On the origin of the Murchison meteorite phosphonates. Implications for pre-biotic chemistry

Abstract

Ab initio calculations, combined with experimental studies on the anaerobic hydrolysis of phosphaalkynes under thermal and photochemical conditions suggest a potential, exogenous source of reduced oxidation state phosphorus for the early Earth.

---

Whilst it is clear that derivatives of orthophosphoric acid (H_xPO₄^3−x; x = 1–3) are the key phosphorus moieties in contemporary biochemistry¹ there is growing opinion that Nature may have exploited lower oxidation state phosphorus species during the early development of life on Earth.² Gulick pointed out as long ago as 1955, that both H-phosphonic (P(III); H₃PO₃) and H-phosphinic (P(I); H₃PO₂) acids could have been accessible on the early Earth if the redox potential of that environment was appropriately reducing.³ More recently Schink and Friedrich have identified sulfate-reducing bacteria which exploit the reverse process, oxidation of H-phosphonic to orthophosphoric acid, as a key step in microbial energy metabolism; 4HPO₃²⁻ + SO₄²⁻ + H⁺ = 4HPO₄²⁻ + HS⁻ for which ΔG° = −87 kcal mol⁻¹, based on sulfate, or −22 kcal mol⁻¹ based on phosphate.⁴ These authors point out that this type of microbial energy metabolism may have operated on the anoxic early Earth should a source of reduced oxidation state phosphorus have been readily available. Herein lies the problem. Within the oxygen-rich environment of the Earth today, essentially all phosphorus sources are based on the fully oxidised orthophosphate. However, the pre-biotic Earth was almost certainly oxygen-poor and probably reducing in potential,⁵ an environment in which lower oxidation state phosphorus species such as phosphonic (H₃PO₃) and phosphinic (H₃PO₂) acids and their derivatives would certainly have been more stable. The question then is how could reduced oxidation state phosphorus have been produced on and/or delivered to the early Earth?

The redox couples H₃PO₄/H₃PO₃ (−276 mV) and HPO₄²⁻/HPO₃²⁻ (−690 mV) suggest that ca. 13 and 32 kcal mol⁻¹ are required to reduce orthophosphate to H-phosphonate depending upon pH.⁶ Whilst such energy input is accessible from electrical discharges associated with early Earth volcanic activity⁷ interest has turned towards an exogenous delivery mechanism of reduced oxidation state phosphorus species as components of extraterrestrial ice-grains, meteorites or cometary fragments.⁸ A key observation in support of this hypothesis was the extraction of methyl and ethylphosphonic acids at nmol g⁻¹ concentrations (Fig. 1d; R = Me, Et) in addition to orthophosphate (at µmol g⁻¹ levels) from the Murchison meteorite.⁹ As found for other molecules of interstellar origin, these phosphonic acids were observed to have enhanced isotopic D/H ratios (δD) of ca. 219 suggesting that they are unlikely to be terrestrial contaminants.¹⁰ We report here the first experimental evidence that hydrolytic modification of interstellar phosphorus molecules may have acted as a source of Murchison-type phosphonates.

Fig. 1 Ab initio G3X level studies on the reaction of P≡CH with H₂O showing both intermediates (lower species) and transition states (upper species; indicated TS1–4). Energy values in kcal mol⁻¹.

Very few phosphorus molecules have been identified within the interstellar medium. The principal ones include PH₃,¹¹ metal rich phosphides such as schreibersite (Fe,Ni)₃P,¹² P≡N¹³ and [P≡C]˙.¹⁴ We are currently exploring the anoxic hydrolytic modification of each of these species and here report our preliminary work on the molecular radical [P≡C]˙. This radical has been detected in cool carbon-rich interstellar gas clouds,¹⁴ such as that surrounding the ageing star IRC+10216, and is presumed to have been formed from interactions between interstellar P⁺ and CH₄.¹⁵ It is a member of the class of unsaturated molecules well-known in interstellar chemistry¹⁶ and is proposed to be a daughter fragment of molecular P≡CH which has been hypothesised to be absent from the gas phase due to condensation on interstellar ice grain surfaces.¹⁴ Moreover, as this radical is the only such interstellar phosphorus species containing the P–C bond, we were especially interested to explore whether this molecule, or more specifically the parent P≡CH, could act as a precursor to the Murchison phosphonic acids.

