Chiral redox-Active isosceles triangles

Article abstract of DOI:10.1021/jacs.6b02086

Designing small-molecule organic redox-active materials, with potential applications in energy storage, has received considerable interest of late. Herein, we report on the synthesis, characterization, and application of two rigid chiral triangles, each of which consist of non-identical pyromellitic diimide (PMDI) and naphthalene diimide (NDI)-based redox-active units. ¹H and ¹³C NMR spectroscopic investigations in solution confirm the lower symmetry (C₂ point group) associated with these two isosceles triangles. Single-crystal X-ray diffraction analyses reveal their rigid triangular prism-like geometries. Unlike previously investigated equilateral triangle containing three identical NDI subunits, both isosceles triangles do not choose to form one-dimensional supramolecular nanotubes by dint of [C-H···O] interaction-driven columnar stacking. The rigid isosceles triangle, composed of one NDI and two PMDI subunits, forms in the presence of N,N-dimethylformamidetwo different types of intermolecular NDI-NDI and NDI-PMDI stacked dimers with opposite helicities in the solid state. Cyclic voltammetry reveals that both isosceles triangles can accept reversibly up to six electrons. Continuous-wave electron paramagnetic resonance and electron-nuclear double-resonance spectroscopic investigations, supported by density functional theory calculations, on the single-electron reduced radical anions of the isosceles triangles confirm the selective sharing of unpaired electrons among adjacent redox-active NDI subunit(s) within both molecules. The isosceles triangles have been employed as electrode-active materials in organic rechargeable lithium-ion batteries. The evaluation of the structure-performance relationships of this series of diimide-based triangles reveals that the increase in the number of NDI subunits, replacing PMDI ones, within the molecules improves the electrochemical cell performance of the batteries.

Full text of DOI:10.1021/jacs.6b02086

Subscriber access provided by University of Pennsylvania Libraries
Article
Chiral Redox-Active Isosceles Triangles
Siva Krishna Mohan Nalluri, Zhichang Liu, Yilei Wu, Keith R. Hermann, Avik Samanta,
Dong Jun Kim, Matthew D. Krzyaniak, Michael R. Wasielewski, and J. Fraser Stoddart
J. Am. Chem. Soc., Just Accepted Manuscript • DOI: 10.1021/jacs.6b02086 • Publication Date (Web): 12 Apr 2016
Downloaded from http://pubs.acs.org on April 12, 2016
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 free 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 accessible to all
readers and 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.
Journal of the American Chemical Society 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 36
Journal of the American Chemical Society
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
Chiral RedoxꢀActive Isosceles Triangles
Siva Krishna Mohan Nalluri,† Zhichang Liu,† Yilei Wu,‡ Keith R. Hermann,† Avik Samanta,† Dong
Jun Kim,† Matthew D. Krzyaniak,‡ Michael R. Wasielewski‡ and J. Fraser Stoddart*,†
†Department of Chemistry, Northwestern University, 2145 Sheridan Road, Evanston, Illinois
60208ꢀ3113, USA
‡Department of Chemistry, ArgonneꢀNorthwestern Solar Energy Research (ANSER) Center,
Northwestern University, 2145 Sheridan Road, Evanston, Illinois 60208ꢀ3113, USA
*Eꢀmail: stoddart@northwestern.edu
MAIN TEXT
*Correspondence Address
Professor J. Fraser Stoddart
Department of Chemistry
Northwestern University
2145 Sheridan Road
Evanston, IL 60208 (USA)
Email: stoddart@northwestern.edu
1
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 2 of 36
1
2
3
4
■ ABSTRACT
5
6
7
8
Designing smallꢀmolecule organic redoxꢀactive materials, with potential applications in energy
storage, has received considerable interest of late. Herein, we report on the synthesis,
characterization and application of two rigid chiral triangles, each of which consist of nonꢀ
identical pyromellitic diimide (PMDI) and naphthalene diimide (NDI)ꢀbased redox active units.
1H and 13C NMR spectroscopic investigations in solution confirm the lower symmetry (C2 point
group) associated with these two isosceles triangles. Singleꢀcrystal Xꢀray diffraction analyses
reveal their rigid triangular prismꢀlike geometries. Unlike previously investigated equilateral
triangle containing three identical NDI subunits, both isosceles triangles do not choose to form
oneꢀdimensional supramolecular nanotubes by dint of [C–H···O] interactionꢀdriven columnar
stacking. The rigid isosceles triangle, composed of one NDI and two PMDI subunits, forms — in
the presence of N,Nꢀdimethylformamide — two different types of intermolecular NDIꢀNDI and
NDIꢀPMDI πꢀπ stacked dimers with opposite helicities in the solidꢀstate. Cyclic voltammetry
reveals that both isosceles triangles can accept reversibly up to six electrons. Continuousꢀwave
electron paramagnetic resonance and electronꢀnuclear double resonance spectroscopic
investigations, supported by density functional theory calculations, on the singleꢀelectron
reduced radical anions of the isosceles triangles confirm the selective sharing of unpaired
electrons among adjacent redoxꢀactive NDI subunit(s) within both molecules. The isosceles
triangles have been employed as electrodeꢀactive materials in organic rechargeable lithiumꢀion
batteries. The evaluation of the structureꢀperformance relationships of this series of diimideꢀ
based triangles reveals that the increase in the number of NDI subunits, replacing PMDI ones,
within the molecules improves the electrochemical cell performance of the batteries.
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
2
ACS Paragon Plus Environment

Page 3 of 36
Journal of the American Chemical Society
1
2
3
4
■ INTRODUCTION
5
6
7
8
Understanding of the geometrical constraints between multiple redoxꢀactive aromatic
building blocks that exhibit efficient electron hopping and delocalization has presented a
considerable challenge to those researchers fabricating1ꢀ5 organic photovoltaic and molecular
electronic devices. The efficiency of intraꢀ as well as intermolecular charge mobilities in organic
semiconductors relies6 on maximizing electronic coupling and minimizing reorganizational
energies associated with charge transport between neighboring πꢀconjugated redoxꢀactive units.
These factors are determined, however, by (i) the relative orientations, (ii) the intermolecular
distances and (iii) the energy matches involving neighboring redoxꢀactive units.7 In this context,
a number of πꢀconjugated aromatic diimides, particularly pyromellitic diimides (PMDIs),
naphthalene diimides (NDIs) and perylene diimides (PDIs), all of which exhibit unique redox
characteristics, have been employed as nꢀtype organic semiconductors with potential applications
in organic fieldꢀeffect transistors,8ꢀ10 photovoltaics,11 and energy storage devices,12ꢀ14 including
lithiumꢀ and sodiumꢀion batteries.15 In particular, πꢀconjugated electronꢀdeficient NDIs16,17 have
been investigated increasingly in a number of fields such as charge transport,18,19
chemosensing,20,21 synthetic membrane transport,22,23 organocatalysis24,25 and biocatalytic charge
transfer assemblies.26,27 Despite continued efforts relating to a number of πꢀconjugated redoxꢀ
active molecules, the exact structureꢀperformance relationship for efficient charge mobilities
remains elusive.28
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
Recently, we reported29 the synthesis of a chiral equilateral triangle (−)ꢀ3NDIꢀꢁ, consisting
of three redoxꢀactive NDI units that can accept reversibly up to six electrons. Both the
experimental observations, as well as theoretical calculations on (−)ꢀ3NDIꢀꢁ, reveal that a
remarkable cyclical throughꢀspace electron delocalization occurs among the overlapping πꢀ
3
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 4 of 36
1
2
3
4
5
6
7
8
9
orbitals of the pairs of adjacent NDI units, as evidenced by the complete sharing of an unpaired
electron in the stable radical anionic state of this equilateral triangle. Another important aspect of
(−)ꢀ3NDIꢀꢁ is the formation30,31 of a supramolecular gel, composed of a dense network of
entangled nanofibers, in (E)ꢀ1,2ꢀdichloroethene, as well as by its solventꢀdependent packing into
a variety of solidꢀstate superstructures — that is, the formation of (i) nonꢀtubular superstructures
in CHCl3, (ii) singleꢀhanded helical tetrameric nanotubes in ClCH2CH2Cl, and (iii) rigid infinite
nonꢀhelical nanotubes in ClCH2CH2Br, ClCH2CH2I, and BrCH2CH2Br — driven by the
columnar stacking of (−)ꢀ3NDIꢀꢁ with a 60° rotational angle between the neighboring triangles
and employing multiple weak [C–H···O] interactions associated with directed halogenꢀhalogen
(X) interactions along the [X···X]ꢀbonded solvent chains inside the nanotubes. In the meantime,
Awaga32 has reported the Xꢀray superstructure of the radical anion of (−)ꢀ3NDIꢀꢁ, formed by
electrochemical reduction, revealing the formation of an impressive K4 structure, driven by the
intermolecular π−π stacking interactions of the NDI radical anions in the equilateral triangle. It
should be emphasized that all of these observations are a consequence of three equivalent NDI
units arranged in a shapeꢀpersistent triangular structure.
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
On the other hand, motivated by the global energy demand in this century, rechargeable
lithiumꢀion batteries (LIBs) have resulted33ꢀ38 in a surge of interest as the stateꢀofꢀtheꢀart power
sources for electronic devices, such as mobile phones and laptops, as well as electric cars
because of their high energy and power densities. A wideꢀvariety of redoxꢀactive organic
materials such as small molecules15,39, polymers12,40,41 and salts,42,43 have been investigated as
active materials in rechargeable LIBs over the past few decades.44 In addition to its remarkable
redoxꢀactive characteristics and high thermal stability, (−)ꢀ3NDIꢀꢁ is only sparingly soluble in
nonꢀaqueous electrolyte solvents. We therefore employed it recently45 as the active material in
4
ACS Paragon Plus Environment

