thermal desorption EI-MS, ESI-MS, APCI-MS and off-line TMAH thermochemolysis/GC-MS. This latter method causes bond cleavage and in situ methylation ...
STRUCTURAL INVESTIGATIONS OF HYDROGEN CYANIDE POLYMERS: NEW INSIGHTS USING TMAH THERMOCHEMOLYSIS/GC-MS ROBERT D. MINARD, PATRICK G. HATCHER and ROBERT C. GOURLEY Departments of Chemistry and Fuel Science, The Pennsylvania State University, University Park, PA 16802, U.S.A.
CLIFFORD N. MATTHEWS Department of Chemistry, University of Illinois, Chicago, IL 60607, U.S.A. (Received 11 November, 1996)
Abstract. Hydrogen cyanide polymers form spontaneously from HCN and traces of base catalysts. It is probable that these polymers played an important role in the early stages of chemical evolution. Nevertheless, their full structural characterization has still not been accomplished. A number of mass spectrometric methods have now been applied to this structural problem including FAB-MS, thermal desorption EI-MS, ESI-MS, APCI-MS and off-line TMAH thermochemolysis/GC-MS. This latter method causes bond cleavage and in situ methylation producing a suite of products which provides valuable insight into the substructural features of HCN polymers and also promises to serve as a sensitive diagnostic tool for detecting the presence of HCN polymers in samples from diverse sources.
1. Introduction Hydrogen cyanide polymers-heterogeneous solids ranging in color from yellow to orange to brown to black-may be among the organic macromolecules most readily formed within the solar system. Their structure is therefore of considerable interest for investigations of prebiotic and extraterrestrial chemistry (Matthews, 1992). HCN is formed in simulation experiments involving a range of gaseous compositions and almost every conceivable energy source. It is required in reactions that form both amino acids and purine or pyrimidine bases. It is involved in several hypothetical pathways for the formation of sugars or the condensation of key monomeric units (amino acids, sugars, phosphate, etc.). The facile polymerization of HCN continues to elude detailed understanding, in spite of its first observation almost 200 yr ago (Proust 1806; Gay-Lussac, 1815). Hydrogen cyanide, whether as a gas, a pure liquid, or in aqueous or organic solution, polymerizes rapidly in the presence of trace amounts of a base catalyst to yield a black solid.
xHCN ! (HCN)x HCN polymers
Corresponding author: Robert Minard, 152 Davey Lab, University Park, PA 16802, U.S.A.
Origins of Life and Evolution of the Biosphere 28: 461–473, 1998. c 1998 Kluwer Academic Publishers. Printed in the Netherlands.
462
R. D. MINARD ET AL.
Complete characterization of this solid material, called ‘azulmin’ in the older literature, has not been possible in spite of extensive efforts (Ferris et al., 1981; Liebman et al., 1994 and references cited therein). This is due to the fact that it is both heterogeneous and reactive. While it can be separated into water-soluble yellowbrown solids and water-insoluble black solids, on standing at room temperature in water, the soluble yellow-brown solids will form some insoluble black solid and the insoluble black solid will partially dissolve to form water-soluble yellow brown solids. Traces of HCN monomer are released from these solids. Thus, it appears that this material is not a stable ‘repeating unit’ polymer in the traditional organopolymer sense, but rather a reactive heteropolymer, or more accurately, a mixture of heteropolymers. Very little is known about the molecular weight distribution of these materials, but since they contain some HCN tetramer, it appears that the mixture contains a broad distribution of molecular weights. Figure 1 shows various structural hypotheses for the primary structural motifs and/or routes of formation for HCN polymers. The original ‘ladder’ structure proposed by Volker (1960) was modified by Umemoto et al. (1987) to better fit NMR data. Ferris et al. (1972, 1978, 1981) do not propose a structure for HCN oligomers or polymers, but have shown that diaminomaleonitrile (DAMN) is formed rapidly and irreversibly from