Reactivity and Survivability of Glycolaldehyde in Simulated Meteorite Impact Experiments
V.P. McCaffrey1*, N.E.B. Zellner2, C.M. Waun1, E.R. Bennett1, and E.K. Earl1
1Department of Chemistry, Albion College, Albion, MI 49224, USA
2Department of Physics, Albion College, Albion, MI 49224, USA
*, 517-629-0622 (Phone), 517-629-0264 (Fax)
Supplemental Materials and Methods
Supplemental Text
Supplemental Figures
Figure S1 – Shock Experiment Targets
Figure S2 – GC of Neat, Unshocked GLA
Figure S3 – Gas chromatogram of GLA/4.6 GPa (shot 3597)
Figure S4 – Gas chromatogram of GLA/9.4 GPa (shot 3603)
Figure S5 – Mass Spectrum of Glycerol, shocked and reference
Figure S6 – Gas Chromatogram (14.0 – 22.0 min) of GLA/4.6 GPa (shot 3597) with corresponding MS.
Figure S7 – Gas Chromatogram (18.5 – 22.0 min) of GLA/9.4 GPa (shot 3603) with corresponding MS.
Figure S8 – GC of authentic, TMS derivatized D-thresose with corresponding MS and structures.
Figure S9 – GC of authentic, TMS derivatized D-erythrose with corresponding MS and structures.
Figure S10 – Gas Chromatogram (20.5 – 22.0 min) of GLA/clay/4.6 GPa (shot 3598) with corresponding MS.
Supplemental Tables
Table S1 – Summary of shock experiments
Table S2 – Retention times
Supplemental References.
Supplemental Materials and Methods
Method: Impact Experiments
The generation of a desired shock stress relies on the impact of a flat projectile onto the flat surface of a target-bearing assembly, ideally with both surfaces parallel. Given the metal composing the target assembly, both the composition of the projectile and its speed can be used to control the peak shock stress. The projectile is usually a metal flyer plate set into the front of a Lexan cylinder, although a projectile without a flyer plate can also be used; flyer plates are typically 22 mm in diameter and 2 mm thick. The projectile's orientation just before impact is determined through orthogonally mounted cameras and a high-speed flash unit.
The sample is placed in a shallow well (10 mm in diameter and 0.76 mm deep) and capped by a thin layer of the desired metal (stainless steel 304). Referring to Figure S1a, the female piece provides the metal cap; the assembled male and female components are threaded into the coupling ring, which is then pressed into a receptacle in a massive metal cylinder (Figure S1b). The sample end of the entire assembly is faced flat on a lathe in stages to prevent overheating of the sample. At the time of the experiment, the entire sample lies just under the surface at one end of the cylinder, which is used primarily to trap the momentum of the projectile (thus minimizing the possibility of damage to the impact chamber by high-speed fragments). It also serves to confine the shocked sample to a limited volume from which it can be extracted following the impact.
Following the experiment, the metal cylinder is mounted in a lathe and the impacted end is machined away very gradually so as to minimize thermal effects from the machining process. Eventually, after enough material has been removed, the metal covering the sample can be pried away and the sample extracted.
The shock-reverberation technique provides a means of attaining high shock stresses at relatively low impact speeds (~2 km s-1). It does so, however, through a series of reflected shocks, a process that is characterized by a lower entropy production than would be generated by a single shock of the same peak stress (Gibbons and Ahrens, 1971). Thus, the shock-induced effects presented and discussed below represent the minimum that would be generated by a single shock of the same amplitude.
As described in Peterson et al. (1997), shocked samples were extracted from the target assembly as quickly as possible after impact in order to limit continuing chemical reactions that could result from residual heat in the target assembly. The vacuum chamber was rapidly repressurized and the target was recovered and machined to provide access to the sample well. During the milling process, the surface of the target holder and assembly were monitored with a remote, hand-held thermometer; surface temperatures of the metal did not rise more than a few degrees above room temperature.
(A) (B)
FIGURE S1. (A) Male (containing the sample well, outlined in black) and female target assembly pieces, which are screwed together via the coupling. (B) Stainless steel target holder, showing the placement of the sample-filled target assembly. The thin gap near the center is typical of the intersection of the flat surface and one of the threads in the coupling.
Control experiments with neat (no mineral matrix added) GLA, a fine-grained white powder (Aldrich), were performed at three impact pressures (Table S1) to ensure that changes seen were due to reactions at the clay surface and not to autocatalysis or reactions with the sample chamber. The GLA was tamped into the sample well and the target assembly was then pressed into the stainless-steel target holder. For masses used, please see Table S1. For each experiment, the entire target was then placed into the impact chamber. The chamber was evacuated to a pressure below 200 mTorr and the target was then impacted by either a Lexan projectile with no flyer plate, an aluminum (AL 2024) flyer plate, or a stainless steel (SS 304) flyer plate, depending on the desired shock stress. Given the composition of the projectiles and the flyer plates, with velocities averaging ~1.1 km s-1, the samples experienced shock pressures ranging from 4.6 to >25 GPa (see Table S1).
