cavity. In general, meniscus lenses are better suited as internal focusing elements in cavity operation. The optical arrangement of this further development of the system is shown schematically in figure 6.The shaded region shows the cross-section of the CO, 10 pm laser beam whilst the dotted region shows the cross-section of the 16 pm Raman shifted beam. Let and r2 be the concave and convex radii of curvature of the meniscus lens 64. The focal length f of the lens is given by

To construct the system as shown in Figure 6 we require that r1 ~ R1 where R1 is the radius of curvature of the beam as shown in the schematic diagram of Figure 4b.In addition, another meniscus lens 68 is constructed as the output element of the system, and we require that its concave surface rm' Rm where Rm is the radius of curvature of the beam as shown in the schematic diagram of Figure 4b The outer convex surface of the lens r01 is constructed so that the output 16 pm beam is made parallel. The r2 convex surface of the lens 64 is antireflection coated at the C02 wavelength at - 10.4 pm . The concave surface of the lens 64 is coated so that it is completely transparent at the C02 laser wavelength and 100 reflecting at the 16 pm wavelength.The concave surface of lens 68 is coated so that it is 100g re- flecting at the C02 laser wavelength and partially reflect ing at the 16 pm wavelength according to the desirable reflectivity for the Raman subcavity. The outer convex surface of the lens 68 is antireflection coated at 16 pm . All surfaces are coated with very hard coating materials. The operation of the system is as follows. The C02 laser operates between mirror 60 and the inner concave surface of the meniscus lens 68, exactly as in Figure 6. The Raman subcavity now operates between the inner concave surfaces of the lenses 64 and 68, the radii of curvature of the two mirrors being r1 - R1 and rm R Rm where R1 and RRm are the radii of the beam as defined in the x-section of Figure 4b.Again, the laser cavity 60-68 and the Raman sub cavity 64-68 have a common focal beam waist at 66, the output 16 pm beam emerging at 69. The cavity construction and its parameters are calculated as before. The C02 laser active medium is placed at 63 and the transverse modes of the laser cavity and the Raman subcavity are controlled with the apertures 61, 65 and 66 or any other apertures placed at suitablepositions. Let us construct, however, the lens 64 to fit the parameters in the example of the laser design describedearlier. We require the focal length of the lens to be f=0.l2 m and from the previous example r1' = R1 = 0.1274 m .Substituting in eq.(9) we find that for a Germanium lens ( = 4.003) r2 =0.0941 m . Thus in order to construct a system operating in exactly the same way as in the example previously given, a Ge meniscus lens 64 must be constructed with the radii of curvature of the two surfaces being r11 =0.1274 m and r2 = 0.0941 m giving the lens focal length to be f = 0.12 in . The Raman gas cell 67 also becomes much easier to construct. An etalon 62 may also be used in the C02 laser cavity for fine tuning. Any suitable material in the infrared can be used for the construction of lenses, mirrors and etalons in the cavity. The ability of the latest technology to make extremely hard coatings in the infrared make the construction of small systems like the one in Figure 6 a very simple task. Note also that the flexibility available in the construction of the system allows for the beam diameters to be made large enough at the positions of the co atings, thus diminishing the intensities, by making the most suitable cavity design.

All cavity designs like the ones shown in Figures 5 and 6 can be made long or short, with large or small beam diameters and with as tight focal beam waists as necessary. The desirable parameters for 16 Fm operation can be attained by proper cavity designs according to the theories and examples described above. In addition, the arrangement shown in Figure 6 may be ideal in attaining threshold at the S(2) rotational level of Deuterium at the 414.71 cm'l shift, as pointed out earlier. Since this shift corresponds to the very strong C02 line P(24) and lies within its 3 GHz bandwidth as explained earlier, a grating may not be needed for tuning, an ordinary mirror with an etalon being sufficient for this purpose in an atmospheric pressure C02 laser.The design and construction of such small systems becomes extremely simple, as well as the production of antiStokes radiation.

As pointed out earlier the construction of small systems emitting 16 Am radiation with output energies below 10 mJ is of paramount importance in the commercial realization of the Molecular Laser Isotope Separation (MLIS) process,main- ly because it has been experimentally established that the selectivity between the two Uranium Hexafluoride (UF6) isotopes at such low pumping powers is excellent even for a straightforward passage of the 16 pm beam through the expansion supercooled molecular gas. The present invention enables very small, easy to build and maintain, highly efficient systems operating at 16 pm to be constructed.It is therefore easy to construct a series of these systems which can be fired parallel to each other through the supercooled UF6 gas along the direction of its expansion. A time delay equivalent to the rate of expansion of the supercooled molecular gas is introduced in between firing each system and thus the UF6 gas is completely uniformly illuminated along the direction of its expansion rendering the whole process an outstanding efficiency in a single highly selective step.

