Three body photodissociation of the water molecule and its implications for prebiotic oxygen production

Oxygen (O2) is scarce in molecular form in the Universe despite oxygen being the third most abundant element. Geological evidence suggests Earth’s primordial atmosphere lacked oxygen and was dominated by H2O, CO2, and N2. Prebiotic O2 has been attributed mainly to CO2 photodissociation by VUV photons producing O atoms that recombine to O2. Direct O2 formation pathways from CO2 photodissociation and dissociative electron attachment have also been identified. Water, a dominant oxygen carrier, has been assumed to photodissociate primarily to OH + H, contributing little to O2. However, the strong correlation of O2 with H2O in the coma of comet 67P suggests water-related O2 formation not captured by current models. The research question addressed here is whether VUV-induced three-body dissociation of H2O producing O atoms is a significant, previously overlooked source of prebiotic oxygen across relevant astrophysical environments, including early Earth, interstellar clouds, and comets.

Prior models considered VUV photodissociation of CO2 followed by three-body recombination as the main abiotic O2 source in primitive atmospheres. Recent work revealed direct O2 production from CO2 photodissociation and via dissociative electron attachment. Observations of abundant O2 in comet 67P’s coma, with a strong O2–H2O correlation, indicate water-linked formation pathways underappreciated by existing photochemical schemes. Extensive experimental and theoretical studies of water photodissociation show that excitation to Ã1B1 (~160 nm) yields H + OH(X) with little internal excitation, while B1A1 (~128 nm) mainly produces OH(X) + H after non-adiabatic transitions, with a minor OH(A) + H channel. Three-body dissociation channels to O(3P) + H + H (Eth = 9.513 eV) and O(1D) + H + H (Eth = 11.480 eV) are energetically accessible at shorter wavelengths. O(3P)+2H formation was detected at Lyman-α (121.57 nm) with small yields, but lack of intense tunable VUV sources hindered quantitative assessment of H2O TBD and its role in O2 formation. The advent of the Dalian VUV free electron laser (DCLS) enables comprehensive exploration across the VUV region.

A supersonic beam of H2O (rotational temperature ~10 K) was photodissociated using the tunable VUV free electron laser at the Dalian Coherent Light Source (DCLS). The VUV-FEL (50–150 nm tuning range, ~30–50 cm−1 bandwidth, up to >100 µJ/pulse at 10 Hz) excited H2O in the 90–110 nm range. H-atom products were detected by H-atom Rydberg tagging time-of-flight (HRTOF). Time-of-flight spectra were recorded with detection axes parallel, perpendicular, and at the magic angle (54.7°) relative to the VUV polarization. TOF spectra were converted to total kinetic energy release (TKER) distributions using EKE = mH(1 + mH/mOH)(d/t)2 with d = 28 cm. Sharp TKER structures were assigned to specific ro-vibrational states of OH(X/A)+H from binary dissociation using energy conservation. Broad TKER features were attributed to three-body dissociation to O(1D)+2H (EKE ≤ ~580 cm−1 at 107.4 nm) and O(3P)+2H (up to ~16,448 cm−1 at 107.4 nm), consistent with energetic limits. The product TKER spectra at 107.4 nm were decomposed into three components—O(1D)+2H, O(3P)+2H, and OH+H—guided by simulations and anisotropy analysis, and integrated to obtain relative H-atom yields and branching ratios. Similar analyses were performed at nine wavelengths: 109.0, 107.4, 106.7, 105.7, 101.3, 98.1, 96.2, 94.5, and 92.0 nm. Angular anisotropies were evaluated from parallel and perpendicular TKER distributions to infer dissociation mechanisms. To assess astrophysical relevance, fragment-dependent photodissociation rates JH2O were computed via J = ∫ Φλ Γλ σλ dλ, using reconstructed early solar VUV photon flux (90–200 nm), H2O photoabsorption cross sections, and measured O-atom quantum yields. A polynomial fit to absorption data was used to interpolate yields over wavelength. Uncertainties in branching ratios are ±10%.