HCN and they present evidence that this tetramer is the precursor that condenses to larger HCN oligomers (Figure 1c). From the standpoint of chemical evolution, the most intriguing hypothesis is that of Matthews (Figure 1b), based on the observation that the water-soluble yellow-brown solids can be hydrolyzed to produce significant yields of -amino acids including glycine, alanine, aspartic acid, glutamic acid, and other protein and non-protein amino acids. Matthews’ model involves HCN polymerization via aminomalononitrile, a trimer of HCN, to form a polyamidine which, on contact with water, yields a heteropolypeptide directly (Matthews, 1992 and references cited therein) without the intervening formation of -amino acids, which are later hydrolysis products. Note that all of these hypotheses involve formation of aminonitrile intermediates, and mechanistic assumptions that the nucleophilic additions of amines to nitriles are reversible, as is the hydrolysis of imines if water is present. The hydrolysis of amidines, however, is not reversible in known examples. Therefore, polymerization of HCN in the presence of water or exposure to water of HCN polymers prepared in the absence of water should produce polymers with different structural features. The effects of temperature, pressure (in both gas and condensed phase reactions), and electrical or UV fluxes have not been fully studied and may also lead to differences in structure and molecular weight distribution. Several instrumental methods have been used for the analysis of these complex polymers, including liquids NMR, differential thermal analysis, pyrolysis GC-MS, fluorescence (Ferris et al., 1981), CPMAS solid-state NMR spectroscopy (McKay et al., 1984), Fourier-transform IR photoacoustic spectroscopy and pyrolysis-mass spectrometry (Liebman et al., 1994). Unfortunately, except for nitrile and amine groups, few of these techniques have definitively substantiated or eliminated the
STRUCTURAL INVESTIGATIONS OF HYDROGEN CYANIDE POLYMERS
463
Figure 1. Formation pathways and/or proposed structures for HCN polymer or polymers: (a) V¨olker (1960) and Umemoto et al. (1987); (b) Matthews (1966); (c) Ferris (1986).
464
R. D. MINARD ET AL.
other functionalities – amidines, amides or imines – hypothesized to be present in substantial amounts in these various structural models. Most suggestive, perhaps, has been the detection of secondary amide structures by solid-state NMR. HCN polymers have been produced under a wide variety of conditions. Table I lists HCN polymer samples (P1, P2, D, and F) from a number of different sources and the conditions under which they were formed. For comparison, two samples of dark brown solids obtained from gaseous electrical (spark) discharge experiments using CH4 /NH3 /H2 O (sample SP1) and CH4 /NH3 (sample SP2) mixtures were analyzed. In order to obtain additional structural information for HCN polymers and to compare the polymeric HCN samples with each other, we have applied a number of mass spectrometric ionization methods including fast atom bombardment mass spectrometry (FAB-MS), thermal desorption/pyrolysis electron ionization mass spectrometry, electrospray (ESI) MS, atmospheric pressure chemical ionization APCI mass spectrometry, and TMAH thermochemolysis/GC-MS. TMAH thermochemolysis has been used to characterize a wide variety of complex heterogeneous materials such as lignin, cutin, humic acids, coal and protein remnants in sediments (McKinney et al., 1995, del Rio et al., 1997). Heating these materials with tetramethylammonium hydroxide (TMAH) at 250 C for 30 min produces methylated bond cleavage products which can be separated and identified by GC-MS, and these products provide great insight into the structure of the original macromolecular material. The in-situ methylation by TMAH makes many of the more polar thermally-assisted bond cleavage products volatile and more amenable to GC-MS analysis. Many of these products are larger substructural units preserving more of the original macromolecular structure and yielding a more complete composite of the original structure.
