Available via license: CC BY 4.0
Content may be subject to copyright.
1
Dynamic Au-C anchors in molecular junctions under oriented external electric
fields
Rajarshi Samajdar1,2†, Hao Yang2,3†, Seungjoo Yi2,3†, Chun-I Wang4^, Michael A.
Pence2,4^, Moeen Meigooni2,5, Seth Putnam4, Xiaolin Liu4, Jitong Ren1,2, Jeffrey S.
Moore2,3,4, Emad Tajkhorshid2,4,5,6,7, Joaquín Rodríguez-López2,4,8, Nicholas E.
Jackson2,4, Charles M. Schroeder1,2,3,4,5,7,8*
1Department of Chemical and Biomolecular Engineering, University of Illinois Urbana-
Champaign, Urbana, Illinois, 61801
2Beckman Institute for Advanced Science and Technology, University of Illinois Urbana-
Champaign, Urbana, Illinois, 61801
3Department of Materials Science and Engineering, University of Illinois Urbana-
Champaign, Urbana, Illinois, 61801
4Department of Chemistry, University of Illinois Urbana-Champaign, Urbana, Illinois,
61801
5Center for Biophysics and Quantitative Biology, University of Illinois Urbana-Champaign,
Urbana, IL 61801
6Department of Biochemistry, University of Illinois at Urbana-Champaign, Urbana, Illinois,
61801, United States
7Department of Bioengineering, University of Illinois at Urbana-Champaign, Urbana,
Illinois, 61801, United States
8Materials Research Laboratory, University of Illinois at Urbana-Champaign, Urbana,
Illinois, 61801, United States
†Contributed equally, co-first author
^Contributed equally, co-second author
*Corresponding author
https://doi.org/10.26434/chemrxiv-2024-g4q8t ORCID: https://orcid.org/0000-0001-7656-8696 Content not peer-reviewed by ChemRxiv. License: CC BY 4.0
2
Abstract
Terminal anchor groups play a key role in controlling the stability and electronic properties
of molecular junctions. Single molecule junctions typically consist of two terminal anchors
linking organic molecules to metal electrodes. Here, we show that p-terphenyl derivatives
containing only a single terminal anchor exhibit conductance features similar to junctions
with two terminal anchors, which arises due to dynamic Au-C bond formation under
oriented external electric fields (OEEFs). A set of p-terphenyl derivatives with one terminal
anchor was prepared and characterized using automated chemical synthesis, single
molecule electronics experiments, molecular dynamics (MD) simulations, and non-
equilibrium Green’s function-density functional theory (NEGF-DFT) calculations. Our
results show that 4-amino-p-terphenyl (PPP) exhibits a distinct and well-defined high
conductance state that is greatly diminished or absent in other p-terphenyl derivatives
with single terminal anchors, whereas a low conductance state is observed in all amino-
p-terphenyl derivatives due to non-covalent dimeric interactions. The electronic properties
of PPP are characterized using a combination of cyclic voltammetry, electrolysis, and
electron spin resonance, revealing that the high conductance state in PPP arises due to
robust Au-C bond formation facilitated by a radical-based rigid resonating structure under
OEEFs. A series of control experiments on junctions with different anchor groups reveals
the role of primary amines in forming dynamic linkages under OEEFs. Overall, these
results suggest that OEEFs can trigger Au-C bond formation leading to high conductance
pathways in organic molecules containing only one terminal anchor. Insights from this
work can be leveraged in the design of molecular electronic devices, particularly in
understanding the mechanisms of molecular binding and junction formation.
https://doi.org/10.26434/chemrxiv-2024-g4q8t ORCID: https://orcid.org/0000-0001-7656-8696 Content not peer-reviewed by ChemRxiv. License: CC BY 4.0
3
Introduction
Molecular junctions generally consist of three main components: a molecular bridge1
metal electrodes2, and terminal anchor groups that electronically couple the bridge to the
electrodes3–6. Terminal anchors3–5 play a key role in controlling the electronic properties
of single molecule junctions via molecular binding, electronic coupling, and junction
stability. One class of robust terminal anchors is characterized by dative bonding
interactions with metal electrodes, including amine6,7, thiol8, pyridine9, and methyl thiol10
groups. However, molecular junctions generally require two terminal anchor groups to
form a closed circuit, which poses synthetic challenges that tend to restrict the chemical
space in molecular junction design. From this view, new binding modalities that avoid the
requirement of two terminal anchors could expand the chemical scope of the toolbox
available for molecular electronics.
In addition to dative anchor groups, covalent anchors have been reported to yield stable
and highly conductive molecular junctions11. However, direct formation of Au-C covalent
bonds in molecular junctions is challenging and typically requires cleavage or reaction of
functional groups such as iodine12, alkyne13,14, or diazonium15,16 to generate Au-C bonds
through in situ reactions. Recently, oriented external electric fields17,18 (OEEFs) have
gained increasing attention for their ability to reorganize the electron distribution of
molecules, stabilize charge-separated resonant forms, and trigger electrochemical
reactions. This strategy can be used to facilitate the formation of direct Au-C covalent
linkages, without the need for cleavage of functional groups, as a new binding mechanism
for single molecule electronics. Nevertheless, the structural and electronic properties that
promote the formation of Au-C contacts under OEEFs is not yet fully understood.
Singly anchored molecules lack a second terminal anchor group to complete a closed
electronic circuit. Prior work has reported that electron transport in singly anchored
molecules occurs by non-covalent interactions such as π−π stacking19,20 and Au-π
interactions21. However, the unanchored terminus is exposed to strong electric fields due
to the close proximity to the electrode surface in the junction, which could induce a
dynamic second anchor via the formation of an Au-C bond18. In addition, understanding
the electron transport behavior in π-conjugated molecules with one terminal anchor group
could provide additional insights into non-covalent intermolecular interactions.
Intermolecular charge transport22 in π-stacked aromatic groups is critical to organic
electronics23. The efficiency of electron transport in π-stacked systems depends on the
electronic coupling and the distance and orientation between neighboring π-stacked
molecules24–26. Although recent work has examined electron transport in singly anchored
organic molecules20,27,28, we lack a complete understanding of the potential for triggered
or dynamic anchor formation for singly anchored molecules under OEEFs.