Method: Chemical Analyses
In all cases, the samples’ organic materials were dissolved in tetrahydrofuran (THF; passed over molecular sieves and alumina to remove water) prior to derivatization and analysis. With the neat GLA, approximately 30 mg of the sample were dissolved in THF and sonicated to ensure complete dissolution. The resulting solution was derivatized with nitrogen-purged BSTFA (N,O-bis(trimethylsilyl)trifluoroacetamide; Fisher Scientific) to form the trimethylsilyl ethers prior to analysis by gas chromatography-mass spectrometry (GC/MS).
Supplemental Text
Analysis of Unshocked Neat GLA
Before analysis, the alcohol groups of the GLA were derivatized with trimethylsilyl (TMS) groups. The resulting silylated compounds have increased volatility and are more easily analyzed by gas chromatography. TMS ethers of sugars and related compounds also have several common and diagnostic fragmentation patterns that allow for more facile identification of unknown compounds. A representative gas chromatogram of the unshocked, silylated GLA in THF is shown in Figure S2. The predominant component of the sample is the 6-membered ring dimer (B) at 19.26 min. Also present are peaks representing the monomer (A) at 14.68 min and the 5-membered ring dimer (C) at 19.32 min. The five- and six- membered ring dimers were identified by comparison of the MS to published spectra (Novina 1984). Small peaks are present at 13.3 and 15.7 min. These are products from side reactions that occur during the functionalization of the GLA. The reaction conditions have been optimized for complete silylation of the hydroxyl groups and no evidence of the mono-silylated GLA dimers was seen.
FIGURE S2. Representative gas chromatogram of neat, silylated GLA showing the three species in equilibrium in THF solution. The peaks correspond to the compounds shown in Figure 1 in the main text. Decane, an internal standard, is marked with a *. All compounds were identified by comparison of their mass spectra to literature mass spectra (Novina 1984).
Analysis of Shocked Neat GLA: Physical Characteristics
When the neat GLA/4.6 GPa sample was recovered, it was a solid that showed signs of darkening around the outer edges of the sample, Table S1. When the target assembly from the shot with GLA/9.4 GPa was opened, the sample was recovered partially as a dark brown liquid that seeped from the solid sample. The viscous liquid sample was collected using cotton swabs and stored separately from the solid sample. Before analysis of the samples, it was noted that they had dried to a solid crust and were no longer viscous. There was no material recovered from the high pressure shot with GLA/25.6 GPa. Most of what remained was a black crust on the walls of the container. It was attempted to obtain measurable material by soaking the sample wells in tetrahydrofuran (THF) before analysis, but no results were obtained.
Shocked Neat GLA: Chemical Analyses
The neat GLA samples at the lowest pressure investigated (4.6 GPa, Figure S3) showed no significant changes in the composition after derivitization and analysis by GC/MS. New peaks at retention times between 16 and 18 minutes were seen, but nothing at retention times longer than 20 minutes was detected, indicating that threose and erythrose were not being formed without the presence of the clay matrix. The new peaks had extremely low abundances and the resulting mass spectra did not lead to conclusive identification. Briefly, fragments of 73, 105, 135, 147, and 239 m/z are seen in the EI-MS of each of the five peaks indicated on the GC (Figure S6). The presence of these fragments suggests that compounds with one or more hydroxyl functional groups are being synthesized in the impact but as can be seen in the GC, at very low abundances (Petersson 1969, 1970).
FIGURE S3. Gas chromatogram of GLA/4.6 GPa (shot 3597). See Figure 1 in the main text for structures of A-C. New compounds are marked with a +.
When the pressure of the shot with neat GLA was increased to 9.6 GPa, however, changes can be seen in the gas chromatogram at longer retention times (Figure S4). In addition to the GLA monomer peak at 14.68 min (not shown) and the dimer peaks at 19.2 min, a new peak at 18.84 min, corresponding to glycerol (Figure S5), was seen. Additionally, there are several new peaks found between 19.75 and 21.25 min labeled with a plus sign (Figure S7). The mass spectra of each of these peaks show 73, 103, and 117 m/z fragments, suggesting that the newly synthesized compounds have one or more hydroxyl functional groups in them. However, none of these peaks had MS or retention times that were consistent with threose or erythrose. Additionally, no peaks for ethylene glycol were found in this sample either.
FIGURE S4. Gas chromatogram of GLA/9.4 GPa (shot 3603). For structures of identified peaks, see Figure 1 in the main text. New peaks in the shocked sample are identified with +. Peaks with mass fragments of 147 and 205 m/z are at 19.83, 20.25, and 20.35 min.
