Method for decomposing sulfur hexafluoride

Abstract

A method for decomposing SF6 comprising the steps (a) providing a solution of an alkali metal hydroxide in an alcohol of formula R–HCOH– R', wherein R and R' independently are selected from H, C1 to C6 – alkyl; cyclohexyl, cyclopentyl, methyl cyclopentyl, or C1 to C6 – alkoxy, (b) contacting the solution with SF6 (c) exposing the solution with SF6 in a closed vessel to electromagnetic radiation at a wavelength of between 240 nm to 300 nm, until SF6 is decomposed.

Description

[0001] METHOD FOR DECOMPOSING SULFUR HEXAFLUORIDE

[0002] The present invention relates to a method for decomposing sulfur hexafluoride SFe.

[0003] BACKGROUND OF THE INVENTION

[0004] Because of its unique chemical and physical properties sulfur hexafluoride (SFe) is used in various industries, including magnesium and aluminum manufacturing, the electronics industry, particle accelerators, optical fiber manufacturing, photovoltaics, military and medical applications. Most of the SFe produced worldwide is used in the power industry, especially in medium and high voltage applications.

[0005] Due to the significant greenhouse potential of SFe, it must not be released into the atmosphere and its use is subject to strict recycling and disposal requirements in many countries. The method of choice for disposal of SFe waste gas is thermal decomposition (pyrolysis) by injection into a reaction furnace at over 1100 °C. During pyrolysis highly reactive components are formed resulting inter alia SO2, SO3, HF which have to be transformed to CaSCh and CaF2 by a sorbent in a second step. Pyrolysis of SFe is very energy intensive and the sorbent is consumed during the reaction, which further increases operating costs. Regarding operational safety, the process is not without problems due to the presence of toxic gases.

[0006] In addition to the thermal decomposition methods, alternative processes have been developed in recent years, in which the decomposition of SFe is carried out by means of non-thermal plasma, photolysis, electrochemistry or reductive-chemical methods. However, these processes have not yet found application because some of these processes have significant disadvantages in terms of decomposition rate, operating and acquisition costs, or operational safety.

[0007] Non-thermal plasma decomposition methods include radio frequency (RF) plasma, microwave plasma, electron beam (EB) plasma, and dielectric barrier discharge (DBD) plasma methods. Even though a lower energy input compared to the thermal decomposition of SFe is required, still highly toxic gases are formed during the decomposition of SFe in the plasma, which must be bound by a suitable sorbent.

[0008] Electrochemical methods for decomposing SFe show low efficiency, low turnover rates, and are largely unexplored. Numerous chemical reactions of strongly reducing substances with SFe are known. In most cases, the decomposition of SFe is initiated by electron transfer to the SFe molecule. Among the most promising methods is the reaction of SFe with alkali metals in liquid ammonia. However, such reactions must be carried out under inert gas conditions and require the use of toxic solvents.

[0009] Direct photolysis of SFe requires high energy UV radiation (< 160 nm). Matrix isolation experiments have shown that the SF6 / H2O system reacts in a xenon matrix under laser irradiation (193 nm) with cleavage of the S-F bond. It has also been shown that SFe can be degraded by UV irradiation (185 nm) of gaseous mixtures of SFe with acetone, styrene or isoprene (Li Huang, et al. “Investigation of a new approach to decompose two potent greenhouse gases: Photoreduction of SF6 and SF5CF3 in the presence of acetone”, Chemosphere, 66 (5), 2007, 833-840; https: / / doi.Org / 10.1016 / j.chemosphere.2006.06.028.). Thereby, the photolytic fragmentation of the additive generates reactive particles which react with SFe. However, these processes are unselective and produce several gaseous and partly toxic products that require appropriate post-treatment. Furthermore, it is known that phosphines, Mg / MgO nanoparticles and TiCh surfaces promote the decomposition of SFe under UV irradiation. Generally, the conversion rates and the removal efficiency with photochemical methods are low and require short wavelength UV radiation.

[0010] EP 4 144 690 Al discloses a process for the decomposition of SFe, which comprises the steps of contacting SFe with a phosphine in the presence of a photosensitizer, followed by irradiation with UV light at 300 - 420 nm. JP 2000279551 A describes a process for decomposing organic chlorine compounds by a combination of photochemical treatment and microbial decomposition. JP 2001 046547 A discloses a method for decomposing organic chlorine compounds by mixing the organic chlorine compound in an organic solvent (e.g., isopropanol) containing NaOH.

[0011] BRIEF DESCRIPTION OF THE INVENTION

[0012] Hence, it is an object of the invention to provide a method for decomposing SFe into non-toxic compounds which is selective, has high yields and has low costs.

[0013] This object is solved by a method for decomposing SFe comprising the steps

[0014] (a) providing a solution of an alkali metal hydroxide in an alcohol of formula R-HCOH- R’, wherein R and R’ independently are selected from H, Ci to Ce - alkyl; cyclohexyl, cyclopentyl, or Ci to Ce - alkyloxy,

[0015] (b) contacting the solution with SFe, and

[0016] (c) exposing the solution with SFe to electromagnetic radiation at a wavelength of between 240 nm to 300 nm, preferably between 270 nm and 290 nm, until SFe is decomposed.