In this work, we investigate the electron transport behavior of singly anchored p-terphenyl
derivatives using a combination of automated chemical synthesis, single molecule
electronics experiments, molecular dynamics (MD) simulations, and non-equilibrium
https://doi.org/10.26434/chemrxiv-2024-g4q8t ORCID: https://orcid.org/0000-0001-7656-8696 Content not peer-reviewed by ChemRxiv. License: CC BY 4.0
4
Green’s function-density functional theory (NEGF-DFT) calculations. Synthesis of p-
terphenyl derivatives was performed using a rapid Suzuki-Miyaura cross coupling
(SMCC) method with reduced reaction time and temperature29. Following synthesis, the
electronic properties of these molecules were characterized using the scanning tunneling
microscope-break junction (STM-BJ) technique30,31. Our results show that 4-amino-p-
terphenyl (PPP) has a surprisingly well-defined high conductance feature despite the
presence of only one terminal anchor, but this high conductance feature is greatly
diminished or absent in all other terphenyl derivatives studied in this work. Our results
further reveal a low conductance feature for all singly anchored amino-p-terphenyl
derivatives studied in this work, which arises due to non-covalent intermolecular
interactions. Flicker noise analysis and machine learning methods such as correlation 2D
analysis and Gaussian mixture modeling (GMM) are used to understand the conductance
behavior. In addition, cyclic voltammetry (CV), bulk-scale electrolysis, and electron spin
resonance (ESR) were used to understand the origin of the high conductance state, which
arises due to Au-C bond formation under OEEFs within the nanogap of molecular
junctions32.
Results and Discussion
Chemical synthesis and characterization
Synthesis of organic molecules for electronics experiments was carried out using Suzuki-
Miyaura cross coupling (SMCC) by leveraging recent advances in iterative automated
synthesis29. Prior SMCC conditions for automated synthesis require more than twelve
hours to prepare one molecule33. Manual synthesis requires even longer time scales to
synthesize a library of small molecules34,35. In this work, we used automated iterative
coupling 33,36 based on a new rapid SMCC method (Supporting Information Section
S1-S3) that significantly decreases the reaction time to ten minutes with high yield29.
An automated small molecule synthesizer (Figure 1a) capable of parallel runs of
deprotections, couplings, and purifications was used, as previously reported33. The
building blocks and protecting groups used in the reaction are shown in Figure 1b. The
synthesis of 4''-methyl-[1,1':4',1''-terphenyl]-4-amine (MPP), 3-fluoro-[1,1':4',1''-
terphenyl]-4-amine (PPF), N-methyl-[1,1':4',1''-terphenyl]-4-amine (PPS), and 4-([1,1’-
biphenyl]-4-yl) pyridine (PPN) was carried out using one step slow release SMCC
chemistry (Figure 1c). 2'-methyl-[1,1':4',1''-terphenyl]-4-amine (PMP) and 3-methyl-
[1,1':4',1''-terphenyl]-4-amine (PPM) were synthesized using a two-step synthesis
process: rapid SMCC followed by slow release SMCC (Figure 1d). Synthesized
molecules were characterized using mass spectrometry and 1H and 13C nuclear magnetic
resonance (NMR) spectrometry (Supplementary Figures 1-12). [1,1':4',1''-terphenyl]-4-
amine (PPP) and 1,1’,4’,1’’-terphenyl-4-thiol (PPT) were purchased from Sigma Aldrich.
https://doi.org/10.26434/chemrxiv-2024-g4q8t ORCID: https://orcid.org/0000-0001-7656-8696 Content not peer-reviewed by ChemRxiv. License: CC BY 4.0
5
In this work, the chemical space (Figure 2a) focuses on p-terphenyl molecules for
molecular electronics due to their rigidity and extended π-conjugation, which results in
low tunneling barriers37–40. We designed a series of singly anchored p-terphenyl
molecules based on the parent molecule PPP by including additional substituents on the
three benzene rings. PPM, PMP, and MPP each contain a methyl group on one of the
aromatic rings, which changes preferred resonance structures, alters ring torsional
Figure 1: Overview of automated chemical synthesis method. (a) Picture of automated
synthesis instrument in our lab for organic molecule synthesis based on Suzuki-
Miyaura cross coupling (SMCC). (b) Summary of molecular building blocks and
boronate protecting groups. (c) One-step slow release SMCC synthesis. (d) Two-step
synthesis, consisting of rapid SMCC followed by slow release SMCC.
https://doi.org/10.26434/chemrxiv-2024-g4q8t ORCID: https://orcid.org/0000-0001-7656-8696 Content not peer-reviewed by ChemRxiv. License: CC BY 4.0
6
angles, or introduces steric hindrance during the formation of molecular junctions.
Molecule PPF was incorporated to investigate the effect of electron withdrawing groups
in contrast to the electron donating nature of the methyl substituted terphenyls. Three
additional control molecules, PPN, PPS, and PPT, are characterized to understand the
role of chemical anchor groups.
Single molecule conductance measurements
The electronic properties of p-terphenyl derivatives were characterized at the single
molecule level using the scanning tunneling microscope-break junction (STM-BJ)
technique30,31. The STM-BJ setup consists of a gold tip electrode that is repeatedly moved
into and out of contact with a gold substrate electrode in a solution containing molecules,
resulting in the continual formation and breakage of single molecule junctions
(Supporting Information Section S1). The STM-BJ instrument is automated, and
experiments are repeated over an ensemble of >5000 molecules for each experiment.
Single molecule conductance data are then analyzed using one- and two-dimensional
(1D and 2D) conductance histograms without data selection. The timescale of a single
STM-BJ pulling trajectory is in the order of milliseconds41, which allows for sampling a
range of molecular conformations42 during a conductance measurement. Prior to
understanding the electron transport behavior of singly anchored terphenyl derivatives,
we characterized the molecular conductance behavior of the molecular analog with two
terminal anchor groups (4,4’-diamino-p-terphenyl), and our results reveal a bimodal
conductance distribution, with a prominent high conductance feature and weak low
conductance feature, consistent with prior literature6 (Supplementary Figure 13).
