Two-Dimensional versus Three-Dimensional Self-Assembly of a Series of 5-Alkoxyisophthalic Acids

Article abstract of DOI:10.1021/acs.langmuir.8b01827

Physisorbed self-assembled monolayers (SAMs) have been suggested as potential models for three-dimensional (3D) crystallization. This work studies the effect of altering the chain length of 5-alkoxyisophthalic acid (C_nISA) on self-assembled mor

Full text of DOI:10.1021/acs.langmuir.8b01827

Subscriber access provided by Kaohsiung Medical University
Interface-Rich Materials and Assemblies
2D versus 3D Self-Assembly of a Series of 5-Alkoxyisophthalic Acids
Esther M. Frederick, José David Cojal González, Jürgen P. Rabe, and Steven L. Bernasek
Langmuir, Just Accepted Manuscript • DOI: 10.1021/acs.langmuir.8b01827 • Publication Date (Web): 15 Aug 2018
Downloaded from http://pubs.acs.org on August 17, 2018
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a service to the research community to expedite the dissemination
of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in
full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully
peer reviewed, but should not be considered the official version of record. They are citable by the
Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore,
the “Just Accepted” Web site may not include all articles that will be published in the journal. After
a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web
site and published as an ASAP article. Note that technical editing may introduce minor changes
to the manuscript text and/or graphics which could affect content, and all legal disclaimers and
ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or
consequences arising from the use of information contained in these “Just Accepted” manuscripts.
is published by the American Chemical Society. 1155 Sixteenth Street N.W.,
Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 10
Langmuir
1
2
3
4
5
6
7
8
9
2D versus 3D SelfꢀAssembly of a Series of 5ꢀAlkoxyisophthalic Acids
Esther Frederick1, José D. Cojal González2, Jürgen P. Rabe2 and Steven L. Bernasek*1,3
AUTHOR ADDRESS 1 Department of Chemistry, Princeton University, Princeton, NJ 08544, USA
2 Department of Physics & IRIS Adlershof, HumboldtꢀUniversität zu Berlin, Dꢀ12489 Berlin, Germany
3 Science Division, YaleꢀNUS College, 138527 Singapore
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
KEYWORDS Self-assembly, solvent effects, crystallization, monolayers, scanning tunneling microscopy, x-ray diffraction,
hydrogen bonding, van der Waals forces, physisorption.
ABSTRACT: Physisorbed selfꢀassembled monolayers (SAMs) have been suggested as potential models for 3D crystallization.
This work studies the effect of altering the chain length of 5ꢀalkoxyisophthalic acid (CnISA) on selfꢀassembled morphology in both
2D and 3D in order to explore the extent comparisons can be drawn between dimensions. Previous studies of 5ꢀalkoxyisophthalic
acid at solidꢀliquid interfaces (2D) reported different morphologies for C5ISA and C6ISAꢀalkoxy chains on the one hand and
C10ISA and C18ISA on the other. Independently, also in 3D a dependence of morphology on chain length has been reported, includꢀ
ing an unexpected inclusion of solvent in the 3D morphology of C6ISA, while the previous reports of 2D selfꢀassembly were driven
only by moleculeꢀmolecule and moleculeꢀsubstrate interactions. However, a complete set of data for comparison has been missing.
Here we report scanning tunneling microscopy (STM) and molecular dynamics (MD) simulations performed for C2ISA selfꢀ
assembled monolayers (SAMs) and STM imaging of C6ISA−C9ISA SAMs, to further examine selfꢀassembly behavior in 2D. In
3D, XRD analysis of C2ISA single crystals was carried out to complete the data set. With a complete set of data, it was observed
that regardless of dimension, short chain length CnISAs formed Hꢀbonding dominated structures, midꢀchain length CnISAs exhibitꢀ
ed solvent dependent morphologies, and long chain length CnISAs displayed vdW dominated solventꢀindependent structures. Howꢀ
ever, the transition point between morphologies occurred at different chain lengths in 2D and 3D regardless of the dominant interꢀ
action. The results of this study inform the design of 2D films and guide the application of knowledge from physisorbed SAMs to
3D systems, including mixedꢀdimensional (2D/3D) van der Waals heterostructures.
Selfꢀassembly is a process by which unorganized molecules
spontaneously organize into selfꢀdetermined arrangements. In
nature, selfꢀassembly plays a key role in the formation of
proteins, cells and advanced biological structures. At the naꢀ
noscale, selfꢀassembly allows molecules that lack coherence in
solution to gain the stability and uniformity needed to operate
as functional nanomaterials when adsorbed on a surface.1 As
the lower size limits of topꢀdown architectures of nanoꢀ
structures are approached, a deeper understanding of how
molecules can be programmed to selfꢀassemble into desired
architectures becomes increasingly important to the future of
nanotechnology.
STM is based on the principle of quantum tunneling, where a
tip brought within tunneling distance of the sample is scanned
across the surface to provide an electron density map that can
be carefully interpreted as surface morphology. The nonꢀ
destructive nature of STM, makes it the best technique to
investigate the morphologies of physisorbed SAMs.
2D selfꢀorganized systems have been assumed to be good
models for 3D crystallization as the confinement of molecules
close to interfaces allows for experimental study of intermoꢀ
lecular forces.6 Extensive reviews on 2D selfꢀassembly are
available covering a range of topics including solvent effects,7
electronic properties,8 chirality/oddꢀeven effects,9 surfaceꢀ
confined supramolecular coordination chemistry,10 substrate
effects,11 molecular templating,1 and control of structure.12 In
an attempt to gain better insight into principles guiding forꢀ
mation of physisorbed SAMs, Plass, Grzesiak and Matzger
attempted to compile a comprehensive database of all strucꢀ
tures as determined by STM studies.13 The present study exꢀ
plores the validity of the assumed connection between 2D
SAM structures and 3D crystals by directly comparing the
selfꢀassembly of a series of 5ꢀalkoxyisophthalic acids at the
solidꢀliquid interface with crystallization from solvent in 3D.
Surface modification using selfꢀassembled monolayers
(SAMs), i.e. well organized single molecule thick layers that
spontaneously form at a solidꢀliquid interface under the approꢀ
priate conditions, is a broad and active field of research. Highꢀ
ly complex structures can be designed by careful molecule
design.2,3 In this study, physisorbed SAMs of a series of 5ꢀ
alkoxyisophthalic acids (CnISAs) were used to probe the relaꢀ
tionship of varying hydrogen bonding (Hꢀbonding) versus van
der Waals (vdW) interactions in 2D and 3D selfꢀassembly.