Note that each laser unit can be as efficient as 8% - plug to light - the whole process becoming outstandingly effic ient. In fact all the laser units can be fired from the same power supply bank provided timing arrangements are made for each one to be fired with the appropriate time delay from each other.Figure 7a shows a schematic arrangement for the MLIS process where five small laser systems at 16 pm , 70 designated by (lil, L2' 2 3 5), L#)# are repetitively appli- ed in the direction of expansion of the supercooled moleclar gas, each laser system being fired with a time delay equivalent to the time taken for the gas to expand from the position of one laser beam to the position of the following laser beam (the expansion velocity of the supercooled molecular gas is approximately 104 cm/sec ). This can easily be achieved by controlling the triggering of the laser systems. The input beams 71 pass through the expansion chamber 72 and irradiate uniformly the US6 supercooled molecular gas at repetitition rates and time delays which match its expansion velocity.The beams emerge at 73. The arrangement can be repeated as many times as necessary. The repetitive irradiation of the supercooled molecular gas along the direction of its expansion using time delays corresponding to the velocity of expansion of the gas can be applied for any optical system or arrangements. Such an arrangement is shown in Figure 7b where the beams 75 from eight small systems 74 (L6 - L13), are split by a beamsplitter 76, each part of the beams (77 and 78) being directed through a molecular gas expansion chamber, 79 and 80, as shown in the figure. Many other practical arrangements are possible. In general, the application of the present invent; ion incorporating a common focal beam waist is unique in that its flexibility in designing it allows the threshold to be attained for any particular Raman gas, even with low gain coefficient, and capable of producing high or low output energies and powers.

Summarizing, the invention of a tunable, high efficiency infrared laser source which can be operated at high or low output powers has been described. It is very versatile and it can be operated at high repetition rates with a minimum of capital and maintenance costs. It can be employed in a multitude of ways and its application to the Molecular Laser Isotope Separation (MLIS) process, renders the latter by far the most efficient industrial route for the commercial separation of the Uranium Isotopes.

Claims (16)

1. A tunable infrared radiation source where the basic concept of the invention involves the positioning of a low Raman gain medium within the cavity of a tunable laser with minimum output coupling and minimum losses, thus enabling the laser pulse to develop fully before achieving Raman threshold, and a subcavity is constructed which can operate at the Raman wavelength through which the output Raman shifted beam is emitted.

2. A tunable infrared radiation source as claimed in claim 1 wherein the laser cavity has no output coupling and thereby the laser pulse after having fully depleted the population inversion within the laser active medium has no other means of escaping and thus traverses the Raman active medium as many times as possible until it is converted into the Raman shifted wavelength by the Raman active medium and subsequently escapes through the Raman subcavity.

3. A tunable infrared radiation source as claimed in claim 1 or claim 2 wherein the pump laser cavity is constru#cted in such a way as to incorporate one or more focal beam waists at the position of the Raman active medium, such beam waists being facilitated by the introduction Wnto the pump laser cavity of lenses or other focusing optical device,the operating optical mode and cavity design being constructed according to Gaussian beam propagation theory.

4. An infrared laser radiation source as claimed in claim 1 or claim 2 or claim 3 wherein the pump laser cavity and the Raman subcavity are both constructed with focal beam waists at the same position within the Raman active medium, the tightness of the focusing governing the intensities which can be reached as well as the interaction length between the pump and Raman shifted beams being very easily controllable by the cavity and subcavity designs, the latter being constructed for any required intensities, energies and interaction length according to Gaussian beam propagation theory.

5. A tunable infrared laser radiation source as claimed in claim 1 or claim 2 or claim 3 or claim 4 wherein use is made of an intracavity lens to accommodate a common focal beam waist position #or the pump laser beam and the Raman subcavity, enabling threshold to be reached even for very low pumping energies, and in addition a development of such a system whose operation and design of the cavity laser mode is governed by the validity condition of inequality (7).

6. A tunable infrared laser radiation source as claimed in claim 1 or claim 2 or claim 4 or claim 5 wherein the Rayleigh ranges of the pump laser beam and the Raman shifted beam having a common focal beam waist position are the same, resulting in the phase fronts being the same with identical beam radii of curvature along the whole of the interaction length, the ratio of the lengths of the pump cavity and the Raman subcavity capable of being adjusted at any required value.