Single molecule conductance traces for PPP unexpectedly show two distinct populations,
as demonstrated by characteristic single molecule conductance traces for high and low
conductance features (Figure 2b). Singly anchored molecules are generally thought to
form active junctions through intermolecular stacking interactions19,20, where two different
molecules, each anchored to a different electrode, form non-covalent dimeric interactions
to complete the circuit. We hypothesized that the low-conductance feature in singly
anchored p-terphenyl derivatives arises due to intermolecular stacked junctions (Figure
2c). However, in the STM-BJ setup, the applied bias between the two electrodes results
in relatively high electric field gradients in the nanoscale junction, thereby giving rise to
an OEEF17,18. We posited that OEEFs can induce a dynamic anchor in molecular
junctions involving the formation of a new second anchor (Figure 2d) without requiring
the cleavage or chemical reaction of additional functional groups. We aimed to
understand if OEEFs could give rise to dynamic covalent anchors resulting in well-
defined43 conductance pathways in singly anchored organic molecules.
https://doi.org/10.26434/chemrxiv-2024-g4q8t ORCID: https://orcid.org/0000-0001-7656-8696 Content not peer-reviewed by ChemRxiv. License: CC BY 4.0
7
We began by characterizing the electron transport properties of amino-p-terphenyl
derivatives in non-polar solvents (1,2,4-trichlorobenzene, TCB) using the STM-BJ
method. Our results show that amino-p-terphenyl derivatives exhibit a characteristic low
conductance feature (Figure 3a and Supplementary Figure 14). Surprisingly, PPP
shows an additional high conductance feature occurring in a significant molecular sub-
population over a large ensemble of molecules (Figure 3b). Unsupervised machine
learning (Supporting Information Section S1), 2D correlation analysis, and Gaussian
mixture modeling (GMM), were used to interpret the two-state conductance behavior
observed for PPP. 2D correlation analysis and GMM indicates that the two conductance
Figure 2: Molecular library and electron transport mechanisms for singly anchored
terphenyl derivatives. (a) Structures of singly anchored organic molecules studied in
this work. (b) Characteristic single molecule conductance traces observed in STM-BJ
experiments for PPP. HC denotes the high conductance state, whereas LC denotes
the low conductance state. (c) Molecular junction schematic with non-covalent dimeric
interactions for PPP. (d) Molecular junction schematic featuring an Au-C bond for PPP.
https://doi.org/10.26434/chemrxiv-2024-g4q8t ORCID: https://orcid.org/0000-0001-7656-8696 Content not peer-reviewed by ChemRxiv. License: CC BY 4.0
8
states are negatively correlated and therefore occur independently in distinct trajectories
during the STM-BJ experiments (Figures 3c, d). These results suggest that the two
conductance populations arise from two different junction configurations.
Flicker noise analysis44 was performed to differentiate between through-bond and
through-space electron transport for the high and the low conductance states
(Supporting Information Section S1). Prior work has shown that the conductance
fluctuations (conductance noise power) exhibit a power law dependence on the mean
conductance G values depending on through-space and through-bond transport
characteristics6,45. Conductance noise is quantified by numerically integrating the
conductance noise power spectral density (PSD) between frequencies of 100 and 1000
Hz22,46. The correlation is quantified by the scaling exponent n of the normalized noise
power (noise power/Gn) versus the average normalized conductance G/G0, where G0 is
the conductance quantum. A scaling exponent n ≈ 2 suggests through-space transmission
whereas an exponent n ≈ 1 corresponds to through-bond transport. These results show
Figure 3: Scanning tunneling microscope-break junction (STM-BJ) measurements of
amino-p-terphenyl derivatives in 1,2,4-trichlorobenzene (TCB) at 250 mV applied bias.
(a) 1D conductance histograms for PPP, PPM, PMP, and MPP. (b) 2D conductance
histogram for PPP. (c), (d) Gaussian mixture modeling and 2D correlation analysis on
PPP revealing anti-correlated high and low conductance states. (e) Flicker noise
analysis of molecular sub-populations corresponding to the high and low conductance
feature for PPP, which suggests through-bond and through-space characteristics for
the high and low conductance features (scaling exponents of n 1.3 and n 2,
respectively).
https://doi.org/10.26434/chemrxiv-2024-g4q8t ORCID: https://orcid.org/0000-0001-7656-8696 Content not peer-reviewed by ChemRxiv. License: CC BY 4.0
9
that the low conductance state in PPP occurs by through-space electron transport,
whereas the high conductance state shows dominant through-bond electron transport
characteristics (Figure 3e). Flicker noise analysis for MPP (Supplementary Figure 15)
is also consistent with through-space electron transport for the low conductance state.
Based on these results, we posit that the low conductance state observed for all amino-
p-terphenyl derivatives in this work arises due to non-covalent dimeric interactions. We
posit that the high conductance state in PPP arises due to the formation of a dynamic
covalent anchor binding to the metal electrode, which is consistent with through-bond
transport for the high conductance state. Based on the structure of PPP, we hypothesize
that the second anchor involves the formation of an Au-C bond under the influence of
OEEFs. To explore this hypothesis further, we pursued a series of additional experiments
and simulations.
2D conductance histograms for amino-p-terphenyl derivatives indicate stark differences
in electron transport pathways for different molecular composition. Notably, the high
conductance state observed in PPP is significantly diminished or absent in all other singly
anchored p-terphenyl molecules studied in this work (Figure 3a and Supplementary
Figure 14), suggesting that molecular substituents disrupt the high conductance pathway
in PPP. Although PPM and PMP show a weak high conductance feature, it is completely
absent in MPP, implying that the presence of a methyl group at the para position on the
aromatic ring completely abolishes the high conductance pathway. OEEFs are known to
stabilize charge resonance structures and catalyze chemical reactions17,18. It is possible
that the quinoidal resonant form of the axisymmetric molecule PPP is relevant for electron
transport, with electron density redistributing towards the para position on the phenyl ring
farthest from the amine anchor.
We hypothesize that the carbon atom at the terminal para position in amino-p-terphenyl
derivatives PPP, PPM, and PMP forms a direct Au-C covalent linkage, creating a single
molecule bridge that completes the circuit and leads to the high conductance pathway
(Figure 2d). The absence of the high conductance feature in MPP is consistent with the
hypothesis that the methyl group at the terminal para position inhibits the interaction
between the molecule and the gold electrode, preventing the formation of the Au-C
covalent linkage. However, the para position remains available for binding in PPM and
PMP, but the substituent groups on these molecules disrupt their symmetry, which
disfavors the formation of the quinoidal resonant form. PMP contains a methyl substituent
on the central ring, which alters the torsional angles and hinders the formation of the
planar quinoidal resonant form, resulting in a weak high conductance feature compared
to PPP.
We further characterized the electronic properties of PPF, which contains an electron
withdrawing group as opposed to the electron donating group in the methyl substituted
terphenyl molecules. PPF exhibits similar electron transport characteristics to PPM
(Supplementary Figure 16), suggesting that the nature of substituent does not
https://doi.org/10.26434/chemrxiv-2024-g4q8t ORCID: https://orcid.org/0000-0001-7656-8696 Content not peer-reviewed by ChemRxiv. License: CC BY 4.0
10
significantly affect the high conductance state. These results indicate that disruption of
symmetry in the amino-p-terphenyl derivatives can inhibit the high conductance state. In
addition, we also characterized the electronic properties of PPN, PPT, and PPS, which
contain thiol, pyridine, and N-methylamino as terminal anchor groups, respectively. The
electron transport behavior of PPN, PPT, and PPS (Supplementary Figure 17) indicates
the absence of a high conductance state in these molecules. PPN exhibits a low
conductance state, which is absent or significantly diminished in PPT and PPS, and likely
arises due to non-covalent dimeric interactions.
