Journal of the American Chemical Society
Article
conformations of each head. In the final step, QM analysis was used to
compare the energy and stability of each conformer. The 100 ns
molecular dynamics simulations were done to generate different
conformations of the heads and estimate the energy barrier for each
head. The three-dimensional structure of heads was optimized
without any geometrical constraints at the B3LYP-D3/6-31G**
level. Antechamber, an inbuilt tool in Amber, was used to calculate
Electrical Measurement. Electrical measurements were con-
2,17
ducted as previously reported.
A gold-coated tip was used to form
a contact with the bottom electrode (Ag substrate) with a
micropositioner (Signatone). A smoothened EGaIn electrode was
brought in contact with the SAM with a micromanipulator to form a
junction (physical contact). The measured diameter of the junction
was used to estimate the area of the junction, assuming a full circular
contact. Current density vs voltage data were collected in a sweeping
bias from −0.5 V to 0.5 V, with a step size of 0.05 V. At least 20 scans
were performed on each junction, and at least 20 junctions were
studied per molecule.
32
RESP charges and to make topology files of all molecules. The all-
atom MD simulations of each head were performed using the
33
AMBER program utilizing the GAFF2 force field. The CPU version
of the sander engine which comes integrated with Amber16 tools was
3
adopted in this study. Each system was minimized over 10 000 steps
using the conjugate gradient (CG) method with a constant dielectric
medium instead of implicit or explicit solvents. The minimized
structures were gradually heated from 0 to 300 K using a constant
NVT ensemble over 500 ps with a Berendsen temperature coupling
and a nonbonded “cutoff” of 999 Å. Under the same NVT ensemble
setup at 300 K with no periodic boundary conditions and 1 fs of time
step, 100 ns of the production run was carried out. After the MD
simulations were performed, the 100 ns of trajectory for each system
was clustered using the “average-linkage” algorithm, generating the
RESULTS AND DISCUSSION
■
To delineate the effect of conformation dynamics at the
electrode−SAM interface, we herein compare the effect of
having open or closed (cyclic) amide head groups paying
special attention to parity (odd−even) effect(s). To mitigate
the effect of gauche defects, we anchor these amides with
either a C or C alkyl thiol such that the overall SAMs are ≥
1
0
11
10,30,36−39
C n-alkylthiol analogue in length (Figure 1d).
34,35
14
top 10 most dominating clustered structures for each system.
In
Besides mitigating gauche defects, the length ensures solid-like
the final step, the most dominating structures were further subjected
to DFT single-point energy calculations to estimate the gas-phase
energy for all possible conformers.
ESP and group volumes were calculated using Spartan Student
v7.27. All head group structures were calculated as the equilibrium
geometry in gas with density functional EDF2 using a polarization
basis setting of 6-31G*. Calculated orbitals and energies were printed
generating both ESP map and volume.
10,30,36−39
(
crystalline) ordering (Figure 1c).
We infer that this
crystalline all-trans extended structure will allow for delineation
of the effect of conformational order and its perturbation to
tunneling of these homologous modifiers (Figure 1d).
Considering the geometric consequences of introducing open
and cyclic dialkyl amide head groups, we observe that though
the amide dipole remains relatively unperturbed these
analogues have significant differences in volume (Figure 1d).
From the electrostatic potential maps, we observe that
Vpyrrolidine < Vdiethyl < Vpiperidine < Vdipropyl. These moieties,
however, have different degrees of conformation freedom that
do not directly relate to the progression in volumetric change,
suggesting that conformational degree of freedom is more
aligned with structures (open vs cyclic) than volumethe
Chemical Synthesis. All molecules were synthesized from the
corresponding brominated carboxylic acids, based on the carbon
chain length (Scheme 1). Conversion of the carboxylic acid was
Scheme 1. Schematic Illustration of the Synthesis of Dialkyl
a
Amides Used in This Study
1
latter being essential in SAM packing. To confirm this
inference, we applied molecular dynamics (MD) simulations
and quantum mechanics (QM) analysis to generate different
head group conformations and compute their energies,
respectively. From simulation results, we confirmed the
existence of a set of cluster structures with similar energies
Table S1), especially in head groups 1, 2, and 4, suggests that it
can easily adopt other conformers, whereas, as expected for
S4). Based on the simulation results, and as expected, we
confirm that the conformational degree of freedom for these
a
A carboxylic acid is converted to an acyl chloride followed by
amidation and installation of the thiol.
achieved by reaction with oxalyl chloride with catalytic N,N-
dimethylformamide. Crude product was able to be used to react
with the correct amines to form the bromo amide product. Following
purification, thiourea allows substitution of the bromine to the thiol.
Dithiothreitol with triethylamine allowed reduction of disulfides that
1
13
Substrate Preparation and SAM Fabrication. The substrates
13
used were prepared in house as previously reported. In brief, a 200
nm Ag thin film was deposited on a fresh Si wafer (111) using e-beam
evaporation. Template stripping was then performed to reveal the
Ag−Si interface.
head groups follows the order: DOFpyrrolidine < DOFpiperidine
<
DOFdiethyl < DOFdipropyl. For brevity, we schematically illustrate
the differences in conformation cluster using pyrrolidine (3)
and piperidine (4) (Figure 1e). For clarity, we summarize
differences in the molecules based on head group volume, head
group conformation degree of freedom, and surface normal
dipole (Figure 1f). Besides the head groups, due to the well-
known odd−even effect, the alkyl space must also be
considered. The C10 alkanethiol backbone, for example, orients
the amide dipole moment at 79.7° relative to the surface
normal, while the C11 alkanethiol backbone orients an
analogous dipole moment 120.3° relative to the surface
normal. This study is, therefore, organized in two sets based
on the underlying spacer alkyl thiol.
To fabricate SAMs on Ag surfaces, we follow the same procedure as
13
previously reported. In brief, the template-stripped substrates were
precleaned with ethanol and then placed in a vial containing 3 mmol
of alkanethiol in 5 mL of ethanol. The substrate and solution were
incubated, under Argon gas, for 3 h. The substrates with SAMs were
then rinsed with ethanol and dried with nitrogen gas.
Preparation of a Smoothened EGaIn Top Electrode. A
polished EGaIn (99.99%, Sigma-Aldrich) conical tip was formed as
previously reported.
syringe onto a surface on which EGaIn would stick. The syringe was
slowly pulled away using a micromanipulator (Narishige You-1) from
the surface to form a conical-shaped tip after breaking the droplet.
The conical tip was smoothened by dipping into 5% acetic acid
2,17
In brief, a droplet of EGaIn pushed from a
(
glacier, 99.7%) solution in ethanol for 2 min.
1
3880
J. Am. Chem. Soc. 2021, 143, 13878−13886
Chen, Jiahao
