Biomedical Engineering Reference
In-Depth Information
for ı
D
1.4 A and 2.0 A. And for the case of
N
D
0, it occurs 11 and 388 times for
ı
D
1.4 A and 2.0 A, but none for ı
D
0. It seems that the effect of the deformation
for ı
2.0 A is considerable for the possibilities of the rare events although the
change of the average value of the water molecules inside is negligible.
For ı
D
2.5 A, the function of free energy is somewhat flat between
N
D
2 and 4,
with a minimum at
N
D
2, due to the frequently rupture of the water chain. No
intermittent filling is found in the range of 0
ı
2.5 A, consistent with that there is
not any minimum at
N
D
0. Although there are many cases with
N
D
0forı
D
2.5 A,
the durations for them are short. We have not found two successive data with
N
D
0,
indicating that the duration for each case with
N
D
0islessthan0.25ps.
In addition, in MD simulations, the update algorithm develops a very slow change
in the center of mass (COM) velocity, and thus in the total kinetic energy of the
system, especially when temperature coupling is used. For the systems with limited
friction, if such changes are not quenched, an appreciable COM motion will be
developed eventually in long runs. Thus, the method of center of mass motion
removal (CMMR) is necessary. During MD simulations, the CMMR method can
be applied at a certain frequency. But for the systems with pressure difference to
evoke a unidirectional transportation, for example, the gating system we discussed
before, the application of CMMR may reduce the effect of such equivalent pressure
difference.
It is clear that this excellent gating behavior results from the one-dimensional
hydrogen bonds between the neighboring water molecules, which shield the external
perturbations. Thus the water permeation across the bio-mimicked nanochannel
with single-filed water molecules inside the nanochannel is effectively resistant to
mechanical noises and sensitive to available mechanical signals.
Further, we consider the effect of CNT length on the behavior of water molecules
in the tube. The CNT we choose is 14.6 A in length and 8.1 A in diameter. Along
the
z
direction, it is embedded in the center of a graphite sheet, which is similar with
the system shown in Fig.
1.3
. A force is also applied to one carbon atom, namely
the forced-atom. Note that the number of carbon rings along the nanotube is 5.5,
which is 0.5 ring longer than the CNT used in previous sections. Interestingly, our
simulation results indicate that the wavelike pattern of water density distribution
becomes much weaker when the length of the perfect CNT is increased to 14.6 A
(see Fig.
1.13
). Especially, the wavelike pattern is quite flat near the center of
the CNT for the unperturbed nanotube. Even near two openings of the CNT, the
wavelike pattern is less obvious than that of the CNT that is 13.4 A. The height
between the first peak and the dip from the left opening of the CNT is only about
0.05 A
1
, which is much smaller than the average value of 0.12 A
1
for the CNT of
13.4 A.
The shape of the CNT plays a role in the structure of the water density
distribution along tube axis as well. Simulation results indicate that the weak
wavelike pattern of water density distribution becomes obvious and shifts slightly
when the CNT is deformed by external force. For ı<1.5
A, the amplitude of