We conducted a series of additional single molecule experiments by varying temperature,
concentration, and applied bias to further understand the high- and low-conductance
populations observed for PPP. Temperature-dependent STM-BJ measurements were
carried out at three different temperatures: 20 0C, 30 0C and 40 0C. Increased temperature
leads to enhanced molecular vibrations, reducing the likelihood of forming intermolecular
junctions during the STM-BJ experiments47. Our results indicate that as the temperature
is increased, the low conductance state is significantly diminished (Figure 4a). These
results are consistent with our hypothesis that the low conductance state arises due to
non-covalent dimeric interactions. On the other hand, the high conductance state is
largely unaffected by the 20ºC temperature increase, suggesting the high conductance
state arises due to a stronger binding mechanism. Concentration-dependent STM-BJ
experiments were also performed for PPP in the range on 0.1-10 mM (Figure 4b). These
results indicate that the low conductance state of PPP exhibits strong concentration
dependence, whereas the high conductance state is concentration independent. These
results are consistent with our hypothesis that the low conductance state arises due to
non-covalent dimeric interactions whereas the high conductance state arises due to
through-bond transport.
Bias dependent STM-BJ experiments (Figure 4c) reveal that at low applied bias (50-150
mV), only the high conductance state is observed. These results indicate that even at
relatively low applied biases, OEEFs are sufficient to induce Au-C linkages. As the applied
bias is increased, the low conductance state emerges, consistent with prior work reporting
that increased bias regulates dimeric interactions in molecular junctions28. The two-state
conductance characteristic of PPP is also observed in a polar solvent such as propylene
carbonate (PC), but with lower counts for each conductance population as the higher
dielectric strength of polar solvents likely reduces the effect of OEEFs18 (Figure 4d). We
also characterized the molecular conductance of PPP in PC in the presence of a reducing
agent (sodium borohydride, NaBH4), which results in the disappearance of the high
conductance state (Supplementary Figure 18). Based on these results, we posit that a
charge separated quinoidal resonant form of PPP is involved for the high conductance
state. Overall, results from STM-BJ experiments indicate that the high conductance in
PPP state involves monomeric through-bond electron transport.
https://doi.org/10.26434/chemrxiv-2024-g4q8t ORCID: https://orcid.org/0000-0001-7656-8696 Content not peer-reviewed by ChemRxiv. License: CC BY 4.0
11
Figure 4: Effect of temperature, concentration, and applied bias on the molecular
conductance of PPP. (a) Temperature-dependent conductance measurements are
consistent with non-covalent dimeric interactions for low conductance state, whereas
the high conductance state is robust to temperature variations. (b) Concentration-
dependent measurements for PPP. (c) Bias-dependent STM-BJ experiments
indicating that the high conductance state is observed at low bias, whereas the high
conductance state has bias dependence. (d) Solvent dependent STM-BJ
measurements indicating that the bimodal conductance distribution is observed in non-
polar (1,2,4-trichlorobenzene, TCB) and polar (propylene carbonate, PC) solvents.
https://doi.org/10.26434/chemrxiv-2024-g4q8t ORCID: https://orcid.org/0000-0001-7656-8696 Content not peer-reviewed by ChemRxiv. License: CC BY 4.0
12
Bulk electrochemistry and spectroscopy
Integrating single molecule measurements with bulk experiments offers a powerful
approach to understand electron transport48,49. We performed a series of bulk
electrochemical and spectroscopy experiments to understand the electronic properties of
PPP, focusing on the origin of the high conductance state. Results from cyclic voltammetry
(CV) experiments show a distinct oxidation wave on the forward scan at approximately
0.65 V vs. Ag/AgCl, with only a small reduction peak at 0.15 V on the return scan (Figure
5a). The absence of a prominent return peak in the CV indicates that a distinct chemical
process occurs after oxidative electron transfer. A second distinct electron transfer event
was observed when the positive limit of the potential window was expanded from 0.8 V
to 1.0 V, however this second electron transfer event leads to the formation of a surface-
bound species (Supplementary Figure 19). Bulk electrolysis was performed to identify
the nature of the products generated upon electrochemical oxidation of PPP. The potential
was maintained at 0.8 V during electrolysis to isolate the first electron transfer event and
avoid possible film formation on the electrode surface. The solution in the working
electrode compartment exhibited dramatic changes in color, changing from colorless to
purple upon oxidation (Figure 5b). All electrochemical experiments were performed in
triplicates, and the integrated charge was used to determine the total number of electrons
transferred during the electrochemical oxidation. Our results (Figure 5c) indicate that one
electron was transferred, which suggests that the electrochemical oxidation of PPP forms
a radical cation species.
Electron spin resonance (ESR) spectroscopy experiments were performed on PPP before
electrolysis, after electrolysis, and after preconcentration of the electrolysis product, as
shown in Figure 5d. The signals at ~3480 G and ~3650 G indicate the presence of a
radical species in the electrolyzed solutions that is not present in the pre-electrolyzed
sample. Based on these results, we posit that the high conductance state of PPP arises
due to a radical species (Supplementary Figure 20), which leads to the formation of a
dynamic anchor based on a covalent Au-C linkage (Figure 2d). Computational modeling
including MD simulations and NEGF-DFT calculations was further pursued to
complement experimental results.
Molecular dynamics (MD) simulations
MD simulations have been used to characterize complex molecular geometries or
conformations adapted by molecular junctions42, enabling comparison between
experimental and theoretical results. Here, we used MD simulations to characterize the
non-covalent dimeric interactions that resemble the low conductance state observed in
the single molecule experiments.
A series of MD simulations was performed to elucidate the dimeric interactions between
methyl substituted amino-p-terphenyl derivatives. For each terphenyl derivative, a pair of
molecules was simulated in TCB and PC solvents, and the separation distance between
https://doi.org/10.26434/chemrxiv-2024-g4q8t ORCID: https://orcid.org/0000-0001-7656-8696 Content not peer-reviewed by ChemRxiv. License: CC BY 4.0
13
their center-of-mass (COM) positions was systematically varied to provide molecular
insights from both energetic and conformational perspectives (Supporting Information
Section S1). The potential mean force (PMF) as a function of the separation distance
between the COM of terphenyl pairs in solution was assessed using umbrella sampling
and the Weighted Histogram Analysis Method (WHAM)50.
