Berendsen thermostat - Computational Chemistry

Berendsen thermostat - Computational Chemistry: This occasion will discuss important aspects of the simulation using NVT ensemble, in which the number of molecules, volume and temperature made permanent. to set the fixed temperature or in accordance with the target temperature of the simulation we need the name of the thermostat.

before defining the thermostat, we need to know first how to define temperature in statistical thermodynamics. in statistical thermodynamics we know that the kinetic energy = 3/2 HCV. while the kinetic energy is defined as = 1/2 mv ^ 2. if we have a banya particles then it is easy to add all the kinetic energy of each particle to get the total kinetic energy. This relationship is based on easily we know that the target temperature will be proportional to the square of the speed. then to regulate the temperature we need to do scaling in speed by using a factor lambda.

lambda = sqrt (T_o / T_system) where T_o is the target temperature while T_system the current temperature of the system. it is thus when T_system> T_o the lambda value <1. lambda factor is used to download speed scaling so that we obtain:

v '= v * lambda

This concept is called with hard scaling so that when the temperature above / below the target temperature then we "force" in order to reset the target temperature. ya remember this can only be done at any stage of equilibration.

Berendsen propose a more "soft" to perform scaling of the speed.

From the above equation we introduce a new constant called the relaxation time. for example, we use relaxation time 100fs or 0.1 ps. oh yes, this relaxation time should be chosen that mediocrity, wearing a relaxation time that is too big it will be tantamount to using a hard scaling or using relaxation time that is too small will produce fluctuations in temperature that is too small and not realistic. remember yes, the Berendsen thermostat, the speed will always be in-scaling. Okay, enough of and hopefully useful!