Isolating Chemical Reaction Mechanism as a Variable with Reactive Coarse-Grained Molecular Dynamics Step-Growth versus Chain-Growth

2025-05-05 0 0 3.45MB 34 页 10玖币

侵权投诉

Isolating Chemical Reaction Mechanism as a

Variable with Reactive Coarse-Grained Molecular

Dynamics: Step-Growth versus Chain-Growth

Polymerization

John J. Karnes,∗,†Todd H. Weisgraber,†Caitlyn C. Cook,†Daniel N. Wang,†

Jonathan C. Crowhurst,†Christina A. Fox,†,‡Bradley S. Harris,¶James S.

Oakdale,†Roland Faller,∗,¶and Maxim Shusteﬀ†

†Lawrence Livermore National Laboratory Livermore, California 94550, United States

‡Department of Materials Science and Engineering, University of California, Davis, Davis,

California 95616, United States

¶Department of Chemical Engineering, University of California, Davis, Davis, California

95616, United States

E-mail: karnes@llnl.gov; rfaller@ucdavis.edu

Abstract

We present a general approach to isolate chemical reaction mechanism as an inde-

pendently controllable variable across chemically distinct systems. Modern approaches

to reduce the computational expense of molecular dynamics simulations often group

multiple atoms into a single “coarse-grained” interaction site, which leads to a loss of

chemical resolution. In this work we convert this shortcoming into a feature and use

identical coarse-grained models to represent molecules that share non-reactive charac-

teristics but react by diﬀerent mechanisms. As a proof of concept we use this approach

arXiv:2210.01758v1 [cond-mat.mtrl-sci] 4 Oct 2022

to simulate and investigate distinct, yet similar, trifunctional isocyanurate resin formu-

lations that polymerize by either chain- or step-growth. Since the underlying molecular

mechanics of these models are identical, all emergent diﬀerences are a function of the

reaction mechanism only. We ﬁnd that the microscopic morphologies resemble related

all-atom simulations and that simulated mechanical testing reasonably agrees with

experiment.

Introduction

Chemical reaction mechanisms, along with intra- and intermolecular forces and other pro-

cessing conditions, dictate the formation of macromolecular structures and materials from

smaller molecular building blocks. Understanding the contributions of these factors permits

their use as design parameters to control the micro- and macroscopic properties of the re-

sulting new material. However, deconvolution of these factors is often not possible since the

molecular mechanics of reactive sites and reaction mechanism are intertwined.

This directs us toward a general question: Can we isolate chemical reaction mechanism

as an independent variable? This is an open and fundamental question in chemical physics.

While diﬀerent reaction mechanisms may be compared experimentally,1–3 these approaches

require unavoidable approximations. Computer simulations do not necessarily share this

limitation and can directly address this thought experiment. In this work, we introduce a

new approach that converts a drawback of coarse-grained computer simulations into a key

feature that enables direct comparison of chemically distinct systems, isolating the chemical

reaction mechanism as an independent variable.

In conventional molecular dynamics (MD) simulations each atom is represented by only a

few parameters and equations that reproduce the intra- and intermolecular interaction ener-

gies for a given system. Integration of Newton’s equations of motion then evolves this system

over time. All-atom models and the resulting simulations naturally mesh with physical and

chemical intuition. However, even with increases in computational power, all-atom classical

MD is still limited in length scale and simulation duration. To loosely quantify, simulating

several 10 nm3boxes of 105atoms, each for hundreds of nanoseconds, taxes the resources

of most researchers. These limitations become more pronounced for high molecular weight

molecules that require both larger simulation boxes and longer simulation times to capture

the correct interactions and dynamics.4,5

Coarse-grained MD (CGMD) approaches that combine multiple atoms into a single rep-

resentative “bead” are a popular approach toward reducing computational expense. Calcu-

lations are accelerated by reducing the number of interaction sites and allowing larger time

steps by removing the fastest motions.6Such techniques have brought on valuable insight

into many ﬁelds of soft matter physics, e.g. entangled dynamics in polymer melts, impact

of curvature and surfaces on biomimetic membranes, etc.7–10 Recently CGMD has been ex-

tended to reactive systems for a range of systems, including the simulated polymerization of

poystyrene,11 polyurethane,12 and silica.13 Notably, CGMD polymerization approaches have

been able to capture characteristic features of the resulting materials, reproducing thermo-

mechanical and morphological properties like the self-assembly into micellar structures or

amorphous clusters, depending on simulation conditions.12,13

When atoms are merged or ‘coarse-grained’ into a single representative site, the potential

energy landscape becomes smoother, more simpliﬁed.14,15 The identity of reactive functional

groups may disappear when merging individual atoms into beads. We may refer to this as

a ‘loss of chemical resolution.’ In this work, we take advantage of both CGMD’s ability to

simulate crosslinked polymer networks and the accompanying loss of chemical resolution. We

represent chemically distinct monomers with similar intramolecular structures but diﬀerent

reactive sites with the same coarse-grained model: a ‘universal monomer’ (UM). Despite the

diﬀerent monomers being represented by precisely the same model, the UM, we preserve the

respective reaction mechanisms native to each monomer species. In practice, we simulate a

box of the ‘universal monomer’ and impart chemical identity by implementing the chemical

reaction mechanism as a modular ‘rule,’ allow the selected reaction to proceed, and ana-

lyze the resulting material. We then repeat with a new liquid box and diﬀerent reaction

mechanism. Since the molecular interactions of the monomer models are precisely the same,

all diﬀerences in the evolution of the microscopic network and properties of the resulting

materials emerge entirely from the diﬀerent reaction mechanisms.

