Laser isotope separation method employing isotopically selective collisional relaxation

Abstract

A method of separating isotopes from polyatomic molecules, in particular 13C from trifluoromethane HCF3, by applying to the polyatomic molecules in the gas phase two infrared laser beams of different frequencies. The first laser has a frequency appropriate to excite a low overtone transition of a light atom stretch vibration and produce vibrationally pre-excited molecules enriched in the desired isotopes, for instance 13C. The second laser has a frequency and energy fluence to selectively induce dissociation of the vibrationally pre-excited molecules by infrared multiphoton excitation. The product of the pressure of the molecules and the time-delay of the second laser pulse relative to the first allows collisional vibrational deactivation of a substantial amount of the vibrationally pre-excited molecules containing non-desired isotope(s), like 12C, before dissociation of the vibrationally excited molecules occurs, while there is no significant collisional vibrational deactivation of pre-excited molecules containing the desired isotope, like 13C. The dissociation products are hence highly enriched in the desired isotope.

Description

FIELD OF THE INVENTION

This invention relates to the separation of a desired isotope from polyatomic molecules containing different isotopes, by applying to the molecules in the gas phase at a predetermined pressure, near infrared radiation of a first pulsed laser and, after a predetermined time-lag which allows a sufficient number of collisions, infrared radiation of a second pulsed laser of different frequency to produce a chemical reaction resulting in a molecule, enriched in the desired isotope, which can be separated from the remainder of the material. The invention is exemplified in a particular by the separation of 13 C isotopes in polaytomic molecules consisting of mostly 12 C isotopes and which contain C—H and C—F bonds.

BACKGROUND OF THE INVENTION

The stable 13 C isotope has been widely used in many applications but until recently in relatively small volume. Recent medical development of the so-called carbon- 13 Diagnostic Breath Test ( 13 C DBT) (U.S. Pat. No. 4,830,010) has dramatically changed the situation. The DBTs are used to assess the condition of organs of the human digestive system. Because of its safety, relative simplicity and wide range of application, the DBT technology has rapidly increased the demand for 13 C.

A limiting factor for the growth in the use of DBTs is the relatively high production cost of highly (>99%) isotopically pure 13 C. The bulk of the 13 C at present is produced by multi-cycle low temperature distillation of CO. This technique is well developed and has nearly reached the maximum of its efficiency, limited by its high energy consumption.

The molecular laser isotope separation (MLIS) approach provides an alternative for production of high purity stable isotopes. The most developed method for MLIS of 13 C is based on infrared multiphoton dissociation (IRMPD) of CF 2 HCl by a pulsed CO 2 laser. This method relies on a 20 cm −1 isotopic shift in the IR absorption spectrum of the 13 C containing molecules relative to 12 C containing molecules for selective absorption and dissociation of 13 CF 2 HCl. The CF 2 dissociation fragments recombine, resulting in stable C 2 F 4 molecules that are separated from parent molecules by distillation.

An example of a recent implementation of this approach by Ivanenko et al. (Applied Physics B, 62, pp. 329-332, 1996) produce a macroscopic enrichment of 13 C using a high-power high repetition rate industrial CO 2 laser. A report by V. Y. Baranov et al., (Proceedings of 4th All-Russian International Scientific Conference on “Physical Chemical Processes at Selection of Atoms and Molecules”, 1999, pp. 12-16) describes a near completed pilot plant in Kaliningrad, Russia, which is designed to produce several tens of kilograms of isotopically pure 13 C a year using the same approach. In both cases, CF 2 HCl is enriched to 30-50% in 13 C by selective IRMPD. In both cases, it is suggested that further enrichment of the products up to 99% 13 C could be accomplished by non-laser techniques such as centrifugation. In another approach, a second stage of laser separation is employed to bring partially enriched product to higher levels of enrichment (Ph. Ma et al., Appl. Phys. B 49 503 (1989)). In the case of 13 C isotope separation using CF 2 HCl as a starting material, the partially enriched product (C 2 F 4 ) is chemically converted to a molecule suitable for the next laser isotope separation cycle (A. P. Dyad'kin, et al.; Proceedings of 4th All-Russian International Scientific Conference on “Physical Chemical Processes at Selection of Atoms and Molecules”, 1999, pp. 17-20). This extra stage complicates the overall process and significantly increases cost of the product.

Under certain conditions, single-laser IRMPD of CF 2 HCl has demonstrated the capability of producing products highly enriched in 13 C in a single stage, but this high degree of enrichment comes at the cost of productivity. The work of Gauthier et al. (Appl. Phys. B. 28, 2, 1982) achieves enrichment to 96%, but this requires operating at low laser fluence and low pressure, both of which decrease he productivity. Reasonable productivity is achieved at only 50% 13 C enrichment, which falls short of the high purity (>99%) required for medical applications.

