Regular Two-Dimensional Arrays of Surface-Mounted Molecular Switches: Switching Monitored by UV-vis and NMR Spectroscopy

Article abstract of DOI:10.1021/jacs.0c01753

Using solid-state 15N NMR spectroscopy, the cis/trans isomerization in a two-dimensional (2-D) array of surface-mounted azobenzene-based switches was detected for the first time. In order to achieve this, a new class of rod-shaped molecular switches, suit

Full text of DOI:10.1021/jacs.0c01753

Subscriber access provided by BIU Pharmacie | Faculté de Pharmacie, Université Paris V
Article
Regular 2-D Arrays of Surface-Mounted Molecular Switches:
Switching Monitored by UV-vis and NMR Spectroscopy
Carina Santos Hurtado, Guillaume Bastien, Milan Mašát, Jakub Radek Sto#ek,
Martin Dracinsky, Igor Ron#evi#, Ivana Císa#ová, Charles T. Rogers, and Ji#í Kaleta
J. Am. Chem. Soc., Just Accepted Manuscript • DOI: 10.1021/jacs.0c01753 • Publication Date (Web): 24 Apr 2020
Downloaded from pubs.acs.org on April 25, 2020
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 32
Journal of the American Chemical Society
1
2
3
4
5
Regular 2-D Arrays of Surface-Mounted Molecular Switches: Switching Monitored by
UV-vis and NMR Spectroscopy
6
7
8
Carina Santos Hurtado,a Guillaume Bastien,a Milan Mašát,a Jakub Radek Štoček,a Martin
Dračínský,a Igor Rončević,a Ivana Císařová,b Charles T. Rogers,c and Jiří Kaletaa*
9
a Institute of Organic Chemistry and Biochemistry of the CAS, Flemingovo nám. 2, 166 10
Prague 6, Czech Republic.
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
b Department of Inorganic Chemistry, Faculty of Science, Charles University in Prague,
Hlavova 2030, 12840 Prague 2, Czech Republic.
c Department of Physics, University of Colorado, Boulder, Colorado, 80309, United States.
ABSTRACT
Using solid-state 15N NMR spectroscopy, the cis/trans isomerization in a 2-D array of
surface-mounted azobenzene-based switches was detected for the first time. In order to
achieve this, a new class of rod-shaped molecular switches, suitable for formation of two-
dimensional regular arrays on large facets of tris(o-phenylenedioxy)cyclotriphosphazene
(TPP) nanocrystals, was synthesized. A mechanochemical approach was used to prepare
corresponding host-guest surface inclusions in a TPP matrix. Comparison of thermal steps in
solution and supramolecular surface inclusions revealed that switching of individual
molecules is not compromised by the close proximity of neighbors.
INTRODUCTION
Self-organization of individual molecular machines like motors, rotors or switches into
regular and well defined two- (2-D)1,2,3,4,5,6,7,8,9 or three-dimensional (3-D)10,11,12,1314,15,16,17,18
arrays is a promising route towards a new generation of smart materials. 2-D assemblies seem
to be particularly interesting because of their possible application in, for example, optics
(OLEDs) or nanoelectronics (memory devices, frequency filters, etc.).19,20 So far, there are
three major pathways leading to such unique systems: (i) self-assembly on metallic surfaces,
(ii) Langmuir-Blodgett technique, and (iii) anchoring of guest molecules on the flat facets of
porous crystalline matrices. Each of these approaches has several advantages, but also a few
drawbacks setting the limits of future applications.
Work presented herein is based on the third approach. The hexagonal form of organic
zeolite-like compound called tris(o-phenylenedioxy)cyctlotriphosphazene (Figure 1) (usually
abbreviated as TPP) has been known to host various organic molecules for
decades.21,22,23,24,25,26 It is an easily accessible white crystalline material which is UV
transparent above ca 280 nm,27 allowing direct measurement of UV-vis spectra of properly
chosen anchored chromophores. Moreover, its perdeuteriated form TPP-d12 allows the study
of the guest molecules by solid-state NMR (deuteriation suppresses the 13C signals of the
13
matrix, simplifying assignment of the guest carbon resonances during C cross-polarization
NMR experiments).5,8,13
Hexagonal TPP has a layered structure where each layer is ~5 Å thick and rotated by
60° towards the neighboring one. These layers contain voids which are superimposed to form
channels with an internal diameter 4.5 - 5 Å. The voids are perpendicular to the layers, are in
triangular arrangement and are located ~11 Å apart (Figure 1).28
- 1 -
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 2 of 32
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
Figure 1. Structure of the TPP molecule (A)
and a top view on the hexagonal form of the
TPP matrix (B). Atom labeling: phosphorus =
orange, oxygen = red, carbon = gray, nitrogen
= blue, and hydrogen = white.
We utilized all these structural properties, took an inspiration in our previous systems
and designed a new class of azobenzene-based molecular switches 1 and 2 that could form
surface inclusions in TPP (Chart 1). In these inclusions, the guest molecules remain only on
the surface of the matrix and are unable to fully penetrate inside the TPP nanocrystals. Letters
“t” and “c”, which are sometimes associated with the numbers of the compounds, refer to
trans and cis isomers.
Compounds 1 and 2 share several structural features: an azobenzene-based molecular
switch (a) that is connected via a linker (b) with a triptycene stopper (c) which is attached to a
linear shaft (d).
(a) The azobenzene-based switch forms the functional heart of these devices.
Azobenzenes are well-known to undergo light-initiated cis/trans isomerization, which can be
easily monitored in solution by for example UV-vis or NMR spectroscopy. In our structures,
the 15N labels facilitate NMR detection, allowing us to follow the switching in a 2-D film
which has not been reported yet. Bulky substituents (chlorine atoms and tert-butyl groups)
increase the van der Waals diameter of the switches (7.7 Å for 1 and 11.2 Å for 2), thus
preventing insertion into the channel in the opposite direction.
(b) A butadiyne linker separates the switches from the triptycene unit. Although the
installation of two conjugated ethynyl units instead of one triple bond prolongs the synthesis,
in the upshot it significantly simplifies the connection of expensive isotopically labeled
azobenzenes with the lower part of the molecule. In general, butadiynes attached to triptycene
bridgehead carbon atoms are much more reactive towards palladium catalyzed coupling
reactions compared to simple ethynyls, because the additional triple bonds effectively
separate them from sterically demanding triptycene units.
- 2 -
ACS Paragon Plus Environment

Page 3 of 32
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
(c) A triptycene unit whose equatorial van der Waals diameter (~11.6 Å) is much
larger than the internal diameter of the TPP channel (4.5 - 5 Å) acts as a stopper that prevents
complete insertion of the guest molecule into the matrix. Moreover, a 9,10-disubstituted
triptycene was chosen due to its high chemical stability, symmetry and easy accessibility in
multigram quantities.8
9
(d) The long shaft serves two major purposes: (i) it firmly anchors the whole
molecular device inside the TPP channel due to its high affinity towards the TPP matrix, and
(ii) it carries two 13C NMR probes (tert-Bu group on the bottom and the bicyclo[1.1.1]pentyl
cage in its central part) whose characteristic carbon atom resonances help to evaluate the
depth of insertion into the matrix.
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
Cl
Cl
N
N
a
15N
15N
b
c
d
1t
2t
Chart 1. Molecular Switches 1t and 2t
Composed of Azobenzene-based Switch
(a), Linker (b), Triptycene Stopper (c),
and Shaft (d).
RESULTS
We describe the synthesis of two molecular switches 1 and 2, single crystal X-ray
structures of six intermediates, formation of surface inclusions with TPP, and characterization
of these materials by solid-state NMR and powder X-ray diffraction (PXRD). Switching of
1
neat structures in solution was independently studied by H NMR and UV-vis spectroscopy,
switching in 2-D films was then monitored by UV-vis. The thermal cis-to-trans step in the
light-induced switching cycle was characterized both in solution and in surface inclusion by
determining enthalpy (ΔH‡) and Gibbs free energy (ΔG‡) of activation. We also provide first
evidence of azobenzene switching in organized 2-D arrays by solid-state 15N NMR
spectroscopy.
- 3 -
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 4 of 32
1
2
3
4
5
6
7
8
Synthesis. Isotopically labeled azobenzenes 3 and 4, that form functional hearts of molecular
machines 1 and 2 were synthesized in three and five steps from commercially available
starting materials (Scheme 1).
Iodiniation of 15N enriched aniline 5 using KI in presence of H2SO4 and H2O2
provided crucial intermediate 6 in high yield.29
9
Second reaction partners, substituted nitrosobenzenes 7 and 8, were prepared using
two different approaches. Direct oxidation of 9 using Oxone (KHSO5) afforded 7 in one step
and acceptable yield.30 Lithiation of 10 using t-BuLi, followed by a reaction with TMSCl
resulted in 11 in an excellent yield. A subsequent nearly quantitative substitution of the TMS
group by NO was performed using NOBF4 in acetonitrile at room temperature and the nitroso
derivative 8 was isolated as a green crystalline solid (Scheme 1).31
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
An acid catalyzed condensation of aniline 6 with nitrosoarenes 7 and 8 yielded
expected azobenzenes 3 and 4 in ca 50 - 80% yields (Scheme 1).
Oxone
Cl
Cl
Cl
Cl
(KHSO5)
Cl
Cl
H2O
CH2Cl2
25 °C
NH2
NO
CH3COOH
N
7
, 30%
15N
9
EtOH
70 °C
15NH2
KI
15NH2
I
H2SO4, H2O2
3
, 52%
CH3OH
25 °C
I
6
, 75%
5
CH3COOH
EtOH
70 °C
N
15N
1. t-BuLi
2. TMSCl
NOBF4
I
Et2O
-78 25 °C
CH3CN
25 °C
Br
TMS
NO
4
, 75%
10
11
8
, 96%
, 90%
Scheme 1. Synthesis of substituted azobenzenes 3 and 4.
Final assembly of molecular switches 1t and 2t is a straightforward and fully modular
approach (Scheme 2).4,8 Both iodides 3 and 4 were cross-coupled with alkyne 128 and silyl
derivatives 13 and 14 were obtained in moderate yields (44 and 59%). Subsequent removal of
the TMS protective group was performed using TBAF in THF at room temperature and
resulted in formation of free acetylenes 15 and 16 in moderate yields again. Sonogashira
cross-coupling between alkynes 15 and 16, and iodide 178 afforded still reasonably well
soluble molecular switches 1t and 2t in nearly quantitative isolated yields (Scheme 2).
- 4 -
ACS Paragon Plus Environment