[0017] The inventors found that SFe can be easily decomposed by a solution of alkali metal hydroxides in alcohols under simultaneous irradiation of the SFe in contact with the solution with UV radiation. The process is technically simple and scalable and has application potential for energy-efficient disposal of SFe on small and large scale.

[0018] The alcohol is oxidized into an aldehyde or ketone and SFe is converted into a sulfite. Based on an example reaction with KOH as alkali metal hydroxide and propan-2-ol as the alcohol, the reaction mechanism is based on the following equations:

[0019] OH 280 nm O (1 )

[0020] Hence, the photochemical decomposition of SFe in a mixture of KOH and secondary alcohols proceeds with oxidation of the alcohol to the ketone and reduction / hydrolysis of SFe to potassium sulfite (K2SO3), potassium thiosulfate (K2S2O3) or diisopropyl sulfite (iP^SCh) and potassium fluoride (KF). The main product containing sulfur is sulfite, as diisopropyl sulfite is hydrolyzed to give sulfite and propan-2-ol under the basic reaction conditions.

[0021] Alternatively, NaOH can be used resulting in Na2SOs, Na2S20s, iPr2SO3 and NaF or mixtures of NaOH and KOH can also be used. A primary alcohol can be used instead of a secondary alcohol. The reaction conditions are set that preferably per 1 Mole of SFe to be decomposed, at least 8 Moles of metal hydroxide are provided in the solution. Preferably, the metal hydroxide is NaOH, KOH or a mixture thereof. The reaction does not require special temperature control. Thus, it can be carried out at room temperature (15 °C to 30 °C). Ideally, the reaction is carried out in a closed vessel.

[0022] Decomposition of SFe in step (c) can be checked by monitoring the SFe consumption in the reaction vessel.

[0023] This invention describes a new method for decomposing SFe selectively and quantitatively into non-toxic compounds, by simply using a primary or secondary alcohol, a base and UV-B or UV-C radiation. The preferred wavelength of the UV radiation is light with a wavelength of between 240 nm to 300 nm, preferably between 270 nm and 290 nm.

[0024] While the reaction of SFe with aqueous solutions of KOH or NaOH in the absence of an alcohol is extremely slow, the addition of the alcohol instead of water as solvent speeds up the reaction considerably, especially when secondary alcohols such as isopropanol are used. The alcohol increases the solubility of SFe in the liquid phase and serves as a reducing agent. Secondary alcohols are particularly effective, because they get reduced to the corresponding ketones during the reaction and the ketones increase the rate of the SFe decomposition, presumably due to the generation of ketyl radicals. As outlined above, no gaseous products are formed because the decomposition in the basic medium is selective. The selective product formation also facilitates the utilization of the reaction products. Other advantages of the invention include low equipment cost due to the simple design, low energy consumption, and the fact that no toxic chemicals are used or formed in the process.

[0025] The alcohol is preferably a secondary alcohol, so R’ and R are each independently selected from Ci to Ce - alkyl; cyclohexyl, cyclopentyl, methyl cyclopentyl, or Ci to Ce - alkoxy. More preferably, R and R’ is selected from Ci to C3 - alkyl; Ci to C3 - alkoxy.

[0026] Most preferably, the alcohol is selected from the group consisting of isopropanol, butan-2-ol, ethanol and mixtures thereof.

[0027] It has been found out that small amounts of a ketone can accelerate the SFe decomposition. Hence, it is preferred that the solution further comprises a ketone, preferably acetone, butanone.

[0028] In a preferred embodiment of the invention, step (a) is carried out in the presence of water. The inventors found that the addition of water to the solution of alkali metal hydroxide in an alcohol result in a biphasic system consisting of the alcohol phase and the aqueous phase comprising alkali metal hydroxide. This biphasic system has the following advantages compared to the monophasic alkali metal hydroxide / alcohol system:

[0029] • The rate of degradation / decomposing of SFe is higher due to lower polarity of the alcohol phase.

[0030] • The reaction becomes more selective as no formation of thiosulfate is observed. The products of the decomposition are metal fluorides and sulfites as the only inorganic salts formed in the reaction.

[0031] • The addition of ketones to accelerate the reaction comes without the sometimes-observed issue of aldol condensation as side reactions.

[0032] • No precipitation of inorganic solids during the reaction is observed. This can facilitate the separation of the inorganic salts. It can also become a key advantage for both the batch process and a continuous flow process

[0033] Preferably, the ratio of water to alcohol is between 1.5 : 1 and 1 : 15, preferably between 1 : 5 and 1 : 10 by volume.

[0034] Definitions:

[0035] Ci to Ce - alkyl means an alkyl group with 1 to 6 carbon atoms. The alkyl group may be a straight-chain alkyl or a branched alkyl.

[0036] Ci to Ce - alkoxy means an alkyl group with 1 to 6 carbon atoms. The alkyl group may be a straight-chain alkyl or a branched alkyl.

