Chemistry Reference
In-Depth Information
Morrow and Maginn
49
also carried out a molecular dynamics study of
[C
4
mim][PF
6
], using an explicit-atom and fully flexible model with a force
field having the following functional form:
X
X
X
2
2
U
¼
k
b
ð
r
r
0
Þ
þ
k
y
ð
y
y
0
Þ
þ
k
w
½
1
þ
cos
ð
n
w
d
Þ
bonds
angles
dihedrals
(
"
#
)
12
s
6
X
X
X
N
1
N
s
q
i
q
j
4
2
ij
r
ij
ij
r
ij
þ
k
c
ð
c
c
0
Þ
þ
4
E
þ
½
5
ij
E
0
r
ij
impropers
i
¼1
j
>1
This force field function contains not only dihedral, Lennard-Jones, and
Coulombic terms, but it also includes bond stretching, angle bending, and
improper ring bending contributions modeled with harmonic potentials.
This is the same form used by Margulis, Stern, and Berne, except they did
not consider improper bending terms. Morrow and Maginn
49
used a some-
what more sophisticated quantum method than previous studies for obtaining
partial charges on each atom. They also utilized Lennard-Jones and dihedral
angle parameters for similar compounds that were available in the CHARMM
database.
42
The force field yielded excellent liquid densities and derivative
properties. In fact, the explicit-atom model did significantly better at matching
the experimental density than did this group's earlier united-atom model.
39
Figure 5 compares the performance of these two models. The united-atom
model underpredicted the molar volume (overpredicted the density) by almost
5%, while the explicit-atommodel was accurate to within 1%. This is consistent
Figure 5
Comparison of the experimental density for [C
4
mim][PF
6
] as a function of
temperature
50
and results obtained using a united-atom model
39
and a more detailed
explicit-atom model.
49