Page 5 of 36
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
organic rechargeable LIBs. Remarkably, the (−)ꢀ3NDIꢀꢁꢀbased organic batteries accessed 95%
of their theoretical capacity (154.8 mAh g−1) at a current rate of 0.1C, while also being able to
operate with a capacity of 58.1 mAh g−1 at ultrahigh rates (100C). Furthermore, a capacity of
71.1 mAh g−1 is maintained at 10C, even after cycling for 300 cycles. Control experiments
carried out on the monomeric reference compound RefꢀNDI led to poor rate capabilities and
cycling performance, attributed45 primarily to the higher solubility of RefꢀNDI in the battery
electrolytes. Despite the fact that a threeꢀdimensional rigid nanoporous framework of the organic
active material plays a role46,47 in achieving high cell performances, a comprehensive
understanding of structureꢀperformance relationships has still to be explored. It remains,
therefore, a puzzle as to what happens to the properties of molecular triangles, if one replaces the
three identical NDI subunits in stepwise fashion with, for instance, PMDI subunits, having
almost identical dimensions to those of NDI subunits, and without disrupting the shapeꢀpersistent
triangular framework. The rational design of this series of diimideꢀbased isosceles triangles is of
particular interest in this present investigation.
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
In the quest to advance our understanding of how the molecular composition of two classes
of nonꢀidentical redoxꢀactive PMDI and NDI units within the isosceles triangles affects the
properties associated with (i) throughꢀspace electron sharing, (ii) their cooperative solidꢀstate
packing and (iii) the remarkable electrochemical cell performance, we have designed and
synthesized (Figure 1) two chiral isosceles triangles—namely, (−)ꢀ1PMDIꢀ2NDIꢀꢁ and (−)ꢀ
2PMDIꢀ1NDIꢀꢁ. We have also characterized and analyzed the spectroscopic, magnetic and
electrochemical properties of (−)ꢀ1PMDIꢀ2NDIꢀꢁ and (−)ꢀ2PMDIꢀ1NDIꢀꢁ and compared our
results with those of the previously reported chiral equilateral triangles—namely, (−)ꢀ3NDIꢀꢁ29
and (−)ꢀ3PMDIꢀꢁ48—as well as the related (Figure 1c) monomeric reference compounds, Refꢀ
5
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 6 of 36
1
2
NDI49 and RefꢀPMDI50. Finally, the structureꢀperformance relationships of batteries fabricated
3
4
5
6
using this series of diimideꢀbased molecular triangles were investigated.
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
■ RESULTS AND DISCUSSION
The two chiral isosceles triangles were prepared by stepwise condensations between
commercially available (RR)ꢀtransꢀ1,2ꢀcyclohexanediamine and two different dianhydride
derivatives—namely, naphthalenetetracarboxylic dianhydride (NDA) and pyromellitic
dianhydride (PMDA). The condensation between (RR)ꢀtransꢀ1,2ꢀcyclohexanediamine and a fiveꢀ
fold excess of either NDA or PMDA in either DMF or acetic acid gave (Figure 1a and b) the
corresponding dimers—namely, (−)ꢀ2NIA2 and (−)ꢀ2PMIA, respectively, in 60‒65% yield. The
condensation of 2 equiv of (RR)ꢀtransꢀ1,2ꢀcyclohexanediamine with 1 equiv of either NDA or
PMDA in the presence of 2 equiv of pꢀTsOH in DMF or acetic acid gave the corresponding
ditosylates of (−)ꢀ1NDIꢀ2NH2 and (−)ꢀ1PMDIꢀ2NH2, respectively. The subsequent extraction of
these salts with a saturated aqueous NaHCO3 solution afforded (Figure 1a and b) the diamines
(−)ꢀ1NDIꢀ2NH2 and (−)ꢀ1PMDIꢀ2NH2 in 60‒65% yield. Macrocyclizations were carried out in
DMF at 130 °C by the cyclocondensation of 1 equiv of (−)ꢀ2NIA with 1 equiv of (−)ꢀ1PMDIꢀ
2NH2 afforded the desired isosceles triangle (−)ꢀ1PMDIꢀ2NDIꢀꢁ in 50% yield, while the
cyclocondensation of 1 equiv of (−)ꢀ2PMIA with 1 equiv of (−)ꢀ1NDIꢀ2NH2 afforded (Figure 1a
and b, see the Supporting Information for more details) the desired isosceles triangle (−)ꢀ
2PMDIꢀ1NDIꢀꢁ in 50% yield. Characterization of both (−)ꢀ1PMDIꢀ2NDIꢀꢁ and (−)ꢀ2PMDIꢀ
1NDIꢀꢁ was achieved by electrospray ionization highꢀresolution mass spectrometry (ESIꢀ
HRMS) which confirmed the presence of the species [M + H]+ in the gas phase at m/z =
989.2779 and 939.2607, respectively.
6
ACS Paragon Plus Environment