2. Experimental Details of the preparation of polymeric HCN samples P1, P2 and spark reaction solids SP1 and SP2 are given in Matthews and Ludicky (1992). Sample D was obtained from the Cornell group who acquired it from an E.I. duPont de Nemours & Co. HCN manufacuring facility (Khare et al., 1994). Sample F was obtained from Fumico, Inc. (Newark DE). Positive fast atom bombardment mass spectrometry (FAB-MS) was carried out using glycerol as the matrix and 5 kV Xe atoms on a Kratos MS-50 mass spectrometer while scanning from m/z 130 to 2000 once every 15 sec. In-source thermal desorption/pyrolysis electron ionization mass spectra (m/z 25 to 1000) were acquired once every 15 sec on a Kratos MS-9/50 mass spectrometer while the sample ( 100 g) was heated at the rate of 20 C min 1 from 25 to 600 in a quartz capillary inserted into an EI source. Positive electrospray ionization, ESI-MS, and atmospheric pressure chemical ionization, APCI-MS, analyses were carried out on 0.1 mg mL 1 solutions of HCN polymers in 50% acetonitrile/water
STRUCTURAL INVESTIGATIONS OF HYDROGEN CYANIDE POLYMERS
465
Table I Summary of thermal desorption/pyrolysis EI-MS and TMAH thermochemolysis/GC-MS analysis data for samples from hydrogen cyanide polymerization and electrical discharge experiments
Sample and source
Thermal desorption/pyrolysis EI-MS Temperatures at desorption maxima ( C) and m/z of ions observed
P1 Black polymer from HCN in CH3 CN with a trace of (C2 H5 )3 N
170 : m/z 108, M+ of (HCN)4 240 : m/z 136 (HCN)5 + H? m/z 164 (HCN)6 + 2H? 430 : m/z 135 (HCN)5 ?
P2 Black solid from neat HCN with a trace of NH3 SP1 brown solid from CH4 + NH3 + H2 O ‘spark’ reaction
SP2 Brown solid from CH4 + NH3 ‘spark’ reaction
D Black HCN polymer from pipes in Dupont HCN production facility (Khare, 1993) F Black HCN polymer from ‘spontaneous’ polymerization of HCN in a metal storage cylinder from Fumico, Inc.
Similar to P1
170 : m/z 27, HCN m/z 44, CO2 m/z 60, CH3 CO2 H complex spectra up to m/z 110 no m/z 108, (HCN)4 no distinct maxima up to 500 Above 400 , complex spectra with ions up to m/z 300; aliphatic series m/z 43, 55, 69; aromatic series 77, 91, 105. no m/z 108, (HCN)4 110 : m/z 44, CO2 m/z 60, CH3 CO2 H 270 : m/z 126, 111, 59, 58, 44 no m/z 108, (HCN)4 200 : m/z 135 (HCN)5 ? m/z 126, 111, 43 no m/z 108, (HCN)4
TMAH thermochemolysis products Products 1–14 as shown in Figure 2
Similar to P1
Products 1–7 trace amount of trimethyl-triazinetrione, 11
Products 1–7 medium amount of trimethyl-triazinetrione, 11. trace of imine 12.
Products 1–7 trimethyl-triazinetrione, 11, and imine, 12 dominate.
Products 1–7 trace of trimethyl-triazinetrione, 11 trace of trimethyladenine, 13, detected
466
R. D. MINARD ET AL.
using an Analytica (Branford, CT) electrospray source on a Kratos MS-50 mass spectrometer. The solution was flow injected at 1 L min 1 while scanning from m/z 50 to 1400 at 10 sec/decade. TMAH thermochemolysis was carried out by placing 0.5–1 mg of sample in a 2-mL glass ampoule adding 0.1 mL of 25% tetramethylammonium hydroxide (TMAH) in methanol. The methanol was removed and the sample sealed under vacuum. The sealed tubes were heated for 30 min at 250 C, cooled, opened and washed out with CH2 Cl2 . The CH2 Cl2 was removed by a stream of N2 , and 0.1 mL of an eicosane internal standard solution (3.2 mg mL 1 in CH2 Cl2 ) was added. GC/MS analysis was carried out using a Hewlett Packard 5972A MSD (Palo Alto, CA). Approximately 0.5 L of a CH2 Cl2 solution was injected onto a J&W DB5 capillary column (30 m 0.25 mm i.d. with 25 m film) in a Hewlett Packard 5890 GC. The column was programmed from 40 C to 300 C at 10 min 1 with an initial hold of 1 min and a final hold of 10 min. Spectra (m/z 35–450) were acquired once each 0.8 sec. Identification of certain peaks was accomplished by computer-aided matching with the NBS Mass Spectral Library (75,000 spectra; NIST, Gaithersburg, MD) or the Wiley Mass Spectral Library (138,000 spectra; Palisades Corp., Newfield, NY).