Our ab initio calculations at the G3X level‡ reveal that hydrolytic modification of P≡CH to methylphosphonic acid via the equation P≡CH + 3H₂O = CH₃P(O)(OH)₂ + H₂ is thermodynamically favourable by ca. 96 kcal mol⁻¹. A more detailed theoretical analysis of the P≡CH/H₂O system reveals that sequential nucleophilic attack of water takes place preferentially at phosphorus via several intermediates. Thermodynamic analysis suggests that H₂O attack at phosphorus should lead to either the P-hydroxyphosphaalkene, CH₂=POH or the more stable (by 5.4 kcal mol⁻¹) methylphosphoroso tautomer CH₃P=O (Fig. 1). Subsequent addition of a second equivalent of H₂O to CH₃P=O affords methyldihydroxyphosphine CH₃P(OH)₂ which would be expected to re-arrange to the more stable (by 11.6 kcal mol⁻¹) tautomer methyl-H-phosphinic acid CH₃P(O)H(OH). However, whilst phosphaalkyne hydrolysis is undoubtedly favourable thermodynamically, significant energy barriers exist along the reaction coordinate. Four of the key transition states have been identified and modelled (Fig. 1). The energy barrier for addition of H₂O to P≡CH is 35.1 kcal mol⁻¹, the gas-phase transition state TS1 possessing significant simultaneous H–C and O–P bond forming interactions as well as both deforming the ∠P–C–H angle from 180° to 136.8° yet elongating the C–P distance by only 0.075 Å (ca. 5%). Clearly an energy barrier of this magnitude would preclude chemical reactions within the interstellar medium unless: (a) alternative gas-phase reagents are involved (e.g., reaction between H₂O⁺ and P≡CH to afford [CH₂=POH]⁺ has zero energy barrier), (b) alternative energy sources (e.g.; photochemical, cosmic ray etc…) and/or (c) some form of surface catalysis on interstellar ice grains were available. We believe that each of these possibilities may be feasible and our ongoing theoretical studies are examining these systems.§