Page 5 of 32
Journal of the American Chemical Society
1
2
3
4
5
6
I
I
Cl
Cl
N
N
15N
15N
7
8
9
Cl
Cl
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
15N
15N
4
3
17
Pd(PPh3)4, CuI
Et3N
17
Pd(PPh3)4, CuI
Et3N
Pd(PPh3)4, CuI
Et3N
Pd(PPh3)4, CuI
Et3N
THF
25 °C
THF
25 °C
THF
25 °C
THF
25 °C
R
R
TMS
13 R
:
14 R
: = TMS, 59%
= TMS, 44%
1. TBAF
2. H2O
1. TBAF
2. H2O
THF, 25 °C
12
15 R
:
16 R
: = H, 100%
= H,
67%
THF, 25 °C
1t
2t
, 90%
, 96%
Scheme 2. Final assembly of molecular switches 1t and 2t.
Preparation and Characterization of Surface Inclusions.
Both switches 1t and 2t were applied into the benzene-free hexagonal TPP-d12 in two
molar concentrations: 5 and 10%. The first concentration corresponds to ca 50% surface
saturation, whereas 10mol % represents the material with all pores occupied by the guest
molecules.10 The surface inclusion compounds are then called as X%Y@TPP-d12, where X
represents the molar percentage of switch Y (Y = 1 or 2). Preparation of such inclusions is a
two-step process that involves (i) a mechanochemical reaction (ball-milling of both
components at defined time), which is followed by (ii) annealing at 70 °C for 24 h.
Solid-state NMR.
Formation of surface inclusions was monitored by solid-state 13C cross-polarization
magic angle spinning nuclear magnetic resonance (13C CP MAS NMR) spectroscopy. Nearly
complete assignment of almost all 13C resonances in the complex structures of both 1t and 2t
in solution (Figure S1 and Figure S2 in the Supporting Information) significantly simplified
the analysis of all surface-bound 2-D materials.
The 31P solid-state NMR experiments provided additional data regarding formation of
surface inclusions (CP MAS technique) and helped to determine the crystal phase of the TPP
matrix using a single-pulse experiment (SPE).
Compound 2t and corresponding surface inclusions were chosen as a representative
example to describe the interpretation of the NMR spectra. The C NMR spectrum of 2t in
THF-d8 solution contains peaks in three major regions (Figure 2A): (i) aliphatic part
containing several characteristic resonances between 20 - 60 ppm, namely nine methyl groups
belonging to two inequivalent t-Bu units with chemical shift at 31.4 ppm, two BCP
bridgehead carbon atoms at 35.0 ppm, two triptycene bridgehead carbon atoms at 53.9 ppm,
and three equivalent methylene carbon atoms representing BCP cage at 59.5 ppm. (ii) Triple
13
- 5 -
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 6 of 32
1
2
3
4
5
6
7
8
9
bond region spanning from ca 75 to 95 ppm. Particularly two peaks from this region with
chemical shift 88.8 and 91.7 ppm deserve special attention. They belong to the triple bond
carbon atoms located next to the BCP bridgeheads and their absence in CP MAS spectra of
the TPP samples is indicating a successful formation of the inclusion.10 (iii) Aromatic region
covering a broad part of the spectrum from ca 110 ppm to 160 ppm contains nearly two
dozens of carbon resonances (Figure 2A).
In the 13C CP MAS NMR spectrum of neat 2t (Figure 2B), several carbon atom
resonances are doubled (BCP methylene bridges at 60.3 and 58.6 ppm, triple bond signals at
89.8 and 88.0 ppm) or are present as broad peaks with distinct shoulders (for example methyl
carbon atoms coming from t-Bu units at ~31 ppm). This indicates the presence of two
crystallographically different environments in the unit cell.
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
The 13C CP MAS NMR spectrum of 5%2t@TPP-d12 (Figure 2C) contains three
characteristic and strong resonances due to the TPP in the aromatic region (~145, ~125, and
~110 ppm). Aromatic carbon atoms of the guest are mostly overlapped with these signals and
thus could not be used for evaluation of this inclusion. In contrast, all aliphatic carbon atoms
are shielded and their signals are therefore slightly upfield shifted (58.9, 53.1, 34.5, and 31.0)
compared to the spectrum of 2t in THF-d8 solution (59.5, 53.9, 35.0, and 31.4). Most of the
C≡C signals give rise to one broad peak with a chemical shift of ~77.2 ppm while the
resonances belonging to carbon atoms next to the BCP bridgeheads are not visible at all.10
The 13C CP MAS NMR spectrum of saturated surface inclusion 10%2t@TPP-d12
(Figure 2D) is nearly identical with 5%2t@TPP-d12. All trends remain preserved and only the
signal-to-noise ratio is improved due to higher guest loading.
The two peaks marked by an asterisk (~20 and ~5 ppm) in Figure 2 (panel C and D)
are spinning sidebands of the two most intense TPP signals (~145 and ~125 ppm). The third
one (~-5 ppm) as well as the corresponding mirror peaks are present in the region that is not
shown in this figure.
13
All C NMR spectra of solid samples also contain signals of residual THF that was
probably originally built into the crystal structure of the guest compound.
- 6 -
ACS Paragon Plus Environment

Page 7 of 32
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
Figure 2. 13C NMR of 2t in THF-d8
solution (A), and solid-state 13C CP MAS
NMR spectra of neat 2t (B), 10%2t@TPP-
d12 (C), and 10%2t@TPP-d12 (D).
Asterisks indicate spinning sidebands.
Both the 31P CP MAS and 31P SPE NMR spectra of 5mol% and 10%2t@TPP-d12
contain one singlet with a chemical shift of ~33 ppm (Figure 3A-D). The CP MAS
experiments confirmed a close proximity of guest molecules carrying hydrogen atoms to the
- 7 -
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 8 of 32
1
2
3
4
5
phosphorus nuclei in the deuteriated TPP matrix, while the presence of one singlet in the SPE
spectra proves its hexagonal form.
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
31
31
Figure 3. P CP MAS NMR: 5%2t@TPP-d12 (A), and 10%2@TPP-d12 (C). P SPE
NMR: 5%2t@TPP-d12 (B), and 10%2t@TPP-d12 (D). Asterisks indicate spinning
sidebands.
Transmission Electron Microscopy (TEM).
One of the resulting samples, 10%2@TPP-d12, was chosen as a representative
example, and subsequently analyzed by transmission electron microscopy (TEM). The sample
consisted of aggregates of flat, disc-like particles with a ~20 nm thickness and ~50 nm
diameter (Figure 4).
- 8 -
ACS Paragon Plus Environment