3. Results and Discussion 3.1. THERMAL DESORPTION/PYROLYSIS EI-MS Heating the samples listed in Table I from room temperature to 600 C in the EI source of a mass spectrometer while acquiring spectra continuously showed different species being sequentially released with increasing temperatures as evidenced by maxima in the single m/z vs. temperature plots for each detectable species. At lower temperatures, evolution of certain species may be due to lower molecular weight molecules, for example the HCN tetramer, DAMN, thermally desorbing from the macromolecular matrix while at higher temperatures, the detected species arise from thermally-assisted bond cleavage or pyrolysis of the macromolecular structure. The former case seems to be true for HCN polymers P1 and P2 which both showed a large amount of HCN tetramer, presumably DAMN (MW = 108), desorbing at low temperature (100–120 ) with possible pentamer or hexamer species coming off at higher temperatures. The presence of DAMN in the black solid is in agreement with the studies by Ferris (1978). None of the other polymeric HCN samples showed the desorption of significant amounts of the tetramer. Two samples, D and SP1 showed intense ions attributable to CO2 and acetic acid evolving at relatively low temperatures, less than 200 C. Khare et al. (1981) had reported low temperature evolution of CO2 and acetic acid and a number of nitriles in the sequential pyrolysis of products (called ‘tholins’) from electrical discharge in CH4 /NH3 /H2 O mixtures. Decarboxylation of -carbonyl carboxylic acids would
STRUCTURAL INVESTIGATIONS OF HYDROGEN CYANIDE POLYMERS
467
take place at these low temperatures. The source of acetic acid was unclear. The mass spectra from samples SP1 and SP2 were much more complex throughout the temperature range, and ions attributable to extended aliphatic structures and aromatic components were present and appeared to be very similar to those reported by Ehrenfreund et al. (1994). Sample F produced ion intensity at m/z 135 which may be attributable to HCN pentamer. Altogether, these data seem to suggest that HCN polymers produced from HCN under non-aqueous conditions with acetonitrile solvent present (P1) or not present (P2), are quite different from the polymeric material produced in spark reactions (SP1 and SP2) and that these, in turn, are different from samples (D and F) obtained under other conditions. 3.2. FAB-MS, ESI-MS AND APCI-MS FAB-MS spectra of these samples run in glycerol and other liquid matrices yielded very low ion intensities in almost all cases and were not very definitive. FAB-MS is subject to great selectivities in terms of sensitivity and it would appear that these materials do not analyze well using FAB. The most positive result was from a freshly prepared P1 sample which yielded ions at m/z 118 [MH+ for (HCN)5 NH4 ?] and m/z = 235 [MH+ for dimer of (HCN)5 -NH4 ?]. Preliminary experiments designed to follow the early stages of HCN oligomerization using ESI-MS and APCI-MS were also disappointing. 3.3. TMAH THERMOCHEMOLYSIS/GC-MS Pyrolysis/GC-MS has been applied to tholins by Ehrenfreund et al. (1994) and lower molecular weight pyrolysis products, primarily HCN, C-2 to C-5 nitriles, pyridine, pyrole, benzene and toluene were observed among the products. Pyrolysis/GCMS of HCN oligomers by Ferris et al. (1981) detected a number of low molecular weight compounds (H2 O, HCN, CH3 CN, CO2 , formamide, (CN)2 and pyridine). They observed that 65% of the sample remained in the residue even after heating to 1000 C indicating that pyrolysis/GC-MS gives a rather incomplete picture of HCN oligomers. Pyrolysis in the presence of added tetramethylammonium hydroxide produces a much broader suite of structurally-significant products for two reasons; the hydroxide can cause cleavage or ‘chemolysis’ of certain functional groups and, at high temperatures, the tetramethylammonium group can methylate polar functional groups producing volatile products which pass through the capillary gas chromatography column. We have found that this procedure can be run at lower temperatures, 250 to 300 C, in sealed tubes, producing identical results, but allowing for better reproducability and quantitation (McKinney et al., 1995) and have called this method TMAH thermochemolysis. The utility of this technique has been demonstrated for a wide variety of biopolymers in complex matrices such as wood or sediments. When TMAH thermochemolysis was applied to polymeric HCN sample P1, a very complex mixture of methylated products was detected by GC-MS as shown in
Figure 2. GC-MS total ion chromatogram of the TMAH thermochemolysis products of HCN polymer P1 with structures of the identified methylated derivatives.