The invention of scanning tunneling microscopy (STM) by
Binnig and Rohrer in 1982,4 enabled an unprecedented level of
microscopic imaging at the nanoscale, making it possible to
visualize selfꢀassembled structures at solidꢀliquid interfaces.5
Prior investigations of CnISA crystallization in both 2 and 3
dimensions provided a preliminary, but incomplete set of data
for comparison of 2D to 3D morphologies. XRD studies by
1
ACS Paragon Plus Environment

Langmuir
Page 2 of 10
Enkelmann, Valiyaveettil, Möessner and Müllen of CnISA tems, specifically testing for similarity in morphology and
crystal lattices14 showed that 3D structure is strongly impacted
by chain length and solvent effects. Molecules with chain
lengths between 6 and 10 carbons formed solvent dependent
crystal structures while those with longer chain length incorꢀ
porate van der Waals interaction to adopt a lamellar architecꢀ
ture. In THF (tetrahydrofuran), a nonꢀhydrogen bonding solꢀ
vent, molecules with chain lengths between 6 and 10 assemꢀ
bled into Hꢀbonded hexamer channels with alkyl chain walls.
However, when methanol (MeOH) was used as a solvent, the
MeOH coꢀadsorbed with the CnISA head groups to form VdW
ribbons with ordered alkyl chains. Longer CnISAs assembled
into ribbons via CnISA chainꢀchain interactions and did not coꢀ
adsorb with either solvent. Solvent and chain length effects
were also observed for the selfꢀassembled 3D structures of 5ꢀ
alkoxyisophthalic acids with first row transitions metals.15
whether the chain length of the transition point remains the
same in 2D as in 3D. While some of the data is present in the
literature, further experiments were needed to complete the
data set required to examine the effect of dimension on selfꢀ
assembly. The available literature data and experimental list
for this study are organized in Table 1.
1
2
3
4
5
6
7
8
Table 1: Literature data (in black) and experiments unique
to this study (in red) for multi-phase 5-alkoxyisophthalic
acid self-assembly.
9
CnISA
2D results
3D results
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
molecule
Short
C2ISA
C2ISA
Chain
Studies by Dickerson et al. of CnISA SAMs16 demonstrated
a similar phenomenon at the solidꢀliquid interface, where
shorter chain length SAMs formed a solely hydrogen bonded
network morphology (C6 and below) while longer chain length
SAMs (C10 and up) formed a lamellar structure with hydrogen
bonding between head groups and VdW interaction between
chains. This prior study did not experimentally explore the
morphology of SAMs with chain lengths 6−10 to determine
the chain length of the transition point. Studying the transition
between SAM morphologies is a critical component for underꢀ
standing the thermodynamics of selfꢀassembly.17 The current
study determined the exact chain length of the transition point
from hexagonal to lamellar morphology to help determine how
molecular interactions affect the architecture, as well as obtain
a more complete picture of the relationship between the 2D
and 3D selfꢀassembly.
Mid Chain
C5ISA, C6ISA:16
Solventꢀdependent
Hꢀbonded network structures:14
C7ISA,C8ISA,
Hꢀbonded network
in THF,
C9ISA
Hꢀbond
guided
lamellar in MeOH
C12ꢀlongerISA:14
Long
Chain
Lengths
C10ISA, C18ISA:16
vdW dominated
lamella
vdW
dominated
lamella
In the present study, the crystal structure of C2ISA has been
obtained and the morphology of a series of C2ISA,
C6ISA−C9ISA SAMs has been determined to supplement
literature results from XRD studies of 5ꢀalkoxyisophthalic
acid crystals14 and STM studies of CnISA SAMs.16 A complete
comparison of short chain length (C2ISA), mid chain length
(C6ISA−C10ISA), and long chain length (C18ISA) ISAs is
provided.
The comparison of selfꢀassembly in multiple phases has litꢀ
erary precedence. A study of C16ISA compared the selfꢀ
assembly of CnISA both in the crystal lattice and at a highly
oriented pyrolytic graphite (HOPG) interface.18 In both the 2D
and the 3D structures, the pure C16ISA formed lamellar strucꢀ
tures of packed interdigitating acids. The HꢀBonding strucꢀ
tures differed due to larger degrees of freedom in 3D, which
allowed the Hꢀbonding heads and alkyl chains to pack in sepaꢀ
rate planes. In 2D, by contrast, the two are confined to the
same plane. C16ISA and C12ISA were further explored in both
EXPERIMENTAL
Synthesis: CnISAs were synthesized according to the scheꢀ
matic below (Figure 1). Sn2 substitution with the appropriate
iodoalkane
was
used
to
convert
dimethyl
5ꢀ
hydroxyisophthalate to dimethyl 5ꢀalkoxyisophthalate. 6 mmol
of the respective alkyl iodide (ThermoFisher Scientific, 98%)
and 6 mmol of K2CO3 dissolved in 7.5 mL acetone were
placed in a three neck round bottom flask under an argon
environment. 1.05 g of dimethyl 5ꢀhydroxyisophthalate (Sigꢀ
maꢀAldrich, Germany, 98%) was dissolved in 7.5 mL acetone
and added via syringe to the flask. Reactants were refluxed at
75° for 24−48 hours. To monitor the reaction progress, aliꢀ
quots were removed, rotovapped and analyzed with mass
spectroscopy. When the reaction was complete, the product
was rotovapped and purified by silica flash column chromaꢀ
tography using a 10:90−20:80 acetone/hexanes eluent. To
form CnISA, the product was further reacted with KOH. The
reaction was monitored as previously with mass spectroscopy.
