Generation and Analysis of Tritium-Substituted Methane

Abstract An unavoidable category of molecular species in large-scale tritium applications, such as nuclear fusion, are tritium-substituted hydrocarbons, which form by radiochemical reactions in the presence of (circulating) tritium and carbon (mainly from the steel of vessels and tubing). Tritium-substituted methane species, CQ4 (with Q = H,D,T), are often the precursor for higher-order reaction chains, and thus are of particular interest. Here we describe the controlled production of CQ4 carried out in the CAPER facility of the Tritium Laboratory Karlsruhe, exploiting catalytic reactions and species enrichment via the CAPER integral permeator. CQ4 was generated in substantial quantities (>1000 cm3 at ~850 mbar, with CQ4content of up to ~20%). The samples were analyzed using laser Raman and mass spectrometry to determine the relative isotopologue composition and to trace the generation of tritiated chain hydrocarbons.


I. INTRODUCTION
Hydrogen chemistry is very versatile and varied, and hydrogen ( 1 H ≡ H) can form molecular compounds with almost all elements in the periodic table.These are encountered in the form of ionic-bond hydrides, but also as covalent-bond compounds, like, e.g., hydrocarbon molecules, water, ammonia, etc., and in multiple metallic interactions.Hydrogen also participates in the adsorption and absorption processes, occupying interstitial voids in crystalline and metallic networks.This wealth of interactions is due to its small size and its electronic structure; note that the chemistry of its isotopes-deuterium ( 2 H ≡ D) and tritium ( 3 H ≡ T)-is equivalent, in principle.
However, for tritium (the radioactive isotope of hydrogen), one must also consider the radiochemical reactions that occur as a consequence of its β-decay 3 H → 3 He + + e − + v − e .The highly energetic decay products, i.e., the helium ions and the β-electrons, lead to the generation of a wealth of secondary ions and radicals through impact interactions with any surrounding atoms/molecules. 1 This peculiarity of tritium modifies its chemical equilibria and reaction products with respect to the two nonradioactive isotopes (H and D), as well as the chemical kinetics of the reactions in which it participates. 2Because of this, the products of these reactions and their concentrations are difficult to predict, in general.
In large-scale tritium applications, such as the Karlsruhe Tritium Neutrino (KATRIN) experiment 3 and the future nuclear fusion experiment ITER, as well as its successor experiments, nonnegligible amounts of shortchain tritiated hydrocarbons are generated.By and large, they contain between one and five C-atoms, which have been detected in circulating Q 2 process gases (with Q ∈ [T, D, H]).This occurs despite the fact that these systems have gas-filtering units, such as permeation cells (or permeators), to remove unwanted non-Q 2 contaminants.
The carbon source for this type of reaction is C-extraction from the steel pipes/vessel walls, with the reactions with tritium taking place on the steel surfaces; note that it is rather unlikely that carbon originates from leaked atmospheric CO 2 or CO after passing through the permeator.Contaminating molecules need to be removed from the process gases because they may cause operational problems in large installations, be deposited in inappropriate places, or they may affect other processes, such as, e.g., the formation of plasmas. 4,5In order to set up the means for the removal/elimination of hydrocarbon contaminants, it is necessary to study their chemistry and to determine their molecular structure, as well as the reactions that these molecules undergo among themselves with the surfaces and other parts of the systems exposed to them.
A good starting point for any in situ/inline study of the tritium-carbon chemistry is the generation of tritiumsubstituted hydrocarbons in sufficient quantities to allow one to perform precision spectroscopy on them, and thus obtain relevant analytical knowledge.The tritiumsubstituted methane species, CQ 4 (with Q ∈ [T, D, H]) are of particular interest since, in general, they constitute the dominant species in tritium circulation systems. 5It should also be noted that CQ 4 may serve as precursors for forming chain hydrocarbon molecules.
Despite being studied for quite some time, detailed spectroscopic knowledge about tritiated methanes is still sparse; this is because the available information corresponds to very low concentrations and very small sample quantities (see, e.g., Refs.1,6, and 7).In those experiments, mixtures of methane and tritium were left to react for up to a few days, and the final products were analyzed by mass spectrometry and Raman spectroscopy.
In order to overcome the problem of only minute amounts of tritiated CQ 4 , we utilized their production in sufficient quantities for high-resolution laser Raman spectroscopy in the CAprice PERmcat (CAPER) facility 8 at the Tritium Laboratory of Karlsruhe (TLK).Quantities of >1000 cm 3 at ~850 mbar, with a CQ 4 content of up to the order of ~20% (conservative estimate), were produced under controlled conditions.In this paper, we describe the production procedure and the detailed analysis of the samples using mass spectrometry and Raman spectroscopy.