Page 7 of 36
Journal of the American Chemical Society
1
2
NMR Spectroscopy. 1H NMR Spectra (Figure 2 and Figures S1ꢀS8) of (−)ꢀ1PMDIꢀ2NDIꢀꢁ
and (−)ꢀ2PMDIꢀ1NDIꢀꢁ recorded in CDCl3 display characteristic resonances for the protons of
both the NDI and PMDI subunits in the downfield regions with δ = 8.2–9.0 and 7.8–8.2 ppm,
respectively, as well as the aliphatic methine (–CH) and methylene (–CH2) protons of the
cyclohexaneꢀlinkers in the upfield regions with δ = 3.0–6.5 and 1.5–3.0 ppm, respectively. In
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
1
particular, H NMR spectrum of (−)ꢀ1PMDIꢀ2NDIꢀꢁ shows (Figure 2b) four sets of signals for
the 8 NDI protons and only one singlet for the 2 PMDI protons, while that of (−)ꢀ2PMDIꢀ1NDIꢀ
ꢁ shows (Figure 2c) two sets of signals for the 4 NDI protons and two doublets for the 4 PMDI
protons. At higher field, three sets of signals were observed for the six methine protons in both
(−)ꢀ1PMDIꢀ2NDIꢀꢁ and (−)ꢀ2PMDIꢀ1NDIꢀꢁ. A comparison of the 1H NMR spectra (Figure 2),
along with 13C NMR spectra (Figure S9), of this series of triangles confirms the lower symmetry
(C2 point group) of the isosceles triangles (−)ꢀ1PMDIꢀ2NDIꢀꢁ and (−)ꢀ2PMDIꢀ1NDIꢀꢁ in
comparison to the equilateral triangles (−)ꢀ3NDIꢀꢁ and (−)ꢀ3PMDIꢀꢁ belonging to the D3 point
1
group. Furthermore, variable temperature H NMR was performed51 (Figure 3) on (−)ꢀ2PMDIꢀ
1NDIꢀꢁ to evaluate the rates of rotation of the PMDI and NDI units around the C–N•••N–C
bond axis of the isosceles triangle. At higher temperatures, coalescence of the two doublets into
one broad signal was observed for the PMDI protons (HC and HC’).52 The free energy of
activation at the coalescence temperature (426 K) was calculated53 to be 23 kcal mol−1.
SingleꢀCrystal Xꢀray Diffraction (XRD) Analyses. In order to gain more insight into the
structural details of these isosceles triangles, we performed singleꢀcrystal Xꢀray diffraction
analyses (Figures 4 and 5) on the crystals obtained (i) by slow vapor diffusion of nꢀhexane into a
3.0 mM solution of (−)ꢀ1PMDIꢀ2NDIꢀꢁ in 1,2ꢀdibromoethane (DBE) and (ii) by slow vapor
diffusion of H2O into a 5.0 mM solution of (−)ꢀ2PMDIꢀ1NDIꢀꢁ in DMF over the course of 3
7
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 8 of 36
1
2
3
4
5
6
7
8
9
days. The solidꢀstate (super)structure (Figure 4) of (−)ꢀ1PMDIꢀ2NDIꢀꢁ reveals the rigid
geometries of the triangular hollow prism in which the tilt angles of the PMDI (23°) and the two
NDI subunits (6 and 14°) are much greater than those29,30 of three NDI subunits (~3°) in the
equilateral triangle (−)ꢀ3NDIꢀꢁ. This observation indicates that the less bulky PMDI unit
facilitates the tilting of all three πꢀsurfaces. In striking contrast to the tubular superstructure of
(−)ꢀ3NDIꢀꢁ in DBE, that of (−)ꢀ1PMDIꢀ2NDIꢀꢁ in DBE reveals (Figure 4cꢀe) neither columnar
stacking of the triangular units to form oneꢀdimensional (1D) nanotubes nor π−π stacking
between two NDI units or PMDI and NDI units. Although the cavity of (−)ꢀ1PMDIꢀ2NDIꢀꢁ is
penetrated (Figure 4a and b) by two [Br···Br]ꢀbonded (3.28 Å) DBE molecules which are
stabilized by [Br···π] interactions (~3.5 Å) with πꢀsurfaces, only one DBE molecule is in antiꢀ
conformation and the other is in the gaucheꢀconformation (Figure 4b), which prohibits the
formation of continuous 1D [Br···Br]ꢀbonded solvent chains that would template the formation
of the nanotubes.30 The replacement of the NDI subunits with PMDI ones within the triangle also
breaks the complementarity of [C−H···O] interactions between two triangles which is crucial for
their columnar stacking.
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
Although we were unable to obtain suitable single crystals of (−)ꢀ2PMDIꢀ1NDIꢀꢁ from
DBE/nꢀhexane for Xꢀray diffraction analysis, the compound did crystallize from the aqueous
DMF system. The solidꢀstate structure of (−)ꢀ2PMDIꢀ1NDIꢀꢁ reveals (Figure 5a) a similar rigid
triangular geometry. Among the three of four asymmetric units in the unit cell, one or two DMF
molecules are bound (Figure 5b, bottom) to their cavities by means of very short π−π overlap
(~3.0 Å) between the amide groups of DMF and the NDI planes. Much to our surprise, these four
isosceles triangles assemble (Figure 5b) into two types of π−π stacking supramolecular dimers—
(i) (P)ꢀhelical dimer [(P)ꢀDimer, Figure 5b, left] with a rightꢀtwisted angle of 57° formed
8
ACS Paragon Plus Environment

Page 9 of 36
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
through π−π stacking (3.4 Å) between the DMFꢀbound NDI plane of one triangle and a PMDI
plane of another triangle, and (ii) (M)ꢀhelical dimer [(M)ꢀDimer, Figure 5b, right] with a leftꢀ
twisted angle of 50° formed through π−π stacking (3.4 Å) between DMFꢀbound NDI planes of
another two triangles. It is worth noting that prior to this work, π−π stacking between NDI units
of molecular triangles had been observed only in the cases of host−guest complexes29
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
−
−
[I3 ⊂3NDIꢀꢁ] of 3NDIꢀꢁ with I3 and the radical anion32 [(−)ꢀ3NDIꢀꢁ]•− of (−)ꢀ3NDIꢀꢁ on
account of the increase in the electron density of the neutral NDI unit as a result of anion binding
and reduction, respectively. In the case of (−)ꢀ2PMDIꢀ1NDIꢀꢁ, it appears that DMF acts in a
−
role similar to that of I3 in increasing the electron density of the NDI unit through very short
π−π stacking which favors the π−π stacking of another facet of the NDI unit with the other πꢀ
plane. The helicity of supramolecular assembles are determined predominantly by the chirality of
the chiral monomer.29,30 In the superstructure of (−)ꢀ2PMDIꢀ1NDIꢀꢁ, however, the assembly of
the homochiral monomers gives rise to two types of dimers with opposite supramolecular
helicities. While the (M)ꢀDimer, formed through NDIꢀNDI π−π stacking, retains (M)ꢀhelicity
which is consistent with that observed29,32 in the π−π stacking assemblies of (−)ꢀ3NDIꢀꢁ, the
(P)ꢀDimer formed through NDIꢀPMDI π−π stacking shows (P)ꢀhelicity. The plausible
mechanism for the coexistence of both these dimers can be explained by the fact that the lessꢀ
bulky PMDI is unable to restrict the effect of the stereogenic centers of the (RR)ꢀtransꢀ1,2ꢀ
cyclohexano groups for specifically forming the corresponding (M)ꢀhelicity. Both (P)ꢀ and (M)ꢀ
Dimers assemble (Figure 5c) to form 1D tapes along the bꢀaxis, respectively, which then pack
(Figure 5d) alternatively in the aꢀc plane to form the crystals. Strikingly, both Xꢀray
superstructures of (−)ꢀ1PMDIꢀ2NDIꢀꢁ and (−)ꢀ2PMDIꢀ1NDIꢀꢁ exhibit a common feature —
that is, there is no columnar stacking of either triangle to form 1D nanotubes, an observation
9
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 10 of 36
1
2
3
4
5
6
7
8
9
which indicates that the breaking of symmetry of the triangles hinders the complementarity of
[C−H···O] hydrogen bonding interactions between the columnar stacked triangles, and
consequently, these observations highlight the necessity of the symmetry of (−)ꢀ3NDIꢀꢁ in order
to provide complementary [C−H···O] interactions for driving the formation of 1D columnar
stacking nanotubes.30
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
UV/Vis Absorption and Circular Dichroism (CD) Spectroscopies. The optical properties
of all the triangles were investigated (Figure 6a) by UV/Vis absorption spectroscopy. The
comparison of the absorption spectra of this series of triangles suggests that the intensity of the
absorption peaks with maxima at 360 and 380 nm decreases gradually on going from (−)ꢀ3NDIꢀ
ꢁ to (−)ꢀ1PMDIꢀ2NDIꢀꢁ to (−)ꢀ2PMDIꢀ1NDIꢀꢁ and eventually disappears in the case of (−)ꢀ
3PMDIꢀꢁ, in line with the decrease in the number of NDI subunits present in the triangle. This
observation is consistent with the characteristic 0←0 and 0←1 vibronic bands corresponding to
the S1←S0 electronic transition of the NDI subunits. On the other hand, the intensity of the
absorption peaks in the region 225−280 nm rises with the increase in the number of PMDI
subunits, an observation which is in good agreement with the characteristic zꢀpolarized π−π*
electronic transitions48 of the PMDI units.
CD spectroscopy (Figure 6b) was also performed in order to probe the chiroptical properties
and Cotton effects generated by the interchromophoric exciton coupling of the neighboring NDI
or PMDI subunits within a given triangle. Characteristically strong negative exciton Cotton
effects are observed (Figure 6b) in the range 350–400 nm, originating from the coupling of the
π–π* electronic transitions polarized along the long zꢀaxis of the NDI subunits (Band I region) in
the case of the triangles (−)ꢀ3NDIꢀꢁ and (−)ꢀ1PMDIꢀ2NDIꢀꢁ, while these strong negative
exciton Cotton effects in this region are not observed in the case of the triangles (−)ꢀ2PMDIꢀ
10
ACS Paragon Plus Environment