Page 9 of 32
Journal of the American Chemical Society
1
2
3
4
Figure 4. TEM of 10%2@TPP-d12.
5
6
Powder X-Ray Diffraction (PXRD).
All four samples have wide and mostly “Lorentzian” shaped diffraction lines that are
common for surface inclusions. The dominant peaks are hexagonal, and so these materials are
characterized by two equivalent in-plane lattice constants a (Å) and a hexagonal layer spacing
c (Å). Both of these parameters including uncertainties are listed in Table 1. Three inclusion
compounds (5%1t@TPP-d12, 10%1t@TPP-d12, and 5%2t@TPP-d12) also contain traces of a
monoclinic TPP polymorph. In contrast, the 10%2t@TPP-d12 appears to be essentially free of
this monoclinic form.
The representative diffractogram of surface inclusion 10%2@TPP-d12 is shown in
Figure 5. The remaining three diffractograms are listed in the Supporting Information (Figure
S15).
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 5. X-ray powder pattern for
10%2t@TPP-d12. The solid black line shows
diffraction data and the red line is the fitting
result assuming hexagonal peak positions and
a background.
Table 1. Hexagonal Lattice Parameters for Surface Inclusions.a
Cmpd.
1t
Inclusion
a (Å)
c (Å)
5%1t@TPP-d12
10%1t@TPP-d12
5%2t@TPP-d12
10%2t@TPP-d12
11.719 ± 0.003
11.740 ± 0.004
11.60 ± 0.01
11.653 ± 0.005
10.107 ± 0.007
10.141 ± 0.004
10.108 ± 0.004
10.100 ± 0.003
2t
a Lattice parameters for hexagonal guest-free TPP: a = 11.454(5) Å and c = 10.160(5) Å.28
Switching in Solution. 1. 1H NMR was used to both quantitatively evaluate the trans/cis ratio
of both isomers of 1 and 2 at the photostationary state, and to determine the basic kinetic
parameters of the thermal cis-to-trans reaction.
All signals of hydrogen atoms in the azobenzene moieties of both 1 and 2 were fully
assigned and used to demonstrate light-initiated switching (Figure 6 and Figure S9).
Irradiation (max = 254 ± 5 nm) of the NMR tube with a THF-d8 solution of 2t (which is used
as an illustrative example) at 30 °C for 15 min leads to the formation of an additional set of
upfield shifted and orange marked signals of hydrogen atoms Ha - Hd that correspond to 2c
(Figure 6B). The 2c/2t ratio at photostationary state, which was obtained by integration of
corresponding hydrogen atom signals, is equal to 4:1 (this ratio is higher comparing to 2:1
obtained for the 1c/1t). Subsequent thermal relaxation that involved keeping the sample in the
- 9 -
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 10 of 32
1
2
3
4
5
dark at 30 °C for 70 h leads to complete disappearance of 2c and only signals corresponding
to starting 2t were present in the spectrum again (Figure 6C).
6
7
Hd
8
9
Hc
Hc
h
T
N
N
15N
15N
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
Hd
Hb
Ha
Hb
Ha
R
R
2c
2t
1
Figure 6. Aromatic part of H NMR spectra
of 2 in THF-d8 at 30 °C before irradiation
(A), at the photostationary state (B), and
after thermal relaxation (C).
2. UV-vis. The thermal step of the interconversion of unstable cis- isomers back to trans-
conformers was independently studied by the UV-vis absorption spectroscopy. The absorption
spectrum of 1t in THF was characterized by four relatively sharp peaks at 285, 300, 318, 332
nm, and a broad one at ~360 nm (Figure 7A, red line). Irradiation (max = 254 ± 5 nm) of this
- 10 -
ACS Paragon Plus Environment

Page 11 of 32
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
solution almost erased the broad peak at ~360 nm and gave rise to a new broad absorption
band at ~460 nm, while also slightly reducing the intensity of the peak at ~285 nm (Figure
7A, blue line). Positions of the remaining absorption peaks characterizing 1c are identical
with 1t. Subsequent thermal isomerization of 1c back to 1t was achieved by keeping the
sample at elevated temperatures for an equivalent of five t1/2 (eg. 9 hours at 40°C, then
overnight at room temperature). The presence of an isosbestic point at 408 nm suggests that
the photochemical isomerization is a clean process (Figure 7B).
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. UV-vis absorption spectra of 1 in
THF at 20 °C before (red) and after (blue)
irradiation (A), and the progress of the
thermal conversion of 1c back to 1t (B).
All UV-vis characteristics of 2 are nearly identical with those for 1 and corresponding
absorption spectra are in the Supporting Information (Figure S10).
Switching in 2-D Film. 1. UV-vis. Not surprisingly, the UV-vis spectra of all four surface
inclusions (5%1@TPP-d12, 10%1@TPP-d12, 5%2@TPP-d12, and 10%2@TPP-d12) suspended
in water are almost the same.
The UV-vis spectrum of the azobenzene moiety in 5%1t@TPP-d12, which was chosen
as a representative example, consists of a broad absorption peak centered at ~375 nm (Figure
8A, blue line). Irradiation of this suspension (max = 254 ± 5 nm) gave a photostationary state
that is characterized by a new broad peak at ~470 nm and the intensity of the original peak at
~375 nm was noticeably reduced (Figure 8A, red line). Isosbestic point is present at ~405 nm.
The progress of the thermal interconversion of 5%1c@TPP-d12 back to 5%1t@TPP-
d12 is depicted on Figure 8B (measured at 40 °C).
- 11 -
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 12 of 32
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
Figure 8. UV-vis spectra of 5%1@TPP-d12
suspended in water before (red) and after
(blue) irradiation (A), and the progress of
thermal
cis/trans
interconversion
(B)
measured at 40 °C.
Since the thermal cis/trans isomerization is the rate determining step of azobenzene
1
switching, the time-dependent H NMR and UV-vis absorption spectroscopy at defined
temperatures (25, 30, 35, 40, 45, 50, 55, 60, and 65 °C) were used to study this process in
solution (Figure 9A). The kinetics in 2-D films on the TPP facets were studied on water
suspended samples using only UV-vis spectroscopy (Figure 9B).
- 12 -
ACS Paragon Plus Environment

Page 13 of 32
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
Figure 9. Eyring plots of the thermal
interconversion of 1c to 1t in: (A) THF
solution determined by UV-vis and H NMR,
and (B) surface inclusions 5%1@TPP-d12, and
10%1@TPP-d12 measured by UV-vis.
1
Basic kinetic parameters characterizing thermal cis- to trans- isomerization for both
switches in solution and 2-D films, namely enthalpy (ΔH‡) and Gibbs free energy (ΔG‡) of
activation, together with the half-life at 20 °C, and a rate constant at the same temperature that
were calculated from the Eyring plots are listed in Table 2.
Experimentally determined values for both switches clearly show that the Gibbs free
energies of activation are nearly identical (~102 ± 2 kJ/mol) in all cases. The enthalpy of
activation is then roughly 100 ± 10 kJ/mol and the calculated half-life at 20 °C is ~50 h.
Table 2. Kinetic parameters for neat 1 and 2 in THF solution, and corresponding surface
inclusions 5%1@TPP-d12, 10%1@TPP-d12, 5%2@TPP-d12, and 10%2@TPP-d12.
Kinetic Parameters
∆G‡
∆H‡
t1/2
k0
(20 °C)
Cmpd.
Method
(kJ/mol)
(kJ/mol)
(20 °C)
UV
3 × 10-6 ± 2 ×
10-6 s-1
103 ± 2
103 ± 2
103 ± 2
101 ± 2
102 ± 2
101 ± 2
102 ± 2
101 ± 2
110 ± 10
90 ± 10
70 ± 30 h
70 ± 30 h
70 ± 30 h
30 ± 15 h
50 ± 20 h
30 ± 15 h
50 ± 20 h
30 ± 15 h
(THF-d8)
1H NMR
(THF-d8)
1
3 × 10-6 ± 2 ×
10-6 s-1
3 × 10-6 ± 2 ×
10-6 s-1
5%1@TPP-d12
10%1@TPP-d12
110 ± 10
100 ± 10
90 ± 10
UV
(susp. in H2O)
6 × 10-6 ± 4 ×
10-6 s-1
UV
4 × 10-6 ± 2 ×
10-6 s-1
(THF-d8)
1H NMR
(THF-d8)
2
6 × 10-6 ± 4 ×
10-6 s-1
80 ± 20
4 × 10-6 ± 2 ×
10-6 s-1
5%2@TPP-d12
10%2@TPP-d12
110 ± 10
100 ± 10
UV
(susp. in H2O)
6 × 10-6 ± 4 ×
10-6 s-1
2. 15N NMR. Selective installation of 15N into the azobenzene moiety in 1 and 2
significantly simplifies its detection in both samples in solution, and particularly in highly
diluted surface inclusions that were analyzed by solid-state NMR technique. The 15N NMR of
2 in THF-d8 reveals two signals: a strong one at 501 ppm that corresponds to 2t and a much
weaker one at 525 ppm belonging to 2c (ca 6% according to integration) (Figure 10A).
- 13 -
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 14 of 32
1
2
3
4
5
6
7
8
9
Irradiation of the sample (max = 254 ± 5 nm) changed the population of both species and 2c
became dominant at the photostationary state (Figure 10B). The 2c/2t ratio was determined as
ca 3:1 based on integration.
The solid-state 15N NMR of the neat 2 revealed that the crystalline sample consists
exclusively of the trans- isomer, as documented by the broad peak at 497 ppm (Figure 10C).
Irradiation (max = 254 ± 5 nm) of this solid sample did not cause any changes in the 15N
NMR 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
The 15N ssNMR spectrum of 10%2@TPP-d12 is characterized by the nitrogen atom
signal at 504 ppm that corresponds to the trans- azobenzene moiety (Figure 10D). Irradiation
(max = 254 ± 5 nm) of this solid material triggered the trans/cis isomerization of azobenzene
units anchored on the surface of TPP nanocrystals. Formation of 2c in the 2-D film was
proven by the appearance of the resonance at 520 ppm (Figure 10E). This peak is slightly
upfield shifted in comparison to 2c in solution.
The N nitrogen atom signals and corresponding switching in both 5 mol % surface
inclusions (5%1@TPP-d12 and 5%2@TPP-d12) were not unequivocally detected due to a very
low signal-to-noise ratio.
15
- 14 -
ACS Paragon Plus Environment