II.A. System Fundamentals
The aim of this experiment was to produce CQ 4 as pure as possible and in sufficient quantities for precision analysis.Based on the aforementioned experiments, we knew that the production of a chemically pure compound was not possible when relying on the hypothetical isotope exchange reaction in a simple binary mixture of gases, i.e., The reactions that actually take place are chemical equilibria that result in a set of products in which the H-atoms in methane are gradually replaced by T-atoms.A quasistable state is reached when the concentrations remain relatively constant over long periods of time, mainly consisting of the chemical family of CQ 4 , in our case Q ∈ [T,H]: Unfortunately, in the presence of tritium, equilibria are not stable but continue to evolve over time due to the tritium activity.If any mixture is stored for a long time, e.g., 18 months in the case of the experiments carried out by McConville and Menke, 1 the observation is that the tritium and carbon concentrations in the gas phase diminish and "deposits" on the surfaces of the sample containers occur.In fact, one could argue that hydrocarbons "polymerize" at the surface.Indeed, such deposits were observed in our experiments, specifically on the windows of the Raman cell used in the analysis setup, although they appeared over much shorter timescales than those reported by McConville and Menke. 1 Note, however, that little substantiated information is available on the nature of such deposits.
Unlike in previous experimental work, we deliberately altered the chemical equilibria to shift and direct them toward the product of highest interest, namely CT 4 .As such, our procedure was completely different than that utilized in earlier experiments.Instead of relying on "passive" exchange reactions and associated equilibration, we used the CAPER facility to perform "catalysisdriven" synthesis.
Here it is worthwhile to recall some construction and operational features of CAPER; for details see, e.g., Ref. 8. CAPER was conceived as the experimental verification of the processing of the tokamak exhaust processing concept for ITER, whose absolute requirement was that the maximum tritium concentration in the flue gas is less than ~3.7 × 10 10 Bq/m 3 .Note that in terms of tritium concentration, this corresponds to the reduction of the tritium concentration by a factor of ~10 6 .CAPER achieves this in three steps in batch operation.In the first step, the effluent gases pass through a Pd/ Ag membrane that is 100% selective for Q 2 in the permeator, where Q 2 (with Q ∈ [T, D, H]) is separated from molecular impurities.Two effluent streams are generated, one containing nearly all of the Q 2 (for later enrichment of tritium), and one in which the non-Q 2 contaminants are removed.In the second step, these contaminants are processed in a closed loop using a combination of heterogeneous catalysis and permeation to extract the Q 2 resulting from the catalysis.The third step is carried out in the PERMCAT reactor, 9 which simultaneously combines the permeation through the Pd/Ag membrane and the catalytic bed in a so-called "isotopic swamping" process, as used in tritium processing for the removal of impurities from the process gas. 9,10he use of the CAPER facility allowed us to incorporate heterogeneous catalysis into the synthesis of CQ 4 through the isotope exchange reactor containing a suitable catalyst.A second important feature of CAPER is that it provides the capability to regulate the flow rates of reagents fed into the reactor.In addition, the permeator provides the possibility of selectively eliminating products of the isotopic exchange reaction, namely Q 2 .In this way, one is able to move the reaction equilibrium in the direction of interest, i.e., ultimately increasing the concentration of CQ 4 in the final mixture.