[0037] Ci to Ce - alkoxy means an alkyl group with 1 to 6 carbon atoms being singularly bonded to oxygen (R-0 with R being the alkyl group which 1 to 6 carbon atoms). The alkyl group of the alkoxy may be a straight-chain alkyl or a branched alkyl.

[0038] DETAILED DESCRIPTION OF THE INVENTION

[0039] Further details of the invention are described below in connection with examples.

[0040] Fig. 1 shows an SFe consumption time plot obtained by the SFe monitor on the SFe uptake by the reaction mixture.

[0041] Fig. 2 shows anJH NMR spectrum (400 MHz) of the collected volatiles in CDCh (Pyrazine was added as internal standard, *CHC13). 8 = 8.52 (s, C4 / / 4N2), 4.70 (sept.,3JHH = 6.2 Hz, OS{OC Me2}2), 3.92 (sept.,3. / HII = 6.1 Hz, HOC Me2), 2.76 (s,2O), 2.09 (s, OC{C#3}2), 1.26 (d,3. / HII = 6.2 Hz, OS{OCH[C#3]2}2), 1.23 (d,3. / nn = 6.2 Hz, OS{OCH[C#3]2}2), 1.11 (d,3JHH = 6.1 Hz, HOCH{C#3}2) ppm.

[0042] Fig. 31 shows an13C{JH} NMR spectrum (101 MHz) of the collected volatiles in CDC13(Pyrazine was added as internal standard, *CDC13). 5 = 207.5 (s, OCMe2), 68.4 (s, OS{OCHMe2}2), 64.0 (s, H0CHMe2), 30.9 (s, OC{CH3}2), 25.2 (s, HOCH{CH3}2), 23.5 (s, OS{OCH[CH3]2}2) ppm.

[0043] Fig. 4 shows the SFe consumption time plot obtained by the SFe monitor demonstrating the influence of the solvent on the SFe uptake by the reaction mixture.

[0044] Fig. 5 shows the SFe consumption time plot obtained by the SFe monitor demonstrating the influence of additives on the SFe uptake by the EtOH / KOH system.

[0045] Fig. 6 shows the SFe consumption time plot obtained by the SFe monitor demonstrating the influence of ketone additives on the SF6 uptake by the reaction mixture.

[0046] Fig. 7 shows the SFe consumption time plot obtained by the SFe monitor demonstrating the influence of the initial gaseous components on the SFe uptake by the reaction mixture.

[0047] Fig. 8 shows the SFe consumption time plot obtained by the SFe monitor demonstrating the influence of the SF6 pressure on the uptake by the reaction mixture.

[0048] Fig. 9 shows the SFe consumption time plot obtained by the SFe monitor demonstrating the influence of the KOH concentration on the SFe uptake by the reaction mixture.

[0049] Fig. 10 shows the SFe consumption time plot obtained by the SFe monitor demonstrating the influence of the light intensity on the SFe uptake by the reaction mixture.

[0050] Fig. 11 shows the SFe consumption time plot obtained by the SFe monitor demonstrating the influence of the KOH concentration on the SFe uptake by the reaction mixture Fig. 12 shows UV-vis spectra of the KOH / iPrOH system.

[0051] Fig. 13 shows UV-vis spectra of the KOH / EtOH system.

[0052] Fig. 14 shows a comparison of the SFe uptake between the KOH / iPrOH system and the biphasic KOH / H2O / iPrOH system.

[0053] Fig. 15 shows UV-vis spectra of the acetone base stability experiment. A) after 24 h in the KOH / iPrOH system, B) after 24 h in the KOH / H2O / iPrOH system, C) HPLC grade isopropanol with 1% acetone.

[0054] Fig. 16 shows a comparison of the SFe uptake between standard conditions and the addition of 4,4-dimethoxybenzophenon and irradiation at 365 nm in the KOH / iPrOH system and the biphasic KOH / H2O / iPrOH system (Reaction setup 2). Fig. 17 shows a comparison of the SFe uptake between standard conditions and the addition of benzophenon and irradiation at 365 nm in the KOH / iPrOH system and the biphasic KOH / FFO / iPrOH system (Reaction setup 2).

[0055] Fig. 18 Initial biphasic KOH / FFO / iPrOH system (left) and the reaction mixture after irradiation for 2 h (right).