One approach to increase the selectivity in laser isotope separation is to use a single-stage two-laser process. U.S. Pat. No. 4,461,686 relates to two-color IR—IR MLIS wherein a first laser excites a non-specified vibrational state and a second laser excites molecules up to a level of a chemical conversion, including dissociation. A similar method has been successfully realized on a laboratory scale by Evseev et al. (Appl. Phys. B36, 93, 1985; Sov. J. Quantum Electron. 18, 385, 1988). While this approach overcomes some of the drawbacks of a single-laser process and achieves relatively high selectivity (S=6000 which corresponds to 13 C enriched to 98.5%), low pressure is still required, limiting the productivity.

One widely known problem of two-laser isotope separation schemes is the possibility of vibrational relaxation of the molecules in the time between the two laser pulses, leading to loss of isotopic selectivity, U.S. Pat. No. 4,461,686 clearly states this problem by specifying a time delay between laser pulses that is shorter than the vibrational relaxation time but longer than the rotational relaxation time of the polyatomic molecules, allowing time for rotational but not for vibrational relaxation.

A number of other two-laser schemes have been employed for separation of various isotopic species, but in most cases, conditions are adjusted to minimize collisions in the time between the two laser pulses and/or the deleterious effects of collisions on the selectivity is explicitly mentioned. In their two-color infrared isotopically selective decomposition of UF 6 , Rabinowitz et al. (Optics Letters 7, 212 (1982)) indicate that they use pressures of less than 10 −7 Torr during runs, ensuring collision free reactions. They clearly state that energy-exchange collisions between the two isotopic species may scramble the selectivity. Using a similar two-color laser isotope separation scheme for SeF 6 , Tiee and Wittig (J. Chem. Phys. 69, 4756 (1978)), state that they use a delay between the two lasers that is short enough so that deleterious energy transfer processes do not have a chance to interfere. In their two-color multiple photon dissociation of CF 3 T, Pateopol and O-Neil (Laser Isotope Separation, SPIE, Vol. 1859, p. 210-218 (1993)) show in FIG. 4 that an increase in pressure, which increases the frequency of collisions, decreases the isotopic selectivity. In a two laser scheme for separation of sulfur isotopes, French patent FR2530966A does not explicitly mention collisions but uses sufficiently low pressure and short time delay such that vibrational relaxation from collisions between the two laser pulses is minimized. In their two laser dissociation scheme for OsO 4 , Ambartzumian et al. (Optics Letters 1, 22 (1977)) do not mention collisions, however the information they provide on the experimental conditions, particularly the low pressure (˜0.3 Torr) suggests that no collisional vibrational relaxation occurs during the process.

A few studies have observed that under certain conditions, collisions seem to enhance the isotopic selectivity. In their single-laser IRMPD of CF 2 HCl for 13 C enrichment, Gauthier et al. (Appl. Phys. B. 28, 2, 1982, FIG. 3 ) demonstrate increasing selectivity with increasing pressure. This increase in selectivity is accompanied with a corresponding decrease in dissociation efficiency (also FIG. 3 ), leading to low values of productivity. In their two-laser IRMPD studies of CF 2 HCl for 13 C enrichment, Evseev et al. (Appl. Phys. B36, 93, 1985; Sov. J. Quantum Electron. 18, 385, 1988) observe modest increase in selectivity both upon increase in the pressure of the working gas as well as upon increasing the delay between the two lasers. They attribute the increased selectivity to different rates of vibrational—vibrational exchange of “hot” ensembles of 12 C and 13 C containing molecules with the ensemble of “cold” unexcited molecules of the main isotope, although they propose no explanation for the rate difference.

We believe that the attribution by Evseev et al. of the pressure and time-delay dependence of the isotopic selectivity to a difference in collisional deactivation rates is essentially correct, although the particular pre-excitation technique that they use, namely CO 2 laser infrared multiphoton excitation (IRMPE), prohibits them from exploiting this effect for simultaneously achieving both high selectivity and high productivity in 13 C isotope separation. IRMPE can either pre-excite molecules to a few low energy vibrational levels when the laser fluence is low, or to a wider distribution of higher energy levels if the laser fluence is high. In both cases, the collisional effect can provide only a limited improvement of selectivity. Indeed, the 6000 maximum isotopic selectivity in their work has been achieved only for relatively low pressure (2.5 Torr) and only for cold molecules (−65° C.). Cooling molecule to temperatures in the range of −60 to −70° C. itself typically increases selectivity of this process by a few times.

The process that is the subject of this present invention makes use of our fundamental understanding of the mechanism of isotopically selective collisional vibrational relaxation to devise a two-laser isotope separation scheme that can make optimal use of this collisional phenomenon. Our experiments show that a selectivity of greater than 9000 can be achieved at room temperature and at pressures greater than 50 Torr.

SUMMARY OF THE INVENTION

An object of the invention is to provide a two-laser infrared multiphoton dissociation process for isotope separation that can produce highly isotopically enriched species in a single stage.