Building upon the above theoretical studies, we reasoned it should be possible to simulate hydrolytic modification of [P≡C]˙ by exploring a model compound of parent P≡CH, since the latter is the simplest member of the well-known phosphaalkyne family of molecules P≡CR, R = organic fragment.¹⁷ We have compared the hydrolytic behaviour of one such model compound, (2,2-dimethylpropylidyne)phosphine¹⁸ P≡CCMe₃ under both thermal and photochemical conditions. Thus, ³¹P{¹H}-NMR analysis of an anaerobic two-phase system comprising P≡CCMe₃ and D₂O (1 ∶ 1000 mole ratio) reveals the gradual evolution of a single major new species after several days (ca. 70% conversion after 6 days) exhibiting a strong 1 ∶ 1 ∶ 1 triplet coupling pattern with δ_P 36.2 ppm and a ¹J_PD of 83 Hz. In addition, ¹³C{¹H}-NMR analysis reveals the expected doublet resonances for CMe₃ quaternary and methyl carbons at δ_C 29.12 (²J_PC 7.5 Hz) and 30.08 ppm (³J_PC 2.7 Hz) respectively. Of most significance however, is the presence of an apparent doublet of quintets of triplets (dqt) coupling pattern at δ_C 41.61 ppm which is consistent with, and simulated by,¶ the suite of coupling constants ¹J_PC 89.0 Hz; ¹J_CD 19.1 Hz and ²J_CD 2.3 Hz, expected for a [DPCD₂] methylene moiety (Fig. 2). These collected data are clearly supportive of this species being ^tBuCD₂P(O)D(OD), resulting from the addition of two equivalents of D₂O to P≡CCMe₃. In addition to ^tBuCD₂P(O)D(OD), as the major product we observe three new species appearing within the P≡CCMe₃/D₂O system over time (<10% after 11 days) displaying respectively a 1 ∶ 1 ∶ 1 triplet ³¹P-NMR resonance at δ_P 36.0 ppm (¹J_PD 84 Hz), a doublet resonance at δ_P 36.6 ppm (¹J_PH 545 Hz) and a singlet resonance at δ_P 30.9 ppm which we assign as ^tBuCD₂P(O)D(OH), ^tBuCD₂P(O)H(OH) and ^tBuCD₂P(O)(OX)₂ (X = H or D) respectively. Negative ion mass spectrometry|| located both [^tBuCD₂P(O)D(O)]⁻ and [^tBuCD₂P(O)H(O)]⁻ at 138.08 (theor: 138.08) and 137.07 respectively whilst also providing further evidence for the presence of the phosphonate [^tBuCD₂P(O)(OH)(O)]⁻ at 153.07 (theor: 153.06). The latter product, neopentylphosphonic acid ^tBuCD₂P(O)(OX)₂ is the direct model for the Murchison phosphonates and we presumed that it formed by either the scavenging of di-oxygen by neopentylphosphinic acid within the sealed tube or by photo-induced reaction with di-oxygen or water. Indeed, we have independently prepared and characterised neopentylphosphinic acid and have confirmed (by ³¹P-NMR) that photochemical irradiation of an oxygen-free aqueous solution of ^tBuCH₂P(O)H(OH) (3 mM; 298 K; 4 h) with a low pressure mercury vapour arc tube (emissions at 254 nm; ca. 500 mW and 185 nm; ca. 40 mW) affords cleanly a 63% conversion to a mixture composed solely of neopentylphosphonic acid ^tBuCH₂P(O)(OH)₂ (δ 29.0, t, ²J_PH 17.5 Hz) and H-phosphonic acid in a 1 ∶ 0.7 molar ratio. Subsequent introduction of air and re-irradiation for 4 h affords complete conversion of the system to neopentylphosphonic acid and phosphoric acid in a 1 ∶ 1.2 molar ratio respectively.

Fig. 2 Partial ¹³C{¹H}-NMR spectrum of ^tBuCD₂P(O)D(OD) emphasising coupling to the [CD₂P(O)D] carbon. Experimental (top; 125.7 MHz; 300 K) and simulated (bottom; gNMR version 4.1.2; Adept Scientific, UK) using ¹J_PC 89.0 Hz; ¹J_CD 19.1 Hz and ²J_CD 2.3 Hz.

Thus, we believe our model studies suggest that condensed interstellar P≡CH is capable of undergoing thermally-induced, anaerobic hydrolysis to alkyl H-phosphinic acids and that phosphonic acids may result from a subsequent photo-induced redox process in aqueous solution. This latter process is effective under anaerobic conditions but leads to more efficient conversions to fully oxidised phosphorus in the presence of di-oxygen. Moreover, our experiments support the conclusion that an interstellar phosphaalkyne hydrolysis mechanism is capable of producing D-enrichment in the resulting phosphonic acids via formation of kinetically non-labile [C–D] linkages. Deuterium incorporated within the hydroxyl or [P–D] groups of either phosphonic or phosphinic acids is envisaged to undergo rapid and complete H/D exchange with terrestrial water sources following Earth-impact of the parent meteorite body thus leading to attenuation of deuterium enrichment at these sites.

This thermal behaviour contrasts significantly however with the behaviour of the P≡CCMe₃/H₂O system under photochemical conditions. Thus, irradiation with a low pressure mercury vapour arc tube of an anoxic mixture of P≡CCMe₃ and oxygen-free water (1 ∶ 25 molar ratio) at a combined vapour pressure of 1 mTorr at 298 K, in a purpose-built photochemical reactor, for a period of one hour affords a white, water-soluble component and a yellow organic-miscible component. The water-soluble material comprises H-phosphonic HP(O)(OH)₂ and orthophosphoric acids P(O)(OH)₃ as the sole phosphorus-containing products in a 15 ∶ 1 molar ratio respectively by ³¹P-NMR spectroscopy. We envisage that the orthophosphoric acid is produced via photolysis of H-phosphonic acid in water, a process we have observed in parallel experiments on the hydrolytic processing of metal phosphides.¹⁹ The organic material contains no phosphorus and returns a C ∶ H ratio of 5.93 ∶ 1 consistent with a C₄H₈ fragment ratio (theor: 6 ∶ 1). However, at least 17% w/w consists of a black, non-volatile ash upon combustion, whose composition is still under investigation. Electrospray mass spectrometric analysis (H₂O ∶ CH₃CN) of this organic material reveals a broad range of fragments from m/z 165–835, with a base peak at m/z 399 suggestive of oligomers. We are currently exploring this photochemical transformation, especially the organic component, in more detail.