468 R. D. MINARD ET AL.
STRUCTURAL INVESTIGATIONS OF HYDROGEN CYANIDE POLYMERS
469
Figure 2. Eleven compounds have been positively identified as their methyl derivatives by comparison of reference spectra and relative retention times: aminoacetonitrile (as the N,N-dimethyl derivative, 1), cyanamide (as its N,N-dimethyl derivative, 2), formamide (as methyl formamide, 3, and as dimethyl formamide, 4), carbamic acid (as the N-methyl methyl carbamate, 5 and its N, N-dimethyl methyl carbamate, 6, derivatives), glycine (as N,N-dimethylglycine, methyl ester, 7), succinimide (as the N-methyl derivative, 9), sym-triazinetrione (as the N,N,N-trimethyl derivative, 11), adenine (as the N,N,N’-trimethyl derivative, 13), and xanthine (as the trimethyl derivative, caffeine, 14). Three other compounds were tentatively identified: alanine (as N,N-dimethylalanine, methyl ester, 8), glutarimide (as N-methyl glutarimide, 10), and the monoimine of sym-triazinetrione (as its N,N,N-trimethyl derivative, 12) based on molecular weight, major ion fragments, and relative retention times. As can be seen in Figure 2, the lower molecular weight compounds 1 to 7 were the major products from TMAH thermochemolysis/GC-MS analysis. Analysis using a different GC temperature program than that employed in Figure 2 led to complete separation of all of these low molecular weight components (Minard and Matthews, 1994). For the most part, these products did not provide much insight into the original structure of P1. Product 2 implied an N-C-N substructure and products 1 and 7 implied an N-C-C-N substructural pattern exists. Of greater consequence was the observation that these compounds were produced in large amounts by all of the samples listed in Table I in roughly the same ratios as shown in Figure 2. Thus, it appears that these TMAH thermochemolysis-derived compounds could be used as diagnostic ‘signatures’ for the presence of HCN polymers and/or their mild hydrolysis products in samples of diverse origin. In addition to these smaller acyclic products, a number of interesting nitrogen heterocycles are observed, compounds 9 to 14. The succinimide 9 and glutarimide 10 are noteworthy in that they imply that reduced -CH2 -CH2 - and -CH2 -CH2 -CH2 substructures are present in the HCN products. Reduction of CN groups to -CH2 is necessary for the formation of the side groups of the amino acids alanine, aspartic acid and glutamic acid that have been observed as acid hydrolysis products of HCN polymers. The sym-triazinetrione derivative 11 was also detected in all of the samples listed in Table I and thus may also be considered as a possible candidate for determining the presence of HCN polymers in a sample. The amount varies considerably between samples, however, and this may be be due to variations in temperature or concentrations of base or water during the polymerization. The imine analogue 12 was also detected in samples P1, P2, SP2, and D. These compounds imply that another possible structural motif, an extended triazine structure (15), may be part of HCN polymers (Scheme 1). The formation of the trimethyl adenine derivative, 13, and trimethyl xanthine derivative, 14, in the TMAH thermochemolysis of samples P1 and P2 implies the presence of these biologically important structures in HCN polymers. Or´o and coworkers (1962) had previously detected adenine from reactions of HCN
Figure 3. Total ion current chromatogram from GC-MS analysis of products from the TMAH thermochemolysis of: (a) HCN polymer P1; and (b) DAMN. Both samples had 25% methanolic solution of TMAH added 12 hr before solvent removal, sealing and heating. Data acquisition was delayed for 7 min and therefore compounds 1 through 8 are not observed. Refer to Figure 2 for the structures of methyl derivatives 9 to 14. The numbers above peaks are the molecular weight of that component. If the number is italicized, then that component is found in both (a) and (b).