According to the invention, this object is achieved by the method as set out below.

In this method, the radiation of the first laser has a predetermined frequency to excite by a single-photon a low overtone vibrational transition of the polyatomic parent molecules, in particular a hydrogen stretch vibration, to produce vibrational pre-excited molecules at a well defined energy enriched in the desired isotope, for instance 13 C.

The radiation of the second laser has a predetermined frequency and predetermined energy fluence to induce selective dissociation of the vibrationally pre-excited excited molecules by infrared multiphoton excitation, in particular of a C-F stretch vibration.

The product of the pressure of the molecules and the time-lag Δt between the pre-excitation by the first laser pulse and the dissociation during the second laser pulse (which results from the effective length of the second laser pulse plus any time delay of the second laser pulse relative to the first laser pulse), is sufficiently high to allow collisional vibrational deactivation of a substantial amount of the vibrationally pre-excited molecules containing non-desired isotope(s), like 12 C, before dissociation of the vibrationally excited molecules occurs while having no significant collisional vibrational deactivation of the pre-excited molecules containing the desired isotope, like 13 C. The dissociation products are hence more highly enriched in the desired isotope as a result of collisions.

Collisions that occur between the two laser pulses and/or during the second pulse are hence used to increase significantly the isotopic selectivity.

As is described more fully below, the use of collisions to significantly increase the isotopic selectivity requires excitation by the first laser to a well defined energy of at least several thousand cm −1 . This is accomplished by direct, single photon excitation of a low overtone (Δv=2 or 3) of a hydrogen atom stretch vibration. The combination of low-overtone excitation by the first laser with isotopically selective collisional deactivation in the time between two laser pulses, followed by selective IRMPD of the pre-excited molecules induced by the second laser represents a unique feature of this invention.

This approach has several important advantages over other implementations of other IRMPD isotope separation schemes. First, vibrational overtone excitation with a continuously tunable laser can reach the maximum selectivity determined by the overlap in the spectra of two isotopic species, while conventional line-tunable CO 2 lasers cannot be sure to hit the point of minimum spectral overlap. Moreover, isotope shifts are in general greater for overtone transitions than for vibrational fundamentals. Secondly, overtone pre-excitation of a light atom stretch vibration can promote molecules directly to the vibrational quasicontinuum with a well defined energy, allowing the parameters of the dissociating laser to be optimized for this energy, preserving the isotopic selectivity gained in the first step. Because the IRMPD processes is applied to molecules already in the vibrational quasicontinuum, efficient dissociation occurs at relatively low (0.5-3 J/cm 2 ) CO 2 laser fluence, avoiding the need to focus the CO 2 laser beam. This permits a great increase of the irradiated volume, since collimated beams can be propagated together for meters, limited only by the beam divergence. While this is an important feature, it is not unique to our process. Most importantly, using vibrational overtone excitation for the first step, followed by collisions of the pre-excited molecules, enhances the isotopic selectivity of the process and at the same time allows higher working pressures where the density of molecules is higher, leading simultaneously to high selectivity (>99% isotopic purity) and reasonable productivity in a single stage process. This can make the process economically feasible and competitive with the current technologies.

Taken together, these factors indicate that the overtone excitation-IRMPD scheme according to the invention should provide a more efficient and selective means of laser isotope separation than previously developed MLIS schemes. The results described below demonstrate that this is indeed the case.

As mentioned above, isotope separation can operate with low fluence laser beams enabling interaction by multiphoton dissociation over a large volume. Consequently, the first and second laser beams can be collimated or slightly diverging or slightly converging beams of low fluence (≦5 J/cm 2 ) overlapping with one another over a substantial portion or all of their respective volumes containing the said polyatomic molecules. The first and second beams can have an angle of divergence/convergence less than 2.0×10 −3 rad.

The method according to the invention is particularly advantageous for separating 13 C isotopes from polyatomic molecules consisting of mostly 12 C isotopes and which contain C—H and C—F bonds, for example molecules of the formula HCF 2 X, wherein X is F, Cl, B or I. There have been a number of other papers and patents that share these working molecules but use a completely different process which does not use collisions to enhance selectivity. We do not claim this class of compounds for laser isotope separation in general, but only as suitable candidates under the specific conditions of our process.

In one example, the molecules are trifluoromethane HCF 3 , the frequency of the first laser is 8753±1 cm −1 or 8549±1 cm −1 , the frequency of the second laser is in the range 1020-1070 cm −1 , and the predetermined energy fluence of the second laser has a value in the range 0.5-5 J/cm 2 depending on the pulse shape of the second laser. Alternatively, for trifluoromethane HCF 3 , the frequency of the first laser is 5936.5±1 cm −1 or 5681±1 cm −1 .

Alternatively, the molecules are CF 2 HCl and the predetermined frequency of the first laser is 5911±5 cm −1 or 8693±2 cm −1 , or the molecules are monofluoromethane CH 3 F.