GENERATION AND ANALYSIS OF TRITIUM-SUBSTITUTED METHANE
Among the other interesting features, catalysts have the ability to direct the reaction toward certain products as well as to improve the reaction kinetics, thus significantly reducing reaction times.The catalyst used in the study described here is a commercial nickel catalyst on Kieselguhr (a porous diatomaceous earth substrate used as a support for catalysts to maximize a catalyst's surface area and activity), with an approximate nickel content of 50%.In general, this type of catalyst is used for methane cracking.
The proposed mechanism of action of the aforementioned type of catalyst is the formation of carbon filaments on the active centers of the catalyst. 11s methane comes into contact with the catalyst, carbon chains grow, keeping nickel at the tip, which sustains the catalysis.The diffusion of carbon on the surface of the catalyst generates carbon deposits that end up deactivating its catalytic action.The operational recovery of this type of catalyst is carried out using hydrogen, among other substances.Hydrogen adsorbs reversibly on nickel, occupying active centers and displacing carbon.It competes with it for the active centers and decreases the efficiency of the cracking reaction, and thus the carbon deposits.Exploiting these two processes, the chemical environment of the catalyst can be modified to either activate or reduce it.Note that, in our case, we used tritium instead of protium.
Hence, if the catalyst has been previously saturated with T 2 , CH 4 will be retained at its surface; therefore, the isotope exchange reactions will be favored instead of the cracking reactions, 10 thus yielding CT x H (4-x) .

II.B. Sample Preparation
Before starting each step of the synthesis, the catalyst has to be activated and reduced so that it is in optimal condition to produce the exchange reaction.For this, a stream of high-purity T 2 (>97%) is passed through the system for at least 18 h at a temperature of 375°C to 400°C.In order to try to minimize the presence of longer-chain hydrocarbons in our samples, the synthesis process proceeds in several stages, as can be seen in the three parts of Fig. 1.The key elements involved in the synthesis are (1) two vessels for high-purity tritium and for the collected reaction mixture, respectively, (2) the permeator, and (3) the exchange reactor.Note that the valves highlighted in green are open and allow for the passage of gases, and those highlighted in white are closed, and thus block passage.Note also that the valve marked SCV constitutes a unidirectional stop-check valve assembly with backflow protection to prevent contamination of the T 2 vessel.
The reaction mixture obtained in each step is retained, as it becomes the starting reagent for the following step.Note that here we refer to the last synthesis run in a series of optimizing trials; the data shown in this paper are from this specific attempt.The relevant CAPER configurations for the substep producing samples 1, 2, and 3 are shown in the top, center, and bottom parts of Fig. 1, respectively.
Accordingly, the starting sample 1 was prepared using a flow rate of CH 4 of ~0.25 mol/h (of purity >99.99%).This flow was mixed with tritium prior to entering the reactor; its flow rate was ~4 mol/h, with tritium purity >97%.The reactor feed was maintained for ~1 h at a temperature of 375°C to 400°C.
In continuous operation, the reaction products were pumped to the permeator where the compounds Q 2 = T 2 , HT, and H 2 were separated from the mixture.The non-Q 2 remainder was transferred to the CQ 4 vessel and stored for the next stage of the synthesis.
Note that with the partial extraction of Q 2 from the mixture, the exchange reaction becomes favorable for the generation of CQ 4 because (1) it shifts the equilibrium toward the products, avoiding the backward reaction toward the reformation of CH 4 [see Eq. ( 2   volume of the sample while increasing the fraction of CQ 4 in the sample.Note also that in order to keep sample 1 at tritium saturation conditions and to favor the isotopic exchange reaction, for storage it was diluted to 50% with high-purity T 2 (>97%).
The production of sample 2 starts with the diluted sample 1; the relevant gas circulation path is shown in the center part of Fig. 1 (note that in this procedural step the permeator is bypassed).In this second step, the gas mixture passes only through the reactor; the reaction time and temperature conditions of the reactor were kept the same as in the first step.All gases leaving the reactor were collected in the CQ 4 vessel to be used in the third step of the synthesis.
The path for the production of sample 3 is shown in the lower part of Fig. 1.The mixture from the previous stage (sample 2) was fed back into the exchange reactor and more tritium was added at a flow rate of ~6 mol/h.Circulation was maintained for ~3 h, keeping the same temperature conditions as before.Note that now, with the permeator in the loop once more, the reaction products were separated into two effluents again; i.e., Q 2 and CQ 4 are directed toward their respective vessels for storage.
Overall, about 0.21 mol of product gas mixture was generated, which is equivalent to ~5 × 10 3 cm 3 for standard pressure and temperature conditions.This was more than enough to conduct an extensive series of analytical measurements to characterize the gas composition and to record high-resolution Raman spectra.