Page 15 of 32
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
- 15 -
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 16 of 32
1
2
3
4
5
6
7
8
9
Figure 10. Solution 15N NMR of 2 in a
THF-d8 solution at 10 °C: before irradiation
(A), and at photostationary state (B). Solid-
state 15N NMR: neat 2 (C), 10%2@TPP-d12
before irradiation (D), and 10%2@TPP-d12
after irradiation (E).
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
Calculations.
Computations were introduced to help to predict the NMR chemical shifts in solid
materials and to provide an independent view on the cis/trans azobenzene isomerization on
the facets of TPP nanocrystals. To make these complex calculations feasible, two major
simplifications were introduced: (i) all guest molecules are inserted in the TPP matrix at the
same depth, and (ii) only the upper parts of molecular switches were taken into account (Chart
1, sections a and b) and the whole lower segments starting with the triptycene units were
replaced by the methyl groups.
(a) Isomerization Energies. For both studied compounds, the calculated isomerization
energies (ΔEtrans – ΔEcis) are slightly lower in periodic arrays than for isolated molecules
(Table 3). In case of compound 1, this packing effect (ΔEperiodic – ΔEisolated) is very small (-
0.33 kJ/mol), which is unsurprising as the shortest distance between atoms in neighboring unit
cells dmin is over 5 Å, translating to very weak array interactions. For the bulkier compound 2,
the difference in isomerization energies is larger and depends on the array geometry. In a
straight array of 2, where the shafts are parallel to the z- axis and the dmin is relatively small
(2.8 and 3.9 Å for trans- and cis- isomers, respectively), the packing effect is larger at -1.59
kJ/mol.
Both 2c@TPP and 2t@TPP appear to be stabilized by tilting (~20○ relative to the z-
axis, shown in Figure 11). In a tilted array, the packing effect is slightly larger (-2.38 kJ/mol
relative to isolated molecules), and the interatomic distances between neighbors are about 0.4
Å longer than in the straight arrays. No tilting was observed in case of 1@TPP.
Table 3. Isomerization energies (ΔE, kJ/mol) of isolated and periodic 1@TPP and 2@TPP,
the shortest interatomic distances between neighbors (dmin, Å) and energies of cis- and trans-
isomers in tilted arrays relative to straight ones (Erel, kJ/mol).
Cmpd.
1@TPP
2@TPPa
2@TPPb
ΔEisolated ΔEperiodic dmin,trans dmin,cis Erel,trans Erel,cis
47.32
42.55
42.55
46.99
40.96
40.17
6.6
2.8
3.2
5.1
3.9
4.3
-
-
0.00
-1.05
0.00
-1.84
a Straight and b tilted array (~20○ relative to the z-axis).
- 16 -
ACS Paragon Plus Environment