- Direct experimental evidence that three-body dissociation (TBD) of H2O is a dominant pathway upon VUV excitation between 90 and 107.4 nm. - Clear observation and assignment of two TBD product channels: O(1D)+2H (intense low-EKE feature with EKEmax ≈ 580 cm−1 at 107.4 nm) and O(3P)+2H (broad feature extending to EKEmax ≈ 16,448 cm−1 at 107.4 nm). - O(1D)+2H channel observed with significant yield for the first time in H2O photodissociation; more than one-third of produced O atoms populate the metastable 1D state. - Branching ratios (TBD vs binary H+OH) at photolysis wavelengths (TBD first; uncertainty ±10%): • 109.0 nm: 0.35 TBD; 0.65 binary • 107.4 nm: 0.67 TBD; 0.33 binary • 106.7 nm: 0.76 TBD; 0.24 binary • 105.7 nm: 0.62 TBD; 0.38 binary • 101.3 nm: 0.77 TBD; 0.23 binary • 98.1 nm: 0.72 TBD; 0.28 binary • 96.2 nm: 0.79 TBD; 0.21 binary • 94.5 nm: 0.86 TBD; 0.14 binary • 92.0 nm: 0.86 TBD; 0.14 binary - Angular anisotropy parameters indicate differing mechanisms: O(1D)+2H is more direct (β ≈ 0.8), whereas O(3P)+2H is more complex with more internal conversion and scrambling (β ≈ 0.2). Observed broad H-atom energy distributions imply predominantly sequential dissociation rather than simultaneous H–H break. - Convolution of early solar VUV flux, H2O absorption cross sections, and measured yields suggests ~21% of H2O photoexcitation events (in the 90–200 nm band) produce O atoms via TBD. - Estimated fragment-specific rates: J_H2O(O) ≈ 5.2 × 10^−5 s^−1 (this work) and J_CO2(O) ≈ 1.8 × 10^−4 s^−1 (literature). Given [H2O] ≈ 10 × [CO2] in certain early Earth scenarios, the O-atom production from H2O could be ~3 times that from CO2 in the same VUV range. - The findings imply that O-atom production from H2O TBD followed by O+O+M → O2 recombination is a significant prebiotic O2 source in environments with intense VUV radiation (e.g., cometary comae, interstellar clouds, early Earth upper atmosphere).

The experiments demonstrate that under VUV irradiation at 90–107.4 nm, H2O predominantly undergoes three-body dissociation to produce O atoms in both 1D and 3P states, rather than the traditionally assumed binary H+OH pathway. The energetic limits and anisotropy of the observed TKER features confirm the assignments and reveal mechanistic differences: a more direct route to O(1D)+2H and a more complex, sequential pathway to O(3P)+2H involving non-adiabatic transitions among excited states and conical intersections. These O atoms can recombine via three-body processes to form O2, providing an alternate abiotic pathway for molecular oxygen formation in VUV-rich environments. Convolution with reconstructed early solar VUV flux and H2O absorption data indicates that roughly one-fifth of H2O photoexcitation events result in O atoms, underscoring the astrophysical significance. In early Earth scenarios where H2O was abundant at high altitudes and the nascent Sun emitted enhanced VUV flux, the H2O route could rival or exceed the CO2 pathway in generating O atoms that subsequently form O2. The sizable yield of reactive O(1D) suggests additional implications for primordial atmospheric chemistry (e.g., oxidation pathways leading to formaldehyde from methane), potentially influencing prebiotic chemical evolution. These insights argue for inclusion of H2O TBD channels in interstellar and planetary photochemical models to more accurately predict oxygen inventories.

This work provides conclusive laboratory evidence that VUV-induced three-body photodissociation of water efficiently produces O atoms, with both O(1D) and O(3P) channels, and that this pathway dominates over binary H+OH fragmentation at 90–107.4 nm. Quantified branching ratios across nine wavelengths and mechanistic analysis from anisotropy and TKER distributions establish mainly sequential dissociation dynamics. When combined with early solar VUV flux and H2O absorption cross sections, the results imply that H2O photochemistry can be a major prebiotic source of O2 through subsequent O-atom recombination in cometary comae, interstellar clouds, and the early Earth’s upper atmosphere. Future work should refine quantum yields across a broader wavelength range, measure state-specific O-atom yields under varied conditions, integrate TBD channels into astrochemical models, and explore implications for other water-rich planetary atmospheres.

- Laboratory conditions (supersonic beam, low rotational temperature, specific polarization geometries) may not fully replicate astrophysical environments. - Branching ratios carry up to ±10% uncertainty; quantum yields were interpolated using polynomial fits to absorption data and limited wavelength sampling. - The early solar VUV flux spectrum is reconstructed and model-dependent; different radiation fields (e.g., ISRF or stellar variability) could adjust absolute rates. - Experiments infer O2 formation indirectly via O-atom production; three-body recombination efficiencies depend on ambient densities and third-body species not probed here. - Only wavelengths between 92 and 109 nm were directly measured; extrapolation outside this range introduces uncertainty. - Mechanistic conclusions (sequential vs concerted) are based on TKER widths and anisotropies without full state-resolved potential energy mapping under all conditions.