II.C. Measurement Protocol
At each step of the synthesis, aliquots of the three produced samples were transferred to tritium-compatible Raman cells and sample cylinders; these were transferred to the TRIHYDE facility of the TLK for analysis.Note that TRIHYDE is a tritium-compatible system designed for gas mixing and analysis.It incorporates a range of measurement system devices, including high-resolution laser Raman spectroscopy and mass spectrometry. 12The samples were measured immediately after they were prepared and were not stored; thus they should constitute a reliable snapshot of the concentrations and majority species at each stage of the synthesis.
First, the Raman spectra were measured using the TRIHYDE-incorporated laser Raman (LARA) measurement instrument.This is assembled around a glove-box appendix 13 containing the Raman sample cell.It comprises a green diode-pumped solid-state laser (GEM, Laser Quantum), with an emission wavelength of λ L = 532 nm, continuous output power of P L = 2 W, and Raman light collection at 90 deg with fiber-bundle coupling to the Raman spectrometer (SP2150 spectrograph with Pixis400B CCD detector, Princeton Instruments).
All of our Raman cells had the same volume of ~7 cm 3 ; they were filled to around ~850 mbar of sample gas.Note that for the Raman spectral data shown in Fig. 2, 20 spectra of 90-s accumulation time each were recorded (and averaged) in the spectral range of interest of 1500 to 4200 cm -1 .
After completion of the Raman measurements, the sample gas was bled from the cell into the mass spectrometer (quadrupole residual gas analyzer HPQ3S, MKS Instruments Inc.) via inlet dosing (UDV 040, Pfeiffer); the measurement pressure was adjusted to about 10 −4 mbar.In the instrument range 1 to 100 atomic mass units (amu), 50 individual spectra of 30-s accumulation each were recorded (and averaged); these mass spectral data are discussed later in Sec.III.B.

