Photochemical reductive homologation of hydrogen cyanide using sulfite and ferrocyanide† †Electronic supplementary information (ESI) available. See DOI: 10.1039/c8cc01499j

On their own, neither sulfite nor ferrocyanide are efficient sources of photochemically-generated electrons for the reductive homologation of hydrogen cyanide, but together they are.


Bisulfite & UV irradiation
Ultraviolet light likely played an important role in chemistry on the surface of early Earth, 1-5 and we (J. X., D. J. R. & J. D. S.) and others have frequently used UV light from a mercury lamp with primary emission at 254 nm to investigate potentially prebiotic photochemistry. Although early solar UV emission was broadband, its influx on the earth would have been constrained due to absorption by atmospheric gases. UV wavelengths below 204 nm would have been persistently screened out due to absorption by atmospheric CO 2 and H 2 O vapour, and SO 2 and H 2 S released by volcanism would have transiently acted as additional UV shields during and shortly after periods of volcanism. 6 In high concentrations, H 2 S strongly absorbs UV wavelengths around 200 nm while SO 2 absorbs around 200 nm and 290 nm. Interestingly, neither species strongly absorbs around 250 nm, meaning that our UV source emits at a wavelength that would have been accessible across a broad range of H 2 S and SO 2 abundances. Although this UV photochemistry is not recapitulated in extant biology, it could have served to lay down supplies of (proto)biomolecules to kick-start life and provision it in its early stages. Biology could then learn different routes to resupply these molecules when prebiotic stores ran out. Biology uses chemistry catalysed by enzymes which is easy to control (by regulating the amount or activity of the enzymes) and tends not to use reaction mechanisms that proceed efficiently in the absence of catalysis. Thus, one would expect biology to make the same products as prebiotic chemistry, but using different transformations or the same transformations using different mechanisms to prebiotic chemistry.
Phosphate was used as a buffer in this study as its presence in numerous (proto)biomolecules (eg. RNA, lipids, ATP and nucleotide cofactors) implies prebiotic and early biotic availability.
The high levels of sulfate in Martian soils strongly suggests that volcanic emissions of SO 2 and H 2 S were common on early Mars. 7,8 Volcanism is still the major source of atmospheric SO 2 and H 2 S on Earth today and it is plausible that volcanic activity was, at least intermittently, much higher on anoxic early Earth, allowing SO 2 to reach higher concentrations. 9 These considerations suggest that SO 2 should be considered as an alternative to H 2 S as an atmospherically-sourced reductant for prebiotic chemistry. Compared to H 2 S, SO 2 is easily concentrated from the atmosphere into groundwater due to the favourability of its hydration and the acidity of the hydrate (pK a = 1.8 for 'sulfurous acid' (SO 2 . xH 2 O), pK a = 7.2 for bisulfite (HSO 3 -)). 7,8 Lastly, it is known from EPR studies 10-12 that UV irradiation of aqueous solutions of sulfite (SO 3 2-) produces sulfite radical anions ( • SO 3 -) and hydrated electrons.

Ferrocyanide
Ferrocyanide ([Fe II (CN) 6 ] 4-) has been considered as a prebiotically plausible solution phase repository of HCN 1 delivered from the atmosphere of early Earth. 13 Free cyanide can be released from solutions of the complex by photoaquation at longer UV wavelengths, 14,15 or, more efficiently, by thermal decomposition of sodium or potassium ferrocyanides in the solid state. 16

Mechanism with sulfite as stoichiometric reductant and ferrocyanide as catalyst/promoter
Ferrocyanide ([Fe(CN) 6 ] 4-) is photoionized (oxidized) to give ferricyanide ([Fe(CN) 6 ] 3-) and a solvated electron as reported in the literature. 17 Assisted by general acids (phosphate monoanion or bisulfite) the hydrated electrons add to HCN 1 or, later on, to the nitrile groups of cyanohydrins to give reduced products. The ferricyanide is then reduced to ferrocyanide by sulfite, affording sulfate (SO 4 2-) as a co-product. Additionally, sulfite might act as a direct source of hydrated electrons during irradiation in which case the sulfite radical anions ( • SO 3 -) could be oxidised by ferricyanide to sulfate. Whatever the relative contributions of the pathways operative, the sulfite effectively functions as a two-electron donor in the presence of ferrocyanide. [18][19][20] However, absent ferrocyanide, the reaction proceeds less efficiently and sulfite can only act as a one-electron reductant -the resultant sulfite radical anion ( • SO 3 -) now dimerising to S 2 O 6 2in the process (see inset box in Scheme 1 in the main text). [10][11][12] The proposed mechanism is also supported by the quantitative data obtained by comparison of the reactions of 11 with 1 equivalent of KCN with and without ferrocyanide (Fig. 3). The reaction involving ferrocyanide afforded a 68% yield (i.e. more than 50%) of reduced products after 1 h of irradiation, which is only stoichiometrically possible with sulfite as a two-electron reductant. The reaction without ferrocyanide could theoretically afford up to 50% of reduced products, however we found that it gave a maximum yield of only 25% of reduced products after 3 h of irradiation.