Our four key conclusions may be summarised as follows. (1) Formation of the Murchison phosphonic acids, and potentially phosphonates on interstellar icy bodies in general, could have occurred via anaerobic hydrolytic modification of condensed molecular species derived from [P≡C]˙. (2) These hydrolyses also allow us to explain the enhanced D ∶ H isotope ratios of interstellar phosphonates. (3) Divergent behaviour in the P≡CR/H₂O system may result from thermal and photochemical stimulation, dependent upon the nature of the group R. (4) The observation of H-phosphonic acid as the chief product from the P≡CR/H₂O system under photolytic conditions is most intriguing given the earlier hypotheses of Schwartz et al.,² Gulick³ and Freidrich and Schink⁴ along with the observation that this species is also the chief product from the anaerobic hydrolysis of meteoritic phosphide minerals.^19,20 It is conceivable that H-phosphonic acid may be a key hydrolytic by-product of more than one interstellar phosphorus molecule and this is something we are actively investigating.²¹

We gratefully acknowledge the Leverhulme Trust (TPK; DEB) and the EC Leonardo Programme (FB). Drs Julie Fisher (University of Leeds) and Anneke Lubben (Bruker Daltonics) are thanked for their valuable help with the gNMR simulations and negative ion mass spectrometry respectively. We thank also Professor Alan Schwartz for most enlightening discussions.