Page 11 of 36
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
1NDIꢀꢁ and (−)ꢀ3PMDIꢀꢁ, because of the absence of more than one neighboring NDI unit to
promote the exciton coupling.54 It should be noted that the negative Cotton effects observed here
are consistent with the absolute (RRRRRR)ꢀconfiguration of the triangles. In the case of (−)ꢀ
3NDIꢀꢁ, the strong negative Cotton effect observed (Figure 6b) at 242 nm originates from the
coupling of the π–π* transitions polarized along the short yꢀaxis of the NDI subunits (Band II
region). On the other hand, the strong negative exciton couplet at 240 and 248 nm in the case of
(−)ꢀ3PMDIꢀꢁ appears as a consequence of the coupling of the π–π* transitions polarized along
both the long zꢀaxis and the short yꢀaxis of the PMDI subunits, respectively. In the case of the
isosceles triangles (−)ꢀ1PMDIꢀ2NDIꢀꢁ and (−)ꢀ2PMDIꢀ1NDIꢀꢁ, the negative exciton Cotton
effects observed in the region 225–240 nm are presumably influenced by the multiple exciton
couplings of the transition dipoles polarized along both the long zꢀ and the short yꢀaxes of the
PMDI and the short yꢀaxis of the NDI subunits.
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
Cyclic Voltammetry (CV). We first performed CV experiments (Figure 7) in CH2Cl2 on the
two isosceles triangles [(−)ꢀ1PMDIꢀ2NDIꢀꢁ and (−)ꢀ2PMDIꢀ1NDIꢀꢁ], and compared these
results with the equilateral triangles [(−)ꢀ3NDIꢀꢁ and (−)ꢀ3PMDIꢀꢁ] as well as with the
monomeric reference compounds (RefꢀNDI and RefꢀPMDI). While the CV of the RefꢀNDI
shows two distinct reversible oneꢀelectron waves with halfꢀwave potentials at –709 and −1131
mV vs Ag/AgCl, corresponding to the formation of the [RefꢀNDI]•− radical anion and the [Refꢀ
NDI]2− dianion, respectively, the CV of (−)ꢀ3NDIꢀꢁ shows clear splittings into six distinct
reversible oneꢀelectron redox processes. Similarly, the CV of the RefꢀPMDI shows a reversible
oneꢀelectron reduction process at −915 mV, corresponding to the formation of the [RefꢀPMDI]•−
radical anion, while the CV of (−)ꢀ3PMDIꢀꢁ also shows clear splittings into three distinct
reversible oneꢀelectron waves.55 Strikingly, the CV of (−)ꢀ1PMDIꢀ2NDIꢀꢁ displays (Figure 7)
11
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 12 of 36
1
2
3
4
5
6
7
8
9
multiple reduction processes involving a total of sixꢀelectrons — namely, (i) two sequential
distinct reversible oneꢀelectron waves at −563 and –708 mV, corresponding to the formation of
two singly reduced [NDI]•− radical anions, (ii) a reversible oneꢀelectron wave at –998 mV,
corresponding to the formation of a singly reduced [PMDI]•− radical anion, (iii) two closely
following reversible oneꢀelectron processes at −1206 and –1313 mV, corresponding to the
formation of two doubly reduced [NDI]2− dianions and (iv) finally, a quasiꢀreversible oneꢀ
electron process at –1776 mV, corresponding to the formation of a doubly reduced [PMDI]2−
dianion. Also, the CV of (−)ꢀ2PMDIꢀ1NDIꢀꢁ displayed (Figure 7) multiple features containing
six electrons in total — namely, (i) a reversible oneꢀelectron reduction process at −564 mV,
corresponding to the formation of a singly reduced [NDI]•− radical anion, (ii) two closely
following reversible oneꢀelectron reduction waves at –881 and –991 mV, corresponding to the
formation of two singly reduced [PMDI]•− radical anions, (iii) a reversible oneꢀelectron
reduction wave at −1215 mV, corresponding to the formation of a singly reduced [NDI]2−
dianion, and (iv) finally, a quasiꢀreversible twoꢀelectron process at –1712 mV, corresponding to
the formation of two doubly reduced [PMDI]2− dianions. Although the first reduction potential
(E1 = ca. −564 mV) of the NDI unit in all of the NDIꢀcontaining triangles [(−)ꢀ3NDIꢀꢁ, (−)ꢀ
1PMDIꢀ2NDIꢀꢁ, (−)ꢀ2PMDIꢀ1NDIꢀꢁ] is nearly constant and shifted by 145 mV towards more
positive potentials compared with that of RefꢀNDI (E1 = −709 mV), the subsequent reduction
potentials can be tuned significantly, thus establishing a set of unique molecular triangles for
structureꢀproperty relationship investigations and the device fabrication of efficient energy
storage. All of these observations suggest that, despite having two types of nonꢀidentical
neighboring redoxꢀactive units within triangular architectures, each molecular triangle can accept
reversibly up to six electrons
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
12
ACS Paragon Plus Environment

Page 13 of 36
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
EPR/ENDOR Spectroscopy. In order to gain more insight into the throughꢀspace electron
sharing among the neighboring redoxꢀactive NDI and PMDI units within the molecular triangles,
continuousꢀwave electron paramagnetic resonance (CWꢀEPR) and electronꢀnuclear double
resonance (ENDOR) spectroscopies were carried out on the monoreduced radical anions of the
triangles generated by adding one equivalent of cobaltocene (CoCp2) as the chemical reductant.
Solutionꢀphase CWꢀEPR spectra of [(−)ꢀ3NDIꢀꢁ]•− and [(−)ꢀ1PMDIꢀ2NDIꢀꢁ]•− monoradical
anions reveal (Figure 8a) a decrease in the spectral width which corresponds to a decrease in the
magnitude of the nuclear hyperfine interactions, relative to that of [RefꢀNDI]•−. In contrast, the
EPR spectrum of [(−)ꢀ2PMDIꢀ1NDIꢀꢁ]•− matches nearly perfect with that of [RefꢀNDI]•−, i.e.,
no significant change in spectral width or number of lines was observed. The CWꢀENDOR
spectra of [(−)ꢀ3NDIꢀꢁ]•− and [(−)ꢀ1PMDIꢀ2NDIꢀꢁ]•− reveal (Figure 8b) an obvious decrease in
the isotropic hyperfine coupling constant (aH) by factors of ca. 3 and 2, respectively, relative to
that of [(−)ꢀ2PMDIꢀ1NDIꢀꢁ]•− and [RefꢀNDI]•−.56 According to the McConnell relation, the
hydrogen isotropic hyperfine coupling constant is proportional to the unpaired spin density at the
adjacent carbon56 and should therefore decrease linearly with the number of NDI subunits which
share the unpaired spin density. The observed decrease in the hydrogen isotropic hyperfine
coupling constants suggests that the unpaired spin density is shared among all of the NDI
subunits present within a given molecular triangle on the timescale (107 s−1) of the electronꢀ
nuclear hyperfine interaction. Moreover, simulation of the CWꢀEPR spectra confirmed that only
the NDI subunits present within the molecular triangles contribute to the spectrum.
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
Density Functional Theory (DFT) Calculations. Previously, we performed29 DFT
calculations on the equilateral triangles: they revealed the presence of delocalized frontier
molecular orbitals (FMOs) located on all of the three symmetrically equivalent NDI subunits in
13
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 14 of 36
1
2
3
4
5
6
7
8
9
the LUMO of (−)ꢀ3NDIꢀꢁ, supporting the CV and EPR results commensurate with cyclic
throughꢀspace electron sharing among all the neighboring NDI subunits. We therefore performed
(Figure 6c and d) similar DFT calculations using a continuum solvent model with the functional
B3LYP/6ꢀ31G* on the two isosceles triangles (−)ꢀ1PMDIꢀ2NDIꢀꢁ and (−)ꢀ2PMDIꢀ1NDIꢀꢁ to
confirm where the delocalized FMOs are located in each case. We observed (Figure 6c and d)
that the LUMO, as well as LUMO+1, of (−)ꢀ1PMDIꢀ2NDIꢀꢁ are located only on the two NDI
subunits present, while the LUMO of (−)ꢀ2PMDIꢀ1NDIꢀꢁ is located only on the one available
NDI subunit. These calculations are consistent with the CV, as well as EPR/ENDOR results,
supporting the throughꢀspace sharing of the unpaired electron selectively by the NDI subunits
present in the case of the monoreduced radical anions [(−)ꢀ1PMDIꢀ2NDIꢀꢁ]•− and [(−)ꢀ2PMDIꢀ
1NDIꢀꢁ]•−. The LUMO+1 of (−)ꢀ2PMDIꢀ1NDIꢀꢁ is located, however, on both the PMDI
subunits present in the triangle, indicating that the two subsequent reductions of the monoradical
anion to diꢀ and triꢀradical anionic states occur on the two electronically communicating PMDI
subunits.
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
SolidꢀState Cyclic Voltammetry. Motivated by our earlier research45 on the remarkable cell
performance given by the equilateral triangle (−)ꢀ3NDIꢀꢁ as an active material in organic
rechargeable LIBs, we decided to investigate the structureꢀperformance relationship of this
unique set of molecular triangles, namely (−)ꢀ3NDIꢀꢁ, (−)ꢀ1PMDIꢀ2NDIꢀꢁ and (−)ꢀ2PMDIꢀ
1NDIꢀꢁ. The electrochemical performance of these triangles was characterized using CR2032ꢀ
type coin cells. The electrodes were prepared by mixing 50 wt% active material [(−)ꢀ1PMDIꢀ
2NDIꢀꢁ or (−)ꢀ2PMDIꢀ1NDIꢀꢁ], 40 wt% carbon black and 10 wt% polyvinylidene fluoride
(PVDF) binder in anhydrous Nꢀmethylꢀ2ꢀpyrrolidone (NMP) to form a wellꢀdispersed slurry,
followed by coating on aluminum foil and drying under vacuum at 120 °C for 24 h.
14
ACS Paragon Plus Environment