The resulting CnISA was purified by separatory funnel extracꢀ
tion. The CnISA was dissolved in 10 mL water, acidified using
1 M HCl and then extracted into diethyl ether. The product
was dried using Na2SO4 for 2 hours, then rotovapped to reꢀ
move remaining solvent and placed on high vacuum overꢀ
night. Resulting fine white powder was analyzed with 1H
NMR and 13C NMR (Bruker 500 AVANCE spectrometer) as
dimensions
coꢀadsorbed
with
pyrazine
and
2.5ꢀ
dimethylpyrazine.19 Additionally, when developing the Twoꢀ
Dimensional Structural Database (2DSD), which compiles
physisorbed SAM structures characterized using STM, Plass
et. al.13 noted similarities between the factors governing 2D
and bulk structures though the actual packing differed. Other
studies have focused on molecules such as polymers,20 organꢀ
ometallic complexes21, discotic liquid crystals22 and biological
materials, such as guanosine,23 to compare the structures in 2D
and 3D, however to the best of our knowledge systematic
studies, comparing the effects of varying the chain length of a
substituent on morphology in 2D versus 3D have not been
done. Continued contributions to the 2DSD with systematic
experimental and theoretical studies are essential for the deꢀ
velopment of predictive models for 2D selfꢀassembly.13,24
This work presents a systematic study comparing multiꢀ
phase assembly of CnISA with a series of alkoxy chain lengths
to analyze the extent 2D SAMs can act as models for 3D sysꢀ
2
ACS Paragon Plus Environment

Page 3 of 10
Langmuir
lations. A distance cutoff of the size of the periodic box was
presented in the supplemental information (See SI 1). Recrysꢀ
1
2
3
4
5
6
7
8
tallization of C2ISA for XRD studies was done according to
used to compute nonꢀbonded interactions. Electrostatic interꢀ
actions were calculated using the Particle mesh Ewald (PME)
method with a space grid of 0.5 Å. As in previous work, a
graphene slab was constructed as a substrate to simulate moꢀ
lecular adsorption on HOPG.31 In all simulations, the substrate
was fixed and its atoms were assigned atom type CG2R61
with no partial charge. Parameters for the C2ISA molecule
were assigned by CGenff analogy. 20 ns MD simulations were
performed for different combinations of molecules per cluster,
including both C2ISA and 1ꢀphenyloctane. A snapshot of the
simulation was taken every 20 ps and rendered according to
the contribution of each atom to the tunneling current.32 STM
simulated MD trajectory images were calculated by averaging
a minimum of 1000 frames.
literature procedures using THF.14
Figure 1: Synthetic scheme for 5ꢀalkoxyisophthalic acid.
9
STM Imaging: Obtaining an ordered structure at the solidꢀ
liquid interface can be a difficult endeavor requiring skill and
experience. Ordered monolayers formed from molecules with
a strong preference for ordering such as the C9ISA were more
easily obtainable relative to others, such as the C2ISA, which
required many attempts. In brief, several ∼1 mM solutions of
the desired CnISA in phenyloctane (Aldrich, 98%) were preꢀ
pared using sequential dilutions sonicated at 60°C for 1 hour.
Monolayer formation was attempted for each slightly varying
concentration by placing a few drops of the respective solution
onto freshly cleaved HOPG (NanoScience Instruments, Reꢀ
search Grade). Samples were placed in the freezer for 1 hour
to allow for monolayer formation. Freezing was not necessary
for monolayer formation, but produced clearer images than
assemblies formed at roomꢀtemperature. After removing from
the freezer, samples sat at room temperature for at least one
hour to equilibrate to ambient conditions before imaging.
Images were obtained with freshly cut Pt/Ir tips (NanoScience
Instruments, 80/20,0.25mm diameter). Reproducible images
from several different samples were obtained for each CnISA.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
RESULTS
STM Results: C2ISA: C2ISA formed SAMs with bright arꢀ
eas arranged in a highly hexagonal pattern (Figure 2) with
long range order (Figure SI 2.1). The average unit cell, incorꢀ
porating both dark and bright areas (as depicted in Figure 2a),
was measured as a= 2.22 ± 0.17 nm, b= 2.14 ± 0.24 nm,
∠
=
a,b
61 ± 3°. The bright spots, corresponding to regions of higher
electron density due to adsorbed molecules, did not reveal
highly resolved molecules. Computational methods (described
in SI 2 and SI 3) were undertaken to gain insight into the unit
cells. A statistical particle analysis (Figure SI 2.2) measured
an average size of 1.5 ± 0.1 nm for the regions of higher elecꢀ
tron density, which is 50% more than the maximum length of
C2ISA (~1 nm). Due to this, each bright spot was assumed to
be a cluster of several molecules, either C2ISA or a coꢀ
adsorption of C2ISA with 1ꢀphenyloctane.
For all except C2ISA, images were obtained using an ambiꢀ
ent Nanosurf AG EasyScan STM system (NanoScience Inꢀ
struments, Liestal, Switzerland) in constant current mode with
various tunneling conditions (Vb = 0.4−0.8 V, It = 0.7−0.9
nA). As this instrument places bias on the tip rather than the
sample, a positive bias was used so electrons tunneled from
the sample to the tip. In the case of C2ISA, a homeꢀbuilt ambiꢀ
ent STM with an Omicron controller was used to obtain imagꢀ
es, with a negative bias voltage placed on the sample (VS) to
ensure electrons tunneled from sample to tip. Height and curꢀ
rent measurements were collected for all scans. Post imaging
analyses were done using SPIP (Image Metrology A/S). Fouꢀ
rier Transform (FT) analyses of the current scan were used to
calibrate images with underlying HOPG. FT analyses of the
corrected height image were used to determine SAM unit cell
parameters. Long range scans and C8ISA images were not
calibrated as the underlying HOPG was not wellꢀenough reꢀ
solved in these images to provide a meaningful calibration.
Proposed structures were modeled with Avogadro.25,26 Monoꢀ
layer formation was confirmed by measuring line profiles over
the edge of the ordered domain onto the bare HOPG surface.
Trimerꢀsolvent clusters were identified as the best candiꢀ
dates for initial MD simulations (Figure 2b). DFT calculations
at the B3LYP/6ꢀ31g(d,p) level showed trimer formation lead
to a gain of 105 kJ/mol, which equals 35 kJ/mol per molecule.
This result is in good agreement with the published value of
60−66 kJ/mol, obtained by measuring the enthalpy of hydroꢀ
gen bonding in carboxylic acid dimers.33 Structures formed by
less than three molecules were not large enough to fit the
measured STM data. Clusters of four molecules were energetꢀ
ically less favorable (~25 kcal/mol per unit cell from MD
simulation analysis) than three molecule clusters. Preliminary
models for clusters of five and six molecules displayed closeꢀ
packing which constrained the molecules to the initial conforꢀ
mation, and therefore were not further considered.
XRD: Crystals were formed by placing saturated THF soluꢀ
tions of the desired CnISA in the freezer for several weeks.
