Page - 16 - in Charge Transport in DNA - Insights from Simulations
Image of the Page - 16 -
Text of the Page - 16 -
TheoreticalBackground
from the reference temperature is correctedbetween two time stepsdependingon
thepre-set timeconstantτ.
dT
dt = 1
τ (Ttarget−T) (2.6)
Thus,wecancalculate the change in temperature
ΔT= Δt
τ (Ttarget−T) (2.7)
as there is adirect relationbetween the temperatureand thevelocitiesof theparti-
cles in the system.
1
2 m 〈
v2 〉
= 3
2 kT (2.8)
The temperature in a system is changedby scaling the velocities of all particles in
the systemwitha scaling factorλ,which is obtained fromthe relation
ΔT= Δt
τ (Ttarget−T)=(λ2−1)T (2.9)
as
λ= √
1+ Δt
τ (
Ttarget
T −1 )
(2.10)
Thecoupling to theexternalheatbathnowdependsonthechoiceof theparameter
τ. With higher values for τ, the temperaturewill converge faster to the reference
temperatureTtarget.
Unfortunately, the Berendsen thermostat does not yield a correct canonical prob-
ability distribution. Amore sophisticated thermostat that rigorously represents a
correct canonical ensemble is the widely used Nosé-Hoover thermostat [58, 59].
Here, a heat reservoir is introduced as an integral part of the system with one
degreeof freedom s, towhichamassQ is associated.
16
Table of contents
- Zusammenfassung 1
- Summary 3
- 1 Introduction 5
- 2 TheoreticalBackground 11
- 3 SimulationSetup 39
- 4 DNAUnderExperimentalConditions 49
- 5 ChargeTransport inStretchedDNA 69
- 6 ChargeTransport inMicrohydratedDNA 79
- 7 AParametrizedModel toSimulateCT inDNA 89
- 8 Conclusion 105
- Appendix 111
- A DNAUnderExperimentalConditions 111
- B CTinMicrohydratedDNA 117
- List ofPublications 137