Full Reaction Pathway Dynamics for Atmospheric Decomposition Reactions: The Photodissociation of H$_2$COO

TL;DR

The paper tackles the problem of characterizing the full reaction-pathway dynamics and product-state distributions in the photodissociation of the smallest Criegee intermediate H$_2$COO. It employs a refined, full-dimensional ML-PES anchored to CASPT2/aVTZ data and conducts extensive nonequilibrium MD on energized H$_2$COO to quantify energy partitioning among product channels. A key finding is the bifurcating CO$_2$+H$_2$ pathway, proceeding via direct and indirect routes through an OCH$_2$O intermediate and formic acid, with distinct vibrational excitations and non-RRKM lifetimes for the intermediates and products; HCOOH dissociation lifetimes fit stretched exponentials with β values signaling non-statistical dynamics. The work demonstrates the necessity of dynamical simulations to capture channel-specific mechanisms and energy redistribution, providing mechanistic insight into Criegee intermediate chemistry and informing atmospheric modeling of the tropospheric H$_2$ budget.

Abstract

Branching ratios for fragmentation channels of important meta- and unstable species are essential for a molecular-level characterization of atmospheric chemistry. Here, the molecular product channels for the decomposition dynamics of the smallest Criegee intermediate, H$_2$COO, are quantitatively investigated. Using a high-quality, full-dimensional machine learned potential energy surface (CASPT2/aug-cc-pVTZ), the translational, rotational, and vibrational energy distributions of the CO$_2$+H$_2$, H$_2$O+CO, and HCO+OH fragmentation channels were analyzed to elucidate partitioning of the available energy. The CO$_2$ + H$_2$ product forms through two different pathways that bifurcate after formation of the OCH$_2$O intermediate. Along the direct pathway, CO$_2$ is preferentially vibrationally excited with H$_2$in its vibrational ground state, whereas for the indirect pathway going through formic acid, H$_2$ can populate levels with $v > 0$. For all product channels passing through energized formic acid, the lifetime distributions are described by stretched exponentials with $β$ ranging from 1.1 to 1.7. This is a clear signature of non-RRKM effects and suggests that the explicit molecular dynamics needs to be followed for a quantitative and realistic description of the photodissociation dynamics.

Figures (15)

• Figure 1: The unimolecular decomposition of H$_2$COO into three bimolecular products: CO$_2$ + H$_2$, H$_2$O + CO, and HCO + OH. The formation of CO$_2$ + H$_2$ proceeds through two competing pathways: the "direct pathway", which connects OCH$_2$O directly to the products via TS5, and the "indirect pathway", in which OCH$_2$O first isomerizes to HCOOH (formic acid) before forming CO$_2$ + H$_2$.
• Figure 2: Energy distributions for the CO$_2$+H$_2$ product channel. Trajectories follow direct (952) and indirect (268) routes. Top panels: fragment CO$_2$; bottom panels: fragment H$_2$. From left to right: translational, rotational, and vibrational energy distributions. The inset in panel F indicates that $v > 0$ can be populated for H$_2$ along the indirect pathway. Note the different scales along the $x-$axes for panels A-C vs. D-F.
• Figure 3: Energy distributions of the product H$_2$O+CO from 664 simulations. The top panels show fragment H$_2$O and bottom show fragment CO. From left to right, the distributions correspond to translational, rotational, and vibrational energies. The red trace in panel E corresponds to a Boltzmann distribution with $T = 5000$ and suggests that the rotational motion of dissociating CO is close to thermal equilibrium.
• Figure 4: Lifetime (main view) and energy (inset) distributions of OCH$_2$O for the direct (black) and indirect (red) channels. In both cases, OCH$_2$O is short-lived, with lifetimes never exceeding 0.2 ps. In the inset, the median energy (dashed vertical lines) are at 0.2 and 2.4 kcal/mol, for the direct and indirect cases, respectively.
• Figure 5: Normalized probability distributions $P(r_{\rm CH},\theta_{\rm HCH})$ while sampling the OCH$_2$O well for each of the direct (left) and indirect (right) channels. The numbers in the contour lines indicate the normalized intensity of geometries. The dashed lines represent the equilibrium structure of OCH$_2$O, and the geometrical criteria used to assign sampled structures to each product channel are listed in Table \ref{['sitab:criteria']}. The red arrows illustrate the reaction pathways from OCH$_2$O to the corresponding reaction pathways.