III.A. Raman Spectroscopy
The sequence of spectra from the start of the catalytic CH 4 + T 2 synthesis (sample 1) to its end (sample 3), shown in Fig. 2, reveals three distinct series of features, which we discuss separately.
First, all three spectra contained the characteristic Q 1 branch lines of the molecular hydrogen isotopologues expected for this type of synthesis, i.e., from T 2 (at ~2466 cm -1 ), HT (at ~3435 cm -1 ), and H 2 (at ~4159 cm -1 ).Their amplitudes differed significantly, indicating the evolution of concentrations for these molecular constituents.While T 2 was present in the starting mixture by default, HT and H 2 are products generated during the synthesis reactions.The source for the generation of these hydrogen isotopologues was CH 4 , in which H-atoms were substituted by T-atoms in the course of the isotope exchange reactions.Thus, in the Raman spectrum corresponding to the initial sample 1, the Raman peak corresponding to the ν 1 band of CH 4 also can be observed (at ~2919 cm -1 ).
This peak diminished as the synthesis progressed; already in the spectrum for sample 2, it had almost completely disappeared, with CH 4 increasingly being converted into CT x H (4-x) .Note the spectral bands displaced toward the larger Raman shift relative to CH 4 ν 1 ; these are associated with the ν 3a bands of the tritiumsubstituted methanes.Note also that the highest content of T 2 , HT, and H 2 , with respect to the rest of the species present in the mixture, was observed for sample 2. This is because during its preparation the permeator was bypassed, which otherwise would have largely removed Q 2 from the mixture.
The set of Raman spectral features in the range 1650 to 1900 cm −1 is dominated by the contribution from the symmetric, normal vibrational mode ν 1 of the CT x H (4-x) family.Note that the peak of the fully tritiated methane, CT 4 , increases at the expense of the other T-methane species as the gas mixture passes through the catalyst and the permeator removes the Q 2 products from the medium, thus preventing hydrogen from reverse substitution.
For further, semiquantitative interpretation of the spectra, one should recall that the Raman signal S R can be approximated by the simplified expression Expression (3) can be used to extract the relative molecular concentrations from the spectra, in principle.However, while for tritium containing Q 2 , the values for the transition probabilities Φ if are known, 12 those for the tritium-substituted methane species CQ 4 are not available; thus, quantitative interpretation is rather difficult at present.
For a better understanding of the evolution in the concentrations of the CT x H (4-x) family, one may zoom into the ν 1 region of the Raman spectra, as shown in Fig. 3.
After the first passage of the sample gas through the catalyst reactor, the predominant species in sample 1 are those with the highest hydrogen content, i.e., CTH 3 and CT 2 H 2 ; but already a substantial quantity of CT 3 H has been generated, together with smaller amounts of CT 4 .
In the preparation of sample 2, during which the gas mixture passes through the catalyst for a second time but bypasses the permeator, the T-methane species, which dominated in sample 1, start to diminish, along with the Recall that for the second step, sample 1 was diluted to 50% in high-purity T 2 .Thus, the dominant species in the chemical environment of both the catalyst and the gas mixture is tritium.Consequently, the action of the catalyst and the balancing with surplus tritium drive the active centers toward saturation with tritium, i.e., effectively exchanging most H-atoms with T-atoms.
In the last step of the synthesis, i.e., in generating sample 3, the relative concentration for the species of highest interest, CT 4 , increases by almost a factor of 3 with respect to the previous phase.The two main reasons for this increase are as follows: 1.The gas fed to the reactor in this phase shows CT 3 H as the majority species within the CT x H (4-x) family, and thus a significant fraction of exchange reactions yields CT 4 .
2. The circulation time, and thus the time available for exchange reactions, is three times longer than in the two previous phases.
The nearly linear relation between the growth in the CT 4 concentration and the time over which exchange reactions progress invites the notion that, at least on the timescale of the experiments discussed here, it is the catalytic conversion that dominates and the influence of tritium decay on the reactions may be treated as being secondary.
It should be stressed that the synthesis process critically depends on the state of the catalyst.For this, optimal reaction conditions of the catalyst (activated or reduced) need to be established at the beginning of each phase.Of equal importance is that an optimal reaction environment in the medium be maintained, i.e., that fresh high-purity T 2 needs to be injected and that the products HT and H 2 , generated by the catalysis reactions, are removed.By this means, the catalyst remains in optimal conditions for longer, and the reaction favors the increase of tritiated methanes.
If the catalyst does not operate under optimal conditions, it would likely deactivate.This would result in mixtures containing lower concentrations of CT x H (4-x) (CT 4 in particular) and higher amounts of hydrocarbon chain molecules; this was observed in some of our earlier synthesis attempts.

III.B. Mass Spectrometry
With regard to the molecular isotopologues Q 2 , it can be seen that sample 2 is the one with the highest quantity of T 2 (at 6 amu), HT (at 4 amu), and H 2 (at 2 amu), and their atomic fragments, as was the case for the Raman spectroscopic data shown in Fig. 4.
For the family of tritium-substituted methane molecules, CT x H (4-x) , the mass spectra show that while all members of the family are present, (semi-) quantification becomes somewhat complex in contrast to the Raman data.This is because of fragmentation into "daughter" molecules in which successively Q-atoms are stripped from the "parent" molecule, leading to the same daughter molecules from different parent molecules.Overall, feature-rich mass spectra result if many, or all, CT x H (4-x) parents are present in the gas mixture (for mass spectral patterns see Ref. 14).
For sample 1, for which hydrogen is still plentiful, all CT x H (4-x) parents and their daughter products are observed, giving rise to a series of incremental mass peaks (Δm = 1 amu) ranging from atomic carbon, C (12 amu),