470 R. D. MINARD ET AL.
STRUCTURAL INVESTIGATIONS OF HYDROGEN CYANIDE POLYMERS
471
Scheme 1.
in ammonium hydroxide. Compound 13 was also detected in sample F. Both compounds 13 and 14 may be present in the other samples, but we were unaware of their formation when these samples were analyzed, and did not look for them at that time. A more thorough analysis of all these samples is planned. For the analysis shown in Figure 2, the methanol was removed under vacuum immediately after the addition of the 25% TMAH methanol solution and the sample thermochemolyzed shortly thereafter. When P1 was allowed to stand at room temperature overnight in the TMAH solution before solvent removal and thermochemolysis, a much greater variety of products was produced and the amount of adenine and xanthine detected was also much greater (Figure 3a). Even more intriguing was the observation that similar treatment of DAMN caused the formation of an opaque dark brown solution which also produced numerous products on thermochemolysis (Figure 3b). This base-catalyzed transformation of DAMN via HCN was first observed by Moser et al. (1968). At least ten of these products, including 9, 10, 11, 13 and 14 are common to HCN polymers and tetramer. It appears that the basic TMAH solution catalyzed extensive transformation of both P1 and DAMN and that there must be some common intermediates since the product distributions were so similar. Many of the products have molecular weights over 200 Da as shown in Figure 3. Note that DAMN produces a more complex mixture than polymeric HCN sample P1 under these conditions. The structural characterization of these compounds should provide considerable insight into the macromolecular structure of HCN polymers. While we have not been able to fully interpret the mass spectral data for most of these products, they all show molecular and fragment ions compatible with methyl analogues of various nitrogen heterocycles including pyrimidines, purines, imidazoles, and pteridines as shown in Scheme 2. The ladder polymers of HCN shown in Figure 1a would seem to be the source of these nitrogen heterocycles, whereas polyamidine components like those shown in Figure 1b could have given rise to the aminoacetonitrile, glycine, and alanine derivatives observed in TMAH thermochemolysis. One intriguing possibility is that some HCN polymers are heterogeneous composites of these two types of structures, consisting of a polyamidine backbone with attached nitrogen heterocycle segments, a hypothetical structural motif reminiscent of the peptide nucleic acids or PNA’s
472
R. D. MINARD ET AL.
Scheme 2. Possible nitrogen heterocycle structures from polymeric HCN sample P1, and HCN tetramer (DAMN) TMAH thermochemolysis products. R groups can be one of the following possibilities: -OH; -CH3 ; -OCH3 ; -CHO; -NH2 ; -H. R’ groups can be =O; -H/-OH; -H/-H.
proposed by Nielsen (Cherny et al., 1993 and Wittung et al., 1994). These could be formed in a great variety of ‘sequences’ which could interact with one another via hydrogen bonds. This model of dynamic, reversible molecular interactions between a myriad of reactive HCN polymers provides an alternative hypothesis for the molecular emergence of life on earth, one which can be tested experimentally through further studies on the structural details of HCN polymers (Matthews, 1994). In conclusion, TMAH thermochemolysis/GC-MS of the dark brown/black materials produced from HCN polymerization, electrical discharge experiments of gaseous NH3 /CH4 /H2 O mixtures and base-catalyzed transformations of DAMN reveals a complex suite of products, many of which are common to all these materials. It seems this new analytical method is a better diagnostic test for the presence of HCN-derived polymers than currently-used IR, NMR, and pyrolysis techniques, though, of course, these techniques can supply other significant structural information. Identification of the resulting TMAH thermochemolysis products has already provided new insight into the substructural motifs that are present in HCN-derived macromolecules. Continued structural elucidation by TMAH thermochemolysis/GC-MS of these and other products, including tholins and organic material of meteoritic origin, should greatly add to our understanding of HCN polymers and their possible roles in chemical evolution.
Acknowledgements We wish to thank Robert Widing for preparing some of the polymeric HCN samples, Michelle Polinko for assisting with the TMAH thermochemolysis analyses, and Bishun Khare for providing sample D. RCG thanks the sponsors of the FlemingMeyer Analytical Chemistry Award for summer research support.
STRUCTURAL INVESTIGATIONS OF HYDROGEN CYANIDE POLYMERS
473
References Cherny, D. Y., Beloservkoskii, B. P., Frank-Kamenstskii, M. D., Egholm, M., Buchardt, O., Berg, R. H. and Nielsen, P. E.: 1993, Proc. Natl. Acad. Sci. USA 90, 1667. del Rio, J. C., McKinney, D. E., Knicker, H., Nanny, M. A., Minard, R. D. and Hatcher, P. G.: 1997, Energy and Fuels (submitted). Ehrenfreund, P., Boon, J. J., Commandeur, J., Sagan, C., Thompson, W. R. and Khare, B.: 1994, Adv. Space Res. 15(3), 335. Ferris, J. P., Donner, D. B. and Lotz, W.: 1972, J. Am. Chem. Soc. 94, 6968. Ferris, J. P. and Edelson, E. H.: 1978, J. Org. Chem. 43, 3989. Ferris, J. P., Edelson, E. H., Auyeung, J. M. and Joshi, P. C.: 1981, J. Mol. Evol. 17, 69. Ferris, J. P. and Hagan, W. J. Jr.: 1984, Tetrahedron 40, 1093. Gay-Lussac, J. L.: 1815, Ann. Chim. Physique (1) 95, 158. Khare, B. N., Sagan, C., Zumberge, J. E., Sklarew, D. S. and Nagy, B.: 1981, Icarus 48, 290. Khare, B. N., Sagan, C., Thompson, W. R., Arakawa, E. T., Suits, F., Callcott, T. A., Williams, M. W., Shrader, S., Ogino, H., Willingham, T. O. and Nagy, B.: 1985, Adv. Space Res. 4, 59. Liebman, S. A., Pesce-Rodriguez, R. A. and Matthews, C. N.: 1994 Adv. Space Res. 15, 71. Moser, R. E., Clagett, A. R. and Matthews, C. N.: 1968, Tetrahedron Letters 13, 1599. Matthews, C. N. and Ludicky, R.: 1992, Adv. Space Res. 12, 1230. Matthews, C. N.: 1992, Origins of Life 21, 421. Matthews, C. N.: 1994, J. Biol. Phys. 20, 275. McKay, R. A., Schaefer, J., Stejskal, E. O., Ludicky, R. and Matthews, C. N.: 1984, Macromolecules 17, 1124. McKinney, D. E., Carson, D. M., Clifford, D. J., Minard, R. D. and Hatcher, P. G.: 1995, J. Anal. and App. Pyrolysis, 34. Minard, R. D. and Matthews, C. N.: 1994, Origins of Life Symposium, 207th National Meeting of the American Chemical Society, San Diego, CA, March 1994. Or´o, J. and Kimball, A. P.: 1962, Arch. Biochem. Biophys. 96, 293. Proust, J. L.: 1807, Gehlen’s J. Chem. Physik, 3, 384; 1806, Ann. Chim. Physique (1), 60, 233. Umemoto, K., Takahashi, M. and Yokata, K.: 1987, Origins Life Evol. Biosphere 17, 283. V¨olker, T. H.: 1960, Angew. Chem. 72, 379. Wittung, P., Nielsen, P. E., Buchardt, O., Egholm, M. and Norden, B.: 1994, Nature 368, 561.