The first predetermined frequency can be produced by stimulated Raman scattering of narrowband tunable radiation of a solid state pulsed laser and the second predetermined frequency is produced by a pulsed CO 2 laser.

The method of the invention can also be applied to the separation of isotopes from other molecules including SiH 4 , SiF 3 H, SiCl 3 H, GeH 4 , and alcohols of the formula R—OH, where R=CH 3 , C 2 H 5 , C 3 H 7 or C 4 H 9 .

The overlapping first and second laser beams can be substantially parallel or can multiple intersect.

BRIEF DESCRIPTION OF DRAWINGS

In the drawings, given by way of example:

FIG. 1 is an energy level schematic for the isotope separation method according to the invention;

FIG. 2 is a diagram of an apparatus used to carry out the isotope separation method according to the invention;

FIG. 3 shows a typical mass-spectrum of C 2 F 4 products enriched with 13 C isotopes by the method according to the invention;

FIG. 4 shows the concentration of 12 C and 13 C in the C 2 F 4 dissociation product obtained in an experimental set-up, as a function of the pressure at zero time-delay between the two pulses (time-lag Δt≈35 ns); and

FIG. 5 shows the concentration of 12 C and 13 C in the C 2 F 4 dissociation product and the percentage of 13 C in the C 2 F 4 dissociation product as a function of time-delay between the overtone excitation laser pulse and the CO 2 laser pulse.

DETAILED DESCRIPTION

This description uses the example of CF 3 H as a working molecule for 13 C isotope separation, although the applicability of the invention is broader in terms of both potential working molecules as well as isotopes that can be separated. The method according to the invention comprises two steps, as shown schematically in FIG. 1 . In the first step, a near-infrared laser pulse pre-excites molecules containing primarily the desired isotope via a low (Δv=2 or 3) overtone transition of the CH stretch vibration. Following this, a CO 2 laser pulse excites a C-F stretch of only the pre-excited molecules, selectively dissociating them by IRMPD, producing CF 2 +HF. The isotopically selected CF 2 radicals are collected after combining to form C 2 F 4 .

Using a narrow bandwidth, continuously tunable laser for pre-excitation, we can choose the excitation frequency to be in exact resonance with an overtone transition of the desired isotopic species, provided the rotational structure of the overtone band is at least partially resolved. Even if the isotopic spectral shift is relatively small, given sufficient resolution it is possible to find parts of the spectrum where there is a minimum of overlap between the different isotopic species. In the case of 13 C isotope separation using CF 3 H as a parent molecule, this pre-excitation usually produces at least 80% 13 C isotopes and at most 20% 12 C isotopes, possibly about 90% 13 C isotopes and about 10% 12 C isotopes.

The second laser beam, the role of which is to dissociate pre-excited molecule via IRMPD, is arranged with a pulse length and/or delay relative to the first laser beam sufficient to allow a substantial amount of vibrationally excited molecules containing 12 C to relax by vibrational state changing collisions to an energy from which they will not be efficiently dissociated. During this time, a substantial amount of the molecules containing 13 C remain excited and available to be dissociated by the second laser pulse. Achieving enrichment to >99% in 13 C isotopes in a single pass requires a substantial difference in the collisional vibrational relaxation rates of pre-excited molecules of different isotopic composition. It is not evident from the published literature that the vibrational relaxation rates of highly vibrationally excited molecules should show a strong isotope dependence. Previous studies of diatomic and triatomic molecules have shown that the detuning of energy levels upon isotopic substitution is sufficient to change the energy gap and hence the energy transfer rate in molecules excited with one quantum of vibrational excitation. For example, Stephenson et al. (J. Chem. Phys. 48, 4790 (1968)) have shown that for the following mixed isotope collisions in CO 2 ,

12 C 16 O 16 O (v 3 =1)+ 12 C 16 O 18 O (v 3 =0)→ 12 C 16 O 16 O (v 3 =0)+ 12 C 16 O 18 O (v 3 =1)

ΔE=18 cm −1

12 C 16 O 16 O (v 3 =1)+ 13 C 16 O 16 O (v 3 =1)→ 12 C 16 O 16 O (v 3 =0)+ 13 C 16 O 16 O (v 3 =1)

ΔE=66 cm −1

the more than three-fold difference in the energy gap results in a difference in vibrational deactivation rate of a factor of 3.