Figure 5. Bulk electrochemistry and spectroscopy experiments on PPP. (a) Cyclic
voltammetry of 1 mM PPP in an electrolyte solution of 0.1 M TBAPF6 in PC. The
potential was swept from 0 V vs. Ag/AgCl to 0.8 V vs. Ag/AgCl in forward and reverse
scans at rates ranging from 25 mV/s up to 1 V/s. (b) Images of a bulk electrolysis
cell before and after electrolysis, showing the dramatic color change in the working
electrode compartment (right) upon oxidation. (c) Bulk electrolysis of 1 mM PPP in
an electrolyte solution of 0.1 M TBAPF6 in PC, at an applied potential of 0.8 V, taken
in triplicate (N = 3). The black trace indicates the mean, and the red error bar
represents 1 standard deviation from the mean. The average number of electrons
transferred was 1.1 ± 0.1. (d) ESR spectra of the 1 mM PPP with no electrolysis,
directly after electrolysis, and after electrolysis with a preconcentration step.
https://doi.org/10.26434/chemrxiv-2024-g4q8t ORCID: https://orcid.org/0000-0001-7656-8696 Content not peer-reviewed by ChemRxiv. License: CC BY 4.0
14
The PMFs of PPP, MPP, and PPM in TCB solution exhibit similar profiles, with the global
minima located at a separation distance of 5.1 Å (Figure 6a). The corresponding binding
energies (Supplementary Table 1) between two terphenyl molecules is 0.63 ± 0.04 kBT
for PPP, 0.68 ± 0.05 kBT for MPP, and 0.79 ± 0.04 kBT for PPM, where T is the absolute
temperature at 300 K. On the other hand, the global minimum of PMP is situated at a
separation distance of 5.9 Å with a binding energy of 0.19 ± 0.05 kBT due to the non-
Figure 6. Molecular dynamics (MD) simulations and non-equilibrium Green’s function-
density functional theory (NEGF-DFT) calculations. (a) Potential mean force (PMF)
profiles for dimers. (b) Distribution of the angle between the long axes of terphenyl
molecules in 1,2,4-tricholorobenzene (TCB). Arrows in the inset represent the long-
axis vector. (c) Experimental and computational transmission values at the Fermi
energy level for the stacked molecular conformations for several amino-p-terphenyl
derivatives. (d) Transmission plots for the high and low conductance state of PPP.
https://doi.org/10.26434/chemrxiv-2024-g4q8t ORCID: https://orcid.org/0000-0001-7656-8696 Content not peer-reviewed by ChemRxiv. License: CC BY 4.0
15
planar intramolecular conformation and steric hinderance caused by its methyl group at
the middle phenyl ring (Supplementary Figure 21). The observed binding energies for
the terphenyl derivatives are below 1 kBT, which is consistent with results from the
temperature-dependent STM-BJ experiments.
We further characterized the structural features of the terphenyl molecules based on the
MD trajectories of the PMF global minimum. Figure 6b shows the angle distribution
between the long axes of terphenyl molecules in TCB solvent, which quantifies the
intermolecular structure between terphenyl dimers. An angle close to 1800 indicates that
the terphenyl molecules are aligned in the same parallel direction, increasing the
likelihood of π-π stacking and providing an efficient conductance pathway. For MPP, a
bimodal distribution is observed which could arise due to the methyl group leading to an
offset in stacking for the dimeric structure. The symmetric geometry of PPP in TCB
promotes a more parallel pair structure, ensuring an effective stacked dimeric structure.
A similar trend in PMFs and conformational features was observed for peptides in PC
solution (Supplementary Figure 22). Molecular conformations generated by MD can be
used in computationally efficient quantum mechanics (QM) calculations to aid in
comparison between theory and experimental results.
Electron transport calculations
Electron transport calculations provide a powerful tool to validate experimentally observed
conductance behavior42,43. NEGF-DFT simulations were performed for methyl substituted
amino-p-terphenyl derivatives in a stacked dimer geometry, corresponding to the low
conductance state characterized by non-covalent intermolecular interactions
(Supporting Information Section S1). For PPP, NEGF-DFT calculations were further
carried out for the proposed high conductance state involving Au-C covalent linkages.
Transmission plots for the stacked dimeric structures for PPP, PPM, and PMP show
qualitatively similar behavior (Supplementary Figure 23). Comparing the transmission
values at the Fermi energy level (Figure 6c) shows that MPP exhibits lower conductance
compared to the other derivatives, in both experiments and NEGF-DFT calculations.
These results suggest that the presence of a methyl group at the terminal para position
on the phenyl ring opposite to the amine can cause an offset in stacked molecular
conformation. These results also support experimental results suggesting that the low
conductance state arises due to non-covalent dimeric interactions.
We also performed NEGF-DFT calculations for the high conductance state of PPP, which
involves the formation of an Au-C covalent bond. Our results indicate that there is a tenfold
difference in computed transmission values between the high and low conductance states
of PPP (Figure 6d), in accordance with our experimentally observed STM-BJ results. The
differences between experimental and computed transmission values could arise due to
the semi-local exchange-correlation functional (PBE) used in our calculations, which
tends to underestimate the HOMO-LUMO gap between the electrodes and molecule. In
addition, the electron transmission calculations are carried out at zero applied bias. The
https://doi.org/10.26434/chemrxiv-2024-g4q8t ORCID: https://orcid.org/0000-0001-7656-8696 Content not peer-reviewed by ChemRxiv. License: CC BY 4.0
16
experimental STM-BJ results indicate an increase in conductance with an increase in
bias. Overall, NEGF-DFT calculations combined with MD simulations support the
hypothesis that the low conductance state arises due to non-covalent intermolecular
interactions and the high conductance state involves the formation of a robust Au-C
covalent linkage.
Conclusions
In this work, the electronic properties of singly anchored amino-p-terphenyl derivatives
are characterized using automated chemical synthesis, single molecule electronics
experiments, bulk scale electrochemistry, MD simulations, and NEGF-DFT calculations.