Page 17 of 32
Journal of the American Chemical Society
1
2
3
4
5
6
Figure 11. Space filling model showing tilted array
of 10%2c@TPP. The triptycene stopper (Chart 1) is
replaced by a methyl group and the TPP matrix was
omitted for clarity.
7
8
9
15
(b) Calculation of the N NMR Shift. Since the NMR chemical shifts describe the local
magnetic fields around active nuclei, periodic systems can be, in principle, approximated by
clusters. The NMR shifts of isomers of 1 and 2 were studied by cutting out a heptamer from
the periodic structure and modeling the shifts of the central unit (details in the Experimental
section) at the LC-TPSS/cc-PVTZ level. The agreement between computationally predicted
nitrogen NMR shifts and experimental results is reasonable (Table 4), especially considering
that the FHMW for experimental peaks in solid-state is about 4 ppm. In case of 1t, there is no
significant difference in shifts when moving from solution to solid-state which is reproduced
in calculations. For 2t, both experiments and calculations show a 2 ppm upfield shift when
moving from solution to the solid state. In case of 2c, the solution-state results are in good
agreement (δexp = 525 ppm; δcalc,sol = 523 ppm), but there is a discrepancy in the solid-state
(δexp = 520 ppm; δcalc = 525 ppm). However, it is worth noting that this result for a tilted array
of 2c is in better agreement than both the one for a straight array (526 ppm) and the one for an
isolated 2c molecule in a vacuum (528 ppm).
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 4 A comparison of experimental (δexp) and LC-TPSS predicted NMR shifts (ppm) for
isomers of 1 and 2 in arrays (δcalc,2-D) or solution (δcalc,sol). Tilted arrays are assumed for 2.
Solid-State
Solution (THF)
Cmpd.
δexp
δcalc
539
513
525
505
δexp
535
516
525
501
δcalc
535
512
523
503
1c
1t
2c
2t
-
516
520
503
DISCUSSION
Synthesis of Molecular Switches. The final assembly of both molecular switches
from substituted azobenzenes and previously developed triptycene- and BCP-based precursors
is a straightforward and fully modular process (Scheme 2) which was developed in our
laboratories earlier and thus does not require any additional comments.
15
In contrast, preparation of selectively N-labeled azobenzenes was accompanied by
many unexpected obstacles (Scheme 1). In general, the condensation of anilines with
nitrosoarenes, which leads to asymmetrically substituted azobenzenes, is not a high-yield
reaction mostly due to the gradual degradation of nitroso compounds and formation of many
undesired side-products. Therefore, the relatively high purchase price of the key building
block, aniline-15N (5), determined its role as the more stable nucleophilic partner during this
transformation. Its selective iodination, which required careful monitoring due to successive
formation of undesired 2,4-diiodoaniline-15N, gave desired 4-iodoaniline-15N (6).
Oxidation of 2,3-dichloroaniline (9) resulted in a complex mixture of compounds,
from which the expected nitroso derivative 7 was isolated using column chromatography in a
low, but still acceptable yield (Scheme 1). Analogous oxidation of the 2,5-di-tert-butylaniline
proceeded in the same manner, but all attempts to isolate the clean nitroso compound 8 from
the crude mixture failed. Derivative 8 was thus prepared from 1-bromo-3,5-di-tert-
butylbenezene (10) in two steps. Surprisingly, both the attempted conversion of 10 to
Grignard reagent using activated magnesium and lithiation by n-BuLi failed and only starting
- 17 -
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 18 of 32
1
2
3
4
5
6
7
8
9
10 was nearly quantitatively recovered. Because of this, the much more reactive t-BuLi was
used to prepare the silyl derivative 11, which was almost quantitatively converted to a
surprisingly stable nitroso derivative 8. This compound forms turquoise crystals (it is one of
the few examples of sterically unprotected nitroso compound that is monomeric also in the
crystalline state - see the Supporting Information for more details) that could be stored for
several weeks at room temperature without visible decomposition.
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
All attempts to use the originally intended pentamethylphenyl instead of later used
3,5-di-tert-butylphenyl as a bulky nonpolar unit in the azobenzene moiety (Chart 1) failed.
Although the corresponding nitrosopentamethylbenzene was finally obtained in an acceptable
yield, the successive reaction with 4-iodoaniline-15N (6) resulted in its complete degradation
1
(as determined by H NMR spectra), and only starting 6 was recovered nearly quantitatively
from the crude reaction mixture (for more details see the Supporting Information).
2-D Arrays of Molecular Switches. Although the employed analytical techniques are
not capable of providing a direct visualization of obtained 2-D assemblies, they do at least
offer a series of mutually complementary data that can help to sketch the whole picture.
NMR Spectroscopy can generally provide crucial information regarding the nature of
the inclusion; it can tell us something about the depth of insertion of the rod-like switches into
the TPP matrix and can sometimes also help to determine the direction in which the guest
molecules entered the TPP channels. And so why do we believe that our systems are surface
inclusions? (i) The TPP is fully deuteriated and thus invisible when using a cross-polarization
NMR technique that is based on polarization transfer (through space) from the hydrogen
atoms to detected nuclei. However, all our 13C and 31P CP MAS NMR spectra contain strong
and characteristic peak(s) of TPP, suggesting a close proximity of protiated guest molecules
to the deuteriated matrix (Figure 2C and 2D). This is also consistent with the fact that (ii)
several characteristic resonances of carbon atoms located in the long shaft (Chart 1) of
molecular switches are upfield shifted due to shielding caused by the insertion into the TPP
matrix. (iii) The 31P SPE NMR spectra clearly show an exclusive presence of hexagonal TPP
(Figure 3B and 3D). Guest-free hexagonal TPP is metastable and easily undergoes
isomerization into a more stable monoclinic polymorph that does not have the characteristic
channels anymore. Insertion of guest molecules into the empty channels stabilizes the
hexagonal phase and stops this undesired phase transition. (iv) Not only the upfield shift of
some characteristic carbon atom resonances, but also their absence can indicate formation of
inclusions. The triple bond carbon atoms with chemical shift ~90 ppm, which are located next
to the BCP bridgeheads are clearly visible and well resolved in the solution spectra (Figure
2A), give rise to well defined peaks in the solid state (Figure 2B), but are absent in surface
inclusions due to insufficient polarization transfer (Figure 2C and 2D). This phenomenon,
which was already mentioned several times earlier,4,8 has been extensively studied only
recently.10
Powder X-ray Diffraction cannot tell us much about the guest molecules, but provides
essential information regarding TPP matrix. Namely, it gives proof (independent from NMR)
that all four surface inclusions consist almost exclusively of hexagonal TPP. Only traces of
the monoclinic polymorph were detected in three samples (5%1@TPP-d12, 10%1@TPP-d12,
and 5%2@TPP-d12).
It is known that the insertion of guest molecules into empty TPP channels is usually
accompanied by expansion of the matrix that is reflected in the in-plane lattice parameter a.
Both molecular switches 1 and 2 in surface inclusions cause expansion of the lattice by 1 - 2%
(Table 1), which is fully consistent with previous observations.4,8,10 Uneven expansion of
several TPP layers close to the surface compared to the rest of the material causes surface
tension which leads to fragmentation of large flat TPP domains. This can also explain
relatively high uniformity of TPP nanocrystals detected by the TEM (Figure 4).
- 18 -
ACS Paragon Plus Environment

Page 19 of 32
Journal of the American Chemical Society
1
2
3
4
5
6
Transition Electron Microscopy revealed that the sample consists of discoidal particles
with ~50 nm in diameter. It means that each flat facet of an ideal defect-free TPP nanocrystal
can accommodate up to ~1100 molecular switches. Please see the Supporting Information of
the reference 4 for detailed information regarding this calculation.
7
Switches in 2-D Assemblies - Do They Maintain Their Functionality? To answer
this fundamental question, basic kinetic parameters like enthalpy (ΔH‡) and Gibbs free energy
(ΔG‡) of activation, together with the half-life at 20 °C were determined for comparison also
for both molecular switches 1 and 2 in solution by 1H NMR and UV-vis spectroscopy (Table
2). The 2-D arrays were studied in water suspensions using UV-vis technique. Comparison of
ΔG‡ and ΔH‡ values shows that the thermal cis/trans- isomerization of anchored switches is a
function of neither density of surface coverage nor presence of bulky substituents (chlorine
atoms or t-Bu groups) on azobenzene units. It means that even switches that are parts of
densely packed 2-D films can operate as freely as the neat molecules in solution.
The computations, which shed a better light on behavior of molecular switches in 2-D
systems, are in good agreement with experimental data. The periodic boundary condition
calculations predict an almost negligible (< 0.5 kJ/mol) effect of packing on the isomerization
energy for 1@TPP, and a somewhat stronger one (~2 kJ/mol) for 2@TPP. In both cases, the
cis- form is more stabilized than the trans-, resulting in the decrease of the isomerization
energy. Since the isomerization transition structures resemble the cis- form32,33 it can be
expected that the array interactions will affect them similarly, leading to no significant
changes in the isomerization rate. In case of 2@TPP, a small tilt relative to the z-axis was
found to stabilize arrays of both the cis- and the trans- isomers. This effect was also quite
small (<2 kcal/mol), suggesting that the shafts can wobble around the normal to the TTP
surface at room temperature.
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
The principal difference between the periodic (TPSS) and heptamer (LC-TPSS)
calculations (where the heptamer represents a small fragment of the 2-D array consisting of
one central molecular switch surrounded by six neighbors) is better treatment of long-range
interactions in the latter level of theory. Both methods give the same trends for packing
effects (Table 5), with their absolute values being ~30% larger at the LC-TPSS level. This
result suggests that the relatively inexpensive TPSS/pob-TZVP method is a good starting
point for the modelling of periodic organic systems.
Overall, the calculations suggest that no major changes in the isomerization energy
occur when inclusions of 1 and 2 are prepared.
Table 5. The differences in isomerization energies (kJ/mol) between the isolated and periodic
structures of 1 and 2, as calculated by the TPSS and LC-TPSS functionals.
ΔEisolated – ΔEarray
TPSS/pob-TZVP LC-TPSS/cc-PVTZ
Cmpd.
1
2
-0.33
-2.38
-0.42
-3.22
15N NMR Spectroscopy. Since chemical shifts of the azobenzene nitrogen atoms are
dependent on the N=N double-bond configuration, the NMR spectroscopy was used for the
first time to detect the cis/trans isomerization in a 2-D material on the solid-gas interface. Due
to a low natural abundance of NMR active 15N nuclei (~0.4%), artificially enriched precursors
were used to prepare selectively labeled azobenzene units in molecular switches and thus
enhance the detection limits and shorten the accumulation times. Presented solid-state NMR
spectra (Figure 10) clearly show presence of a cis- isomer characterized by a peak at ~520
ppm after irradiation (max = 254 ± 5 nm) of the solid sample. This peak completely vanishes
with time and can be repetitively re-called by irradiation. Relatively low concentration of the
- 19 -
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 20 of 32
1
2
3
4
5
6
cis- isomer compared to the starting trans- conformer might be due to (i) insufficient
irradiation of the solid powder material and/or (ii) long spectral accumulation time due to low
concentration of guest molecules.
15
Experimental N NMR chemical shifts in both solution and solid-state correlate very
7
well with theoretical values. The cluster approach for calculating solid-state NMR reproduced
the 15N signals well, with a mean unsigned error (MUE) of 2.7 ppm. A possible source of this
error is inclusion of only the single array geometry in the final NMR calculation; it was
previously shown that very large molecular clusters are necessary for accurate predictions of
NMR parameters.34,35 Furthermore, the effects of zero-point vibrations and molecular
dynamics at finite temperatures may be also important,36 but were neglected in the
calculations.
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
SUMMARY
A new generation of rod-like molecular switches based on substituted azobenzenes was
designed, synthesized and subsequently applied into a TPP matrix to get corresponding
surface inclusions. Exact positions of individual switches within such 2-D arrays were
dictated by the crystal structure of the matrix. Selective introduction of 15N into the
azobenzene moiety allowed us to detect the cis/trans isomerization for the first time in the 2-
15
D film by N solid-state MAS NMR technique. Time-resolved and temperature dependent
UV-vis spectroscopy was used to determine enthalpy (ΔH‡) and Gibbs free energy (ΔG‡) of
activation, together with the half-life at 20 °C. It was found that the kinetic parameters for the
thermal cis/trans isomerization are independent on the density of surface coverage or
bulkiness of used azobenzene units. Thus, the surface-mounted molecular switches can
operate as easily as individual molecules in solution.
EXPERIMENTAL SECTION
General conditions, laboratory equipment, instruments and techniques mentioned in
sections Materials, Procedures, X-Ray Diffraction, Powder X-Ray Diffraction, Transmission
Electron Microscopy, and UV-vis Analysis were used in the same manner as in our earlier
works.5,8,14
Materials. All reactions were performed in a round-bottom or Schlenk flasks under a
positive pressure of argon, unless otherwise noted. Heated reactions were conducted in an oil
bath. Reagents were obtained from commercial suppliers and were used without purification.
Solvents were dried (THF and ether over sodium with a benzophenone as an indicator,
triethylamine and CH2Cl2 over CaH2) and distilled under argon prior to use. Compounds 12,8
17,8 and TPP-d129 were prepared in accordance with previous publications.
Inclusion Compounds. All surface inclusions were marked as X%Y@TPP-d12, where
X represents the molar percentage of molecular switch Y.
All four samples of surface inclusions were prepared from neat components (solvent-
free hexagonal TPP and molecular switches 1t and 2t) according to our previously published
procedure.5
Procedures. Reactions were monitored by the thin-layer chromatography (TLC) using
aluminum sheets pre-coated with silica gel (Silica gel 60 F254). TLC spots were visualized by
exposure to UV light (254 or 366 nm) and/or by immersion in a 5% solution of
phosphomolybdic acid in ethanol (TLC plates were subsequently developed by heating at 200
°C). Solvents were removed under reduced pressure on a rotary evaporator with a water bath
heated to 55 °C. Column chromatography was done using silica gel (70 - 230 mesh, pore size
60 Å).
Chemical shifts of proton (1H) and carbon (13C) nuclear magnetic resonance (NMR)
spectra are reported in parts-per-million (ppm) on the δ scale with following solvents used as
- 20 -
ACS Paragon Plus Environment