536
DÍAZ BARRERO et al.It is worth noting that the aforementioned filament phenomenon manifests itself in the mass spectra, too.As the overall interaction time between the catalyst and the mixture increases, so does the concentration of carbon chain molecules (number of C-atoms n > 1).These can be saturated hydrocarbons, i.e., C n H (2n+2) (with only C-C single bonds in the chain ≡ alkanes); unsaturated hydrocarbons, i.e., C n H 2n (with one C=C double bond in the chain ≡ alkenes); or C n H (2n-2) (with one C≡C triple bond in the chain ≡ alkynes).Of course, more than one unsaturated bond may be encountered, giving rise to a range of additional hydrocarbon families.
For the particular family of molecules with two C-atoms (n = 2), C 2 Q (6-2z) (with z indicating the number of unsaturated bonds), the following interesting feature is observed.In the mass spectrum of sample 3, the peak with mass 36 amu stands out, exhibiting quite a large amplitude that has evolved over the duration of the synthesis process.It can be assigned to the chemically rather stable molecule C 2 T 4 , i.e., fully tritiated ethylene.While exact quantification was not possible, its concentration seems to be substantial in comparison to the signal amplitudes of the other hydrocarbon molecules.This leads us to think that some of the lower-intensity Raman peaks, which we have not yet identified in the spectra, might actually originate from the aforementioned C 2 T 4 .

IV. DISCUSSION
As suspected, quite from the beginning a chemically pure compound could not be obtained (in principle we had wished for "pure" CT 4 ).However, the multistep synthesis strategy did allow us to shift the chemical equilibria and direct the reaction products toward the species with the highest tritium content, the majority being CT 3 H and CT 4 with almost similar magnitude (compare the amplitudes in both the Raman spectra shown in Figs. 2 and 3).The concentrations of the different species can be estimated from the Raman spectral data; by and large, these were corroborated in the mass spectrometry analysis.
As stated earlier, the species concentrations can be reasonably well quantified from the (integrated) Q 1 branch Raman data based on Eq. (3).However, for this the Raman transition probabilities Φ if for each species need to be known.Unfortunately, these are not known for the tritiumsubstituted methanes.Regardless, one can make at least a crude estimate for the CQ 4 content in sample 3. Using the observed relative transition probability Φ rel for the nontritiated species CH It should be possible to improve this yield by tweaking the synthesis design to keep the catalyst in better activation and reduction conditions.This might be done either (1) by decreasing the flow rate feeding the exchange reactor or (2) by finding strategies to more effectively prevent the formation of carbon filaments on the catalyst.It is in these amorphous carbon filaments where most of the CH 4 introduced into the system in the first step of the synthesis is lost (remember that these catalysts are used precisely for methane cracking).Some of it remains on the catalyst and is removed each time the catalyst is regenerated between synthesis steps.Some of the filament fragments were observed in the mass spectra as the families of C 3 Q (8-2z) , C 4 Q (10-2z) , C 5 Q (12-2z) , etc.
Both the amount of the mixture (~5-6 × 10 3 cm 3 ) and the concentration of CT x H (4-x) were significantly higher than those utilized in previous studies.This gave us the opportunity to record Raman spectra with further increased resolution in order to isolate individual Q 1 branches of the symmetric normal mode of vibration ν 1 of each of the species present in the mixture.With this one should be able to estimate the amount of each of them more rigorously.