[0056] 1. Materials and Methods

[0057] 1.1. Materials

[0058] Sulfur hexafluoride 3.0 (99.9%) was generously donated by the company DILO GmbH. All other compounds were purchased from commercial sources and used as received if not stated otherwise. The alcohols used and their commercial source are as follows: methanol (HPLC grade 99.8%; FisherSci); ethanol (technical grade 96%, contains 1% isopropanol, Donauchem); / / -propanol (p.a. 99.5%; Sigma Aldrich), isopropanol (technical grade 98%; VWR), isopropanol (HPLC grade 99.9%; Sigma Aldrich), / / -butanol (99.9%; Sigma Aldrich), .scc-butanol (ReagentPlus, 99%, Sigma Aldrich), 2-methyl-2-pentanol (99%; Sigma Aldrich), / / -hexanol (98%; VWR), ethylene glycol (spectrophotometric grade 99%; Sigma Aldrich), l-methoxypropan-2-ol (ReagentPlus 99.5%; Sigma Aldrich), 2 -propoxy ethanol (98%; TCI), acetone (ACS reagent grade 99.5%; Sigma Aldrich), butanone (HPLC grade 99%, Sigma Aldrich). Potassium hydroxide was obtained from Sigma Aldrich (pellets, >85%). The water content of potassium hydroxide was determined by titration using an aqueous HC1 solution (1 M) and bromothymol blue.

[0059] 1.2. Light sources

[0060] All photolysis experiments were carried out using quartz glassware. The following LED light sources were used for the irradiation experiments: VCClilte LED Typ VAOL-SAlxAx-SA (585 nm) and Seoulviosys LED Typ CUD1KFMA (310 nm) within a custom-build irradiation setup, EvoluChem™ LED 405PF (405 nm) and EvoluChem™ LED 365PF (365 nm) within the EvoluChem PhotoRedOx Box™, Led-Tech XBT-3535-UV LED (280 nm) in a self-built setup. In addition, a high intensity mercury-xenon lamp (Hamamatsu 200W Mercury Xenon 365nm wide band L9566-06A) was used with a spectral distribution of 240 to 550 nm.

[0061] 1.3. Analytical Methods

[0062] ’H,13C and19F NMR spectroscopy was performed on Bruker AVANCE I 400 or Bruker ACANCE III 400 spectrometers. Chemical shifts are given in parts per million (ppm) relative to SiMe4 (XH,13C) or CFCh (19F) and they were referenced to the residual solvent signals (CDCI3: 'H <5H = 7.26,13C be = 77.16). Chemical shifts (3) are reported in ppm. NMR multiplicities are abbreviated as follows: s = singlet, d = doublet, sept = septet. For the determination of yields using NMR spectroscopy, quantitative19F NMR (19F qNMR) and 'H NMR experiments were performed with increased relaxations time (DI = 25 seconds).

[0063] Capillary electrophoresis (CE) was performed for ion analysis using a capillary electrophoresis apparatus from Agilent, a C4D conductivity detector from TraceDec and a capillary from Polymicro Technologies with an inner diameter of 50 pm.

[0064] UV-vis absorption spectroscopy was carried out on a PerkinElmer LAMBDA XLS+ spectrophotometer using standard quartz UV / vis cuvettes (d = 1 cm). pH determination was carried out using an inoLab® pH 7110 meter. Prior to measurements a two-point calibration was done using two technical puffers (buffer solution pH 7.00 and buffer solution pH 4.00 from Carl Roth).

[0065] Monitoring of the SFe consumption was done by a self-built electrical pressure control unit (SFe-monitoring device or SFe-monitor). The readout and pressure control were done using an Arduino Nano E / A-Board with an ATmega328 microcontroller. The pressure regulation unit consists of an inlet tube of 3 bar SFe from gas storage cylinder that is attached to a magnetic valve. The outlet of the magnetic valve is connected to a piezo-resistive silicon pressure sensor and the reaction vessel using a T-piece adapter. When the pressure in the reaction vessel drops below a given threshold, the magnetic valve opens for a fixed time interval of 10 ps. The number of cycles thus corresponds to a defined SFe consumption in the reaction. A calibration experiment reveals a consumption of SFe per cycle of 0.04 mmol at a given SFe storage cylinder pressure of 3 bar and reaction vessel pressure of 2 bar (Equation a). The resulting data (saved as a .csv file) consists of the time, the number of cycles of the magnetic valve, and the pressure in the reaction vessel. The data was imported to OriginPro and the number of cycles of the magnetic valve was plotted against the reaction time. n(SF6consumption) (a) n(SF6per cycle) = numbers of cycles of the magnetic valve

[0066] 1.4. Experimental Procedures Three different reaction setups were used for the photolysis experiments depending on the volume of the reaction mixture.

[0067] Reaction Setup 1 was used for the irradiation of a volume of 1 mL. A quartz NMR tube (diameter 5 mm) containing the reaction mixture was pressurized with 3 bar SFe and the sealed tube was subsequently irradiated for a given time.

[0068] Reaction Setup 2 was used for the irradiation of a volume of 15 mL. A quartz tube (diameter 17 mm) containing the reaction mixture under 2 bar SFe atmosphere was irradiated for a given time and the SFe consumption was monitored using the SFe-monitor.

[0069] Reaction Setup 3 was used for the irradiation of a volume of 80 mL. A borosilicate glass tube (diameter 42 mm) containing the reaction mixture was sealed with a quartz sight glass. The SFe pressure was set to 2 bar and the consumption was monitored using the SFe-monitor.