This approach toward isolating reaction mechanism as a variable borrows elements from

cellular automata (CA), as deﬁned by Rucker, where a discrete set of elements are each in

a ‘state’ and these elements’ states are updated in parallel using a homogenous and local

‘rule’.16 In our simulations the UM’s reactive sites may be considered as elements in binary

on/oﬀ (reacted/unreacted) states and the reaction mechanisms are analogous to CA rules.

There is not a large formalistic distance between these reactive CGMD simulations and

CA simulations of related physical systems17–19 and we regard the present work as a new

approach toward designing reactive CGMD that incorporates the underlying philosophies

of CA and exhibits the same interesting emergent behaviors.16,20–22 By implementing the

reaction mechanism as a ‘rule,’ we may rapidly quantify the resulting polymer’s micro- and

macroscopic properties as a function of polymerization mechanism alone.

The rapidly evolving additive manufacturing (AM) landscape provides a recent exam-

ple where reaction mechanism is used as a design variable. Compare Volumetric Additive

Manufacturing (VAM)23–25 with two-photon polymerization (2PP).26,27 Both approaches ac-

complish three-dimensional (3D) printed structures via photopolymerization. Whereas 2PP

sequentially scans a tightly-focused sub-micron laser focal volume to write a structure in 3D

space, VAM concurrently solidiﬁes all points within a 3D object by illuminating a rotating

volume of photosensitive resin with a dynamically evolving light pattern.24 These techniques

use vastly diﬀerent irradiation times: a few nanoseconds of femtosecond pulses in 2PP and

minutes of illumination in VAM. Techniques like 2PP are enabled by the fast reaction kinetics

of multi-functional acrylate resins, and nascent techniques like VAM naturally borrow feed-

stocks from more mature AM technologies like 2PP during the development phase. However,

the longer irradiation time of VAM allows using resins with slow kinetics, thus enabling ex-

ploration of new chemistries. Since reaction mechanism profoundly eﬀects the printed part’s

material properties, chemical reaction mechanism becomes a design parameter.3In this work,

Cook et. al. successfully replaced the pendant acrylate groups in their photoresin with thiol

and alkene functionalities, eﬀectively switching the reaction mechanism within their VAM

instrument from radical-initiated chain-growth to a radical step-growth ‘click’ chemistry.3

Swapping the functional groups thus resulted in more mechanically-robust printed parts by

photopolymerization-based additive manufacturing.

We select this use case as proof-of-concept to demonstrate our new computational ap-

proach and show that computer simulations are well-positioned to rapidly explore this design

space and avoid lengthy and costly formulate-cure-test development cycles. Our work be-

gins with the same three trifunctional isocyanurate monomers used by Cook et al. and uses

the UM approach to isolate diﬀerences that emerge from changes in reaction mechanism

and resulting microscopic network topologies. This isolation is possible since the UM ap-

proach ignores diﬀerences in the monomers’ molecular mechanics introduced by the diﬀerent

functional groups.

Many contemporary polymer simulations focus on a speciﬁc polymer or class of poly-

mers,28–32 and we use similar analytical approaches to investigate and validate the work

presented here. However, the methodology we introduce with the UM approach is more

closely related to studies where researchers systematically disabled components of MD force

ﬁelds. For example, one sets partial charges of a molecule or chemical moiety to zero to

quantify the contribution of electrostatic interactions to local intermolecular ordering, struc-

ture, and dynamics.32,33 Gissinger and co-workers disabled chain-chain linking in reactive

all-atom MD simulations of styrene, resulting in polystyrenes with polydispersity and cyclic

structures unlike that of the experimental system.34 Although nonphysical, this in silico

deconstruction reveals the contributions of force ﬁeld components to emergent physical and

chemical properties.

In the remainder of this work we provide a detailed overview of the UM approach but

文档加载中……请稍候！
如果长时间未打开，您也可以点击刷新试试。

下载文档到电脑，查找使用更方便

10 玖币 0人已下载

立即下载

摘要：

IsolatingChemicalReactionMechanismasaVariablewithReactiveCoarse-GrainedMolecularDynamics:Step-GrowthversusChain-GrowthPolymerizationJohnJ.Karnes,,yToddH.Weisgraber,yCaitlynC.Cook,yDanielN.Wang,yJonathanC.Crowhurst,yChristinaA.Fox,y,zBradleyS.Harris,{JamesS.Oakdale,yRolandFaller,,{andMaximShusteyy...

展开>> 收起<<

Isolating Chemical Reaction Mechanism as a Variable with Reactive Coarse-Grained Molecular Dynamics Step-Growth versus Chain-Growth.pdf

共34页,预览5页

还剩页未读，继续阅读

声明：本站为文档C2C交易模式，即用户上传的文档直接被用户下载，本站只是中间服务平台，本站所有文档下载所得的收益归上传人(含作者)所有。玖贝云文库仅提供信息存储空间，仅对用户上传内容的表现方式做保护处理，对上载内容本身不做任何修改或编辑。若文档所含内容侵犯了您的版权或隐私，请立即通知玖贝云文库，我们立即给予删除！

Isolating Chemical Reaction Mechanism as a Variable with Reactive Coarse-Grained Molecular Dynamics Step-Growth versus Chain-Growth

相关推荐

开通VIP享超值会员特权

作者详情

相关内容

热门标签

举报选择: