Yaohong Zhou1, Haidong Wang1. 1. Key Laboratory for Thermal Science and Power Engineering of Ministry of Education, Department of Engineering Mechanics, Tsinghua University, Beijing 100084, China.
Abstract
In recent years, sensing technology based on nanopores has become one of the trustworthy options for characterization and even identification of a single biomolecule. In nanopore based DNA sequencing technology, the DNA strand in the electrolyte solution passes through the nanopore under an applied bias electric field. Commonly, the ionic current signals carrying the sequence information are difficult to detect effectively due to the fast translocation speed of the DNA strand, so that slowing down the translocation speed is expected to make the signals easier to distinguish and improve the sequencing accuracy. Modifying the nanopore structure is one of the effective methods. Through all-atom molecular dynamics simulations, we designed an asymmetric double-layer graphene nanopore structure to regulate the translocation speed of a single carbon chain. The structure consists of two nanopores with different sizes located on two layers. The simulation results indicate that the asymmetric nanopore structure will affect the chain's translocation speed and the ionic current value. When the single carbon chain passes from the smaller pore to the larger pore, the translocation time is significantly prolonged, which is about three times as long as the chain passing from the larger pore to the smaller pore. These results provide a new idea for designing more accurate and effective single-molecule solid-state nanopore sensors.
In recent years, sensing technology based on nanopores has become one of the trustworthy options for characterization and even identification of a single biomolecule. In nanopore based DNA sequencing technology, the DNA strand in the electrolyte solution passes through the nanopore under an applied bias electric field. Commonly, the ionic current signals carrying the sequence information are difficult to detect effectively due to the fast translocation speed of the DNA strand, so that slowing down the translocation speed is expected to make the signals easier to distinguish and improve the sequencing accuracy. Modifying the nanopore structure is one of the effective methods. Through all-atom molecular dynamics simulations, we designed an asymmetric double-layer graphene nanopore structure to regulate the translocation speed of a single carbon chain. The structure consists of two nanopores with different sizes located on two layers. The simulation results indicate that the asymmetric nanopore structure will affect the chain's translocation speed and the ionic current value. When the single carbon chain passes from the smaller pore to the larger pore, the translocation time is significantly prolonged, which is about three times as long as the chain passing from the larger pore to the smaller pore. These results provide a new idea for designing more accurate and effective single-molecule solid-state nanopore sensors.
Nanopore technology has
emerged as a significant research tool
for single-molecule analysis and detection.[1−3] One of the most
prominent applications is DNA sequencing.[4] This technology typically[5,6] involves applying an
applied bias electric field to the electrolyte surrounding the nanopore.
A single DNA strand and ions are electrophoretically driven by the
electric field. The ordered ion transportation will form a relatively
stable ionic current. When the DNA strand passes through the nanopore,
it will block the transport of ions. The four kinds of bases in DNA
are different in structure, so the blocking effect of each base on
ions is different,[7] resulting in different
ionic current signals. The sequence information can be read by monitoring
the signals of the ionic current during the whole process of DNA passing
through the nanopore. The continuous development of this technology
is conducive to finally realizing the real-time, high-throughput,
and low-cost sequencing of DNA.[8−10] Nanopore technology can be divided
into biological nanopore technology and solid-state nanopore technology
according to different nanopore materials,[11] among which the biological nanopore membrane usually has a larger
thickness (about dozens of nanometers), much larger than the distance
between adjacent bases in the DNA strand (only about 0.3–0.5
nm).[12] As a result, it is difficult to
distinguish a single base directly. The solid-state nanopore has superior
mechanical strength and chemical stability, especially monolayer graphene
which has a thickness of 0.334 nm, close to the distance between adjacent
bases in the DNA strand. The graphene nanopore is expected to have
a higher spatial sensing resolution than the biological nanopore,
which becomes a better choice for direct discrimination of a single
DNA base;[13] but on the other hand, one
of the biggest challenges in real applications of the graphene nanopore[13,14] is that the DNA strand passes through the nanopore very fast due
to its atomic thickness, often at an average rate of more than one
nucleotide per microsecond. It is too fast to clearly distinguish
the four bases of a single DNA strand by monitoring the ionic current
signal.[15,16]Consequently, slowing down the translocation
speed has become one
of the urgent problems to overcome in order to further improve the
sensing accuracy and reliability of a solid-state nanopore. A series
of methods have been tried to reduce and control the translocation
speed, such as controlling solvent viscosity,[17] increasing contact friction,[18] modifying
nanopore structure,[19−25] modifying DNA chain,[26] controlling electric
field,[27] etc. If we focus on the idea of
regulating the translocation by modifying the nanopore structure,
it usually includes physical and chemical modification methods, such
as changing the size and shape of the nanopore,[28] chemical functionalization,[29] or surface charge modification.[30] General
methods of physical structure modification can prolong the translocation
time by increasing the thickness of nanopores,[31] for example, using multilayer nanopore structures. A special
kind of multilayer nanopore structure is geometrically asymmetric.
Asymmetric nanopore structures are different in shapes, and a conical
nanopore is one of the typical asymmetric structures. The influence
mechanism of this common asymmetric structure on the translocation
process of the DNA chain has been reflected in both experimental[32,33] and simulation studies.[34,35] However, the conical
nanopores mentioned in these studies are usually thick. However, increasing
the nanopore thickness will also reduce the spatial resolution of
single-molecule detection technology, which is not the best solution.
Therefore, we proposed an asymmetric double-layer nanopore structure,
which can effectively prolong the translocation time and improve the
detection accuracy by using a double-layer graphene, maintaining the
high spatial resolution of the solid-state nanopore sensor.In this work, we proposed a new asymmetric double-layer graphene
nanopore structure and carried out extensive all-atom molecular dynamics
simulations to calculate the translocation time of a single carbon
chain passing through the nanopore structure and the corresponding
ionic current. The asymmetric nanopore structure consists of two nanopores
of different sizes. The larger pore is named the L-pore, and the energy
barrier when the single carbon chain passes through the L-pore first
is smaller than that when it passes through the smaller pore first,
which is conducive to the translocation event of the single carbon
chain. The smaller pore is named the S-pore, and the translocation
time can be regulated by changing the diameter of the S-pore. Furthermore,
the orientation of the single carbon chain passing through the nanopore
structure is changed by switching the positions of these two nanopores,
and then the effects of the orientation of the single carbon chain
passing through the asymmetric nanopore structure on translocation
properties and ion transport were revealed. Based on the simulation
results, we found that the translocation time of a single carbon chain
passing the nanopore from the S-pore to the L-pore can be significantly
prolonged and benefit the characterization of molecular structure.
This provides an idea for the optimal design and preparation of the
solid-state nanopore structure for a longer molecule detection time
and higher DNA sequencing accuracy.
Computational
Methods
System Setup
Figure depicts a schematic of the molecular dynamics
simulation model. A box of the KCl solution is divided into cis and
trans parts by inserting a membrane of an asymmetric double-layer
graphene nanopore structure. The box size is 50 × 50 × 102
Å along the x, y, and z directions. Periodic boundary conditions are applied along
all three axis directions. The simulation domain contains 2000 water
molecules and 300 KCl molecules to keep the KCl concentration at about
2.0 mol/L. The ionic solution environment was kept the same for all
the simulation cases, and the influence of the ionic solution environment
on the translocation of the carbon chain through the nanopore was
consistent in all cases. Two graphene nanopores with different diameters
were obtained by removing different numbers of carbon atoms from the
centers of two graphene layers. In recent years, there have also been
simulation studies using Langevin Molecular Dynamics,[36−38] Monte Carlo,[39−41] and other methods to simulate the translocation process
of polymers through a nanopore structure. In these studies, the polymeric
chain is just like the DNA strand in nanopore sequencing technology,
and the influence of the chain length and pore size on the translocation
time has been revealed. In this work, the carbon chain can be seen
as the basic skeleton of a DNA molecule.[42] In order to simplify the model, a single carbon chain was used to
replace the real DNA strand. The chain is made up of 10 butadiene
molecules. The molecular dynamics simulation package LAMMPS[43,44] was adopted to simulate the translocation of a single carbon chain
across the asymmetric double-layer graphene nanopores. The reactive
force field (ReaxFF[45]) was utilized to
describe the interaction between C, H, O, K, and Cl atoms in the simulation.
Figure 1
Schematic
diagram of the simulation model for a single carbon chain
passing through an asymmetric double-layer graphene nanopore.
Schematic
diagram of the simulation model for a single carbon chain
passing through an asymmetric double-layer graphene nanopore.
ReaxFF Reactive Force Field
Method and Data
Analysis
In recent years, the development of the reactive
force field has made molecular dynamics simulation more practical,
and complex hydrocarbon reactions can be better simulated[46] with ReaxFF. New reactive force field[45] parameters have been trained against quantum
mechanical (QM) calculations related to water binding energies, hydration
energies, and energies of proton transfer, which is suitable for water
and electrolyte systems. The water (H/O) parameters have been widely
used in various systems, including proteins,[47,48] metal oxides,[49−51] and organic molecules.[52] It can describe the interaction between water molecules and ions
such as K+ and Cl–. Therefore, this reactive
force field can be well applied to our simulated system. The LAMMPS
software package was performed to simulate the trajectories of all
atoms with the time step of 0.1 fs. First, the conjugate gradient
(CG) algorithm was used to minimize the energy of the simulated system.
Then, the simulated system was relaxed under the NPT ensemble for
50 ps to maintain a constant temperature of 300 K and control the
pressure to 1 atm. After the relaxation process, the double-layer
graphene membrane was fixed in the middle of the simulated system,
and the external electric field was applied in the Z-direction (E) and Y-direction (E), respectively, which covered the entire box. Then, the simulated
system ran for 250 ps under the NVT ensemble. In order to improve
the efficiency of computation, each carbon atom in the single carbon
chain was given a positive charge of one unit. This is similar to
negatively charged DNA molecule,[53] which
is more visibly affected by electric fields. The cation and anion
will move under the application of the electric field. The ionic current
can be calculated from the simulated atomic trajectories.The
position coordinates of all atoms in the simulated system were recorded
every femtosecond. The ionic current was computed[54] aswhere z and q are the z coordinate and the charge of atom i,
respectively.
Here, Δt is set to 1 fs, and L is the length of the simulated system
in the z direction. Furthermore, the ionic current
calculated by the above equation was the average value of the instantaneous
ionic current during the Δt.
Results and Discussion
From the L-Pore to the
S-Pore
We
have made asymmetric nanopores in two graphene layers; the larger
pore is named L-pore, while the smaller pore is named S-pore. In the
case of the translocation from the L-pore to the S-pore, Table summarizes the translocation
status of the single carbon chain when changing the diameter of the
S-pore (D) under
different electric field intensities. Figure shows the simulation model and the translocation
process when the single carbon chain passes through the L-pore first.
The diameter of the L-pore (D) was fixed at 3.6 nm. Eight groups of S-pores of different
sizes were set up, with diameters of 2.0, 2.2, 2.4, 2.6, 2.8, 3.0,
3.2, and 3.4 nm, respectively.
Table 1
Translocation of
the Single Carbon
Chain while Changing the Diameter of the S-Pore under Different Z-Direction and Y-Direction Electric Field
Intensities
DL-pore (nm)
DS-pore (nm)
Ez (V/Å)
Ey (V/Å)
translocation occurred
3.6
2.0
0.15
0.1
yes
3.6
2.2
0.15
0.1
yes
3.6
2.4
0.08
0.05
no
3.6
2.6
0.05
0.05
no
3.6
2.8
0.03
0.05
no
3.6
3.0
0.1
0.05
yes
3.6
3.2
0.1
0.05
yes
3.6
3.4
0.1
0.05
yes
3.6
3.6
0.1
0.05
yes
Figure 2
Molecular dynamics simulation of the single carbon chain passing
through the L-S type nanopore structure under an applied bias electric
field. (a) The simulation model: the single carbon chain is shown
in yellow, other carbon atoms are shown in gray, the hydrogen atoms
are shown in white, the oxygen atoms are shown in red, the potassium
ions are shown in purple, and the chloride ions are shown in green.
(b)–(g) together demonstrate the translocation process of the
single carbon chain passing through the nanopore structure.
Molecular dynamics simulation of the single carbon chain passing
through the L-S type nanopore structure under an applied bias electric
field. (a) The simulation model: the single carbon chain is shown
in yellow, other carbon atoms are shown in gray, the hydrogen atoms
are shown in white, the oxygen atoms are shown in red, the potassium
ions are shown in purple, and the chloride ions are shown in green.
(b)–(g) together demonstrate the translocation process of the
single carbon chain passing through the nanopore structure.As shown in Table , when the diameter of the L-pore is constant, the
smaller the diameter
of the S-pore is, and the greater the electric field intensity is
required to drive the translocation of the single carbon chain, especially
in the Z-direction. It means that a larger barrier
needs to be overcome for the single carbon chain to pass through the
nanopores. The effect of the pore diameter on the energy barrier is
consistent with other computational and experimental studies.[55−57] The value of E largely
determines the driving force of the single carbon chain; if E is not large enough to overcome
the resistance in the simulated system, it is difficult for the single
carbon chain to pass through the nanopores. The resistance consists
of liquid viscosity resistance, interaction between the the single
carbon chain and graphene nanopores, and charge repulsion.[58] It is worth noting that the application of the Y-direction electric field can effectively bring the single
carbon chain closer to the position of the nanopores. Simulation results
show that without the application of the Y-direction
electric field, the single carbon chain is easily adsorbed on the
surface of graphene, which means that it is difficult for the chain
to pass through the nanopore.Under the same electric field
parameters, with different diameters
of the S-pore, the duration of the single carbon chain passing through
the nanopores is different; as shown in Figure (a), the duration decreases nonlinearly as
the diameter of the S-pore increases linearly. This indicates that
for the two nanopores in the simulated system, the regulation of the
duration can be realized by only changing the diameter of one of the
nanopores, and the duration is sensitive to the change of pore diameter.
However, this feedback is not significantly reflected in the simulation
results of the ionic current. As shown in Figure (b), under the same electric field parameters,
the smaller the diameter of the S-pore is, the smaller the overall
level of the ionic current is; but the difference between ionic currents
is not significant, because the diameter of the S-pore does not change
much. In addition, the variation trend of the simulated ionic current
is similar to that in ref (58). That is, after the electric field is applied, the ionic
current will step up to a maximum value and then drop to a relatively
stable value, which may also include the blocking value caused by
the volume occupying effect[59] of the single
carbon chain. The ionic current in Figure (b) was calculated at the moment when the
electric field was applied.
Figure 3
(a) The duration for the translocation of the
single carbon chain
when the diameter of the S-pore is 3.0, 3.2, 3.4, and 3.6 nm, E = 0.1 V/Å, and E = 0.05 V/Å. (b) The
ionic current during the translocation of the single carbon chain
when the diameter of the S-pore is 3.0, 3.2, and 3.4 nm, E= 0.1 V/Å, and E = 0.05 V/Å.
(a) The duration for the translocation of the
single carbon chain
when the diameter of the S-pore is 3.0, 3.2, 3.4, and 3.6 nm, E = 0.1 V/Å, and E = 0.05 V/Å. (b) The
ionic current during the translocation of the single carbon chain
when the diameter of the S-pore is 3.0, 3.2, and 3.4 nm, E= 0.1 V/Å, and E = 0.05 V/Å.
From the S-Pore to the L-Pore
In
the case of the translocation from the S-pore to the L-pore, the single
carbon chain can pass through the asymmetric double-layer nanopore
when E = 0.1 V/Å
and E = 0.05 V/Å,
regardless of the S-pore’s diameter, as shown in Table . It suggests that the electric
field restriction is less stringent when passing through the nanopore
from the S-pore to the L-pore than that in the opposite direction.
Here, the diameter of the L-pore was also fixed as 3.6 nm. The atomic
trajectory results obtained by the simulation indicate that, when
the single carbon chain passes through the L-pore first (as shown
in Figure ), the ion
flux is larger, so the space near the S-pore will accumulate K faster, and the space near
the L-pore will accumulate Cl faster. As a result, the single carbon chain will be constrained
by the greater resistance. When the S-pore is not large enough, it
is difficult for the single carbon chain to complete the translocation
due to the excessive resistance. However, if the chain passes through
the S-pore first (as shown in Figure ), the ion flux will decrease, which helps to reduce
resistance caused by ion accumulation. This makes it easier for the
carbon chain to pass through. Figure compares the duration of the translocation of a single
molecule in two cases.
Table 2
Translocation of
the Single Carbon
Chain while Changing the Diameter of the S-Pore at E = 0.1 V/Å and E = 0.05 V/Å
DS-pore (nm)
DL-pore (nm)
Ez (V/Å)
Ey (V/Å)
translocation occurred
2.0
3.6
0.1
0.05
yes
2.2
3.6
0.1
0.05
yes
2.4
3.6
0.1
0.05
yes
2.6
3.6
0.1
0.05
yes
2.8
3.6
0.1
0.05
yes
3.0
3.6
0.1
0.05
yes
3.2
3.6
0.1
0.05
yes
Figure 4
Molecular dynamics
simulation of the single carbon chain passing
through the S-L type nanopore structure under an applied bias electric
field. (a) The simulation model: the single carbon chain is shown
in blue, other carbon atoms are shown in gray, the hydrogen atoms
are shown in white, the oxygen atoms are shown in red, the potassium
ions are shown in purple, and the chloride ions are shown in green.
(b)–(g) together demonstrate the translocation process of the
single carbon chain passing through the nanopore structure.
Figure 5
Comparison of the translocation duration in two cases,
where the
diameter of the S-pore is 3.0, 3.2, and 3.4 nm, E = 0.1 V/Å, and E = 0.05 V/Å.
Molecular dynamics
simulation of the single carbon chain passing
through the S-L type nanopore structure under an applied bias electric
field. (a) The simulation model: the single carbon chain is shown
in blue, other carbon atoms are shown in gray, the hydrogen atoms
are shown in white, the oxygen atoms are shown in red, the potassium
ions are shown in purple, and the chloride ions are shown in green.
(b)–(g) together demonstrate the translocation process of the
single carbon chain passing through the nanopore structure.Comparison of the translocation duration in two cases,
where the
diameter of the S-pore is 3.0, 3.2, and 3.4 nm, E = 0.1 V/Å, and E = 0.05 V/Å.As shown in Figure , it is obvious that when the single carbon chain passes through
the S-pore first, the duration of the translocation is longer. The
effect of the S-pore’s diameter on the duration is similar
to that of passing through the L-pore first. This suggests that the
duration can be regulated by controlling the diameter of the S-pore
in both cases. Moreover, the case of passing through the S-pore first
can better prolong the duration time, which is about three times as
long as that passing through the L-pore first. In this case, the S-pore
can also reduce ion flux and help the single carbon chain to better
complete the translocation. However, the reduction of the ion flux
also brings some problems. When the single carbon chain passes through
the nanopores, the ion transport will be blocked, reflecting the volume
occupying effect.[60,61] As the ion flux is lower if passing
through the S-pore first, the volume occupying effect will be weakened.
This will also result in a smaller overall level of the ionic current. Figure (a) shows the ionic
current calculation results when the diameter of the S-pore is 2.2,
2.4, 2.6, 2.8, 3.0, and 3.2 nm, respectively, E = 0.1 V/Å, and E = 0.05 V/Å. Figure (b) shows the ionic current when the single
carbon chain in the system is removed. Comparing the ionic current
results in Figure (a) and Figure (b),
the blocking effect of the single carbon chain on the ionic current
will be clearly revealed. Figure (c) and Figure (d) show the ionic current blocking of the single carbon chain
passing through the nanopores in two cases (in two opposite directions)
when the S-pore’s diameter is 3.0 nm.
Figure 6
Ionic current of the
single carbon chain passing through double-layer
graphene. (a) When the single carbon chain existed in the simulated
system. The ionic current calculation results when the diameter of
the S-pore is 2.2, 2.4, 2.6, 2.8, 3.0, and 3.2 nm, respectively, E = 0.1 V/Å, and E = 0.05 V/Å. (b) When
the single carbon chain was removed. The ionic current calculation
results when the diameter of the S-pore is 2.2, 2.4, 2.6, 2.8, 3.0,
and 3.2 nm, respectively, E = 0.1 V/Å, and E = 0.05 V/Å. (c) When the single carbon chain passes through
the S-pore first. The calculation results of the ionic current with
and without the single carbon chain in the simulated system, when
the S-pore’s diameter is 3.0 nm, E = 0.1 V/Å, and E = 0.05 V/Å. (d) When the single carbon chain
passes through the L-pore first. The calculation results of the ionic
current with and without the single carbon chain in the simulated
system, when the S-pore’s diameter is 3.0 nm, E = 0.1 V/Å, and E = 0.05 V/Å.
Ionic current of the
single carbon chain passing through double-layer
graphene. (a) When the single carbon chain existed in the simulated
system. The ionic current calculation results when the diameter of
the S-pore is 2.2, 2.4, 2.6, 2.8, 3.0, and 3.2 nm, respectively, E = 0.1 V/Å, and E = 0.05 V/Å. (b) When
the single carbon chain was removed. The ionic current calculation
results when the diameter of the S-pore is 2.2, 2.4, 2.6, 2.8, 3.0,
and 3.2 nm, respectively, E = 0.1 V/Å, and E = 0.05 V/Å. (c) When the single carbon chain passes through
the S-pore first. The calculation results of the ionic current with
and without the single carbon chain in the simulated system, when
the S-pore’s diameter is 3.0 nm, E = 0.1 V/Å, and E = 0.05 V/Å. (d) When the single carbon chain
passes through the L-pore first. The calculation results of the ionic
current with and without the single carbon chain in the simulated
system, when the S-pore’s diameter is 3.0 nm, E = 0.1 V/Å, and E = 0.05 V/Å.Figure (a) and Figure (b) suggest that,
regardless of whether there is a single carbon chain in the simulated
system, the larger the S-pore’s diameter is, the larger the
overall level of the ionic current is. The ionic current decreases
with the decrease of the S-pore’s diameter.[62] However, if comparing Figure (a) and Figure (b), it can be seen that when the single
carbon chain is removed, the volume occupying effect disappears, which
leads to the increase of the ion flux, so the ionic current drops
faster, which shows up as a steeper curve in Figure (b). Similar to the results mentioned above,
due to the small difference in the S-pore’s diameter, there
is no obvious difference in the calculated ionic current, but the
difference still can be seen. Figure (c) shows the current blockage phenomenon caused by
the volume occupying effect in the case of passing through the S-pore
first. Compare the results as shown in Figure (d). The current blockage is weaker than
the case of passing through the L-pore first. This indicates that
the S-pore’s diameter constrains the ion flux, and the smaller
the ion flux is limited, the weaker the current blockage phenomenon
will become. It means that when the single carbon chain passes in
the direction from the S-pore to the L-pore, the resistance of ion
accumulation will be reduced, and the volume occupying effect is weakened;
but it is worth noting that the translocation event can still be detected
even though the volume occupying effect is weakened.
Conclusions
In this work, molecular dynamics simulation
was used to study the
translocation properties of the single carbon chain in two types of
asymmetric double-layer graphene nanopores. The incidence of translocation
events based on S-L type nanopores (from the S-pore to the L-pore)
was better than that based on L-S type nanopores (from the L-pore
to the S-pore). The translocation of the single charged carbon chain
is affected by the accumulation of ions on either side of the nanopores.
In the case of L-S type nanopores, if the translocation of the chain
is still not completed until the ion concentration reaches a high
level, it will be difficult for the chain to pass through the nanopore
due to the higher charge repulsion attribute to the faster ion accumulation;
but the S-L type nanopore structure can reduce the ion flux, thus
slowing down the ion accumulation rate, which is beneficial to the
translocation. Regardless of the L-S or S-L type nanopore structure,
the duration of the single carbon chain passing through the nanopore
can be regulated by only controlling the size of the nanopore in one
layer of the graphene membrane. The duration obtained by using the
S-L type nanopore structure is about three times as long as that of
the L-S type nanopore structure. It has a good inspiration for improving
the single-molecule nanopore detection technology by using the asymmetric
nanopore structure. The S-L type nanopore structure is more conducive
to reduce the translocation speed, even though the ionic current will
be reduced as well. Moreover, the reduced ionic current has limited
influence on the identification of a single molecule. The proposed
asymmetric nanopore structure and simulation results provide new ideas
for the design and preparation of solid-state nanopore structures
for prolonging detection time and achieving higher accuracy.
Authors: Jurgen Fritz; Emily B Cooper; Suzanne Gaudet; Peter K Sorger; Scott R Manalis Journal: Proc Natl Acad Sci U S A Date: 2002-10-17 Impact factor: 11.205
Authors: Swati Bhattacharya; Ian M Derrington; Mikhail Pavlenok; Michael Niederweis; Jens H Gundlach; Aleksei Aksimentiev Journal: ACS Nano Date: 2012-07-13 Impact factor: 15.881