One cannot expect a similar fractional difference in the energy gap to occur in a polyatomic molecule excited to a vibrational overtone level, however. For example, consider the specific case of a CF 3 H molecule prepared in the v CH =3 level prior to collision with other CF 3 H molecules that are vibrationally unexcited. For the 13 C species colliding with unexcited 12 C molecules, the energetics of this process would be given by

13 CF 3 H(v CH =3)+ 12 CF 3 H(v CH =0)→ 13 CF 3 H(v CH =2)+ 12 CF 3 H(v CH =1)

ΔE=−254 cm −1

while for vibrationally excited 12 C molecules colliding with unexcited 12 C molecules one would have

12 CF 3 H(v CH =3)+ 12 CF 3 H(v CH =0)→ 12 CF 3 H(v CH =2)+ 12 CF 3 H(v CH =1)

ΔE=−240 cm −1

The difference in the energy gap of these two processes comes from the isotope effect on the difference in the v=3 to v=2 energy spacing. For both isotopic species, the 3→2 de-excitation process will be a few hundred cm −1 out of resonance with the 0→1 excitation process because of the large anharmonicity of the C-H stretch. The small fractional difference in the energy deficit of the process between the 13 C and 12 C species is significantly less than the simple case of CO 2 shown above. It is difficult to imagine that such a small fractional difference in energy gap could lead to a substantial difference in vibrational energy transfer rates in CF 3 H.

The key to understanding the phenomenon of collisional enhanced isotope selectivity exploited in this invention is to realize that the initially prepared state of CF 3 H is not a pure v CH =3 stretch state. Vibrational state-mixing in these molecules has been extensively studied and is well understood (J. Segall et al., J. Chem. Phys. 86, 634 (1987)). Each overtone level with N CH stretch quanta is characterized by a polyad of N+1 strongly coupled states consisting of combinations of stretch and bend excitations. Thus for N=3, one has the group of coupled states (designated |v s , v b > for the number of stretch and bend quanta respectively) |3,0>, |2,2>, |1,4>, and |0,6>. The eigenstates in this energy region can be expressed as linear combinations of these zeroth-order states:

|j>=c 1 |3,0>+c 2 |2,2>+c 3 |1,4>+c 4 |0,6>

Higher resolution spectroscopy of the CH stretch overtone levels in this molecule reveals that these “first-order” states are weakly coupled to other nearby states (O. V. Boyarkin and T. R. Rizzo, J. Chem. Phys. 105, 6285 (1996)). In the limit that all the available states within a small energy region about the v CH =3 level are coupled in some measure, a state in this energy region can be represented as the following mixture $$ &LeftBracketingBar; j ⟩ = c 3 ⁢ &LeftBracketingBar; 3 ⟩ ⁢ &LeftBracketingBar; 0 ⟩ ⁢   ⁢ &LeftBracketingBar; 0 ⟩ ⁢   ⁢ … ⁢   ⁢ &LeftBracketingBar; 0 ⟩ + &LeftBracketingBar; 2 ⟩ ⁢   ⁢ ∑ i   ⁢   ⁢ c 2 ⁢ i ⁢   ⁢ ∏ k = 2 9 ⁢   ⁢ &LeftBracketingBar; v ki ⟩ + &LeftBracketingBar; 1 ⟩ ⁢ ∑ i   ⁢   ⁢ c 1 ⁢ i ⁢   ⁢ ∏ k = 2 9 ⁢   ⁢ &LeftBracketingBar; v ki ⟩ + &LeftBracketingBar; 0 ⟩ ⁢ ∑ i   ⁢   ⁢ c 0 ⁢ i ⁢   ⁢ ∏ k = 2 9 ⁢ &LeftBracketingBar; v ki ⟩

where the zeroth-order states are grouped by their number of CH stretch quanta. A simple state count would reveal that the most numerous zeroth-order states in this mixture are those with 0 quanta of CH stretch mode. In fact, the average occupation number of all the different vibrational modes of such a mixed state with ˜9000 cm −1 of energy is between 0 and 1.

Let us now consider the effect of state-mixing on the collisional energy transfer process. In a sense, one can expect this mixed state to behave more like a molecule with one quantum of vibration. In this case, the relevant processes for us to consider are now

13 CF 3 H (v=1)+ 12 CF 3 H (v=0)→ 13 CF 3 H (v=0)+ 12 CF 3 H (v=1)

and

12 CF 3 H (v=1)+ 12 CF 3 H (v=0)→ 12 CF 3 H (v=0)+ 12 CF 3 H (v=1)

where the vibrational quantum number v could refer to any vibrational mode. This process will be more nearly resonant as in the case of CO 2 presented above. For collisions between 12 C containing molecules, the only energy deficit will be due to the cumulative effect of the off-diagonal anharmonicities between the C-H stretch mode and the other modes of the molecule. For collisions between 13 C and 12 C containing molecules, the energy deficit will be the sum of the these off-diagonal anharmonicities together with the isotope shifts. For every mode, the isotope shift will make the collisions of 13 C containing molecules be further off-resonant than the 12 C containing molecules. Since most off-diagonal anharmonicities tend to be small, the isotope shift should represent a substantial fraction of the energy gap for the energy transfer process. In view of the steep dependence of the v—v transfer rate on the energy gap, this would lead to a substantial difference in relaxation rates for vibrationally excited 13 CF 3 H and 12 CF 3 H.