Single molecule experiments reveal two well-defined conductance pathways in some
amino-p-terphenyl derivatives due to the formation of robust Au-C covalent linkages under
OEEFs. Bulk electrochemistry and spectroscopy experiments show that a radical cation
state occurs when PPP is exposed to an electric field, which facilitates the formation of
robust Au-C covalent linkages. Our results show that the formation of Au-C linkages is
favored when the charge separated resonant state experiences minimal disruption. The
introduction of substituents on the three benzene rings within the terphenyl system
disrupts this state, hindering the high-conductance electron transport pathway. In
addition, our work highlights the importance of non-covalent dimeric interactions in
molecular electronics that can be leveraged for the design of materials for bulk scale
measurements. MD simulations are used to understand the stacking conformations for
various amino-p-terphenyl derivatives, and NEGF-DFT calculations are carried out to
understand the electron transport behavior observed in single molecule experiments.
Overall, our work highlights the role of OEEFs in generating dynamic anchor groups in
molecular junctions without requiring the cleavage or conversion of functional groups,
which enhances the chemical toolbox available for constructing molecular electronics.
https://doi.org/10.26434/chemrxiv-2024-g4q8t ORCID: https://orcid.org/0000-0001-7656-8696 Content not peer-reviewed by ChemRxiv. License: CC BY 4.0
17
References
1. Chen, Hongliang, and J. Fraser Stoddart. "From molecular to supramolecular
electronics." Nature Reviews Materials 6, 804-828 (2021).
2. Li, Songsong, Hao Yu, Xinyi Chen, Andrew A. Gewirth, Jeffrey S. Moore, and Charles
M. Schroeder. "Covalent Ag–C bonding contacts from unprotected terminal acetylenes
for molecular junctions." Nano letters 20, 5490-5495 (2020).
3. Hong, Wenjing, David Zsolt Manrique, Pavel Moreno-García, Murat Gulcur, Artem
Mishchenko, Colin J. Lambert, Martin R. Bryce, and Thomas Wandlowski. "Single
molecular conductance of tolanes: experimental and theoretical study on the junction
evolution dependent on the anchoring group." Journal of the American Chemical
Society 134, 2292-2304 (2012).
4. Kaliginedi, Veerabhadrarao, Alexander V. Rudnev, Pavel Moreno-García, Masoud
Baghernejad, Cancan Huang, Wenjing Hong, and Thomas Wandlowski. "Promising
anchoring groups for single-molecule conductance measurements." Physical
Chemistry Chemical Physics 16, 23529-23539 (2014).
5. Ward, Jonathan S., Andrea Vezzoli, Charlie Wells, Steven Bailey, Samuel P. Jarvis,
Colin J. Lambert, Craig Robertson, Richard J. Nichols, and Simon J. Higgins. "A
Systematic Study of Methyl Carbodithioate Esters as Effective Gold Contact Groups
for Single‐Molecule Electronics." Angewandte Chemie 136, e202403577 (2024).
6. Magyarkuti, András, Olgun Adak, Andras Halbritter, and Latha Venkataraman.
"Electronic and mechanical characteristics of stacked dimer molecular
junctions." Nanoscale 10, 3362-3368 (2018).
7. Quek, Su Ying, Maria Kamenetska, Michael L. Steigerwald, Hyoung Joon Choi,
Steven G. Louie, Mark S. Hybertsen, J. B. Neaton, and Latha Venkataraman.
"Mechanically controlled binary conductance switching of a single-molecule
junction." Nature nanotechnology 4, 230-234 (2009).
8. Inkpen, Michael S., Zhen–Fei Liu, Haixing Li, Luis M. Campos, Jeffrey B. Neaton, and
Latha Venkataraman. "Non-chemisorbed gold–sulfur binding prevails in self-
assembled monolayers." Nature chemistry 11, 351-358 (2019).
9. Kamenetska, M., Su Ying Quek, A. C. Whalley, M. L. Steigerwald, H. J. Choi, Steven
G. Louie, C. Nuckolls, M. S. Hybertsen, J. B. Neaton, and L. Venkataraman.
"Conductance and geometry of pyridine-linked single-molecule junctions." Journal of
the American Chemical Society 132, 6817-6821 (2010).
10. Yu, Hao, Songsong Li, Kenneth E. Schwieter, Yun Liu, Boran Sun, Jeffrey S. Moore,
and Charles M. Schroeder. "Charge transport in sequence-defined conjugated
oligomers." Journal of the American Chemical Society 142, 4852-4861 (2020).
11. Cheng, Z-L., Rachid Skouta, Hector Vazquez, Jonathan R. Widawsky, Severin
Schneebeli, W. Chen, Mark S. Hybertsen, Ronald Breslow, and Latha Venkataraman.
"In situ formation of highly conducting covalent Au–C contacts for single-molecule
junctions." Nature nanotechnology 6, 353-357 (2011).
12. Starr, Rachel L., Tianren Fu, Evan A. Doud, Ilana Stone, Xavier Roy, and Latha
Venkataraman. "Gold–carbon contacts from oxidative addition of aryl iodides." Journal
of the American Chemical Society 142, 7128-7133 (2020).
https://doi.org/10.26434/chemrxiv-2024-g4q8t ORCID: https://orcid.org/0000-0001-7656-8696 Content not peer-reviewed by ChemRxiv. License: CC BY 4.0
18
13. Bejarano, Francesc, Ignacio Jose Olavarria-Contreras, Andrea Droghetti, Ivan
Rungger, Alexander Rudnev, Diego Gutiérrez, Marta Mas-Torrent et al. "Robust
organic radical molecular junctions using acetylene terminated groups for C–Au bond
formation." Journal of the American Chemical Society 140, 1691-1696 (2018).
14. Pla-Vilanova, Pepita, Albert C. Aragonès, Simone Ciampi, Fausto Sanz, Nadim
Darwish, and Ismael Diez-Perez. "The spontaneous formation of single-molecule
junctions via terminal alkynes." Nanotechnology 26, 381001 (2015).
15. Peiris, Chandramalika R., Yan B. Vogel, Anton P. Le Brun, Albert C. Aragonès,
Michelle L. Coote, Ismael Díez-Pérez, Simone Ciampi, and Nadim Darwish. "Metal–
single-molecule–semiconductor junctions formed by a radical reaction bridging gold
and silicon electrodes." Journal of the American Chemical Society 141, 14788-14797
(2019).
16. Hines, Thomas, Ismael Díez-Pérez, Hisao Nakamura, Tomomi Shimazaki, Yoshihiro
Asai, and Nongjian Tao. "Controlling formation of single-molecule junctions by
electrochemical reduction of diazonium terminal groups." Journal of the American
Chemical Society 135, 3319-3322 (2013).
17. Huang, Xiaoyan, Chun Tang, Jieqiong Li, Li-Chuan Chen, Jueting Zheng, Pei Zhang,
Jiabo Le et al. "Electric field–induced selective catalysis of single-molecule
reaction." Science Advances 5, eaaw3072 (2019).