Structures were obtained using Bruker D8 Venture and anaꢀ
lyzed using the SQUEEZE method in the program PLATON
which compensated27 for the scattering from the disordered
solvent. Images of the structures were created using Vesta.28
Figure 2: (a) 20 × 20 nm STM height image of C2ISA in 1ꢀ
phenyloctane assembled on HOPG. White arrows show the orienꢀ
tation of the graphite underneath (VS = ꢀ0.30 V, It = 0.40 nA). (b)
C2ISA can arrange in 4 different inꢀplane trimers. The orange
circle represents the average size of the bright spots
MD: Molecular dynamics simulations were performed usꢀ
ing NAMD29 (2.10 build for Linux) with CHARMM general
force field (CGenff).30 A Langevin thermostat with a 1 psꢀ1
damping coefficient and a 2 fs time step was used in all simuꢀ
3
ACS Paragon Plus Environment

Langmuir
A tunneling current simulation method, previously shown to
Page 4 of 10
C9ISA images had the best resolution and will be used to
1
2
3
4
5
6
7
8
accurately reproduce STM images of soft supramolecular
networks,32 was used to account for transitory nonꢀtrimer
intermediate structures formed during the simulation. For the
starting configuration, four trimers and eight solvent moleꢀ
cules (filling the gaps in between trimers) were placed in a
hexagonal box with periodic boundary conditions to reproduce
an infinite system. The hexagonal box was developed using
the unit cell parameters of Figure 2a, leading to a periodic box
of four unit cells as displayed in Figure 3b. The xyz positions
of all components were recorded every 20 ps in snapshot files,
with every simulation running for 20 ns. The MD tunneling
current simulation (Figure 3b) was then produced by averagꢀ
ing the 1000 snapshot files collected throughout the simulaꢀ
tion. Eight different 20 ns MD simulations were performed to
take into account all possible chirality combinations four
C2ISAs can form when adsorbed parallel to the substrate (Figꢀ
ure 2b). The resulting images were then averaged to produce a
MD simulation (Figure 3a) representative of all possible trimer
chirality combinations. The average of these MD simulations
was the closest fit to experimental STM image, with a simuꢀ
lated unit cell of a = 2.21 ± 0.10 nm, b = 2.21 ± 0.10 nm,
describe the SAM formation results at these chain lengths.
C9ISA formed highly ordered SAMs with lamellar patterns of
alternating dark and light stripes. As mentioned in the experiꢀ
mental section, light stripes were attributed to regions of
greater electron density, and as such represent the aromatic
head groups which have greater electron density as compared
to alkyl chains, while dark stripes were attributed to the alkyl
chains. After calibration to HOPG, the FFT of the height imꢀ
age displayed two sets of bright spots which shared one comꢀ
mon spot (red) representing two related parallelogram cell
structures (see Figure 5). In both unit cells; a = the shorter side
running along the bright stripe, b = the side spanning the disꢀ
tance between the bright lines and ∠a,b = the angle between the
two sides. The cyanꢀred set corresponded to the larger unit
cell spanning the distance between stripes, with unit cell valꢀ
ues of a= 0.6 ± 0.3 nm, b= 3.3 ± 0.3 nm,
greenꢀred set corresponded to the unit cell spanning halfway
between with a change to the b parameter only, with b = 1.3 ±
0.3 nm. The large error in the a parameter value demonstrates
that the spacing along the bright stripes cannot be confidently
assigned. The resolution of the images only allows for estabꢀ
lishing the b parameter value.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
∠
= 83 ± 4°. The
a,b
∠
a,b = 59 ± 1°. This result suggests that the hexagonal experiꢀ
mental pattern is due to conformational freedom in the arꢀ
rangement of energetically favorable and hydrogenꢀbond
driven trimers. See SI 3 for a more detailed breakdown of the
MD simulations.
Figure 5: C9ISA unit cell and FFT. (a) 15 x 15 nm C9ISA height
image after calibration to HOPG with inset FFT. The FFT disꢀ
played two sets of bright spots, which shared one common spot
(red) (Vb = 0.76 V, It = 0.77 nA). The white arrows indicate symꢀ
metry directions of the underlying HOPG. (b) ~60 x 60 nm STM
height image of C9ISA. The 2D hexagonal network displayed
long range order, further demonstrated in the inset FFT (Vb = 0.66
V, It = 0.80 nA).
Figure 3: (a) 20 × 20 nm simulated STM measurement calculated
by averaging the MD trajectories of the 8 potential trimer combiꢀ
nations. (b) Render of the tunneling current for a 5×5 periodic
simulation box (10 × 10 unit cells) after 20 ns of MD simulation
at 300K. One periodic box (2 × 2 unit cells) is greyed with the
molecular structure imposed. The hydrogen bonds in the models
go beyond the periodic box.
The first suggested structure involved interdigitated adjaꢀ
cent C9ISA molecules with an Hꢀbonded backbone as in preꢀ
vious work. This model yielded lamellae of 2.1 nm, which was
1.3 nm too small to fit the measurements. The proposed model
that best fit this data consisted of a unit cell containing two
C9ISA groups laid tail to tail without alkyl chain interdigitaꢀ
tion and Hꢀbonding between adjacent head groups (see Figure
6). As this model was energetically unfavorable, incorporaꢀ
tion of phenyloctane solvent molecules was proposed as a
stabilizing factor. The exact orientation of the molecules could
not be determined. As the head groups were typically not
visible, it is most likely that the alkyl chains are lying flat on
the substrate with the head groups tilted upward from the
surface. Moreover, one of the symmetry axes of the underlyꢀ
ing HOPG is oriented about 5° with respect to the vector b.
This suggests that the alkyl chains lie along this symmetry
axis, which would also maximize the vdW interactions beꢀ
tween the alkyl chains and the substrate. The smaller structure
could then be explained as the repeating unit of a single pheꢀ
C7ISA – C9ISA: Long range ordered lamellar SAMs were
observed for C7ISA−C9ISA SAMs (Figure 4 and Figure 5b).
Figure 4: ~60 x 60 nm STM height images of (a) C7ISA (Vb =
0.48V, It = 1.0 nA) and ( b) C8ISA (Vb = 0.70V, It = 0.77 nA)
displayed long range ordered SAMs, further demonstrated in the
inset FFT.
4
ACS Paragon Plus Environment

Page 5 of 10
Langmuir
nyloctane and C9ISA molecule. The larger structure, containꢀ
ing two C9ISA molecules, was taken as the unit cell as its
1
2
3
replication produced a schematic that lined up well with exꢀ
perimental measurements.
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 7: 10 nm x 10ꢀnm C7ISA height image with unit cell (Vb =
0.72 V, It = 0.70 nA). The white arrows indicate the orientation of
the underlying HOPG.
C10ISA: Due to the surprising solvent incorporation in
C6ISA−C9ISA SAMs, C10ISA images from previously pubꢀ
lished work were reꢀexamined. A reꢀanalysis demonstrated bꢀ
values of 2.6 ± 0.3 nm nm which matches the alkoxyꢀalkoxy
C10ISA chain interdigitation model proposed previously by
Dickerson et al.16 Therefore, the transition point between
solvent dependent and solventꢀindependent morphologies was
found to be C10ISA.
Figure 6: 15 x 15 nm C9ISA STM image with (a) overlaid proꢀ
posed structural model and (b) structural model (Vb = 0.76 V, It =
0.76 nA). The arrows on the model correspond to the circled spots
in the FFT in Figure 5. The underlying HOPG lattice is indicated
in Figure 5a.