Page 15 of 36
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
The solidꢀstate CV of (−)ꢀ1PMDIꢀ2NDIꢀꢁ displayed (Figure 9a) three sets of peaks,
centered at 2.50, 2.43 and 2.10 V vs Li/Li+, presumably corresponding to (i) the twoꢀelectron
(mono reduction of two NDI subunits to two [NDI]•− radical anions), (ii) the threeꢀelectron (first
reduction of PMDI subunit to [PMDI]•− radical anion and subsequent reduction of two [NDI]•−
radical anions to [NDI]2− dianions) and (iii) oneꢀelectron (subsequent reduction of [PMDI]•−
radical anion to [PMDI]2− dianion) redox processes, respectively. Also, the solidꢀstate CV of
(−)ꢀ2PMDIꢀ1NDIꢀꢁ displayed (Figure 9d) only two sets of peaks, centered at 2.43 and 2.19 V vs
Li/Li+, presumably corresponding to the fourꢀelectron (double reduction of NDI subunits to
[NDI]2− dianions and the monoreduction of the two PMDI subunits to two [PMDI]•− radical
anions) and twoꢀelectron (subsequent reduction of two [PMDI]•− radical anion subunits to two
[PMDI]2− dianions) redox processes. In addition, the comparison of the solidꢀstate CV (Figure
9a and d) with the solutionꢀstate CV (Figure 7) of (−)ꢀ1PMDIꢀ2NDIꢀꢁ and (−)ꢀ2PMDIꢀ1NDIꢀꢁ
reveals significant overlapping of the peaks in the solidꢀstate. This overlap leads to a decrease in
the total number of peaks observed and some of these peaks are not wellꢀresolved. It is
interesting to note (Figure 9a and d) that the areas of the oxidation and reduction peaks as well as
the peak potentials in the solidꢀstate CV remained almost unchanged, with a slight decrease in
the current, during cycling for both (−)ꢀ1PMDIꢀ2NDIꢀꢁ and (−)ꢀ2PMDIꢀ1NDIꢀꢁ, suggesting
that all the redox reactions are highly reversible even in the solidꢀstate. In line with the CV
results obtained previously45 for (−)ꢀ3NDIꢀꢁ, the good electrochemical stability and reversibility
of (−)ꢀ1PMDIꢀ2NDIꢀꢁ and (−)ꢀ2PMDIꢀ1NDIꢀꢁ obtained here prompted us to evaluate the
structureꢀperformance relationship of these triangles as active materials in organic rechargeable
LIBs.
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
15
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 16 of 36
1
2
Electrochemical Cell Performance. We investigated57 the galvanostatic measurements of
the batteries at different current rates of 0.5 and 1.0C.58 In the case of (−)ꢀ1PMDIꢀ2NDIꢀꢁ, a
threeꢀstep charge/discharge profile was noted upon careful observation at rates of 0.5 and 1.0C
(Figure 9b and c). The not so wellꢀdefined plateaus displayed in the charge/discharge profile of
the (−)ꢀ1PMDIꢀ2NDIꢀꢁ battery at about 2.3 and 2.0 V match the CV results (Figure 9a). The
experimental charge and discharge capacities (Figure 9b) of the (−)ꢀ1PMDIꢀ2NDIꢀꢁ battery at
0.5C are 132.2 and 133.9 mAh g−1, corresponding to 81 and 82% of the theoretical capacity
(162.6 mAh g−1) of (−)ꢀ1PMDIꢀ2NDIꢀꢁ, respectively. The cycling performance of (−)ꢀ1PMDIꢀ
2NDIꢀꢁ battery was also tested (Figure S10) at current rates of 0.5 and 1.0C. While a slight
discharge capacity decay was observed during the initial cycling, the rate of decay is improved
when the rate is increased from 0.5 to 1.0C (Figure 9 and Figure S10).
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
Similar experiments performed on the (−)ꢀ2PMDIꢀ1NDIꢀꢁ battery also revealed (Figure 9e
and f) a more wellꢀdefined charge/discharge profile upon careful observation at rates of 0.5 and
1.0C. Surprisingly, the experimental charge and discharge capacities of (−)ꢀ2PMDIꢀ1NDIꢀꢁ
battery at 0.5C are 39.4 and 40.1 mAh g−1, corresponding to only 23.0 and 23.5% of the
theoretical capacity (171.3 mAh g−1) of (−)ꢀ2PMDIꢀ1NDIꢀꢁ. By contrast, the cycling
performance of the (−)ꢀ2PMDIꢀ1NDIꢀꢁ battery showed (Figure S10) a rapid decay of the
capacity. The poor performance observed for the (−)ꢀ2PMDIꢀ1NDIꢀꢁ battery, compared with
that of the (−)ꢀ1PMDIꢀ2NDIꢀꢁ battery, could be attributed to the increased solubility59,60 of the
neutral as well as the reduced states of the (−)ꢀ2PMDIꢀ1NDIꢀꢁ based active material under the
galvanostatic measurement conditions.
Overall, the comparisons of the rate performance and cycling stability of this series of
diimideꢀbased batteries containing (−)ꢀ3NDIꢀꢁ, (−)ꢀ1PMDIꢀ2NDIꢀꢁ and (−)ꢀ2PMDIꢀ1NDIꢀꢁ
16
ACS Paragon Plus Environment