Page 21 of 32
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
internal references: THF-d8 (δ = 1.72 and 3.58 ppm for 1H, and δ = 25.40 and 67.50 ppm for
13C) and CHCl3 (δ = 7.26 ppm for 1H and δ = 77.0 ppm for 13C). The following abbreviations
are used in connection with NMR: s = singlet, d = doublet, m = multiplet, br = broad. Melting
points were measured in open glass capillaries and reported values are uncorrected.
High-resolution mass spectra (HRMS) with electron impact (EI) were measured on a
time-of-flight mass spectrometer. HRMS with electrospray ionization (ESI) or using
atmospheric-pressure chemical ionization (APCI) were acquired on an instrument combining
Orbitrap and linear ion trap.
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
Powder X-ray Diffraction. The X-ray powder diffraction data were collected at room
temperature on Bruker D2 Phaser (Bruker, Germany) diffractometer with parafocusing Bragg-
Brentano geometry using Cu Kα radiation (λ = 0.15418 nm, U = 30 kV, I = 10 mA) and a
LYNXEYE XE line detector. Data were collected over the angular range 5-90° (2θ) with a
step size 0.018° (2θ) and 0.8 s/step acquisition time. The samples were loaded on a silicon
holder (6° off cut from (111)).
X-Ray Diffraction. The X-ray data were collected on Bruker D8 VENTURE Kappa
Duo PHOTON100 by IμS micro-focus sealed tube either MoKα (λ= 0.71073) (3, 4, 7, S18,
and S20) or CuKα (λ = 1.54178 Å) (8) at low temperature. All structures were solved and
refined as was already described in detail in our previous publications.5,8
There are two exceptions from standard structure determination procedure. Diffraction
intensities of S20 were corrected for non-merohedric twinning during data reduction.
The precision of 4 was hampered by disorder of one phenyl moiety over four
positions. The restriction on positions and displacement parameters were applied on
disordered atoms to ensure proper geometry of phenyl ring.
Also -NO moiety in 8 is disordered in all three symmetrically independent molecules
in unit cell. The occupation factors and coordinates of split positions were freely refined.
Transmission Electron Microscopy. The sample of 10%2@TPP-d12 was prepared
according to previously published procedure.5
Calculations. Structures of two-dimensional arrays of 1@TPP and 2@TPP (simplified
by replacing the triptycene unit and TPP matrix with a methyl group) were optimized using
periodic boundary conditions, with a unit cell corresponding to the crystal structure of TTP (a
= b = 11.454 Å;  = 120○). The TPSS functional37 was used with the pob-TZVP basis set,
which is optimized for the treatment of periodic systems.38 To improve SCF convergence, a
large initial damping factor (3-5) was used. These optimizations were done using Turbomole
7.1.39
Final NMR shifts were determined by cutting out a heptamer (with approximate C6
symmetry) from the optimized structure of the array (shown in red on Figure S16 in the
Supporting Information). The edge units of this cluster were then frozen, and the central
molecule (blue) was re-optimized (its terminal methyl group was also frozen to preserve drift)
at the LC-TPSS/cc-PVTZ level. Finally, the NMR shifts of the central unit were determined at
the same level of theory by using the GIAO method. The LC-TPSS functional was chosen
because of its treatment of long-range interactions40 and good performance for modeling
nitrogen NMR,41 and the cc-PVTZ basis set according to suggestion by Iron.42 Solution-state
NMR was obtained using the same level of theory and an implicit (PCM43) description of the
solvent. Absolute NMR shifts were scaled (see Supporting Information). These calculations
were done using Gaussian 09 software.44
13
15
31
Solid-State NMR. The C, N and P solid-state NMR spectra were acquired on a
JEOL ECZ600R spectrometer working on following frequencies: 150.9 MHz for 13C, 60.8
MHz for 15N and 243.0 MHz for 31P. Solid samples were loaded into 3.2 mm rotors and
measurements were performed at 18 kHz magic angle spinning (MAS) rate. All spectra were
measured using cross polarization (CP) with a ramped amplitude shape pulse. Only
- 21 -
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 22 of 32
1
2
3
4
5
6
7
8
9
phosphorus spectra were also measured with a single-pulse experiment. Following contact
13
31
15
times for CP experiments were used: 5 ms for C and P, and 10 ms for N. The chemical
shifts were referenced as follows: 15N to crystalline α-glycine (δ = 34.1 ppm for nitrogen), 13C
to crystalline sodium 3-(trimethylsilyl)propane-1-sulfonate abbreviated as DSS (δ = 0.0 ppm
for CH3 carbon), and 31P to (NH4)2HPO4 (δ = 1.37 ppm). The relaxation delays were
estimated from 1H saturation recovery experiments and ranged from 0.5 s to 5 s. To suppress
the cis-to-trans isomerization while measuring the 15N spectra of the inclusion after
irradiation, the measurement was taken at -20 °C and the MAS speed was reduced to 12 kHz.
Note that frictional heating can increase the sample temperature significantly.
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 Analysis. All samples were treated in the same manner and measured on the
same instrument as described in our earlier work.5
Illumination of the surface inclusions for 15N solid-state NMR experiments was
performed with the 20 W UV-A lamp with maximum at 368 nm. The solid sample was kept at
6-8 °C during the irradiation and manipulation; the irradiation time was 3 hours and the
sample was spread to wide area and manually mixed every 30 min to maximize the exposure.
Synthesis.
Molecular Switch 1. To a Schlenk flask was added 15 (40 mg, 0.067 mmol), 17 (35 mg,
0.067 mmol), Pd(PPh3)4 (3 mg, 0.003 mmol, 4 mol %), and CuI (0.4 mg, 0.002 mmol, 3 mol
%). The reaction flask was evacuated eight times and filled with argon. Subsequently, THF (5
mL) and Et3N (2 mL) were added from a syringe, and the slightly yellowish solution was
stirred for 5 h at room temperature. A dense orange solid precipitated during this time. Solid
material was separated by centrifugation (5 min, 2500 rpm) and carefully washed with
methanol (4 × 5 mL), acetone (4 × 5 mL), THF (1 × 5 mL), and ether (2 × 5 mL). Solvents
were removed by centrifugation (5 min, 2500 rpm) after each washing. Compound 1t was
isolated as a slightly orange solid (64 mg, 0.064 mmol, 96%).
Mp > 280 °C (dec.). 1H NMR (500 MHz, THF-d8): δ 1.34 (s, 9H), 2.46 (s, 6H), 7.15 –
7.17 (m, 6H), 7.41 – 7.44 (m, 3H), 7.45 – 7.47 (m, 4H), 7.54 – 7.57 (m, 2H), 7.57 – 7.59 (m,
2H), 7.63 – 7.65 (m, 2H), 7.68 – 7.71 (m, 2H), 7.73 – 7.76 (m, 6H), 7.89 – 7.92 (m, 2H), 8.06
– 8.08 (m, 2H). 13C {1H} NMR (125 MHz, THF-d8): δ 31.7, 31.8, 32.1, 35.3, 54.3, 59.9, 75.5,
76.7, 77.9, 78.0, 78.9, 79.1, 79.3, 79.7, 80.5, 81.1, 89.1, 92.0, 116.8 (d, J = 5 Hz), 121.9,
122.7, 123.05, 123.10, 124.6 (d, J = 4 Hz), 125.7, 126.0, 126.6, 127.07, 127.09, 127.35,
127.44, 128.8, 132.8, 132.9, 133.6, 134.8 (d, J = 1 Hz), 135.2, 138.3, 141.8, 144.05, 144.11,
151.46 (d, J = 6 Hz), 151.53, 153.7 (d, J = 1 Hz). 15N NMR (50 MHz, THF-d8): δ 516.2 ppm.
IR (KBr): 3069, 3029, 2962, 2913, 2876, 2223, 2149, 1607, 1595, 1563, 1496, 1459, 1453,
1429, 1416, 1363, 1304, 1264, 1201, 1178, 1150, 1140, 1101, 1052, 1033, 1011, 1004, 945,
843, 822, 785, 751, 728, 704, 639, 618, 574, 545, 492 cm-1. MS, m/z (%): 998.3 (100, M + H).
HRMS (APCI) m/z: [M + H]+ Calcd for C71H47Cl2N15N+ 998.3081; Found 998.3084. Anal.
Calcd. for C71H46Cl2N15N: C, 85.36; H, 4.64; N, 2.90. Found: C, 85.27; H, 4.60; N, 2.89.
Molecular Switch 2. To a Schlenk flask was added 16 (75 mg, 0.116 mmol), 17 (61 mg,
0.116 mmol), Pd(PPh3)4 (5 mg, 0.005 mmol, 4 mol %), and CuI (1 mg, 0.003 mmol, 3 mol
%). The reaction vessel was evacuated five times and filled with argon. Subsequently, THF (4
mL) and Et3N (2 mL) were added from a syringe and the slightly yellowish solution was
stirred for 18 h at room temperature. A dense orange solid precipitated during this time. Solid
material was separated by centrifugation (5 min, 2500 rpm) and carefully washed with
methanol (4 × 5 mL), acetone (4 × 5 mL), THF (1 × 5 mL), and ether (2 × 5 mL). Solvents
were removed by centrifugation (5 min, 2500 rpm) after each washing. Compound 2t was
obtained as a slightly orange solid (108 mg, 0.104 mmol, 90%).
Mp > 205 °C (dec.). 1H NMR (500 MHz, THF-d8): δ 1.34 (s, 9H), 1.41 (s, 18H), 2.46
(s, 6H), 7.15 – 7.17 (m, 6H), 7.41 – 7.44 (m, 2H), 7.45 – 7.47 (m, 2H), 7.54 – 7.57 (m, 2H),
- 22 -
ACS Paragon Plus Environment