[0070] 2. Photochemical degradation of SFe in basic solutions

[0071] 2.1. Screening of the base with isopropanol as the solvent using Reaction Setup 1

[0072] Preliminary screening of LiOH, NaOH, KOH, and Ca(OH)2 shows that the SFe degradation is most effective when KOH is used as the base.

[0073] 2.2. Screening of the solvent with KOH as the base using Reaction Setup 1

[0074] General Procedure: Potassium hydroxide powder (30.0 mg, 0.464 mmol) was dissolved in the solvent (1 ml) indicated in Table 1 and the solution was transferred into an NMR tube. The tube was pressurized with SFe (3 bar), vigorously shaken and irradiated with light at 280 nm for one hour. The resulting reaction mixture was transferred into a vial and the volatiles were evaporated under reduced pressure to afford a solid residue consisting of a mixture of inorganic salts (e.g. KF, K2SO3). The residue was dissolved in water and the content of KF was determined using quantitative19F NMR spectroscopy with potassium tritiate (10.0 mg, 0.0531 mmol) as internal standard. Assignment of the19F NMR resonances: 5 = 58.5 (SFe), -78.6 (OTf ), -119.1 (F") ppm. As solvent water and different alcohols were used. The results are summarized in Table 1.

[0075] Table 1 : Yields based on fluoride formation, determined by19F qNMR spectroscopy, relative to the amount of KOH used assuming that 8 KOH are consumed per SFe molecule according to the following equation: 8 KOH + SFe + alcohol 6 KF + K2SO3 + 5 H2O + aldehyde / ketone.

[0076] 2.3 Irradiation wavelength dependence of SFe degradation using the KOH / iPrOH system A quartz glass NMR tube was charged with a solution of potassium hydroxide (30.0 mg, 0.464 mmol) in technical grade isopropanol (1 ml). Then, the quartz NMR tube was pressurized with SFe (3 bar), shaken vigorously, and irradiated with light at different wavelength for 1 h. The resulting reaction mixture was analysed by19F NMR spectroscopy showing that no reaction took place using light at 585 nm, 405 nm or 365 nm, while the fluoride resonance was detected for irradiation experiments with light at 310 nm, 280 nm and 240-550 nm. To determine the reaction yields, the reaction mixture was transferred into a vial and the volatiles were evaporated under reduced pressure. The resulting mixture of inorganic salts (e.g. KF, K2SO3) was dissolved in water and the fluoride content was determined using quantitative19F NMR spectroscopy with potassium tritiate (10.0 mg, 0.0531 mmol) as internal standard. The results of experiments using different wavelengths for the electromagnetic radiation is summarized in Table 2.

[0077] Table 2: Yields based on fluoride formation, determined by19F qNMR spectroscopy, relative to the amount of KOH used assuming that 8 KOH are consumed per SFe molecule according t acetone.

[0078] 2.4. Analysis of SFe degradation reaction products using the KOH / iPrOH system

[0079] Using Reaction Setup 2, a quartz tube (diameter 17 mm) was charged with a solution of potassium hydroxide (1.00 g, 15.5 mmol) in technical grade isopropanol (15 ml). The solution was degassed by four freeze-pump-thaw cycles and the reaction tube was pressurized with SFe (2 bar). The reaction mixture was irradiated for 2 h with light at 280 nm and the consumption of SFe was monitored using the SFe-monitor. A white precipitate was formed during the reaction. The volatiles were distilled off at 100 °C in vacuo for 3 hours. The collected volatile components and the remaining solid components were analyzed as follows:

[0080] Volatile components: An aliquot of 0.1 mL of the resulting solution was dissolved in CDCI3 and analyzed by NMR spectroscopy: No F-containing compound was detected via19F NMR spectroscopy;1H and13C NMR spectroscopy showed the resonances of isopropanol, acetone and diisopropyl sulphite. The total amount of these compounds was determined by integration ofXH NMR resonances using pyrazine (10.0 mg, 0.125 mmol) as internal standard: Solid components: The mixture of inorganic salts was analysed by NMR spectroscopy, capillary electrophoresis (CE) and by a pH meter.

[0081] For the analysis by NMR spectroscopy, 52 mg of the mixture of inorganic salts was dissolved in water and potassium tritiate (10.0 mg, 0.0531 mmol) was added as internal standard. No resonances of C-containing compounds were detected by13C {1H } NMR spectroscopy. The amount of fluoride formed in the reaction was determined using19F qNMR spectroscopy:

[0082] To determine the total amount of ions formed in the reaction, 10 mg of the mixture of inorganic salts was diluted with 0.2 L water to afford a solution with a concentration of 50 mg / L. The solution was analyzed by capillary electrophoresis (CE):

[0083] *calculated according to the following equation

[0084] P(Ion) ■ F(dilution) ■ 94 n(Ion) = - — — -

[0085] M(Ion)

[0086] To determine the pH of the stock solution of inorganic salts, 0.5 ml of the stock solution was diluted to 20 ml with water and a pH of 8.2 was measured using a pH meter. In a separate experiment, a solution of potassium sulfite with similar concentration showed a pH of 8.9. To further confirm that all potassium hydroxide had been consumed in the SFe decomposition reaction, 0.1 ml of potassium hydroxide (0.1 M) was added to the stock solution of inorganic salts which resulted in a pH of 9.8.