18. Lin, Junfeng, Yaxin Lv, Kai Song, Xuwei Song, Hongjun Zang, Pingwu Du, Yaping
Zang, and Daoben Zhu. "Cleavage of non-polar C (sp 2)‒C (sp 2) bonds in
cycloparaphenylenes via electric field-catalyzed electrophilic aromatic
substitution." Nature Communications 14, 293 (2023).
19. Martín, Santiago, Iain Grace, Martin R. Bryce, Changsheng Wang, Rukkiat Jitchati,
Andrei S. Batsanov, Simon J. Higgins, Colin J. Lambert, and Richard J. Nichols.
"Identifying diversity in nanoscale electrical break junctions." Journal of the American
Chemical Society 132, 9157-9164 (2010).
20. Frisenda, Riccardo, Vera AEC Janssen, Ferdinand C. Grozema, Herre SJ Van Der
Zant, and Nicolas Renaud. "Mechanically controlled quantum interference in individual
π-stacked dimers." Nature chemistry 8, 1099-1104 (2016).
21. Wei, Yujing, Liang Li, Julia E. Greenwald, and Latha Venkataraman. "Voltage-
modulated van der Waals interaction in single-molecule junctions." Nano Letters 23,
567-572 (2023).
22. Yu, Hao, Jialing Li, Songsong Li, Yun Liu, Nicholas E. Jackson, Jeffrey S. Moore, and
Charles M. Schroeder. "Efficient intermolecular charge transport in π-stacked
pyridinium dimers using cucurbit [8] uril supramolecular complexes." Journal of the
American Chemical Society 144, 3162-3173 (2022).
23. Coropceanu, Veaceslav, Jérôme Cornil, Demetrio A. da Silva Filho, Yoann Olivier,
Robert Silbey, and Jean-Luc Brédas. "Charge transport in organic
semiconductors." Chemical reviews 107, 926-952 (2007).
24. Yi, Yuanping, Veaceslav Coropceanu, and Jean-Luc Brédas. "A comparative
theoretical study of exciton-dissociation and charge-recombination processes in
https://doi.org/10.26434/chemrxiv-2024-g4q8t ORCID: https://orcid.org/0000-0001-7656-8696 Content not peer-reviewed by ChemRxiv. License: CC BY 4.0
19
oligothiophene/fullerene and oligothiophene/perylenediimide complexes for organic
solar cells." Journal of Materials Chemistry 21, 1479-1486 (2011).
25. Solomon, Gemma C., Carmen Herrmann, Josh Vura-Weis, Michael R. Wasielewski,
and Mark A. Ratner. "The chameleonic nature of electron transport through π-stacked
systems." Journal of the American Chemical Society 132, 7887-7889 (2010).
26. Delgado, M. Carmen Ruiz, Eung-Gun Kim, Demétrio A. da Silva Filho, and Jean-Luc
Bredas. "Tuning the charge-transport parameters of perylene diimide single crystals
via end and/or core functionalization: a density functional theory
investigation." Journal of the American Chemical Society 132, 3375-3387 (2010).
27. Pan, Xiaoyun, Enrique Montes, Wudmir Y. Rojas, Brent Lawson, Héctor Vázquez, and
Maria Kamenetska. "Cooperative Self-Assembly of Dimer Junctions Driven by π
Stacking Leads to Conductance Enhancement." Nano Letters 23, 6937-6943 (2023).
28. Tang, Yongxiang, Yu Zhou, Dahai Zhou, Yaorong Chen, Zongyuan Xiao, Jia Shi,
Junyang Liu, and Wenjing Hong. "Electric field-induced assembly in single-stacking
terphenyl junctions." Journal of the American Chemical Society 142, 19101-19109
(2020).
29. Wang, Wesley, Nicholas H. Angello, Daniel J. Blair, Theodore Tyrikos-Ergas, William
H. Krueger, Kameron NS Medine, Antonio J. LaPorte, Joshua M. Berger, and Martin
D. Burke. "Rapid automated iterative small-molecule synthesis." Nature Synthesis, 1-
8 (2024).
30. Li, Songsong, Hao Yu, Kenneth Schwieter, Kejia Chen, Bo Li, Yun Liu, Jeffrey S.
Moore, and Charles M. Schroeder. "Charge transport and quantum interference
effects in oxazole-terminated conjugated oligomers." Journal of the American
Chemical Society 141, 16079-16084 (2019).
31. Li, Bo, Hao Yu, Elena C. Montoto, Yun Liu, Songsong Li, Kenneth Schwieter, Joaquín
Rodríguez-López, Jeffrey S. Moore, and Charles M. Schroeder. "Intrachain charge
transport through conjugated donor–acceptor oligomers." ACS Applied Electronic
Materials 1, 7-12 (2018).
32. Ciampi, Simone, Nadim Darwish, Heather M. Aitken, Ismael Díez-Pérez, and Michelle
L. Coote. "Harnessing electrostatic catalysis in single molecule, electrochemical and
chemical systems: a rapidly growing experimental tool box." Chemical Society
Reviews 47, 5146-5164 (2018).
33. Li, Songsong, Edward R. Jira, Nicholas H. Angello, Jialing Li, Hao Yu, Jeffrey S.
Moore, Ying Diao, Martin D. Burke, and Charles M. Schroeder. "Using automated
synthesis to understand the role of side chains on molecular charge transport." Nature
communications 13, 2102 (2022).
34. Gillis, Eric P., and Martin D. Burke. "A simple and modular strategy for small molecule
synthesis: iterative Suzuki− Miyaura coupling of B-protected haloboronic acid building
blocks." Journal of the American Chemical Society 129, 6716-6717 (2007).
35. Ryan, Seán TJ, Ryan M. Young, James J. Henkelis, Nema Hafezi, Nicolaas A.
Vermeulen, Andreas Hennig, Edward J. Dale et al. "Energy and electron transfer
dynamics within a series of perylene diimide/cyclophane systems." Journal of the
American Chemical Society 137, 15299-15307 (2015).
https://doi.org/10.26434/chemrxiv-2024-g4q8t ORCID: https://orcid.org/0000-0001-7656-8696 Content not peer-reviewed by ChemRxiv. License: CC BY 4.0
20
36. Angello, Nicholas H., David M. Friday, Changhyun Hwang, Seungjoo Yi, Austin H.
Cheng, Tiara C. Torres-Flores, Edward R. Jira et al. "Closed-loop transfer enables
artificial intelligence to yield chemical knowledge." Nature (2024), 1-8.
