Organized by GPT-6-Astra. SPONGE 2.0 is under active, iterative development; features, parameters, and documentation may change. Consult the GitHub repository’s docs and skills for the latest version information.

Source snapshot for this page.

Thermostat Parameters

SPONGE requires a thermostat in nvt and npt modes.

For TOML input, prefer the grouped form:

mode = "nvt"
target_temperature = 300.0

[thermostat]
mode = "middle_langevin"
tau = 0.1
seed = 2026

The parser also accepts the flat compatibility keys thermostat, thermostat_tau, and thermostat_seed, but new documentation should use the grouped TOML form above.

Thermostat Selection

Parameter	Scope	Type	Default	Description
mode	thermostat	string	required for NVT/NPT	Thermostat algorithm

Supported values:

Value	Description
"middle_langevin"	Middle Langevin dynamics
"langevin"	Deprecated alias of middle_langevin
"andersen"	Andersen thermostat
"berendsen_thermostat"	Berendsen thermostat
"bussi_thermostat"	Bussi velocity-rescaling thermostat
"nose_hoover_chain"	Nose-Hoover chain thermostat

Top-level Parameters

Parameter	Scope	Type	Default	Description
target_temperature	top-level	float	300.0	Target temperature (K)
velocity_max	top-level	float	disabled	Velocity cap used by some thermostats

target_temperature is not part of [thermostat]. It stays at the top level and is also the quantity affected by temperature schedules.

Common Thermostat Parameters

Parameter	Scope	Type	Default	Description
tau	thermostat	float	1.0	Coupling time constant in ps
seed	thermostat	int	random	Random seed for stochastic thermostats

Notes:

• seed is used by middle_langevin, andersen, and bussi_thermostat
• berendsen_thermostat and nose_hoover_chain do not read seed
• middle_langevin converts tau internally to a friction coefficient gamma = 1 / tau
• andersen converts tau to an internal collision update interval

Algorithm-specific Parameters

middle_langevin

[thermostat]
mode = "middle_langevin"
tau = 0.1
seed = 2026

Uses tau and seed in the thermostat scope, plus the top-level velocity_max.

andersen

[thermostat]
mode = "andersen"
tau = 0.1
seed = 2026

Uses tau and seed in the thermostat scope, plus the top-level velocity_max.

berendsen_thermostat

[thermostat]
mode = "berendsen_thermostat"
tau = 0.1

Uses only tau in the thermostat scope.

bussi_thermostat

[thermostat]
mode = "bussi_thermostat"
tau = 0.1
seed = 2026

Uses tau and seed in the thermostat scope. New H5 checkpoints store the Bussi random stream as typed portable Philox state; scalar standard-library engine text is accepted only when reading older checkpoints.

nose_hoover_chain

[thermostat]
mode = "nose_hoover_chain"
tau = 0.2

[nose_hoover_chain]
length = 3
restart_input = "nhc_restart.in"
restart_output = "nhc_restart.out"
crd = "nhc_crd.dat"
vel = "nhc_vel.dat"

Additional parameters:

Parameter	Scope	Type	Default	Description
length	nose_hoover_chain	int	1	Chain length
restart_input	nose_hoover_chain	string	-	Read initial chain state from file
restart_output	nose_hoover_chain	string	-	Write chain state to file
crd	nose_hoover_chain	string	-	Write chain coordinate trajectory
vel	nose_hoover_chain	string	-	Write chain velocity trajectory

nose_hoover_chain also reads the top-level velocity_max.

Temperature Schedule

Temperature schedules are controlled by top-level keys, not by [thermostat]:

Parameter	Type	Description
target_temperature_schedule_mode	string	"step" or "linear"
target_temperature_schedule_steps	array	Inline schedule: [{ step = N, value = T }, ...]
target_temperature_schedule_file	string	Path to a TOML schedule file

Inline target_temperature_schedule_steps only works when the main mdin file is TOML.

target_temperature = 300.0
target_temperature_schedule_mode = "linear"
target_temperature_schedule_steps = [
    { step = 0, value = 100.0 },
    { step = 10000, value = 300.0 },
    { step = 50000, value = 300.0 },
]