Analysis of the GLA/25.6 GPa shot (sample 3596, Table S1) was not possible because no material was recovered due to decomposition of the sample. Only a black residue was left on the inside of the target assembly. This material was not recovered or analyzed. This complete loss of the GLA is consistent with similar studies of amino acids, in which Bertrand et al. showed a steep decline in the survival of amino acids at pressures > 220 kbar (22 GPa; Bertrand et al. 2009).
After analysis of the shocked neat GLA samples, few changes were seen in the gas chromatograms. As the pressure of the impact was increased, additional peaks were seen in the chromatogram but due to their extremely low abundances, conclusive identification of these new compounds was not possible. Analysis of the mass spectra of the new peaks showed fragmentation patterns that are consistent with compounds containing OTMS groups in the derivatized compounds. Current studies in our lab are aimed at elucidating the exact structure of these new compounds.
FIGURE S5. A. Mass spectrum of the authentic sample of tri-TMS derivatized glycerol. B. Mass spectrum of 18.95 min peak (Figure S4) from unshocked GLA/clay.
FIGURE S6: Gas Chromatogram (14.0 – 22.0 min) of GLA/4.6 GPa (shot 3597) with corresponding MS.
16.36 min
16.77 min
FIGURE S6: cont.
17.44 min
17.54 min
17.91 min
FIGURE S7: Gas Chromatogram (18.5 – 22.0 min) of GLA/9.4 GPa (shot 3603) with corresponding MS.
19.83 min
20.25 min
FIGURE S7: cont.
20.35 min
20.50 min
21.25 min
Figure S8: GC of authentic, TMS derivatized D-thresose with corresponding MS and structures.
20.81 min:
20.88 min:
20.98 min:
Figure S9: GC of authentic, TMS derivatized D-erythrose with corresponding MS.
20.84 min: impurity
20.90 min: erythrose
21.15 min: erythrose
FIGURE S9: cont.
21.50 min: impurity
21.79 min: impurity
Figure S10: GC (20.5 – 22.0 min) of GLA/clay/4.6 GPa (shot 3598) with corresponding MS.
20.81 min assigned to threose
20.88 min assigned to threose
20.91 min assigned to erythrose
FIGURE S10: cont.
20.97 min
21.16 min assigned to erythrose
21.67 min unidentified
21.83 min unidentified
Table S1. Summary of shock experiments (August 2011, JSC EIL). ND = not measured.
Shot # / Sample / Starting Mass(g) / Density / Recovered Mass
(g) / % Recovered / v
(km/sec) / P
(GPa) / Target/
Flyer Plate
3597 / GLA / 0.062 / 1.1277 / 0.06102 / 98.4 / 1.113 / 4.6 / SS 304/
No flyer plate
3603 / GLA / 0.0546 / 0.9931 / 0.05180 / 94.9 / 0.885 / 9.4 / SS 304/
Al 2024
3596 / GLA / 0.0616 / 1.1205 / 0 / 0 / 1.183 / 25.6 / SS 304/
SS 304
3598 / GLA +
Clay / 0.0903 / 1.6425 / 0.0852 / 94.3 / 1.127 / 4.6 / SS 304/
No flyer plate
3604 / GLA +
Clay / 0.1128 / 1.7394 / 0.1160 / 94.5 / 1.012 / 12.2 / SS 304/
Al 2024
3622 / GLA +
Clay / ND / ND / ND / ND / 1.163 / 25.1 / SS 304/
SS 304
Table S2. Retention times of selected peaks in authentic samples of organic compounds and shocked GLA and GLA/clay mixtures.
Sample / RT (min) / RT (min) / RT (min) / RT (min) / RT (min)Ethylene Glycol / 12.53
Glycolic Acid / 15.35
Glycerol / 18.90
D-Threose / 20.81, 20.88, 20.98
D-Erythrose / 20.92, 21.15
Shot 3603
(GLA only,
9.4 GPa) / 18.86
Shot 3598
(GLA/Clay
4.6 GPa) / 15.27 / 18.92 / 20.81, 20.88, 20.97 / 20.91, 21.16
Shot 3604
(GLA/Clay
12.2 GPa) / 15.26 / 18.92 / 20.81, 20.87, 20.96 / 20.92, 21.15
Shot 3622
(GLA/Clay
25.1 GPa) / 12.51 / 20.81, 20.87, 20.97 / 20.91, 21.15
Supplemental References
Bertrand M, van der Gaast S, Vilas F, Hörz F, Haynes G, Chabin A, Brack A and Westall F (2009) The fate of amino acids during simulated meteoritic impact. Astrobio 9:943-951.
Gibbons RV and Ahrens T J (1971) Shock metamorphism of silicate glasses. J Geophys Res 76:5489-5498.