Notes and references

1. H. Lodish, A. Berk, S. L. Zipursky, P. Matsudaira, D. Baltimore and J. Darnell, Molecular Cell Biology, W. H. Freeman, New York, 4th edn, 2000.
2. R. M. De Graaf, J. Visscher and A. W. Schwartz, Nature, 1995, 378, 474.
3. A. Gulick, Am. Sci., 1955, 43, 479.
4. B. Schink and M. Friedrich, Nature, 2000, 406, 37.
5. See for example; F. Westall, Astrophys. Space Sci. Libr., 2004, 305, 287.
6. N. N. Greenwood and A. Earnshaw, Chemistry of the Elements, Pergamon Press, Oxford, UK, 1984.
7. D. Glindemann, R. M. De Graaf and A. W. Schwartz, Origins Life Evol. Biosphere, 1999, 29, 555.
8. C. Chyba and C. Sagan, Nature, 1992, 355, 125.
9. G. W. Cooper, W. M. Onwo and J. R. Cronin, Geochim. Cosmochim. Acta, 1992, 56, 4109.
10. G. W. Cooper, M. H. Thiemens, T. L. Jackson and S. Chang, Science, 1997, 277, 1072.
11. J. P. Ferris, A. Bossard and H. Khwaja, J. Am. Chem. Soc., 1984, 106, 318.
12. S. N. Britvin, N. S. Rudashevsky, S. V. Krivovichev, P. C. Burns and Y. S. Polekhovsky, Am. Mineral., 2002, 87, 1245.
13. L. M. Ziurys, Astrophys. J., 1987, 321, L81.
14. B. E. Turner, T. Tsuji, J. Bally, M. Guelin and J. Cernicharo, Astrophys. J., 1990, 365, 569.
15. J. R. Flores, J. Phys. Org. Chem., 1996, 9, 419.
16. See for example, H. W. Kroto, Chem. Soc. Rev., 1982, 11, 435.
17. Phosphorus: The Carbon Copy, ed. K. B. Dillon, F. Mathey and J. F. Nixon, Wiley, Chichester, 1998 and references therein.
18. G. Becker, H. Schmidt, G. Uhl and W. Uhl, Inorg. Synth., 1990, 27, 243.
19. D. E. Bryant and T. P. Kee, Chem. Commun., 2006 10.1039/b602651f.
20. M. A. Pasek and D. S. Lauretta, Astrobiology, 2005, 5, 515.
21. A highly relevant paper on this issue has just appeared; E. Macia, Chem. Soc. Rev., 2005, 34, 691.
22. M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuseria, M. A. Robb, J. R. Cheeseman, J. A. Montgomery, Jr., T. Vreven, K. N. Kudin, J. C. Burant, J. M. Millam, S. S. Iyengar, J. Tomasi, V. Barone, B. Mennucci, M. Cossi, G. Scalmani, N. Rega, G. A. Petersson, H. Nakatsuji, M. Hada, M. Ehara, K. Toyota, R. Fukuda, J. Hasegawa, M. Ishida, T. Nakajima, Y. Honda, O. Kitao, H. Nakai, M. Klene, X. Li, J. E. Knox, H. P. Hratchian, J. B. Cross, V. Bakken, C. Adamo, J. Jaramillo, R. Gomperts, R. E. Stratmann, O. Yazyev, A. J. Austin, R. Cammi, C. Pomelli, J. Ochterski, P. Y. Ayala, K. Morokuma, G. A. Voth, P. Salvador, J. J. Dannenberg, V. G. Zakrzewski, S. Dapprich, A. D. Daniels, M. C. Strain, O. Farkas, D. K. Malick, A. D. Rabuck, K. Raghavachari, J. B. Foresman, J. V. Ortiz, Q. Cui, A. G. Baboul, S. Clifford, J. Cioslowski, B. B. Stefanov, G. Liu, A. Liashenko, P. Piskorz, I. Komaromi, R. L. Martin, D. J. Fox, T. Keith, M. A. Al-Laham, C. Y. Peng, A. Nanayakkara, M. Challacombe, P. M. W. Gill, B. G. Johnson, W. Chen, M. W. Wong, C. Gonzalez and J. A. Pople, GAUSSIAN 03 (Revision B.05), Gaussian, Inc., Wallingford, CT, 2004.
23. L. A. Curtiss, P. C. Redfern, K. Raghavachari and J. A. Pople, J. Chem. Phys., 2001, 114, 108.

---

Footnotes

† Electronic supplementary information (ESI) available: Synthetic methods, characterisation, NMR and mass spectrometric analyses and computational analyses. See DOI: 10.1039/b517497j

‡ All calculations were carried out using the Gaussian 03 suite of programs.²² Geometry optimisations and vibrational frequency calculations were performed at B3LYP/6-311+G(3 df, 2 p) level. Scale factor of 0.9854 was used for zero-point energy corrections. The accurate energetic parameters were evaluated using G3X model chemistry, which approximates the electron correlation level of QCISD(T,Full)/G3XLarge.²³

§ We are focusing especially upon the possibility of surface transition metal catalysis since preliminary DFT and coupled cluster calculations suggest that P≡CH binds to the known interstellar ions Fe⁺ and Mg⁺ with binding energies of the order of 24 kcal mol⁻¹. In both cases, strong Fe–P and Fe–C binding is involved such that the ∠P–C–M angles lie within the 85–90° range. This bonding obviously leads to structural distortions which should activate the phosphaalkyne towards hydrolytic modification and may suggest a reason for the non-detection of un-modified P≡CH in IRC+10216. More detailed studies on metal-binding and possible activation of P≡CH towards hydrolysis are in progress.

¶ The [CD₂P(O)D] portion of the spectrum has been simulated using gNMR version 4.1.2 (Adept Scientific, UK), see ESI for details.

|| Samples were analysed by injection of 0.1 ml of a ca. mM aqueous solution into a Bruker microTOF instrument operating under electrospray conditions with a 1 ∶ 1 v/v acetonitrile ∶ water mobile phase at 0.6 ml min⁻¹. The instrument was calibrated externally using sodium formate.

---