Table of Contents
Fetching ...

Tuning molecular thermal conductance through endgroup modification and halogen substitution

Jonathan J. Wang, Dvira Segal

TL;DR

This work demonstrates that intrinsic molecular phononic heat transport in carbon-backbone chains can be tuned via endgroup engineering and halogen substitution, analyzed through a three-stage workflow that couples ab initio MD data with machine-learned interatomic potentials to enable efficient nonequilibrium MD. Key findings show CH$_3$ and NH$_2$ endgroups maximize $G_{ m th}$, while heavy halogen substitutions and fluorination suppress conductance, with $G_{ m th}$ becoming nearly length-independent for $N \ge 8$ carbon atoms. The approach provides structure–function insights into mode matching and scattering mechanisms that govern phonon transport in molecules, albeit with limitations related to metal-contact effects and generalizability of ML potentials across chemistries. The results offer a versatile chemical route to control phononic heat flow in single molecules, with potential implications for molecular thermal management design.

Abstract

We demonstrate tuning of the phononic thermal conductance in single molecules with carbon-chain backbones through modifications of terminal groups and halogen substitution of hydrogen atoms. Our simulations focus on intrinsic molecular properties, and we employ a workflow based on {\it ab initio} molecular dynamics, enabling the training and development of machine-learned interatomic potentials. These potentials are subsequently used in classical nonequilibrium molecular dynamics simulations to extract thermal conductance coefficients. Replacing terminal methyl groups with amine, sulfur, or halogen substituents leads to pronounced changes in thermal conductance: bromine-terminated chains exhibit the lowest conductance, whereas amine and methyl-terminated chains show the highest. Additionally, single-atom substitution of hydrogen by fluorine or other halogens along the alkane backbone significantly reduces thermal transport. Finally, our simulations of the length dependence of thermal conductance in alkane chains containing 3-12 carbon atoms reveal its saturation beyond eight carbon atoms. Together, our findings show that simple chemical modifications offer a versatile route to controlling phononic heat flow in single molecules.

Tuning molecular thermal conductance through endgroup modification and halogen substitution

TL;DR

This work demonstrates that intrinsic molecular phononic heat transport in carbon-backbone chains can be tuned via endgroup engineering and halogen substitution, analyzed through a three-stage workflow that couples ab initio MD data with machine-learned interatomic potentials to enable efficient nonequilibrium MD. Key findings show CH and NH endgroups maximize , while heavy halogen substitutions and fluorination suppress conductance, with becoming nearly length-independent for carbon atoms. The approach provides structure–function insights into mode matching and scattering mechanisms that govern phonon transport in molecules, albeit with limitations related to metal-contact effects and generalizability of ML potentials across chemistries. The results offer a versatile chemical route to control phononic heat flow in single molecules, with potential implications for molecular thermal management design.

Abstract

We demonstrate tuning of the phononic thermal conductance in single molecules with carbon-chain backbones through modifications of terminal groups and halogen substitution of hydrogen atoms. Our simulations focus on intrinsic molecular properties, and we employ a workflow based on {\it ab initio} molecular dynamics, enabling the training and development of machine-learned interatomic potentials. These potentials are subsequently used in classical nonequilibrium molecular dynamics simulations to extract thermal conductance coefficients. Replacing terminal methyl groups with amine, sulfur, or halogen substituents leads to pronounced changes in thermal conductance: bromine-terminated chains exhibit the lowest conductance, whereas amine and methyl-terminated chains show the highest. Additionally, single-atom substitution of hydrogen by fluorine or other halogens along the alkane backbone significantly reduces thermal transport. Finally, our simulations of the length dependence of thermal conductance in alkane chains containing 3-12 carbon atoms reveal its saturation beyond eight carbon atoms. Together, our findings show that simple chemical modifications offer a versatile route to controlling phononic heat flow in single molecules.

Paper Structure

This paper contains 9 sections, 9 figures.

Figures (9)

  • Figure 1: The gold–alkanedithiol–gold single-molecule junction represents a typical setup for experiments and MD simulations such as in Refs. JW-HeatMDJW-HeatFluc. In the schematic, the gray regions at both ends indicate fixed atoms, while the red and blue regions correspond to thermostatted atoms maintained at target temperatures, $T_h$ and $T_c$, respectively. Heat current $J$ develops through the system, allowing the thermal conductance to be determined. The purple box highlights the molecule, which is the primary focus of this study. In this work, we investigate heat flow considering the molecular structure alone, with the thermostats applied directly to the molecular terminal atoms, indicated by red and blue arrows.
  • Figure 2: Schematic of a decane ( XC$_{10}$H$_{20}$X) system used to examine the effect of different molecular endgroups (X). In the NEMD production run, Langevin thermostats are applied to the central atom of each endgroup, maintaining a hot temperature $T_h$ and a cold temperature $T_c$ at opposite ends of the molecule. Endgroups are depicted to relative scale.
  • Figure 3: (a) Thermal conductance results of C$_{10}$H$_{20}$ alkane chains with different thermostatted contact groups at the ends. (b)-(e) Select temperature profiles from individual production runs for endgroups of (b) Br, (c) SH, (d) NH$_2$, and (e) CH$_3$. Temperature difference $\Delta T$ are calculated as the difference between temperatures of terminal atoms.
  • Figure 4: (a) Thermal conductance of carbon-based molecules with F substitution. (Left-to right) As a reference point we begin from pure decane that has H atoms (gray). A single fluorine substitution on the alkane backbone is probed at two locations (L1 and L2, hashed orange). Other structures are a chain with two F atoms substituting H at the same C atom (orange), and a fully fluorinated perfluorodecane (yellow). (b) The structures of C$_{10}$H$_{21}$F differing by F substitution location L1 and L2. The F atom is colored in yellow; red/blue colorings indicate thermostatted C atoms.
  • Figure 5: Thermal conductance of alkane chains with a single substitution on the C$_{10}$ backbone at the center of the chain (position L1 from Fig. \ref{['Fig4']}(b)). Molecules are C$_{10}$H$_{21}$X where X stands for H (no substitution); CH$_3$ group; a halogen atom.
  • ...and 4 more figures