Page 23 of 32
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
7.57 – 7.59 (m, 2H), 7.63 – 7.65 (m, 2H), 7.67 (m, 1H), 7.72 – 7.76 (m, 6H), 7.85 – 7.88 (m,
4H), 7.99 – 8.01 (m, 2H). 13C {1H} NMR (125 MHz, THF-d8): δ 31.7, 31.8, 32.0, 35.3, 35.8,
54.1, 59.8, 75.5, 75.9, 77.90, 77.91, 78.7, 79.0, 79.2, 79.3, 80.4, 81.0, 89.1, 92.0, 118.4 (d, J =
4 Hz), 121.8, 122.6, 123.1, 124.0 (d, J = 3 Hz), 124.4, 125.7, 126.6, 127.1, 127.3, 127.5,
132.86, 132.94, 133.6, 134.7, 138.2, 141.7, 143.98, 143.99, 151.4, 152.9, 153.7 (d, J = 5 Hz),
153.8 (d, J = 2 Hz). 15N NMR (50 MHz, THF-d8): δ 501.4 ppm. IR (KBr): 3070, 3033, 2963,
2911, 2869, 2228, 2155, 1607, 1493, 1477, 1459, 1453, 1393, 1363, 1303, 1268, 1246, 1209,
1171, 1140, 1113, 1100, 1032, 1004, 946, 901, 884, 840, 822, 752, 725, 699, 639, 615, 575,
546, 535, 493 cm-1. MS, m/z (%): 1042.5 (100, M + H). HRMS (APCI) m/z: [M + H]+ Calcd
for C79H65N15N+ 1042.5113; Found 1042.5111. Anal. Calcd. for C79H64N15N + THF in 2:1
ratio: C, 90.25; H, 6.31; N, 2.69. Found: C, 90.45; H, 6.22; N, 2.60.
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
(E/Z)-1-(2,3-Dichlorophenyl)-2-(4-iodophenyl)diazene-2-15N (3). To a solution of 6 (100
mg, 0.46 mmol) and 7 (121 mg, 0.69 mmol) in 94% ethanol (5.5 mL) was added glacial acetic
acid (50 μL, 0.87 mmol) under an argon atmosphere at room temperature. The crude mixture
was stirred 2 h at 70 °C, then it was quenched with water (20 mL), and extracted with CH2Cl2
(3 × 30 mL). Combined organic phases were dried over Na2SO4, and solvent was removed
under reduced pressure. Column chromatography on a silica gel (hexane) afforded 3 as a
mixture of E/Z isomers (3:2) in form of an orange solid (90 mg, 0.24 mmol, 52%).
Mp: 134.1 - 136.4 °C. (Z)-1-(2,3-Dichlorophenyl)-2-(4-iodophenyl)diazene-2-15N:
1H NMR (400 MHz, CDCl3):  6.24 - 6.26 (m, 1H), 6.64 - 6.68 (m, 2H), 7.01 - 7.05 (m, 2H),
1
7.61 - 7.63 (m, 2H). (E)-1-(2,3-Dichlorophenyl)-2-(4-iodophenyl)diazene-2-15N: H NMR
(400 MHz, CDCl3):  7.27 - 7.31 (m, 1H), 7.58 (m, 2H), 7.70 (m, 2H), 7.88 - 7.90 (m, 2H).
13C {1H} NMR (100 MHz, CDCl3):  99.2, 115.9 (d, J = 5 Hz), 125.1 (d, J = 4 Hz), 127.5,
132.4, 134.0, 134.5, 138.7 (d, J = 2 Hz), 150.3, 152.0. IR (KBr): 3080, 2924, 2852,
2556,1951, 1909, 1812, 1777, 1695, 1652, 1576, 1565, 1477, 1419, 1390, 1154, 1053, 1004,
830, 788, 495 cm-1. MS, m/z (%): 376.9 (60, M), 231 (62), 202.9 (100). MS, m/z (%): 376.9
(60, M), 231 (62), 202.9 (100). HRMS (EI) m/z: [M]+ Calcd for C12H7N15NCl2I+ 376.9001;
Found 376.9002. Anal. Calcd. for C12H7N15NCl2I: C, 39.70; H, 1.94; N, 3.86. Found: C,
40.09; H, 1.90.
(E)-1-(3,5-Di-tert-butylphenyl)-2-(4-iodophenyl)diazene-2-15N (4). To a solution of 6 (258
mg, 1.18 mmol) and 8 (387 mg, 1.76 mmol) in 94% ethanol (20 mL) was added glacial acetic
acid (100 μL, 1.74 mmol) under argon atmosphere. The solution was stirred 2 h at 70 °C. The
crude reaction mixture was quenched with water (30 mL) and extracted with CH2Cl2 (3 × 50
mL). Combined organic phases were dried over MgSO4, the dried solution was filtered, and
the filtrate concentrated to dryness. Column chromatography on silica gel (hexane/ethyl
acetate – 1:1) gave 4 as an orange solid (370 mg, 0.88 mmol, 75%).
Mp 134.9 - 136.6 °C. 1H NMR (400 MHz, CDCl3):  1.40 (s, 18H), 7.58 (m, 1H), 7.63
- 7.68 (m, 2H), 7.77 (m, 2H), 7.84 - 7.88 (m, 2 H). 13C {1H} NMR (400 MHz, CDCl3):  31.6,
35.3, 97.3, 117.5 (d, J = 4 Hz), 124.5 (d, J = 4 Hz), 125.9, 138.4 (d, J = 2 Hz), 152.1, 152.3
(d, J = 2 Hz), 152.5 (d, J = 5 Hz). IR (KBr): 2949, 2901, 2865, 1642, 1601, 1580, 1569, 1477,
1458, 1444, 1427, 1417, 1392, 1364, 1314, 1300, 1270, 1244, 1214, 1201, 1170, 1109, 1093,
1053, 1027, 1007, 974, 908, 899, 884, 825, 715, 707, 699, 629, 572, 501, 479 cm-1. MS, m/z
(%): 422.1 (13, M + H), 421.1 (57, M), 231.9 (10), 202.9 (36, PhI), 190.2 (15), 189.2 (100,
PhtBu2), 147.1 (8), 133.1 (23), 126.9 (39), 76.0 (6), 57.1 (18). HRMS (EI) m/z: [M]+ Calcd for
C20H25IN15N+ 421.1033; Found 421.1034. Anal. Calcd. for C20H25IN15N: C, 57.01; H, 5.98;
N, 6.66; I, 30.12. Found: C, 56.89; H, 5.95; N, 6.60; I, 30.43.
- 23 -
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 24 of 32
1
2
3
4
5
6
7
8
9
15N-Iodoaniline (6). Previously published procedure describing synthesis of non-labeled
derivative29 was adapted as follows: Two necked 50 mL flask was charged with MeOH (25
mL) and concentrated H2SO4 (1.16 mL, 22 mmol). To the solution were added 5 (400 μL, 1.8
mmol), KI (720 mg, 4.3 mmol) and finally 30% H2O2 (860 μL, 8.5 mmol). Then the reaction
mixture was stirred at room temperature and monitored by GC for 15 h. (Sample preparation:
200 μL was taken every 2 h from the crude reaction mixture, then it was diluted with 200 μL
CH2Cl2 and the organic phase was washed with 200 μL of concentrated aqueous NaHSO3,
200 μL of water, and dried over Na2SO4. The solvent was evaporated; the sample was diluted
in 100 μL of CH2Cl2, and injected in to the GC.) The crude reaction mixture was diluted with
CH2Cl2 (60 mL) and the organic phase was washed with concentrated NaHSO3 (2 × 30 mL),
water (3 × 30 mL), and dried over Na2SO4. The dried solution was filtered and the filtrate
concentrated to dryness. Subsequent column chromatography on silica gel (CH2Cl2) provided
6 as a light brown solid (300 mg, 1.36 mmol, 75%).
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
Mp 63.1 - 64.2 °C. 1H NMR (400 MHz, CDCl3):  3.67 (s, 2H), 6.47 (d, J = 9 Hz, 2H),
7.41 (d, J = 9 Hz, 2H). 13C {1H} NMR (100 MHz, CDCl3): 79.5, 117.4 (d, J = 2.8 Hz), 138.1
(d, J = 1.4 Hz), 146.2 (d, J = 11.6 Hz), 146.1. IR (KBr): 3406, 3300, 3200, 3059, 2701, 1870,
1630, 1583, 1483, 1418, 1275, 1178, 1175, 1127, 1078, 1175, 1127, 1078, 1057, 999, 946,
926, 816, 689, 584, 498, 410, 401 cm-1. MS, m/z (%): 221 (100, M). HRMS (ESI+) m/z: [M +
15
H]+ Calcd for C6H7 NI+ 220.9588; Found 220.9588. Anal. Calcd. for C6H6I15N: C, 32.75; H,
2.75; N, 6.82. Found: C, 33.10; H, 2.94; N, 6.96.
1,2-Dichloronitrosobenzene (7). To a solution of 9 (3.60 mL, 31 mmol) in CH2Cl2 (60 mL)
was added a solution of Oxone® (2 g, 62 mmol) in H2O (150 mL). The two-phase system was
vigorously stirred at room temperature until complete consumption of the starting 9 (usually
18-22 h). Progress of the oxidation was monitored by GC. (Sample preparation: 200 μL was
taken directly from the crude reaction mixture every 2 h. Solvents were evaporated, 200 μL of
CH2Cl2 was added and sample was injected in to the GC apparatus.) After that time, the
organic layer was separated and the aqueous phase was extracted with CH2Cl2 (3 × 40 mL).
The combined organic phases were washed with 1M HCl (2 × 70 mL), water (3 × 40 mL),
and dried over Na2SO4. Solvent was removed under reduced pressure and column
chromatography on silica gel (hexane/CH2Cl2 - 4:1) yield 7 as a green oil that solidified in a
freezer (1.654 g, 9 mmol, 30%).
Mp 91.9 - 93.9 °C. 1H NMR (400 MHz, CDCl3):  6.18 (dd, J1 = 8 Hz, J2 = 2 Hz, 1H),
7.27 (dd, J1 = 2 Hz, J2 = 2 Hz, 1H), 7.85 (dd, J1 = 8 Hz, J2 = 2 Hz, 1H). 13C {1H} NMR (100
MHz, CDCl3):  107.0, 126.9, 136.1, 136.6, 141.5, 161.2. IR (KBr): 3081, 3069,2925, 2696,
1937, 1863, 1794, 1684, 1609, 1576, 1504, 1474, 1445, 1431, 1274, 1231, 1183, 1152, 1106,
1052, 1006, 969, 913, 892, 786, 744, 714, 656, 569, 536, 469, 426 cm-1. MS, m/z (%): 175.0
(75, M), 145.0 (100, M - NO), 109.0 (45), 74.0 (20). HRMS (EI) m/z: [M]+ Calcd for
C6H3Cl2NO+ 174.9592; Found 174.9593. Anal. Calcd. for C6H3Cl2NO: C, 40.95; H, 1.72; N,
7.96. Found: C,41.29; H, 1.87; N, 7.73.
1,3-Di-tert-butyl-5-nitrosobenzene (8). To a solution of 11 (500 mg, 1.90 mmol) in
acetonitrile (35 mL) in a 100 mL three necked flask was added NOBF4 (445 mg, 3.81 mmol)
at room temperature. Originally colorless reaction mixture turned green immediately. After 30
minutes, the reaction mixture was poured into 1:1 mixture of water and CH2Cl2 (100 mL).
The organic phase was separated and dried over MgSO4. The dried solution was filtered and
the filtrate concentrated to dryness. Column chromatography on silica gel (hexane/ethyl
acetate - 3:1) yielded 8 as a light green oil which solidified upon cooling (400 mg, 1.82 mmol,
96%).
- 24 -
ACS Paragon Plus Environment

