Efficient “Middle” Thermostat Scheme
for the Quantum/Classical Canonical
Ensemble via Molecular Dynamics
Xinzijian Liu, Kangyu Yan and Jian Liu
Abstract We have recently developed a unified “middle” thermostat scheme for
rationally designing molecular dynamics (MD)/path integral molecular dynamics (PIMD) algorithms for efficiently sampling the configuration space for the
classical/quantum canonical ensemble with or without constraints. It significantly
improves the time interval by a factor of 4–10 to achieve the same accuracy for
structural and configuration-dependent thermodynamic properties for MD (as well
as for any thermodynamic properties for PIMD). It has been implemented in AMBER
(2018/2019 version), which is available for MD/PIMD simulations with force fields,
QM/MM, or ab initio methods.
Keywords Canonical ensemble · Molecular dynamics · Path integral molecular
dynamics · Thermostat algorithms · “Middle” thermostat scheme · Quantum
statistics · Classical statistics · Holonomic constraints · Multi-time-step
techniques · Isokinetic constraints · Sampling efficiency
1 Introduction
Molecular dynamics (MD) has been widely used in chemistry, biology, materials,
environmental science, and other scientific fields [1, 2]. When nuclear quantum
effects are important, MD can be implemented to perform imaginary time path integral (PIMD) by virtue of the ring-polymer isomorphism [3, 4], which in principle
offers a numerically exact tool for practically studying quantum statistical properties
in molecular systems [4–8] where quantum exchange effects are negligible. Many
cases of interest involve the canonical ensemble, of which the number of particles
(N), volume (V) and temperature (T) are constant.
Xinzijian Liu and Kangyu Yan: Both authors contributed equally to the work.
X. Liu · K. Yan · J. Liu (B)
Beijing National Laboratory for Molecular Sciences, Institute of Theoretical and Computational
Chemistry, College of Chemistry and Molecular Engineering, Peking University, Beijing 100871,
China
e-mail: jianliupku@pku.edu.cn
© Springer Nature Switzerland AG 2020
L. Mammino et al. (eds.), Advances in Quantum Systems in Chemistry,
Physics, and Biology, Progress in Theoretical Chemistry and Physics 32,
https://doi.org/10.1007/978-3-030-34941-7_13
257
for the Quantum/Classical Canonical
Ensemble via Molecular Dynamics
Xinzijian Liu, Kangyu Yan and Jian Liu
Abstract We have recently developed a unified “middle” thermostat scheme for
rationally designing molecular dynamics (MD)/path integral molecular dynamics (PIMD) algorithms for efficiently sampling the configuration space for the
classical/quantum canonical ensemble with or without constraints. It significantly
improves the time interval by a factor of 4–10 to achieve the same accuracy for
structural and configuration-dependent thermodynamic properties for MD (as well
as for any thermodynamic properties for PIMD). It has been implemented in AMBER
(2018/2019 version), which is available for MD/PIMD simulations with force fields,
QM/MM, or ab initio methods.
Keywords Canonical ensemble · Molecular dynamics · Path integral molecular
dynamics · Thermostat algorithms · “Middle” thermostat scheme · Quantum
statistics · Classical statistics · Holonomic constraints · Multi-time-step
techniques · Isokinetic constraints · Sampling efficiency
1 Introduction
Molecular dynamics (MD) has been widely used in chemistry, biology, materials,
environmental science, and other scientific fields [1, 2]. When nuclear quantum
effects are important, MD can be implemented to perform imaginary time path integral (PIMD) by virtue of the ring-polymer isomorphism [3, 4], which in principle
offers a numerically exact tool for practically studying quantum statistical properties
in molecular systems [4–8] where quantum exchange effects are negligible. Many
cases of interest involve the canonical ensemble, of which the number of particles
(N), volume (V) and temperature (T) are constant.
Xinzijian Liu and Kangyu Yan: Both authors contributed equally to the work.
X. Liu · K. Yan · J. Liu (B)
Beijing National Laboratory for Molecular Sciences, Institute of Theoretical and Computational
Chemistry, College of Chemistry and Molecular Engineering, Peking University, Beijing 100871,
China
e-mail: jianliupku@pku.edu.cn
© Springer Nature Switzerland AG 2020
L. Mammino et al. (eds.), Advances in Quantum Systems in Chemistry,
Physics, and Biology, Progress in Theoretical Chemistry and Physics 32,
https://doi.org/10.1007/978-3-030-34941-7_13
257