[0087] The characterization data of the volatile and solid components suggest that three equations need to be considered in the photochemical SFe degradation reaction: OH 280 nm O (1 )

[0088] The weighting of the individual reaction equations is based on results of the CE using the relative amounts of the S-containing compounds and the total amount of fluoride:

[0089] 11.91 mmol

[0090] Coverall) = - 7 - = 1-99 mmol

[0091] 6

[0092] Coverall) = «(SO32-) + n(S2032-) ■ 2 + n(iPrO)2SO) (5.) n(iPrO)2SO) = 1.99 mmol — 1.42 mmol — 0.06 mmol ■ 2 = 0.47

[0093] Table 3: Weighting of the reaction equations (2)-(4).

[0094] 6 KF + 0.71 K2SO3+ 0.03 K2S2O3+

[0095] OH 280 nm

[0096] 7.48 KOH + 1 SF6+ 1.55 I - *

[0097] 0.24 5.08 H2O

[0098] By considering the weighting scheme of

[0099] Table 3 and based on the initial amount of KOH (15.5 mmol), the theoretical yield of KF is calculated to be 12.06 mmol, which is in agreement with the experimental values obtained by19F qNMR spectroscopy (12.3 mmol) and by CE (11.9 mmol). ncaic.(KF) = 1.42 mmol ■ 6 + 0.06 mmol ■ 12 + 0.47 mmol ■ 6 = 12.06 mmol By considering the weighting scheme of

[0100] Table 3 and based on the initial amount of KOH (15.5 mmol), the theoretical yield of acetone is calculated to be 2.13 mmol, which is in agreement with the experimental values obtained by *H NMR spectroscopy (1.96 mmol). ncaic.(acetone)= n(SO32-) ■ 1 + n(S2O32-) ■ 4 + n(iPrO)2SO) ■ 1 (7.) ncalc(acetone) = 1.42 mmol ■ 1 + 0.06 mmol ■ 4 + 0.47 mmol ■ 1 = 2.13 mmol

[0101] The results are shown in Fig. 1, Fig. 2 and Fig. 3.

[0102] 2.5 Influence of the solvent on the SFe degradation rate using KOH as the base and Reaction Setup 2

[0103] General Procedure: A quartz tube (diameter 17 mm) was charged with a solution of potassium hydroxide (1.00 g, 15.5 mmol) in the solvent (15 ml) indicated in Table 4. The solution was degassed by four freeze-pump-thaw cycles and the reaction tube was pressurized with SFe (2 bar). The reaction mixture was irradiated for 2 h with light at 280 nm and the consumption of SFe was monitored using the SFe-monitor. A white precipitate was formed during the reaction. The resulting suspension was transferred into a 100 mL round bottom flask using small amounts of water to ensure complete transfer of the precipitate. The volatiles were evaporated under reduced pressure to afford a mixture of inorganic salts (e.g. KF, K2SO3). The residue was dissolved in water until a volume of 5 mL was obtained. An aliquot of 1 ml of this aqueous solution was analysed by quantitative19F NMR spectroscopy using potassium tritiate (10.0 mg, 0.0531 mmol) as internal standard. Assignment of the19F NMR resonances: 5 = -78.6 (OTE), -119.1 (F’) ppm.

[0104] Table 4: Yields based on fluoride formation, determined by19F qNMR spectroscopy, relative to the amount of KOH used assuming that 8 KOH are consumed per SFe molecule according to the following equation: 8 KOH + SFe + alcohol 6 KF + K2SO3 + 5 H2O + aldehyde / ketone.

[0105] The SFe consumption as a function of time for different solvents is shown in Fig. 4. Different solvents and solvent mixtures are shown with different symbols at 5-minute intervals. Table 5: Estimated yields based on fluoride formation, determined by19F qNMR spectroscopy, relative to the amount of KOH used assuming that 8 KOH are consumed per SFe molecule according to the following equation: 8 KOH + SFe + EtOH 6 KF + K2SO3 + 5 H2O + acetaldehyde. The SFe consumption as a function of time for different additives is shown in Fig. 5. The SFe uptake by the EtOH / KOH system is represented. Different additives are shown with different symbols at 5-minute intervals.

[0106] Table 6: Estimated yields based on fluoride formation, determined by19F qNMR spectroscopy, relative to the amount of KOH used assuming that 8 KOH are consumed per SFe molecule according to the following equation: 8 KOH + SFe + sec-butanol -> 6 KF + K2SO3 + 5 H2O + butanone.

[0107] The SFe consumption as a function of time for different ketone additives is shown in Fig. 6. Different additives are shown with different symbols at 5-minute intervals.

[0108] 2.6. Optimization of the reaction conditions for the KOH / iPrOH system using Reaction Setup 2

[0109] General Procedure: A quartz tube (diameter 17 mm) was charged with a solution of potassium hydroxide (see Table 7) in technical grade isopropanol (15 ml). The solution was degassed by four freeze-pump-thaw cycles and the reaction tube was pressurized with SFe (see Table 7). The reaction mixture was irradiated for 2 h with light at 280 nm and the consumption of SFe was monitored using the SFe-monitor. A white precipitate was formed during the reaction. The resulting suspension was transferred into a 100 mL round bottom flask using small amounts of water to ensure complete transfer of the precipitate. The volatiles were evaporated under reduced pressure to afford a mixture of inorganic salts (e.g. KF, K2SO3). The residue was dissolved in water until a volume of 5 mL was obtained. An aliquot of 1 ml of this aqueous solution was analysed by quantitative19F NMR spectroscopy using potassium tritiate (10.0 mg, 0.0531 mmol) as internal standard. Assignment of the19F NMR resonances: 5 = -78.6 (OTE), -119.1 (F’) ppm.

[0110] Table 7: Estimated yields based on fluoride formation, determined by19F qNMR spectroscopy, relative to the amount of KOH used assuming that 7.48 KOH are consumed per SFe molecule according to the following equation: 7.48 KOH + SFe + 1.55 iPrOH 6 KF + 0.71 K2SO3 + 0.03 K2S2O3 +0.24 (iPrO)2SO + 1.07 acetone + 5.08 H2O.

[0111] The SFe consumption as a function of time for different initial gaseous components is shown in Fig. 7. Different gaseous components are shown with different symbols at 5-minute intervals.

[0112] The SFe consumption as a function of time for different SFe pressures is shown in Fig. 8 (symbols at 5-minute intervals are shown). Note that the amount of SFe transferred per cycle to the reaction vessel is lower when the SFe pressure increases.

[0113] The SFe consumption as a function of time for different KOH concentrations is shown in Fig. 9 (symbols shown at 5-minute intervals).

[0114] The SFe consumption as a function of time for different light intensities is shown in Fig. 10 (symbols shown at 5-minute intervals).

[0115] 2.7. Scaleup of the SFe degradation with the KOH / iPrOH system using Reaction Setup 3 A borosilicate glass tube (diameter 42 mm) equipped with a quartz sight glass was charged with a solution of potassium hydroxide in technical grade isopropanol (80 ml). The reaction vessel was pressurized with SFe (1 bar) and irradiated with light at 280 nm for the indicated period of time. The consumption of SFe was monitored using the SFe-monitor. A white precipitate formed during the reaction. The reaction mixture was transferred into a 250 mL round bottom flask using small amounts of water to ensure complete transfer of the precipitate. The volatiles were evaporated under reduced pressure to afford a mixture of inorganic salts (e.g. KF, K2SO3). The residue was dissolved in water until a volume of 40 mL was obtained. An aliquot of 1 ml of this mixture was analysed by quantitative19F NMR spectroscopy using potassium tritiate (10 mg, 0.05 mmol) as internal standard. Assignment of the19F NMR resonances: 5 = -78.6 (OT ), -119.1 (F") ppm.

[0116] Table 8: Estimated yields based on fluoride formation, determined by19F qNMR spectroscopy, relative to the amount of KOH used assumingthat 8 KOH are consumed per SFe molecule according to the following equation: 8 KOH + SFe + iPrOH 6 KF + K2SO3 + 5 H2O + acetone.

[0117] *Due to the limited solubility of KOH in iPrOH, the initial reaction mixture was a suspension.

[0118] The SFe-consumption as a function of time is shown in Fig. 11. Note that the initial reaction mixture was a suspension in the case of the 15 g KOH experiment due to the limited solubility. The interval indicated by the symbols is every 3 hours.

[0119] 2.8. Experimental procedure of the system with water

[0120] Experiments according to the invention were performed in which step (a) was carried out in the presence of water. One experiment was performed using Reaction Setup 2 using a mixture of water (2 ml) and HPLC grade isopropanol (13 ml) as solvent (KOH: 1.0 g, 16.8 mmol; SFe pressure: 2 bar; irradiation time: 2 h). The resulting biphasic system consists of an isopropanol layer on top of the concentrated aqueous solution of KOH. The photochemical decomposition was carried out with vigorous stirring of the mixture. In contrast to the KOH / iPrOH system, no precipitate was formed during the reaction. The volatiles were distilled off at ambient temperature in vacuo for 4 hours and they were analysed by GC-MS and1H NMR spectroscopy. The remaining solid was analysed by19F qNMR and capillary electrophoresis.

[0121] The addition of water to the solution of alkali metal hydroxide - alcohol results in a biphasic system consisting of the alcohol phase and the aqueous phase comprising alkali metal hydroxide. As can be seen in Fig. 14, the rate of degradation / decomposition of SFe is higher due to lower polarity of the alcohol phase. Furthermore, the decomposition reaction turned out to be more selective as no formation of thiosulfate was observed but only metal fluorides and sulfites are formed (see Table 9) The addition of acetone as a ketone accelerated the reaction further (see Fig. 15). As is demonstrated in Fig. 18, no inorganic solids are formed during the reaction.

[0122] Table 9: Overview of the analysis results of the products from the SFe decomposition reaction using the biphasic KOH / f O / iPrOH system. (- = not detectible with the method, X = not detected, = detected)

[0123] By contrast to the KOH / iPrOH system, no thiosulfate is formed. Yield: 93%.

[0124] 2.9. Experimental procedures including photosensitizer:

[0125] The experimental setups as described in Chapters 2.4 and 2.8 were repeated and in step a) a ketone-based photosensitizer (benzophenone (Fig. 17) or 4,4 ’-dimethoxybenzophenone (Fig. 16)) was added to the alkali metal hydroxide / alcohol or in the biphasic alkali metal hydroxide / water / alcohol system. Using the additional photosensitizers redshifts the irradiation wavelength (in the present case the ketone-based photosensitizers allowed irradiation with light at 365 nm).

[0126] However, the reaction turned out to be not autocatalytic anymore as in the examples according to the invention and the reaction is therefore slower (see Fig. 16 and Fig. 17) despite a higher radiation power of the 365 nm light source. Furthermore, more dominant side reactions were observed, presumably aldol condensations, but maybe also with the photocatalyst itself. In addition, the cost of the photocatalyst increased the cost for the SFe decomposition.

[0127] 3. UV-vis absorption spectroscopy

[0128] 3.1. UV-vis spectra of the KOH / iPrOH system

[0129] A standard quartz UV-vis cuvette (d = 1 cm) was charged with a solution of KOH (90 mg, 1.6 mmol) in HPLC grade isopropanol (3 mL), the cuvette was pressurized with 3 bar SFe, and the cuvette was irradiated with light at 280 nm for 1 hour.

[0130] A series of UV-vis spectra were recorded for the KOH / iPrOH system confirming the formation of acetone during the irradiation reaction:

[0131] Fig. 12 shows UV-vis spectra of the KOH / iPrOH system. A) The initial KOH / iPrOH solution under SFe atmosphere, B) The reaction solution after irradiation with light at 280 nm for 1 hour and subsequent filtration, C) HPLC grade iPrOH with 1% acetone.

[0132] 3.2. UV-vis spectra of the KOH / EtOH system

[0133] A standard quartz UV-vis cuvette (d = 1 cm) was charged with a solution of KOH (90 mg, 1.6 mmol) in ethanol (abs.) (3 mL), the cuvette was pressurized with 3 bar SFe, and the cuvette was irradiated with light at 280 nm for 1 hour.

[0134] Fig. 13 shows UV-vis spectra of the KOHZEtOH system. A) The initial KOH / EtOH solution under SF6 atmosphere, B) The reaction solution after irradiation with light at 280 nm for 1 hour and subsequent filtration.

Claims

CLAIMS1. A method for decomposing SFe comprising the steps(a) providing a solution of an alkali metal hydroxide in an alcohol of formula R-HCOH- R’, wherein R and R’ independently are selected from H, Ci to Ce - alkyl; cyclohexyl, cyclopentyl, methyl cyclopentyl, or Ci to Ce - alkoxy(b) contacting the solution with SFe(c) exposing the solution with SFe in a closed vessel to electromagnetic radiation at a wavelength of between 240 nm to 300 nm, until SFe is decomposed.

2. A method according to claim 1, wherein the light has a wavelength of between 270 nm to 290 nm.

3. A method according to claim 1 or claim 2, wherein per 1 Mole of SFe to be decomposed, at least 8 Moles of metal hydroxide are provided in the solution.

4. A method according to claim 3, wherein per 1 Mole of SFe to be decomposed, at least 8,8 Moles of metal hydroxide are provided in the solution.

5. A method according to one of claims 1 to 4, wherein the metal hydroxide is NaOH, KOH or a mixture thereof.

6. A method according to one of claims 1 to 5, wherein R is H and R’ is selected from Ci to Ce - alkyl; Ci to Ce - alkyloxy.

7. A method according to one of claims 1 to 6, wherein the alcohol of formula R- C(OH)RIR2 is selected from the group consisting of isopropanol, butan-2-ol, ethanol and mixtures thereof.

8. A method according to one of claims 1 to 7, wherein the solution further comprises a ketone, preferably acetone or butanone.

9. A method according to one of claims 1 to 8, wherein the closed vessel is permeable to electromagnetic radiation at wavelengths between 240 nm to 300 nm and the source for electromagnetic radiation is located outside the vessel.

10. A method according to one of claims 1 to 9, wherein SFe is dosed into the vessel at a pressure of up to 10 bar, preferably up to 3 bar.

11. A method according to one of claims 1 to 10, wherein SFe is continuously dosed into the vessel.

12. A method according to one of claims 1 to 11, wherein the decomposition of SFe is monitored in step (c).

13. A method according to one of claims 1 to 12, wherein step (a) is carried out in the presence of water.

14. A method according to claim 13, wherein the ratio of water to alcohol is between 1.5 : 1 and 1 : 15, preferably betweenl : 5 and 1 : 10 by volume.

Images