Page 17 of 36
Journal of the American Chemical Society
1
2
suggest that the electrochemical cell performance of the batteries follows45,61 the order (−)ꢀ
3NDIꢀꢁ > (−)ꢀ1PMDIꢀ2NDIꢀꢁ > (−)ꢀ2PMDIꢀ1NDIꢀꢁ. From this data, it can be concluded that
the replacement of the NDI with PMDI subunits within the molecular triangle can dramatically
increase the solubility of the resultant active material in the battery electrolytes. These
observations are in good agreement with the literature report14 which suggests that the PMDI
derivatives are noted to deliver lower capacities compared with their NDI counterparts, on
account of the larger electron affinity of the NDI units.
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
■ CONCLUSIONS
In summary, two chiral isosceles triangles (−)ꢀ1PMDIꢀ2NDIꢀꢁ and (−)ꢀ2PMDIꢀ1NDIꢀꢁ
1
have been synthesized. Comparison of the H NMR spectra of these two triangles with those of
the previously characterized29,48 equilateral triangles [(−)ꢀ3NDIꢀꢁ and (−)ꢀ3PMDIꢀꢁ] is
consistent with the lower symmetries (C2 point groups) of these two isosceles triangles. The
solidꢀstate (super)structures of the isosceles triangles, grown from both 1,2ꢀdibromoethane/nꢀ
hexane and DMF/H2O, reveal the formation of 2D layerꢀlike superstructures, confirming the lack
of the ability of the isosceles triangles to form extended 1D tubular superstructures previously
observed30 in the case of (−)ꢀ3NDIꢀꢁ. Specifically, (−)ꢀ2PMDIꢀ1NDIꢀꢁ, in the presence of
DMF, forms two different types of intermolecular NDIꢀNDI and NDIꢀPMDI πꢀπ stacking dimers
[(P)ꢀ and (M)ꢀDimers] with opposite supramolecular helicities. Cyclic voltammetry suggests
that throughꢀspace electronic communication between the identical NDIꢀNDI or PMDIꢀPMDI
units in the isosceles triangles produces up to six accessible redox states. It should be noted that
both the EPR and ENDOR experiments, supported by DFT calculations, in the case of
monoreduced radical anions of the isosceles triangles indicate that the unpaired electron is shared
selectively among the adjacent NDI — but not PMDI — units within the triangular geometries.
17
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 18 of 36
1
2
3
4
5
6
7
8
9
The isosceles triangles, when employed as active materials in organic rechargeable lithiumꢀion
batteries, were shown to have capacities of 140.1 and 58.1 mAh gꢀ1 for (−)ꢀ1PMDIꢀ2NDIꢀꢁ and
(−)ꢀ2PMDIꢀ1NDIꢀꢁ, respectively. When compared with the remarkable cell performance we
have observed45 for (−)ꢀ3NDIꢀꢁ, a good rate performance and cycling stability were noted for
(−)ꢀ1PMDIꢀ2NDIꢀꢁ, while a relatively poor electrochemical performance was recorded for (−)ꢀ
2PMDIꢀ1NDIꢀꢁ, mainly on account of the good solubility of the active material (−)ꢀ2PMDIꢀ
1NDIꢀꢁ in the battery electrolytes. The structureꢀperformance relationships for this series of
molecular triangles reveal that the electrochemical cell performance improves as the number of
NDI subunits is increased within the triangles. Probing these relationships is providing us with
increased understanding on how the rational design of redoxꢀactive organic molecules affects
their electron transport properties, thus paving the way for the fabrication of next generation
organic electronics and energy storage devices.
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
■ ASSOCIATED CONTENT
Supporting Information
Experimental details, including synthesis, NMR and supportive figures. This material is available
free of charge via the Internet at http://pubs.acs.org.
■ AUTHOR INFORMATION
Corresponding Author
*stoddart@northwestern.edu
Notes
The authors declare no competing financial interest.
■ ACKNOWLEDGEMENTS
18
ACS Paragon Plus Environment

Page 19 of 36
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
This research is part (Project #34ꢀ948) of the Joint Center of Excellence in Integrated Nanoꢀ
Systems (JCIN) at King Abdulaziz City of Science and Technology (KACST) and Northwestern
University (NU). This work was supported by the Chemical Sciences, Geosciences, and
Biosciences Division, Office of Basic Energy Sciences, DOE under grant no. DEꢀFG02ꢀ
99ER14999 (M.R.W.). The authors would like to thank both KACST and NU for their continued
support of this research.
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
■ REFERENCES
(1)
Susumu, K.; Frail, P. R.; Angiolillo, P. J.; Therien, M. J. J. Am. Chem. Soc. 2006, 128,
8380.
(2)
Wu, Y.; Nalluri, S. K. M.; Young, R. M.; Krzyaniak, M. D.; Margulies, E. A.; Stoddart,
J. F.; Wasielewski, M. R. Angew. Chem., Int. Ed. 2015, 54, 11971.
Sun, D.; Rosokha, S. V.; Kochi, J. K. Angew. Chem., Int. Ed. 2005, 44, 5133.
Rosokha, S. V.; Neretin, I. S.; Sun, D.; Kochi, J. K. J. Am. Chem. Soc. 2006, 128, 9394.
Guo, X.; Facchetti, A.; Marks, T. J. Chem. Rev. 2014, 114, 8943.
Wang, C.; Dong, H.; Hu, W.; Liu, Y.; Zhu, D. Chem. Rev. 2012, 112, 2208.
Brédas, J.ꢀL.; Calbert, J. P.; da Silva Filho, D.; Cornil, J. Proc. Natl. Acad. Sci. U.S.A.
2002, 99, 5804.
(3)
(4)
(5)
(6)
(7)
(8)
(9)
Katz, H. E.; Lovinger, A. J.; Johnson, J.; Kloc, C.; Siegrist, T.; Li, W.; Lin, Y. Y.;
Dodabalapur, A. Nature 2000, 404, 478.
He, T.; Stolte, M.; Würthner, F. Adv. Mater. 2013, 25, 6951.
(10) Zheng, Q.; Huang, J.; Sarjeant, A.; Katz, H. E. J. Am. Chem. Soc. 2008, 130, 14410.
19
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 20 of 36
1
2
3
4
5
6
7
8
9
(11) Huang, H.; Zhou, N.; Ortiz, R. P.; Chen, Z.; Loser, S.; Zhang, S.; Guo, X.; Casado, J.;
López Navarrete, J. T.; Yu, X.; Facchetti, A.; Marks, T. J. Adv. Funct. Mater. 2014, 24,
2782.
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
(12) Han, X.; Chang, C.; Yuan, L.; Sun, T.; Sun, J. Adv. Mater. 2007, 19, 1616.
(13) Wang, H.ꢀg.; Yuan, S.; Ma, D.ꢀl.; Huang, X.ꢀl.; Meng, F.ꢀl.; Zhang, X.ꢀb. Adv. Energy
Mater. 2014, 4, 1301651.
(14) Song, Z.; Zhan, H.; Zhou, Y. Angew. Chem., Int. Ed. 2010, 49, 8444.
(15) Häupler, B.; Wild, A.; Schubert, U. S. Adv. Energy Mater. 2015, 5, 1402034.
(16) Bhosale, S. V.; Jani, C. H.; Langford, S. J. Chem. Soc. Rev. 2008, 37, 331.
(17) Suraru, S.ꢀL.; Würthner, F. Angew. Chem., Int. Ed. 2014, 53, 7428.
(18) Sakai, N.; Bhosale, R.; Emery, D.; Mareda, J.; Matile, S. J. Am. Chem. Soc. 2010, 132,
6923.
(19) Zhan, X.; Facchetti, A.; Barlow, S.; Marks, T. J.; Ratner, M. A.; Wasielewski, M. R.;
Marder, S. R. Adv. Mater. 2011, 23, 268.
(20) Míšek, J.; Vargas Jentzsch, A.; Sakurai, S.ꢀi.; Emery, D.; Mareda, J.; Matile, S. Angew.
Chem., Int. Ed. 2010, 49, 7680.
(21) SantosꢀFigueroa, L. E.; Moragues, M. E.; Climent, E.; Agostini, A.; MartinezꢀManez, R.;
Sancenon, F. Chem. Soc. Rev. 2013, 42, 3489.
(22) Dawson, R. E.; Hennig, A.; Weimann, D. P.; Emery, D.; Ravikumar, V.; Montenegro, J.;
Takeuchi, T.; Gabutti, S.; Mayor, M.; Mareda, J.; Schalley, C. A.; Matile, S. Nat. Chem.
2010, 2, 533.
(23) Vargas Jentzsch, A.; Hennig, A.; Mareda, J.; Matile, S. Acc. Chem. Res. 2013, 46, 2791.
20
ACS Paragon Plus Environment