This model for the detailed mechanism of the difference in collisional relaxation rates between two isotopomers allow us to chose conditions in which we can fully exploit this phenomenon for isotope separation. In order to use collisional enhancement of the isotopic selectivity effectively, the first laser excitation step needs to fulfil several requirements:

(1) It must pre-excite molecules to a sufficiently high energy such that the vibrational states are substantially mixed in the sense described above. All molecules exhibit such state-mixing, but the energy at which it occurs is molecule specific.

(2) It must pre-excite molecules to a sufficiently high energy such that the second laser can selectively dissociate pre-excited molecules and not unexcited molecules.

(3) It must pre-excite molecules to a well defined energy, such that the parameters of the second laser can be optimized to dissociate selectively those molecules that have not undergone substantial vibration energy relaxation by collisions. This requires single-photon and not multiple-photon excitation.

If condition (1) is not fulfilled, the collisional vibrational relaxation rates of the two isotopes of a highly excited molecule would not be sufficiently different to enhance the selectivity of the isotope separation process. If either condition (2) or condition (3) is not fulfilled, the second laser will not be able to selectively dissociate the desired isotope, even if the vibrational relaxation rates are different. Previous two-laser isotope separation techniques do not fulfill these requirements and hence are not able to achieve both high selectivity and reasonable productivity simultaneously.

The following Example illustrates experiments underlying the invention and its implementation on laboratory scale.

EXAMPLE

Exemplary of this new approach for isotope separation, trifluoromethane, CF 3 H, is used as the parent molecule for 13 C isotope separation. Trifluoromethane was chosen for the following reasons:

(1) It has an IR active light atom vibration associated with the carbon atom ( 13 C-H stretch, υ 1 =3025.3 cm −1 ), through which a substantial amount of vibrational energy can be deposited into a molecule via a low overtone transition. The isotopic shift in the 3υ 1 band of 39.7 cm −1 is appreciable.

(2) The vibrational states accessed by a CH stretch overtone excitation are substantially mixed.

(3) It has another IR active vibration with a fundamental frequency shifted to the high frequency side of a conventional CO 2 laser ( 13 C-F stretch, υ 5 =1132.4 cm −1 ). One can expect that the optimum frequency for IRMPD of vibrationally excited molecules lies within tuning range of CO 2 laser.

(4) The IRLAPS detection technique has been successfully implemented for studying the overtone spectroscopy of this molecule, and we know that IRMPD of vibrationally excited CF 3 H can be fulfilled with high selectivity. See for example papers by Boyarkin/Settle/Rizzo in Ber. Bunseges. Phys. Chem. 99, 504 (1995) and by Boyarkin/Rizzo in J. Chem. Phys. 105, 6285 (1996).

(5) The lowest dissociation channel for CF 3 H produces CF 2 and HF. Because the CF 2 fragment is the same as that from the IRMPD of CF 2 HCl, we can take advantage of the large body of work on CF 2 collection in this highly studied system.

FIG. 2 schematically illustrates an experimental apparatus, comprising a cylindrical glass cell 1 (which is shown in two-dimensional projection, 2 cm in diameter, 50 cm long) with BaF 2 windows filled with CF 3 H to a specific pressure as measured by a capacitance manometer.

A 20-25 mJ pre-excitation laser pulse 2 , generated by Raman shifting a 90 mJ pulse from a Nd:YAG pumped dye laser (Spectra Physics GCR-270, Lumonics HP 500) in high pressure H 2 , promotes CF 3 H molecules in cell 1 at 7.5 mbar pressure to the 3υ 1 (CH stretch) level via the Q-branch at 8753 cm −1 . This laser pulse 2 , which has a duration of 5-6 nanoseconds, is focused into the center of cell 1 by an F=+120 cm lens 4 , giving an estimated maximum fluence of 1-2 J/cm 2 at the beam waist.

After a fixed delay, a pulse 3 from the CO 2 laser (Lumonics, TEA-850) arrives and selectively dissociates the vibrationally pre-excited molecules via IRMPD to produce CF 2 and HF. This beam 3 is first truncated to a size of 5-10 mm diameter by passing it through an adjustable iris and then focused by an F=+75 cm lens 5 to a 2×2 mm 2 beam waist. Its fluence can be varied over a wide range. This pulse consists of a peak of 150 ns FWHM followed by a 2-3 μsec tail carrying more than 60% of the total pulse energy. The two laser beams 2,3 (i.e. the pre-excitation laser and CO 2 laser) are combined on a 10 mm thick BaF 2 Pellin-Broca prism 6 and enter the cell 1 from the same side.

The CF 2 dissociation products eventually recombine to form C 2 F 4 , which is sampled from cell 1 at 7. The relative concentrations of the C 2 F 4 with different carbon isotopes are measured in quadrupole mass-spectrometer 8 at atomic masses 100 ( 12 C 2 F 4 ), 101 ( 12 CF 2 13 CF 2 ) and 102 ( 13 C 2 F 4 ). The productivity of the process for 13 C is determined as twice the integral of the signal at mass 102 plus the integral of signal at mass 101 . Correspondingly, the productivity of 12 C is twice the integral of the signal at mass 100 plus the integral of signal at mass 101 . The percentage of 13 C is determined as a ratio of the 13 C productivity to the total productivity of the process.