The hypothesis that a transition point would be found was
substantiated. However, the transition point was significantly
more complex than what might be expected based on intermoꢀ
lecular CnISA interactions and substrateꢀmolecule interactions
alone. The unexpected incorporation of solvent at the transiꢀ
tion point leads to a whole new set of questions to be explored
in future work. What is the dependence of this transition point
on the solvent? Can the transition point be pushed to higher
and lower chain lengths by changing the chain length of the
solvent? While solvent dependence on physisorbed SAM
morphology has been previously explored,6,34ꢀ39 the ability to
achieve solventꢀindependent SAM morphologies by altering
the chainꢀlength of a moiety on an integral SAM component is
an interesting discovery.
C7ISA and C8ISA also displayed lamellar SAMs. The disꢀ
tance between the bright lines in the C7ISA was measured as
3.0 ± 0.6 nm (Figure 7). These values were too large to fit the
intermolecular alkyl chain interdigitation model of 1.7 nm.
Also of note, in the C7ISA, several bright spots appeared halfꢀ
way between the lamellae (see Figure 7 and SI 4). These spots
could potentially be attributed to disorder of the incorporated
solvent molecules leading to random head groups serendipiꢀ
tously oriented in a manner that could be imaged. C8ISA imꢀ
ages were not of high enough resolution to confidently caliꢀ
brate to the underlying HOPG. Both the uncalibrated image
and attempted calibrations led to bꢀvalues of 3.1 ± 0.9 nm. The
lower end of the error 2.2 nm value is still higher than the
expected 1.9 nm. As the C8ISA falls between two solvent
incorporated lamellae, it is most likely that the C8ISA SAM
also involves phenyloctane incorporation.
XRD Results: XRD structure determinations showed
C2ISA formed P21/c space group crystals with unit cell parameꢀ
ters of a = 12.02 Å b = 12.46 Å c = 16.76 Å and angles α = γ =
90° and β = 107.77° (see SI for CIF file). Crystal structure
views along the a and c axes displayed four sheets of moleꢀ
cules layered one upon another without chemical connection
(Figure 8). The view along the cꢀaxis showed benzene ring
alignment for the middle two molecules, strongly suggesting
the layers were held together by π−π stacking forces. As preꢀ
viously observed in Enkelmann’s14,40 work, these structures
were dominated by Hꢀbonding and consisted of stacked layers
of tapeꢀlike sheets. The current studies provide unit cell measꢀ
urements and a detailed analysis not seen in previous work.
5
ACS Paragon Plus Environment

Langmuir
Page 6 of 10
and no involvement of chainꢀchain vdW interactions, which
would contribute 36−42 kJ/mol.
1
2
3
4
5
6
7
8
9
At midꢀchain lengths between C7ISA−C9ISA in SAMs and
C6ISA−C12ISA in 3D, solvent dependent morphologies were
observed. The surprising inclusion of solvent in SAMs of
C7−C9ISA indicates the formation of midꢀchain length SAMs
is more complex than had been expected based on previous
CnISA SAM work.16 The observation can be explained by the
following: SAM formation occurs when the gain in enthalpy
due to intermolecular and surfaceꢀmolecule interactions overꢀ
comes the otherwise unfavorable ordering. There are similar
enthalpy gains for a neat or mixed CnISA/phenyloctane SAMs
for C7ISA−C9ISA with a maximum difference of one vdW
interaction per molecule. The decrease in entropy that must be
overcome for SAM formation is less for a mixed monolayer
SAM which incorporates multiple species and therefore has
more variety than a neat SAM. It is also important to note that
mixed SAMs formed at chain lengths (n = 7−9) that matched
closely to the solvent chain length of 8 carbons. Therefore, the
mixed monolayer formation for C7ISA and C9ISA is attributed
to the lower entropic cost of formation combined with similar
enthalpy gains and optimal packing as in the neat SAMs. As
no neat SAMs were observed, it is likely that the enthalpy
gains are not high enough to overcome the entropic cost of
forming a neat monolayer.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 8: 3D crystal structure of C2ISA viewed along the a axis
(left) and cꢀaxis (right).
The view along the bꢀaxis (Figure 9) showed that each layer
consisted of two molecules per each unit cell. Within the layer,
molecules formed carboxylic acid dimers held together by
hydrogen bonding intermolecular interactions between head
groups. Adjacent dimers were further Hꢀbonded to neighborꢀ
ing dimers on both sides to form ribbons. The ribbons were
not connected to each other, which agreed with previous reꢀ
sults.
Above C10ISAs, both 2D and 3D assemblies exhibited laꢀ
mellar morphologies without solvent inclusion. At these chain
lengths, the forces governing the assembly are similar. Howꢀ
ever as previously noted by Enkelmann et al.,14 the details of
the Hꢀbonding arrangements are dimension dependent. As
before, the SAM morphology can be explained by the balance
between entropic cost of ordering and enthalpy gains from
intermolecular and surface interactions. The unfavorable orꢀ
dering of a neat monolayer is overcome by the increased vdW
interactions at chain lengths of 10 carbons and above. The
energetic difference between whether CnISA form a neat and
mixed SAM in phenyloctane on HOPG is on the order of 3
additional vdW interactions (C10ISA−C10ISA versus
C7ISA−C7ISA), which is at most 49 kJ/mol.
The goal of finding a quantitative transition point between
differing morphologies was achieved. However, the unexꢀ
pected inclusion of the solvent at the transition point meant
this transition point was more complicated than the expected
CnISA−CnISA interactions. Two transition points were obꢀ
served. The first was the transition from a Hꢀbonded C6ISA
SAM to a solventꢀdependent lamellar C7ISA SAM, while the
second was the transition from a solventꢀdependent lamellar
C10ISA SAM to a solventꢀindependent C11ISA SAM.
Figure 9: 3D crystal structure of C2ISA viewed along the bꢀaxis.
Top: bꢀaxis view of the full fourꢀlayered unit cell with hydrogens
omitted for clarity. Bottom: A single layer.
DISCUSSION
To summarize, while details of the structures were dimenꢀ
sionality dependent, at a given chain length, there was a strong
similarity for the trends in forces that governed selfꢀassembly
of CnISA regardless of dimension (see Table 2).
Interestingly, the transitions between morphologies ocꢀ
curred at different chain lengths in 2D SAMs than in 3D crysꢀ
tals. It is important to emphasize that the experimental condiꢀ
tions for 2D SAMs used phenyloctane as a solvent, while the
literature and experiment results for 3D crystals used MeOH
and/or THF. Therefore, it is not yet known if the difference in
the chain length of morphology transition is due to substrate
effects or due to solvent interactions. Experiments with the
same solvent at these chain lengths would be of use for further
quantifying these relationships and determining the extent to
which the substrate stabilized the interactions. Additionally,
flow micro calorimetry experiments and molecular mechanics
At chain lengths below C7ISA in SAMs and below C7 ISA
in 3D, assembly was driven by intermolecular Hꢀbonding
interactions without VdW contributions. Published values for
van der Waals chainꢀchain interactions are 6−7 kJ/mol per
carbon16,41 and enthalpy of hydrogen bonding in carboxylic
dimers determined from dimer dissociation energies are 60−66
kJ/mol.33 In SAMs, the observed morphology involved CnISA
molecules with 60−66 kJ/mol Hꢀbonding intermolecular interꢀ
actions plus unquantified substrate, solvent and entropy values
6
ACS Paragon Plus Environment

Page 7 of 10
Langmuir
generalizable model for surface assembly has thus far not been
simulations to quantify the enthalpy of adsorption as done by
Barnard and Matzger42 would be useful to further understand
the behavior of the system.
possible, the study of complex systems such as those in this
work provide the detailed information necessary for insight
into the patterns of selfꢀassembly across different systems. A
deep understanding of similarities in selfꢀassembly behavior is
necessary for developing a generalizable theoretical frameꢀ
work.
1
2
3
4
5
6
It is of great interest to further this work in collaboration
with theorists. Devising a model from this experimental data
that can be used to predict future structures could provide
insight into generalizable rules for selfꢀassembly. While a
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Table 2: Summary of literature and experimental structure results for multiphase CnISA self-assembly
CnISA molecule
2D results
3D results
Short Chain Lengths
HꢀBonding governed structures
Mid Chain Lengths
Solvent dependent structures
Long Chain Lengths
vdW dominated structures, no solvent effects
CONCLUSIONS
chain structures, starting at C10ISA in 2D and at C13 in 3D,
exhibited solvent independent lamellar morphologies.
In conclusion, 5ꢀalkoxyisopthalic acids were used as model
systems for comparing 2D and 3D selfꢀassembly. Similar
trends were observed for the forces governing selfꢀassembly,
however the chain lengths of morphological transitions and the
finer details of the assemblies differed. This work demonꢀ
strates that 2D systems may be used to inform the behavior of
forces in 3D selfꢀassembly, however care must be taken when
working in transition regions.
Interestingly, the transitions between morphologies ocꢀ
curred at different chain lengths in 2D SAMs than in 3D crysꢀ
tals. Further experiments using the same solvent in both diꢀ
mensions and exploring other substrates would be of interest
to further elucidate the substrate contribution to SAM stability.
The unexpected solvent dependent morphologies observed for
midꢀchain length 2D SAMs leads to an interest in exploring
whether solventꢀindependent film morphologies are attainable
by altering the chainꢀlength of a moiety on an integral SAM
component. This work adds to the continual exploration of
physisorbed SAMs, which is crucial to building the data needꢀ
ed to develop models for predicting SAM and thin film morꢀ
phology, including mixedꢀdimensional (2D/3D) van der Waals
heterostructures.43
Shorter chain CnISA molecules exhibited Hꢀbonding morꢀ
phologies in both phases, with network structures for
C2ISA−C6ISA SAMs and dimer tape structures for
C2ISA−C5ISA 3D crystals. The midꢀchain length molecules,
C6ISA−C9ISA SAMs and C7−C12ISA 3D crystals exhibited
solvent dependent morphologies, with SAMs forming wide
lamella with unexpected phenyloctaneꢀCnISAs alkyl chain
interdigitation and 3D structures forming layered ring or laꢀ
mellar structures in THF and MeOH respectively.14 The long
7
ACS Paragon Plus Environment

Langmuir
Page 8 of 10
7. (a) Elemans, J. A. A. W.; De Feyter, S. Structure and Function
ASSOCIATED CONTENT
1
2
3
4
5
6
7
8
Revealed with Submolecular Resolution at the LiquidꢀSolid Inꢀ
terface. Soft Matter 2009, 721ꢀ735. (b) Yang, Y.; Wang, C.
Solvent Effects on TwoꢀDimensional Molecular Selfꢀ
Assemblies Investigated by Using Scanning Tunneling Microsꢀ
copy. Curr.Opin. Colloid Interface Sci. 2009, 14, 135 ꢀ 147.
8. Amabilino, D. B.; De Feyter, S.; Lazzaroni, R.; GomarꢀNadal, E.;
Veciana, J.; Rovira, C.; AbdelꢀMottaleb, M. M.; Mamdouh,
W.; Iavicoli, P.; Psychogyiopoulou, K.; Linares, M.; Minoia,
A.; Xu, H.; and Puigmart, J. Monolayer SelfꢀAssembly at Liqꢀ
uidꢀSolid Interfaces: Chirality and Electronic Properties of
Molecules at Surfaces. J. Phys. Condens. Matter 2008, 20,
184003.
9. (a) Tao, F.; Bernasek, S. L. Understanding OddꢀEven Effects in
Organic SelfꢀAssembled Monolayers. Chem. Rev. 2007, 107,
1408ꢀ1453. (b) De Feyter, S.; De Schryver, F. C. Selfꢀ
Assembly at the Liquid/Solid Interface: STM Reveals. J. Phys.
Chem. B 2005, 109, 4290ꢀ4302.
10. (a) Lin, N.; Stepanow, S.; Ruben, M.; Barth, J. SurfaceꢀConfined
Supramolecular Coordination Chemistry. In Templates in
Chemistry III; Broekmann, P., Dötz, K. and Schalley, C., Eds.;
Springer Berlin / Heidelberg: 2009; Vol. 287, pp 1ꢀ44.
11. Barth, J. V. Molecular Architectonic on Metal Surfaces. Annu.
Rev. Phys. Chem. , 375.
12. De Feyter, S.; De Schryver, F.C. TwoꢀDimensional Supramolecuꢀ
lar SelfꢀAssembly Probed by Scanning Tunneling Microscopy.
Chem. Soc. Rev. 2003, 32, 139ꢀ150.
13. Plass, K. E.; Grzesiak, A. L.; Matzger, A. J. Molecular Packing
and Symmetry of TwoꢀDimensional Crystals. Acc. Chem. Res
2007, 40, 287ꢀ293.
14. Enkelmann, V.; Valiyaveettil, S.; Möessner, G.; Müllen, K. Selfꢀ
Assembly of 5ꢀAlkoxyisophthalic Acids: Alkyl Chain Length
Dependence for the Formation of ChannelꢀType and Sheetꢀ
Type Structures. Supramol. Sci. 1995, 2, 3ꢀ7.
15. McCormick, L.J.; Morris, S.A.; Slawin, A.M.Z.; Teat, S.J.; Morꢀ
ris, R.E. Coordination Polymers of 5ꢀAlkoxy Isophthalic Acids.
Cryst. Growth Des., 2016, 16 (10), 5771–5780.
16. Dickerson, P. N.; Hibbard, A. M.; Oncel, N.; Bernasek, S. L.
HydrogenꢀBonding versus van der Waals Interactions in Selfꢀ
Assembled Monolayers of Substituted Isophthalic Acids.
Langmuir 2010, 18155.
17. Blunt, M. O.; Adisoejoso, J.; Tahara, K.; Katayama, K.; Van, d.
A.; Tobe, Y.; De Feyter, S. TemperatureꢀInduced Structural
Phase Transitions in a TwoꢀDimensional SelfꢀAssembled Netꢀ
work. J. Am. Chem. Soc. 2013, 135, 12068ꢀ12075.
18. Valiyaveettil, S.; Mullen, K. Multiphase SelfꢀAssembly of 5ꢀ
Alkoxyisophthalic Acid and its Applications. New J. Chem.
1998, 22, 89ꢀ95.
19. EichhorstꢀGerner, K.; Stabel, A.; Moessner, G.; Declerq, D.;
Valiyaveettil, S.; Enkelmann, V.; Müllen, K.; Rabe, J. P. Selfꢀ
Assembly of a TwoꢀComponent HydrogenꢀBonded Network:
Comparison of the TwoꢀDimensional Structure Observed by
Scanning Tunneling Microscopy and the ThreeꢀDimensional
Crystal Lattice. Angewandte Chemie International Edition in
English 1996, 35, 1492ꢀ1495.
20. Samorí, P.; Francke, V.; Enkelmann, V.; Müllen, K.; Rabe, J. P.
Synthesis and Solid State Structures of Functionalized Pheꢀ
nyleneethynylene Trimers in 2D and 3D. - Chem. Mater. 2003,
15, 1032 ꢀ 1039.
Supporting Information. CnISAs NMR data, C2ISA MD simulaꢀ
tion information, C2ISA CIF file. This material is available free of
charge via the Internet at http://pubs.acs.org.
AUTHOR INFORMATION
Corresponding Author
*sberna@princeton.edu
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Author Contributions
The manuscript was written through contributions of all authors.
All authors have given approval to the final version of the manuꢀ
script.
ACKNOWLEDGMENT
This work was partially supported by the National Science Founꢀ
dation, Division of Chemistry, CHEꢀ1213216, the Division of
Materials Research, DMRꢀ1506989 and the Deutsche Forꢀ
schungsgemeinschaft, SFBꢀ765. ETF acknowledges the support
of Humboldtꢀtalent travel grant within the PrincetonꢀHumboldt
University Profile Partnership, NSFꢀGRFP for funding and Proꢀ
fessor Jeffrey Schwartz and Travis Shaw for useful discussions,
and Erin Grey for ChemDraw edits. We acknowledge Phil Jeffrey,
PhD, for XRD structures.
ABBREVIATIONS
CnISA, 5ꢀalkoxyisophthalic acids; SAM, selfꢀassembled monoꢀ
layer; Hꢀbonding, hydrogen bonding; vdW, van der Waals; STM,
scanning tunneling microscopy; HOPG, highly oriented pyrolytic
graphite.
REFERENCES
1. ShanꢀShan, L.; Northrop, B. H.; Yuan, Q.; Wan, L. J.; Stang, P. J.
Surface Confined Metallosupramolecular Architectures: Forꢀ
mation and Scanning Tunneling Microscopy Characterization.
Acc. Chem. Res. 2009, 249.
2. Ahn, S.; Matzger, A. J. Six Different Assemblies from One Buildꢀ
ing Block: TwoꢀDimensional Crystallization of an Amide Amꢀ
phiphile. J. Am. Chem. Soc. 2010, 132, 11364ꢀ11371.
3. Xue, Y.; Zimmt, M. B. Patterned Monolayer SelfꢀAssembly Proꢀ
grammed by Side Chain Shape: FourꢀComponent Gratings. J.
Am. Chem. Soc. 2012, 134, 4513ꢀ4516.
4. Binnig, G.; Rohrer, H. Scanning Tunneling Microscopy ꢀ From
Birth to Adolescence. Rev. Mod. Phys. 1987, 59, 615ꢀ625.
5. (a) Mizutani, W.; Shigeno, M.; Ono, M.; Kajimura, K. Voltageꢀ
Dependent Scanning Tunneling Microscopy Images of Liquid
Crystals on Graphite. Appl. Phys. Lett. 1990, 56, 1974ꢀ1976.
(b) Rabe, J.P.; Buchholz, S. Commensurability and Mobility in
TwoꢀDimensional Molecular Patterns on Graphite. Science,
1991, 253 (5018), 424ꢀ427. (c) Rabe, J.P.; Buchholz, S. Direct
Observation of Molecular Structure and Dynamics at the Interꢀ
face Between a Solid Wall and an Organic Solution by Scanꢀ
ning Tunneling Microscopy. Phys. Rev. Lett. 1991, 66(16),
2096ꢀ2099.
6. Liu, P.; Miao, X.; Li, Z.; Zha, B.; Deng, W. Twoꢀdimensional Selfꢀ
assembly of Singleꢀ, Polyꢀ and Coꢀcrystals at the Liquid/Solid
Interface. Cryst. Eng. Comm. 2014, 41, 9690 ꢀ 9696.
8
ACS Paragon Plus Environment

Page 9 of 10
Langmuir
21. Tong, M.; Zheng, S.; Chen, X. SelfꢀAssembly of Twoꢀ and Threeꢀ
Dimensional Coordination Networks with Hexamethylenetetꢀ
works of Dehydrobenzo[12]annulene Derivatives via Alkyl
Chain Interdigitation. J. Am. Chem. Soc. 2006, 128, 16613ꢀ
16625.
1
2
3
4
5
6
7
8
ramine and Different Silver(I) Salts. Chemistry – A European
Journal 2000, 6, 3729ꢀ3738.
37. Tao, F.; Bernasek, S. L. Complexity in the SelfꢀAssembly of
Bifunctional Molecules on HOPG: The Influence of Solvent
Functionality on SelfꢀAssembled Structures. Langmuir 2007,
23, 3513ꢀ3522.
22. Ruan, P.; Xiao, B.; Ni, H. L.; Hu, P.; Wang, B. Q.; Zhao, K. Q.;
Zeng, Q. D.; Wang, C. The Influence of Alkyl Chain Substituꢀ
tion Pattern on the Twoꢀ and ThreeꢀDimensional Selfꢀ
Assembly of Truxenone Discogens. Liquid Crystals 2014,
41:8, 1152 ꢀ 1161.