Page 21 of 36
Journal of the American Chemical Society
1
2
3
4
5
6
(24) Zhao, Y.; Domoto, Y.; Orentas, E.; Beuchat, C.; Emery, D.; Mareda, J.; Sakai, N.;
Matile, S. Angew. Chem., Int. Ed. 2013, 52, 9940.
7
8
9
(25) Zhao, Y.; Sakai, N.; Matile, S. Nat. Commun. 2014, 5, 3911.
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
(26) Nalluri, S. K. M.; Berdugo, C.; Javid, N.; Frederix, P. W. J. M.; Ulijn, R. V. Angew.
Chem., Int. Ed. 2014, 53, 5882.
(27) Berdugo, C.; Nalluri, S. K. M.; Javid, N.; Escuder, B.; Miravet, J. F.; Ulijn, R. V. ACS
Appl. Mater. Interfaces 2015.
(28) MasꢀTorrent, M.; Rovira, C. Chem. Rev. 2011, 111, 4833.
(29) Schneebeli, S. T.; Frasconi, M.; Liu, Z.; Wu, Y.; Gardner, D. M.; Strutt, N. L.; Cheng, C.;
Carmieli, R.; Wasielewski, M. R.; Stoddart, J. F. Angew. Chem., Int. Ed. 2013, 52, 13100.
(30) Liu, Z.; Liu, G.; Wu, Y.; Cao, D.; Sun, J.; Schneebeli, S. T.; Nassar, M. S.; Mirkin, C. A.;
Stoddart, J. F. J. Am. Chem. Soc. 2014, 136, 16651.
(31) Liu, Z.; Sun, J.; Zhou, Y.; Zhang, Y.; Wu, Y.; Nalluri, S. K. M.; Wang, Y.; Samanta, A.;
Mirkin, C. A.; Schatz, G. C.; Stoddart, J. F. J. Org. Chem. 2016, 81, 2581.
(32) Mizuno, A.; Shuku, Y.; Suizu, R.; Matsushita, M. M.; Tsuchiizu, M.; Reta Mañeru, D.;
Illas, F.; Robert, V.; Awaga, K. J. Am. Chem. Soc. 2015, 137, 7612.
(33) Whittingham, M. S. Chem. Rev. 2004, 104, 4271.
(34) Armand, M.; Tarascon, J. M. Nature 2008, 451, 652.
(35) Bruce, P. G.; Scrosati, B.; Tarascon, J.ꢀM. Angew. Chem., Int. Ed. 2008, 47, 2930.
(36) Goodenough, J. B.; Kim, Y. Chem. Mater. 2010, 22, 587.
(37) Manthiram, A.; Fu, Y.; Chung, S.ꢀH.; Zu, C.; Su, Y.ꢀS. Chem. Rev. 2014, 114, 11751.
(38) Tarascon, J. M. Phil Trans. R. Soc. A 2010, 368, 3227.
(39) Tobishima, S. i.; Yamaki, J. i.; Yamaji, A. J. Electrochem. Soc. 1984, 131, 57.
21
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 22 of 36
1
2
3
4
5
6
(40) Le Gall, T.; Reiman, K. H.; Grossel, M. C.; Owen, J. R. J. Power Sources 2003, 119–
121, 316.
7
8
9
(41) Novák, P.; Müller, K.; Santhanam, K. S. V.; Haas, O. Chem. Rev. 1997, 97, 207.
(42) Walker, W.; Grugeon, S.; Mentre, O.; Laruelle, S.; Tarascon, J.ꢀM.; Wudl, F. J. Am.
Chem. Soc. 2010, 132, 6517.
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
(43) Armand, M.; Grugeon, S.; Vezin, H.; Laruelle, S.; Ribiere, P.; Poizot, P.; Tarascon, J. M.
Nat. Mater. 2009, 8, 120.
(44) Song, Z.; Zhou, H. Energy & Environmental Science 2013, 6, 2280.
(45) Chen, D.; Avestro, A.ꢀJ.; Chen, Z.; Sun, J.; Wang, S.; Xiao, M.; Erno, Z.; Algaradah, M.
M.; Nassar, M. S.; Amine, K.; Meng, Y.; Stoddart, J. F. Adv. Mater. 2015, 27, 2907.
(46) Nishimura, S.ꢀi.; Kobayashi, G.; Ohoyama, K.; Kanno, R.; Yashima, M.; Yamada, A.
Nat. Mater. 2008, 7, 707.
(47) Pasta, M.; Wessells, C. D.; Liu, N.; Nelson, J.; McDowell, M. T.; Huggins, R. A.; Toney,
M. F.; Cui, Y. Nat. Commun. 2014, 5, Article 3007.
(48) Gawroński, J.; Brzostowska, M.; Gawrońska, K.; Koput, J.; Rychlewska, U.; Skowronek,
P.; Nordén, B. Chem. Eur. J. 2002, 8, 2484.
(49) Shukla, D.; Nelson, S. F.; Freeman, D. C.; Rajeswaran, M.; Ahearn, W. G.; Meyer, D.
M.; Carey, J. T. Chem. Mater. 2008, 20, 7486.
(50) Gawroński, J.; Brzostowska, M.; Kacprzak, K.; Kołbon, H.; Skowronek, P. Chirality
2000, 12, 263.
(51) Since the two PMDI protons in (−)ꢀ1PMDIꢀ2NDIꢀꢁ are homotopic, it is not possible to
1
probe the rate of rotation of the PMDI unit about its C–N•••N–C axis by dynamic H
NMR spectroscopy. Likewise in (−)ꢀ3PMDIꢀꢁ, the six PMDI protons are homotopic and
22
ACS Paragon Plus Environment

Page 23 of 36
Journal of the American Chemical Society
1
2
1
3
4
so once again it is not possible by dynamic H NMR spectroscopy to probe the rate of
5
6
rotation of the PMDI units about their C–N•••N–C axes.
7
8
9
(52) The significant upfield shifts of the resonances for the heterotopic PMDI protons (HC and
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
1
HC’) in the H NMR spectra recorded at different temperatures (298–426 K) in
CD3SOCD3 could be a result of changes in intramolecular shielding effects from
neighboring PMDI and NDI units. .
(53) The kinetic and thermodynamic parameters associated with the rotation of the PMDI
units around the C–N•••N–C axis of the isosceles triangle (−)ꢀ2PMDIꢀ1NDIꢀꢁ were
estimated by the peak separation observed (ꢁνex ~ 11 Hz) for the PMDI protons (HC and
HC’) at low temperatures. The rate constant kc at the coalescence temperature Tc (426 K)
was calculated (Sutherland, I. O. Annu. Rep. NMR Spectrosc. 1971, 4, 71–235)
employing the approximation that kc = π(ꢁνex)/(2)1/2 and found to be ~ 24 s−1. By using
#
the Eyring equation, ꢁGc = RTcln(kch/kbTc), in which R is the gas constant, h is Planck’s
constant, and kb is Boltzmann’s constant, this rate constant (kc) corresponds to a ꢁGc#
value of ~ 23 kcal mol−1. There is slight line broadening of the signals for the NDI
protons (HA/B) upon increasing the temperature.
(54) The shape of the CD signal at 380 nm is unsymmetrical on account of the nearby CD
signal at 360 nm.
(55) The second set of reduction potentials for [RefꢀPMDI] and [(−)ꢀ3PMDIꢀꢁ],
corresponding to the generation of [RefꢀPMDI]2– dianion and [(−)ꢀ3PMDIꢀꢁ]6–
hexaanion, respectively, are too close to that of the solvent (CH2Cl2) and hence the
second redox potentials could not be accessed.
(56) McConnell, H. M. J. Chem. Phys. 1956, 24, 764.
23
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 24 of 36
1
2
3
4
5
6
7
8
9
(57) A current rate of nC refers to achieving full charge or discharge of the battery in 1/n h.
(58) The theoretical capacity of each triangle was calculated by assuming a reversible six
electron process under the battery operation conditions.
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
(59) Gao, J.; Lowe, M. A.; Conte, S.; Burkhardt, S. E.; Abruña, H. D. Chem. Eur. J. 2012, 18,
8521.
(60) Shimizu, A.; Kuramoto, H.; Tsujii, Y.; Nokami, T.; Inatomi, Y.; Hojo, N.; Suzuki, H.;
Yoshida, J.ꢀi. J. Power Sources 2014, 260, 211.
(61) Preliminary experiments carried out on the equilateral triangle [(−)ꢀ3PMDIꢀꢁ] battery
revealed that the symmetric nature of [(−)ꢀ3PMDIꢀꢁ] exhibits, somewhat unexpectedly,
improved battery performance. We are currently exploring why this battery is not
following the trend predicted on the basis of the other three batteries.
24
ACS Paragon Plus Environment