7. A tunable 16 pm laser radiation source as claimed in any preceding claim, with very high efficiency at high repetition rates and with minimum capital and maintenance costs for UF6 isotope separation, consisting of a C02 tunable laser with no output coupling and minimum-losses within thecåv- ity of which is incorporated a Raman active medium with low Raman gain coefficient at 16 pm (in this case Para-hydrogen gas at the optimum temperature and pressure) so that the laser pulse fully develops before any Raman oscillation sets in, and a subcavity is constructed which can operate atithe Raman wavelength (16 pm) using a thin Germanium etalon at the Brewster angle or any other suitable dichroic device within the main C02 cavity, the output 16 pm beam being emitted through the output mirror of this subcavity.

8. A tunable 16 pm laser radiation source as claimed in claim 7 or any preceding claim wherein a Raman subcavity containing Para-hydrogen gas is incorporated within the main cavity of a tunable C02 laser with no output coupling and minimum losses using a thin Germanium etalon at the Brewster angle or any other suitable dichroic device so that the C02 laser operates with its polarization perpendicular to that of the Raman shifted 16 pm subcavity wavelength, and wherein the Raman subcavity may accommodate etalons, gratings, filters or any other optical elements, the output 16 pm beam emerging through the output mirror of the Raman subcavity.

9. An infrared laser radiation source as claimed in any preceding claim wherein the entire pump beam energy can be converted into a Raman shifted beam which can be directly emitted from the system and where tunability of the output beam is achieved simply by substituting the pump cavity mirror by a grating and tuning the pump beam to produce the required Raman shifted wavelength, the cavity being capable of accommodating an etalon or any other optical elements which may be required for finer or specific operations of the system, the damage threshold of the elements in the cavity being easily controllable by the cavity design without affecting at all the required Raman threshold intensity.

10. A tunable infrared laser radiation -source as claimed in claim 5 or claim 7 or any other preceding claim wherein the intracavity lens is a meniscus lens which is constructed to fit the design of the system in such a way that its concave surface can be used as the 100 mirror for the Raman subcavity, the concave surface of another specially constructed meniscus lens serving at the same time as the output mirror of the Raman subcavity and a 100 mirror of the pump laser cavity, the surfaces of the system being appropriately coated for the operation of the system with hard coatings, substantially as described herein and with 'ref er- ence to Figure 5b.

11. A tunable 16 pm radiation source as claimed in any preceding claim wherein the Raman active medium is Deuterium at the optimum temperature and pressure so that the required output Raman wavelength for Uranium Hexafluoride > @ isotope separation at 628.31 cm'l can be obtained from the frequency shift of the 5(2) rotational level of Deuterium at 414.71 cm'l corresponding to the very strong C02 laser line P(24) of the 001-020 transition at 1043.163316 cm#1 and lies within its 3 GHz gain linewidth at atmospheric pressure, requiring only an etalon for the cavity to be tuned to the desired wavelength, enabling an extremely simple system for Uranium Hexafluoride isotope separation to be constructed.

12. A tunable infrared radiation source as claimed in any preceding claim wherein any infrared laser system can be used as the pump laser and any Raman active medium can be used within the Raman subcavity in order to produce a required wavelength for any one particular application.

13. A tunable infrared radiation source as claimed in any preceding claim wherein the system can be modified to produce up conversion of the pump laser radiation and emit an antiStokes Raman beam by making the output mirror of the Raman subcavity highly reflecting at the Stokes wavelength and more transparent at the antiStokes wavelength or by making any other change to the Raman subcavity whereby it enables the interaction of the pump pulse with the high intensity Stokes beam within the Raman medium to generate a high intensity antiStokes beam which emerges from the output mirror of the Raman subcavity.

14. An infrared laser radiation system as claimed in any preceding claim wherein the said system forms part of aseries of such small systems which are repetitively applied to the Uranium Hexafluoride (UF6) supercooled molecular gas in the molecular laser isotope separation (MLIS) process along the direction of its expansion and fired with a time delay between them equivalent to the time taken for the gas to expand from the position of one laser beam to the position of the following laser beam.

15. A tunable infrared laser radiation system as claimed in any preceding claim wherein the said system or part of the said system or method or concept described in any preceding claim is or forms part of any system or process for Uranium Hexafluoride (UF6) isotope separation or enrichment, or of any other system for the separation or enrichment of any other isotopic species.

16. A tunable infrared laser radiation source substantially as described herein with reference to Figure 1, Figure 5, or Figures 1, 5 and 6 of the accompanying drawings.