General methods
All reagents and deuterated solvents used for reactions and spiking experiments were purchased from Sigma-Aldrich or Acros Organics and were used without further purification. All photochemical reactions were carried out in Norell Suprasil quartz NMR tubes or Spectrosil quartz cuvettes purchased from Sigma-Aldrich using Hg lamps with principal emission at 254 nm in a Rayonet photochemical chamber reactor RPR-200, acquired from The Southern New England Ultraviolet Company. A Mettler Toledo SevenEasy pH Meter S20 was used to monitor the pH, and deoxygenation of solution was achieved by sparging anhydrous argon through the solution for 15-20 min. All unknown compounds in the reaction mixtures were confirmed by spiking experiments with authentic compounds either purchased from Sigma-Aldrich or synthesized in house using conventional synthetic chemistry. 1 H and 13 C NMR spectra were acquired using a Bruker Ultrashield 400 Plus operating at 400.1 MHz and 100.6 MHz respectively. Samples consisting of H 2 O/D 2 O mixtures were analyzed using HOD suppression to collect 1 H NMR data. The quantitative 13 C NMR spectra were acquired with inverse-gated decoupling, with a 90˚ excitation pulse and an inter-pulse delay of 70 seconds. 13

1,2-Dihydroxypropanenitrile (7)
15.0 mg of glycoaldehyde dimer 2 2 (0.13 mmol) was mixed with 16.0 mg of potassium cyanide (0.25 mmol) in 0.5 mL of degassed water (containing 10% of D 2 O). The pH of the mixture was adjusted to 7 with 1M HCl before being transferred to an NMR tube to record the 1 H and 13 C NMR spectra of 1,2-dihydroxypropanenitrile 7. 1

2-Hydroxypropanenitrile (17) and 1-hydroxyethane-1-sulfonate (21)
14.0 μl of acetaldehyde 18 (0.25 mmol) was mixed with 31.5 mg of Na 2 SO 3 (0.25 mmol) in 0.5 mL of degassed water (containing 10% of D 2 O). The pH of the mixture was adjusted to 7 with 1M HCl before being transferred to an NMR tube to record the 1 H and 13 C NMR spectra of 1-hydroxyethane-1-sulfonate 21. to 7 with 1M HCl. The mixture was subject to irradiation at 254 nm for 5 h. The pH of the mixture was adjusted to 9.2 with 1M NaOH and the mixture was kept at room temperature overnight before being spiked with authentic standards.

Photoreduction of 13 C-labelled hydroxymethanesulfonate (11) with 13 C-labelled KCN
8.0 μl of 13 C-labelled formaldehyde solution (0.05 mmol, 20 wt. % in water, from Sigma-Aldrich) was mixed with 12.6 mg of Na 2 SO 3 (0.10 mmol) in 0.5 mL of degassed water (containing 10% of D 2 O). The pH of the mixture was adjusted to 7 with 1M HCl. 6.6 mg of 13 C-labelled KCN (0.10 mmol) and 12.0 mg of NaH 2 PO 4 were added to the mixture. The pH was readjusted to 7 with 1M HCl. The mixture was transferred to a quartz NMR tube and irradiated at 254nm for 12.5 h in total. The sample was taken out of the photochemical reactor at intervals to record 1 H and 13 C NMR spectra. At the end of the reaction, 2 mg (0.03 mmol) of 13 C-labelled sodium formate was added to the mixture to enable quantification of the reaction products in quantitative 13 C NMR spectra. Argon was sparged through the solution at the end to remove excess HCN from the mixture.

Photoreduction of 13 C-labelled hydroxymethanesulfonate (11) with 13 C-labelled KCN with or without ferrocyanide
11.2 μl of a solution of 13 C-labelled formaldehyde (0.08 mmol, 20 wt. % in water, from Sigma-Aldrich) was mixed with 9.5 mg of Na 2 SO 3 (0.08 mmol) in 3.0 mL of degassed water (containing 10% of D 2 O). The pH of the mixture was adjusted to 7 with 1M HCl. 5.0 mg of 13 C-labelled KCN (0.08 mmol) and 36.0 mg of NaH 2 PO 4 (0.30 mmol) were added to the mixture. The pH was readjusted to 7 with 1M HCl. The mixture was divided into two parts, into one of which was added 1.6 mg of K 4 Fe(CN) 6 .3H 2 O (0.004 mmol, 10 mol%). Both solutions in quartz NMR tubes were irradiated in the photochemical reactor side by side and taken out at intervals to record 13 C spectra.  -10  0  10  20  30  40  50  60  70  80  90  100  110  120  130  140  150  160  170 δ / ppm

13 C NMR time course of the photoreduction of 13 C-labelled KCN with sodium sulfite
Photoreduction of the mixture of 240 mM 13 C-labelled hydroxymethanesulfonate 11, 240 mM 13 C-labelled KCN and 240 mM NaH 2 PO 4 at pH 7. a) 13 C NMR spectrum of the mixture after 2.5 h irradiation; b) 13 C NMR spectrum of the mixture after 5 h irradiation.

13 C NMR spectra of the spiking experiments for the photoreduction of 13 C-labelled KCN with sodium sulfite
a) 13 C NMR spectrum of reaction mixture at pH 9.2 overnight after irradiation for 5 h; b) as a), spiked with unlabelled glycolonitrile 6; c) as b), spiked with unlabelled aminoacetonitrile 14; d) as c), spiked with unlabelled iminodiacetonitrile 15 Signals from the spiking compounds are indicated with blue arrows, as shown above.

13 C NMR spectra of the spiking experiments for the photoreduction of 13 C-labelled hydroxymethanesulfonate (11) with 13 C-labelled KCN
a) 13 C NMR spectrum of reaction mixture after irradiation for 12.5 h and sparging with argon; b) as a), spiked with unlabelled glycoaldehyde 2; c) as b), spiked with unlabelled glycolonitrile 6; d) as c), spiked with unlabelled iminodiacetonitrile 15. Signals from the spiking compounds are indicated with blue arrows, as shown above.

1 H NMR time course of the photoreductions of hydroxymethanesulfonate (11) with KCN with or without ferrocyanide
Photoreduction