37. Wakamatsu, Satoshi, Uichi Akiba, and Masamichi Fujihira. "Electronic tunneling
through a single molecule embedded in self-assembled monolayer matrices." Colloids
and Surfaces A: Physicochemical and Engineering Aspects 198, 785-790 (2002).
38. Duan, Lili, and Simon J. Garrett. "An investigation of rigid p-methylterphenyl thiol self-
assembled monolayers on Au (111) using reflection− absorption infrared spectroscopy
and scanning tunneling microscopy." The Journal of Physical Chemistry B 105, 9812-
9816 (2001).
39. Heimel, Georg, Dieter Somitsch, Peter Knoll, and Egbert Zojer. "Ab initio study of
vibrational anharmonic coupling effects in oligo (para-phenylenes)." The Journal of
chemical physics 116, 10921-10931 (2002).
40. Maya, F., & Tour, J. M. Synthesis of terphenyl oligomers as molecular electronic device
candidates. Tetrahedron 60, 81-92 (2004).
41. Feng, Anni, Yu Zhou, Mohammed AY Al-Shebami, Lichuan Chen, Zhichao Pan, Wei
Xu, Shiqiang Zhao et al. "σ–σ Stacked supramolecular junctions." Nature
Chemistry 14, 1158-1164 (2022).
42. Samajdar, Rajarshi, Moeen Meigooni, Hao Yang, Jialing Li, Xiaolin Liu, Nicholas E.
Jackson, Martín A. Mosquera, Emad Tajkhorshid, and Charles M. Schroeder.
"Secondary structure determines electron transport in peptides." Proceedings of the
National Academy of Sciences 121, e2403324121 (2024).
43. Liu, Xiaolin, Hao Yang, Hassan Harb, Rajarshi Samajdar, Toby J. Woods, Oliver Lin,
Qian Chen et al. "Shape-persistent ladder molecules exhibit nanogap-independent
conductance in single-molecule junctions." Nature Chemistry, 1-9 (2024).
44. Adak, Olgun, Ethan Rosenthal, Jeffery Meisner, Erick F. Andrade, Abhay N.
Pasupathy, Colin Nuckolls, Mark S. Hybertsen, and Latha Venkataraman. "Flicker
noise as a probe of electronic interaction at metal–single molecule interfaces." Nano
letters 15, 4143-4149 (2015).
45. Fu, Tianren, Shanelle Smith, María Camarasa-Gómez, Xiaofang Yu, Jiayi Xue, Colin
Nuckolls, Ferdinand Evers, Latha Venkataraman, and Sujun Wei. "Enhanced coupling
through π-stacking in imidazole-based molecular junctions." Chemical Science 10,
9998-10002 (2019).
46. Adak, Olgun, Ethan Rosenthal, Jeffery Meisner, Erick F. Andrade, Abhay N.
Pasupathy, Colin Nuckolls, Mark S. Hybertsen, and Latha Venkataraman. "Flicker
noise as a probe of electronic interaction at metal–single molecule interfaces." Nano
letters 15, 4143-4149 (2015).
47. Tan, Ying, Jialing Li, Songsong Li, Hao Yang, Teng Chi, Stephen B. Shiring, Kangying
Liu, Brett M. Savoie, Bryan W. Boudouris, and Charles M. Schroeder. "Enhanced
electron transport in nonconjugated radical oligomers occurs by tunneling." Nano
letters 23, 5951-5958 (2023).
48. Bard, Allen J., and Larry R. Faulkner. "Fundamentals and Applications, New York:
Wiley, 2001." 1364-1365 (2002).
https://doi.org/10.26434/chemrxiv-2024-g4q8t ORCID: https://orcid.org/0000-0001-7656-8696 Content not peer-reviewed by ChemRxiv. License: CC BY 4.0
21
49. Li, Jialing, Sanja Pudar, Hao Yu, Songsong Li, Jeffrey S. Moore, Joaquín Rodríguez-
López, Nicholas E. Jackson, and Charles M. Schroeder. "Reversible switching of
molecular conductance in viologens is controlled by the electrochemical
environment." The Journal of Physical Chemistry C 125, 21862-21872 (2021).
50. Kumar, Shankar, et al. "The weighted histogram analysis method for free‐energy
calculations on biomolecules. I. The method." Journal of computational chemistry 13,
1011-1021 (1992).
https://doi.org/10.26434/chemrxiv-2024-g4q8t ORCID: https://orcid.org/0000-0001-7656-8696 Content not peer-reviewed by ChemRxiv. License: CC BY 4.0
22
Acknowledgements
This work was supported by the U.S. Department of Energy, Office of Science, Basic
Energy Sciences under Award No. DE-SC0022035 for H.Y., X.L., M.M., J.S.M., E.T., and
C.M.S., the National Science Foundation under Award 2227399 and the Army Research
Office under Cooperative Agreement Number W911NF-22-2-0246 for R.S. and C.M.S.
The views and conclusions contained in this document are those of the authors and
should not be interpreted as representing the official policies, either expressed or implied,
of the Army Research Office or the U.S. Government. The U.S. Government is authorized
to reproduce and distribute reprints for Government purposes notwithstanding any
copyright notation herein. This work was supported by the Molecule Maker Lab Institute,
an AI Research Institutes program supported by the US National Science Foundation
under grant no. 2019897 for S.Y. We gratefully acknowledge Furong Sun and the UIUC
Mass Spectrometry Lab. S.P. is grateful for support from the National Science Foundation
Graduate Research Fellowship. M.P. is grateful for the support of Beckman Institute for
Advanced Science and Technology Graduate Fellows Program with support from the
Arnold and Mabel Beckman Foundation.
Author contributions
R.S., H.Y., S.Y., and C.M.S. conceived this study. R.S. and H.Y. performed STM-BJ
experiments, data analysis, and unsupervised machine learning. S.Y. performed chemical
synthesis and characterization. C.W performed MD simulations. M.P. and S.P. carried out
bulk scale electrochemistry, electrolysis, and spectroscopy experiments. R.S. performed
the NEGF-DFT calculations. M.M., X.L., J.R. assisted in experiments and simulations.
The manuscript was written by R.S., H.Y. and C.M.S. with contribution from all authors.
Competing Interests
The authors declare no competing interests.
Additional Information
Supplementary information contains supplementary figures, supplementary tables, and
supplementary text.
https://doi.org/10.26434/chemrxiv-2024-g4q8t ORCID: https://orcid.org/0000-0001-7656-8696 Content not peer-reviewed by ChemRxiv. License: CC BY 4.0