Page 25 of 36
Journal of the American Chemical Society
1
2
3
4
5
Captions of Figures
6
7
8
9
Figure 1. Stereospecific stepwise preparation of chiral isosceles triangles a) (−)ꢀ1PMDIꢀ2NDIꢀꢁ
and b) (−)ꢀ2PMDIꢀ1NDIꢀꢁ, from (RR)ꢀtransꢀ1,2ꢀcyclohexanediamine, naphthalenetetraꢀ
carboxylic dianhydride (NDA) and pyromellitic dianhydride (PMDA). c) Structural formulas of
the monomeric reference compounds (RefꢀNDI and RefꢀPMDI) and the chiral equilateral
triangles [(−)ꢀ3NDIꢀꢁ and (−)ꢀ3PMDIꢀꢁ]. The aromatic NDI and PMDI subunits within the
molecular triangles are shown in blue and magenta, respectively.
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
1
Figure 2. A comparison of the annotated H NMR spectra (500 MHz, CDCl3, 298 K) of the
molecular triangles a) (−)ꢀ3NDIꢀꢁ, b) (−)ꢀ1PMDIꢀ2NDIꢀꢁ, c) (−)ꢀ2PMDIꢀ1NDIꢀꢁ and d) (−)ꢀ
3PMDIꢀꢁ. The signals corresponding to the aromatic protons of the NDI and PMDI subunits,
and those corresponding to the aliphatic protons of the cyclohexano linkers are annotated in blue,
magenta and black, respectively. The unambiguous assignment of the axial and equatorial
protons of the cyclohexano linkers is presented in the Supporting Information.
Figure 3. Variable temperature 1H NMR spectra (600 MHz, CD3SOCD3) in the aromatic region
of the isosceles triangle (−)ꢀ2PMDIꢀ1NDIꢀꢁ. The coalescence temperature of the resonances for
the PMDI protons (HC and HC’) is observed at 426 K.
Figure 4. Solidꢀstate (super)structures of (−)ꢀ1PMDIꢀ2NDIꢀꢁ. (a and b) Plan and sideꢀon views
of the inclusion complex formed between (−)ꢀ1PMDIꢀ2NDIꢀꢁ and two [Br···Br]ꢀbonded 1,2ꢀ
dibromoethane which are depicted in spaceꢀfilling representations. (cꢀe) Three views of the
packing superstructure of (−)ꢀ1PMDIꢀ2NDIꢀꢁ, where the 1,2ꢀdibromoethane molecules and H
atoms are omitted for the sake of clarity. The four nonequivalent units (−)ꢀ1PMDIꢀ2NDIꢀꢁ are
highlighted in red, green, blue and magenta, respectively.
Figure 5. Solidꢀstate (super)structures of (−)ꢀ2PMDIꢀ1NDIꢀꢁ. a) Tubular representation of the
triangular (−)ꢀ2PMDIꢀ1NDIꢀꢁ. b) Top: Two different types of π−π stacking dimers [(P)ꢀ and
(M)ꢀDimers] exhibiting opposite supramolecular helicities: the NDI and PMDI moieties not
involved in π−π stacking are omitted for the sake of clarity. Bottom: The location of DMF
molecules inside the triangles and the π−π stacking of NDI•••NDI as well as PMDI•••NDI. For
the sake of clarity, the (P)ꢀDimer is depicted in blue and the (M)ꢀDimer is depicted in green. (c
and d) Plan and sideꢀon views of the superstructure of both the (P)ꢀ and (M)ꢀDimers. Red
hatched lines indicate the π−π stacking interactions between the aromatic subunits.
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 26 of 36
1
2
3
4
5
6
7
8
9
Figure 6. a) UV/Vis Absorption and b) CD spectra of (−)ꢀ3NDIꢀꢁ (black traces), (−)ꢀ1PMDIꢀ
2NDIꢀꢁ (red traces), (−)ꢀ2PMDIꢀ1NDIꢀꢁ (green traces) and (−)ꢀ3PMDIꢀꢁ (blue traces)
recorded in CH2Cl2 at 298 K. Graphical representations of the DFTꢀcalculated (B3LYP/6ꢀ
31G**) FMOs of the molecular triangles showing the available c) HOMO−1, HOMO, LUMO
and LUMO+1 levels of (−)ꢀ1PMDIꢀ2NDIꢀꢁ and d) HOMO, LUMO and LUMO+1 levels of (−)ꢀ
2PMDIꢀ1NDIꢀꢁ. FMO refers to the frontier molecular orbital, HOMO refers to the highest
occupied molecular orbital and LUMO refers to the lowest unoccupied molecular orbital. Orbital
isosurfaces are illustrated at 0.004 electrons Bohr−3.
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. Solutionꢀstate CVs (0.5 mM in CH2Cl2, 100 mM TBAPF6, 50 mVs−1, 298 K) of Refꢀ
NDI, (−)ꢀ3NDIꢀꢁ, (−)ꢀ1PMDIꢀ2NDIꢀꢁ, (−)ꢀ2PMDIꢀ1NDIꢀꢁ, (−)ꢀ3PMDIꢀꢁ and RefꢀPMDI.
Halfꢀwave peak potentials (E1/2) are shown in mV.
1
Figure 8. a) CWꢀEPR and b) H ENDOR spectra (0.25 mM in CH2Cl2, 298 K) of [RefꢀNDI]•−,
[(−)ꢀ3NDIꢀꢁ]•−, [(−)ꢀ1PMDIꢀ2NDIꢀꢁ]•− and [(−)ꢀ2PMDIꢀ1NDIꢀꢁ]•−, formed by the
monoreduction of their corresponding neutral states by adding 1 equiv of cobaltocene. Overlay
between the experimental spectra (black traces) and their simulated spectra (red traces). Isotropic
nitrogen and proton hyperfine coupling constants were obtained from simulation of the CWꢀEPR
and 1H ENDOR, respectively.
Figure 9. Solidꢀstate CVs and charge/discharge properties of the (−)ꢀ1PMDIꢀ2NDIꢀꢁ and (−)ꢀ
2PMDIꢀ1NDIꢀꢁ batteries. CV profiles of the first, second and fifth cycles recorded at a scan rate
of 0.05 mVs−1 for the a) (−)ꢀ1PMDIꢀ2NDIꢀꢁ battery and d) (−)ꢀ2PMDIꢀ1NDIꢀꢁ battery.
Charge/discharge profiles for the first, fifth and hundredth cycles at rates of b) 0.5C and c) 1.0C
for the (−)ꢀ1PMDIꢀ2NDIꢀꢁ battery. Charge/discharge profiles for the first, fifth and hundredth
cycles at rates of e) 0.5C and f) 1.0C for the (−)ꢀ2PMDIꢀ1NDIꢀꢁ battery.
ACS Paragon Plus Environment

Page 27 of 36
Journal of the American Chemical Society
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
a)
b)
c)
Figure 1
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 28 of 36
1
2
3
4
5
6
7
8
O
O
a)
b)
HA/B
HA
HB
N
N
HDO x
OO
HB
(−)ꢀ3NDIꢀ∆
O
HC
O
N
HA
N
O
O
9
HD/E
HF/G
O
O
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
N
N
O
HC
HE/F
HD
HG
O
(−)ꢀ1PMDIꢀ2NDIꢀ∆
HC
HA/A′/B/B′
HJ/K/L/M/N/O
HQ
HD
HE
HF
HP
HG HH
HI
c)
d)
(−)ꢀ2PMDIꢀ1NDIꢀ∆
HC/C′
HA/B
HQ
HI/J/K/L/M/N/O
HD
HP
HE
HF
HG
HH
(−)ꢀ3PMDIꢀ∆
HA
HB
HC/D/E
HF
δ / ppm
Figure 2
ACS Paragon Plus Environment

Page 29 of 36
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
HA/B HA/B
HC/C’
426 K
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
415 K
404 K
393 K
(−)ꢀ2PMDIꢀ1NDIꢀ∆
382 K
371 K
355 K
333 K
316 K
298 K
δ / ppm
Figure 3
ACS Paragon Plus Environment