FIG. 3 represents a typical mass-spectrum of C 2 F 4 products generated by the procedure described above using the apparatus of FIG. 2 . The pressure of the CF 3 H sample in cell 1 is 7.5 mbar, the time-delay between the end of the pre-excitation pulse 2 and the beginning of the dissociating CO 2 laser pulse 3 is 75 ns, and the CO 2 laser fluence is adjusted to 2.5 J/cm 2 . The spectrum has been obtained after 10 min irradiation of the sample with the lasers operating at 10 Hz laser repetition rate. The observed ratio of signals at masses 100 - 102 corresponds to >99% 13 C concentration in the C 2 F 4 product. The estimated dissociation yield for each pair of laser pulses is 1.5-3% of all 13 CF 3 H molecules within the irradiated volume. These results represent a record in 13 C performance by MLIS.

The importance of collisional vibrational deactivation of CF 3 H molecules pre-excited to the 3υ 1 level is illustrated by FIGS. 4 and 5 .

FIG. 4 represents the pressure dependence of the concentration of 12 C (triangles, left hand scale; experiment (a)) and 13 C (squares, left hand scale; experiment (b)) in the C 2 F 4 dissociation product. In experiment (a), the wavelength of the first laser was tuned such that the amount of pre-excited 12 CF 3 H at 10 mbar is about the same as the amount of pre-excited 13 CF 3 H in experiment (b) at 10 mbar.

The experiments of FIG. 4 have been performed at zero time-delay between the two laser pulses and with at CO 2 laser fluence of 7 J/cm 2 . In this case, the time-lag Δt=about 35 ns, as explained below for FIG. 5 . All other parameters are as indicated above. One can see that the peak concentration of 12 C and 13 C in the C 2 F 4 occurs at different pressure, and this is a result of more rapid collisional vibrational deactivation of the pre-excited 12 CF 3 H as compared to 13 CF 3 H during the CO 2 laser pulse. The total number of the pre-excited molecules grows linearly with pressure, but competition with collisional deactivation reduces the number of pre-excited CF 3 H that can be dissociated at a given fluence of the CO 2 laser. Consequently, the amount of C 2 F 4 produced initially increases with increasing pressure but then drops. The rate of this drop is different for different carbon isotopes.

FIG. 5 shows the concentration of 12 C (triangles, left hand scale; experiment (a)) and 13 C (squares, left hand scale; experiment (b)) in the C 2 F 4 dissociation product and the percentage of 13 C in the C 2 F 4 dissociation product (circles, right-hand scale; experiment (b)) as a function of time-delay between the overtone excitation laser pulse and the CO 2 laser pulse. In experiment (a), the wavelength of the first laser was tuned such that about the same amount of 12 CF 3 H was pre-excited as the amount of 13 CF 3 H in experiment (b). Because these measurements are made in two different experiments, the percentage of 13 C cannot simply be calculated from the concentrations of both isotopes shown in the Figure.

FIG. 5 illustrates the productivity for the two isotopes and the percentage of 13 C in the C 2 F 4 product as a function of the time-delay between the two laser pulses at 5 mBar CF 3 H pressure and CO 2 laser fluence of 3.5 J/cm −2 . It is clear that the percentage of 13 C in C 2 F 4 grows with increasing delay up to about 200 ns—that is with increasing number of vibrationally deactivating collisions. One can see that because of this deactivation, the total amount of produced C 2 F 4 drops as a function of time delay, but it does so at different rates for the two different isotopes of carbon.

Thus, both the final percentage of 13 C in the C 2 F 4 product and the productivity of the 13 C separation process can be controlled by the parameter P·Δt, where P is CF 3 H pressure and Δt is a time between the pre-excitation by the first laser pulse and the dissociation during the second laser pulse. This implies a time-delay between the two laser pulses and an effective duration of the dissociating pulse (i.e., the time the pulse is on before dissociation, which we have determined to be about 35 ns for the CO 2 laser pulse shape used here). Production of C 2 F 4 highly enriched in 13 C therefore requires this parameter, P·Δt, to be large enough for near all vibrationally pre-excited 12 CF 3 H to be collisionally deactivated. For a quantitative estimate of the optimal parameter P·Δt, values for the vibrational deactivation constants from the 3υ 1 level have been determined experimentally to be about 13 K 3 =1.5 μs mbar and 12 K 3 =4.5 μs mbar for 13 CF 3 H and 12 CF 3 H respectively.

Another aspect illustrated by this example is the relatively low fluences of the pump and the dissociating radiation required for the process to be highly selective while retaining a reasonably high level of productivity. This permits the use of collimated rather than focused laser beams, provided the pulse energies of two lasers are high enough. This allows the volume in which both laser energy fluences are in the optimal range to be much larger than what can be achieved with focused beams.