V. CONCLUSION
Our results show that CQ 4 can be produced in large enough quantities from an individual production batch so that one can perform precision spectroscopy on them, and thus obtain relevant analytical knowledge.Therefore, a wealth of experiments can be performed using the same starting sample.Such experiments may address aspects of different analysis techniques (as demonstrated here), or the gas may be used for specific chemical interaction scenarios, such as, for example, exposure of particular surfaces to tritiated methane.But this work also shows that our approach, while rather successful, is still not yet perfect.Since the production of CQ 4 via the CAPER facility had to compete with its priority use for the KATRIN experiment, the time available for incremental improvements of the CQ 4 processing procedures was rather limited.However, with the experience gained thus far, we envisage coordinating our future efforts better with the KATRIN service periods during which CAPER can be utilized for different tasks.In this way, we expect to gain even better control over the distribution of CQ 4 isotopologues in the mixture as was possible to date.Even better for the production of pure CT 4 , which would be ideal, would be to use a H-free production path.This, however, would require the use of a different chemical reactor setup, which would be a long-term goal since at present the perpetual use of CAPER for other experiments prevents us from implementing structural changes.
Equally, our Raman analysis warrants attention as well.This is because the TRIHYDE Raman analysis tool was designed to analyze Q 2 mixtures rather than perform high-resolution spectroscopy.Thus, in particular, it is planned to add a high-resolution Raman spectrometer that would allow for full spectral resolution of all CT x H (4-x) species; at present, parts of the spectral features do overlap.
In this context, we have just begun theoretical quantum calculations on the structure of the Raman vibrationrotation spectra of each species in the CT x H (4-x) family.The aim is to be able to subtract these theoretical calculations from the experimental spectra in order to both reduce the non-Q 1 branch background and to better attribute the contribution of each species in the family to the overall Raman spectrum (unfortunately, individual S 1 , R 1 , P 1 , and O 1 branches overlap and cannot be easily separated).With the help of the pure synthetic Raman spectra, in comparison to the mixed experimental Raman spectra, we hope to be able to disentangle and quantify all individual CT x H (4-x) species in our gas mixtures.
)], and (2) it reduces the 532 DÍAZ BARRERO et al. • GENERATION AND ANALYSIS OF TRITIUM-SUBSTITUTED METHANE FUSION SCIENCE AND TECHNOLOGY • VOLUME 80 • APRIL-MAY 2024

Fig. 1 .
Fig. 1.Simplified schematic diagram of the CAPER subsections used in the production/enrichment of CQ 4 (note that under certain reaction conditions, longer-chain T-hydrocarbons are generated as well).The gas circulation paths for the individual, sequential phases are indicated by arrows; they are determined by closing/opening relevant valves.The valve marked SCV is the unidirectional stop-check valve assembly.For further details, see text.
GENERATION AND ANALYSIS OF TRITIUM-SUBSTITUTED METHANE • DÍAZ BARRERO et al. 533 FUSION SCIENCE AND TECHNOLOGY • VOLUME 80 • APRIL-MAY 2024

Fig. 2 .
Fig. 2. Data of the analysis of the Raman spectra for the sequence samples 1, 2, and 3 recorded after the individual CQ 4 enrichment steps; for details, see text.

534
DÍAZ BARRERO et al. • GENERATION AND ANALYSIS OF TRITIUM-SUBSTITUTED METHANE where e ν L = laser transition energy e ν R = Raman transition energies Φ if = Raman transition probability N i = molecular particle density I L = excitation laser power.

Fig. 3 .
Fig. 3. Detail of the group of peaks corresponding to the symmetric normal mode of vibration ν 1 of the CT x H (4-x) family of compounds.

Fig. 4 .
Fig. 4. Data of the analysis of the mass spectra for the sequence samples 1, 2, and 3 recorded after the individual CQ 4 enrichment steps; for details, see text.
GENERATION AND ANALYSIS OF TRITIUM-SUBSTITUTED METHANE • DÍAZ BARRERO et al. 537 FUSION SCIENCE AND TECHNOLOGY • VOLUME 80 • APRIL-MAY 2024 • DÍAZ BARRERO et al. 531 FUSION SCIENCE AND TECHNOLOGY• VOLUME 80 • APRIL-MAY 2024 • GENERATION AND ANALYSIS OF TRITIUM-SUBSTITUTED METHANE to fully tritiated methane, CT 4 (24 amu).The exception is mass 23 amu, for which neither a parent nor a daughter fragment of CT x H (4-x) exists.As the synthesis process progresses, the mass spectra become sparse, until for sample 3, basically only two parents and two of their daughter fragments prevail, namely, CT 4 (24 amu)/CT 3 (21 amu) and CT 3 H (22 amu)/CT 2 H (19 amu).This confirms the findings from the Raman spectra quite well.
one can estimate the CQ 4 -content of sample 3 to be ~20%.From this estimate, the yield for this isotope exchange reaction, i.e., for the conversion of CH 4 to CT 3 H and CT 4 (the two majority species in the resulting mixture), is approximately ~16%.