23. Gottarelli, G.; Masiero, S.; Mezzina, E.; Pieraccini, S.; Rabe, J. P.;
Samorí, P.; Spada, G. P. The SelfꢀAssembly of Lipophilic
Guanosine Derivatives in Solution and on Solid Surfaces.
Chemistry – A European Journal 2000, 6, 3242ꢀ3248.
24. Palma, C.; Cecchini, M.; Samori, P. Predicting SelfꢀAssembly:
From Empirism to Determinism. . Chem. Soc. Rev. 2012, 41,
3713 ꢀ 3730.
25. Hanwell, M. D.; Curtis, D. E.; Lonie, D. C.; Vandermeersch, T.;
Zurek, E.; Hutchinson, G. R. Avogadro: An Advanced Semanꢀ
tic Chemical Editor, Visualization, and Analysis Platform. J.
Cheminform. 2012, 4:17.
26. Avogadro: An OpenꢀSource Molecular Builder and Visualization
Tool. Version 1.1.1. http://avogadro.openmolecules.net/.
27. Spek, A. L. Structure Validation in Chemical Crystallography.
Acta Crystallographica Section D 2009, 65, 148ꢀ155.
28. Momma, K.; Izumi, F. VESTA: A ThreeꢀDimensional Visualizaꢀ
tion System for Electronic and Structural Analysis. J. Appl.
Crystallogr. 2008, 41, 653ꢀ658.
29. Phillips, J.; Braun, R.; Wang, W.; Gumbart, J.; Tajkhorshid, E.;
Villa, E.; Chipot, C.; Skeel, R.; Kalé, L.; Schulten, K. Scalable
Molecular Dynamics with NAMD. J. Comput. Chem. 2005, 16,
1781ꢀ1802.
30. Vanommeslaeghe, K.; Hatcher, E.; Acharya, C.; Kundu, S.;
Zhong, S.; Shim, J.; Darian, E.; Guvench, O.; Lopes, P.; Voroꢀ
byov, I.; Mackerell, A. D. CHARMM General Force Field: A
Force Field for DrugꢀLike Molecules Compatible with the
CHARMM AllꢀAtom Additive Biological Force Fields. J.
Comput. Chem. 2010, 31, 671ꢀ690.
31. Björk, J.; Hanke, F.; Palma, C.; Samori, P.; Cecchini, M.; Persson,
M. Adsorption of Aromatic and AntiꢀAromatic Systems on
Graphene through π−π Stacking. J. Phys. Chem. Lett. 2010, 1,
3407 ꢀ 3412.
32. Palma, C.; Ciesielski, A.; Oner, M. A.; Schaeffer, G.; Lehn, J.;
Barth, J.; Samori, P. TwoꢀDimensional Soft Supramolecular
Networks. Chem. Commun. 2015, 51, 17297 ꢀ 17300.
33. Winkler, A.; Hess, P. Study of the Energetics and Dynamics of
Hydrogen Bond Formation in Aliphatic Carboxylic Acid Vaꢀ
pors by Resonant Photoacoustic Spectroscopy. J. Am. Chem.
Soc. 1994, 116, 9233ꢀ9240.
38. Shao, X.; Luo, X.; Hu, X.; Wu, K. Solvent Effect on Selfꢀ
Assembled
Structures
of
3,8ꢀBisꢀhexadecyloxyꢀ
benzo[c]cinnoline on Highly Oriented Pyrolytic Graphite. J
Phys Chem B 2006, 110, 1288ꢀ1293.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
39. Chen, X.; Luais, E.; Darwish, N.; Ciampi, S.; Thordarson, P.;
Gooding, J. J. Studies on the Effect of Solvents on Selfꢀ
Assembled Monolayers Formed from Organophosphonic Acids
on Indium Tin Oxide. Langmuir 2012, 28, 9487ꢀ9495.
40. Valiyaveettil, S.; Enkelmann, V.; Moeßner, G.; Müllen, K. Apꢀ
proaches to Supramolecular Structures with Various Topoloꢀ
gies in the Crystal Lattice. Macromolecular Symposia 1996,
102, 165ꢀ173.
41. Lavrich, D. J.; Wetterer, S. M.; Bernasek, S. L.; Scoles, G. Phyꢀ
sisorption and Chemisorption of Alkanethiols and Alkyl Sulꢀ
fides on Au(111). J. Phys. Chem. B 1998, 102, 3456ꢀ3465.
42. Barnard, R. A.; Matzger, A. J. Functional Group Effects on the
Enthalpy of Adsorption for SelfꢀAssembly at the Soluꢀ
tion/Graphite Interface. Langmuir 2014, 30, 7388ꢀ7394.
43. Schwarz, D.; Noda1, Y.; Klouda, J.; SchwarzováꢀPecková, K.;
Tarábek, J.; Rybáček, J.; Janoušek, J.; Simon, F.; Opanasenꢀ
ko, M.V.; Čejka, J.; Acharjya, A.; Schmidt, J.; Selve, S.;
ReiterꢀScherer, V.; Severin, N.; Rabe, J.P.; Ecorchard, P.; He,
J.; Polozij, M.; Nachtigall, P.; Bojdys, M.J. Twinned Growth of
MetalꢀFree, TriazineꢀBased Photocatalyst Films as Mixedꢀ
Dimensional (2D/3D) van der Waals Heterostructures. Adv.
Mater. 2017, 1703399.
34. Kampschulte, L.; Werblowsky, T. L.; Kishore, R. S. K.;
Schmittel, M.; Heckl, W. M.; Lackinger, M. Thermodynamical
Equilibrium of Binary Supramolecular Networks at the Liquidꢀ
Solid Interface. J. Am. Chem. Soc. 2008, 130, 8502ꢀ8507.
35. Tahara, K.; Okuhata, S.; Adisoejoso, J.; Lei, S.; Fujita, T.; De
Feyter, S.; Tobe, Y. 2D Networks of RhombicꢀShaped Fused
Dehydrobenzo[12]annulenes: Structural Variations under Conꢀ
centration Control. J. Am. Chem. Soc. 2009, 131 (48), 17583 ꢀ
17590.
36. Tahara, K.; Furukawa, S.; Ujiꢀi, H.; Uchino, T.; Ichikawa, T.;
Zhang, J.; Mamdouh, W.; Sonoda, M.; De Schryver, F. C.; De
Feyter, S.; Tobe, Y. TwoꢀDimensional Porous Molecular Netꢀ
9
ACS Paragon Plus Environment

Langmuir
Page 10 of 10
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Table of Contents Graphic
10
ACS Paragon Plus Environment