The difference between collimated and focused beams is illustrated as follows. Suppose the pre-excitation laser delivers 0.5 J pulse energy in an about 8 mm beam with divergence 2·10 −3 . This gives an energy fluence of about 1 J/cm 2 . This can be achieved, for example, by stimulated Raman scattering of output of an alexandrite solid state laser in high pressure H 2 . A dissociating beam of the same diameter and divergence and with 2-3 J/cm 2 energy fluence can be produced by a TEA CO 2 laser. These two beams can be overlapped for a length of up to 3.3 meters before the fluences will drop to a half of their initial values because of divergence. This gives about a 0.25 liter irradiated volume where the process occurs. After each pass, the beams can be slightly recollimated and sent again to the reactor to increase the active volume by several times. This can be compared with a typical active volume of a few mm 3 achieved in experiments with focused laser beams. Thus, the low laser fluences required for the described process allows irradiation of large volumes by collimated beams.

Claims

1. A method of separating a desired isotope from a starting material comprising polyatomic molecules containing different isotopes, by applying to said molecules in the gas phase, at a predetermined pressure P, infrared radiation of a first pulsed laser to excite by a single photon an overtone transition of a light-atom stretch vibration of said molecules to produce vibrationally pre-excited molecules enriched in the desired isotope, and, infrared radiation of a second pulsed laser of different frequency to induce a chemical reaction resulting in a molecule constituting a dissociation product, enriched in the desired isotope, which can be separated from the starting material, wherein:the radiation of the first laser has a predetermined frequency optimized to select the desired isotope via a single-photon transition; the first laser pre-excites the polyatomic molecules to an excited vibrational level that is high enough in energy to allow an increase in isotopic selectivity by collisions; the radiation of the second laser selectively induces dissociation of the vibrationally pre-excited molecules by infrared multiphoton excitation, at a time Δt after the pre-excitation by the first laser pulse; the pressure, P, and the time, Δt, between the pre-excitation by the first laser pulse and the dissociation during the second laser pulse have a pressure-time product P·Δt, the magnitude of this pressure-time product P·Δt being sufficiently high to allow collisional vibrational deactivation of a substantial amount of the vibrationally pre-excited molecules containing non-desired isotope(s) before dissociation of the vibrationally excited molecules occurs while achieving significantly less collisional vibrational deactivation of the pre-excited molecules containing the desired isotope; whereby said collisions that occur in said time Δt between pre-excitation by the first laser pulse and the dissociation during the second laser pulse enhance the isotopic selectivity of the isotopic separation method, such that said dissociation products is highly enriched in the desired isotope.

2. The method of claim 1 wherein the polyatomic molecules of the starting material consist mostly of 12C isotopes and further contain 13C isotopes, said polyatomic molecules containing C—H and C—F bonds, said method separating the 13C isotopes from the polyatomic molecules.

3. The method of claim 2, wherein the polyatomic molecules of the starting material are of the formula HCF2X, where X is F, Cl, B or I.

4. The method of claim 2, wherein the polyatomic molecules of the starting material are trifluoromethane CHF3 and the predetermined frequency of the first laser is 8753±1 cm−1 or 8549±1 cm−1 and the predetermined frequency of the second laser is in the range 1020-1070 cm−1.

5. The method of claim 4, wherein the first laser produces at least 80% of said pre-excited molecules containing 13C and at most 20% of said pre-excited molecules containing 12C and the magnitude of said pressure-time product P·Δt is sufficient to achieve said dissociation products enriched to >95% in 13C isotopes for each pair of the laser pulses.

6. The method of claim 5, wherein the first laser produces about 90% of said pre-excited molecules containing 13C and at most 10% of said pre-excited molecules containing 12C and the magnitude of said pressure-time product P·Δt is sufficient to achieve said dissociation products enriched to >99% in 13C isotopes for each pair of the laser pulses.

7. The method of claim 2, wherein the polyatomic molecules of the starting material are trifluoromethane HCF3 and the predetermined frequency of the first laser is 5936.5±1 cm−1 or 5681±1 cm−1.

8. The method of claim 2, wherein the polyatomic molecules of the starting material are CF2HCl and the predetermined frequency of the first laser is 5911±5 cm−1 or 8693±2 cm−1.

9. The method of claim 2, wherein the polyatomic molecules of the starting materials are monofluoromethane CH3F.

10. The method of claim 1, wherein the polyatomic molecules of the starting material are selected from SiH4, SiF3H, SiCl3H, GeH4 and alcohols of the formula R—OH where R=CH3, C2H5, C3H7 or C4H9.

11. The method of claim 1 wherein the first and second lasers are collimated or slightly diverging or slightly converging beams of low fluence overlapping with one another over a substantial portion or all of their respective volumes containing the polyatomic molecules.