Page 25 of 32
Journal of the American Chemical Society
1
2
3
4
5
1
Mp 93.9 - 96.0 °C. H NMR (400 MHz, CDCl3):  1.40 (s, 18H), 7.77 (m, 2H), 7.89
(m, 1H). C {1H} NMR (400 MHz, CDCl3):  31.4, 35.3, 116.1, 117.9, 128.9, 129.9, 152.5,
13
167.3. IR (KBr): 2965, 2906, 2868, 2749, 2721, 1787, 1605, 1583, 1530, 1490, 1479, 1463,
1438, 1396, 1365, 1346, 1325, 1291, 1278, 1247, 1203, 1150, 1100, 1027, 998, 931, 894, 868,
853, 815, 767, 751, 744, 718, 697, 515, 431 cm-1. MS, m/z (%): 220.2 (100, M + H). HRMS
(EI) m/z: [M + H]+ Calcd for C14H22NO+ 220.1696; Found 220.1706. Anal. Calcd. for
C14H21NO: C, 76.67; H, 9.65; N, 6.39. Found: C, 76.41; H, 9.59; N, 6.06.
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
(3,5-Di-tert-butylphenyl)trimethylsilane (11). To a solution of 10 (1.00 g, 3.71 mmol) in
ether (30 mL) was added dropwise a solution of tert-BuLi (1.7 M, 4.37 mL, 7.43 mmol) at -78
°C. After 30 minutes, TMSCl (1.0 mL, 8.17 mmol) was added. A white precipitate appeared
and the reaction mixture was stirred 1 h at room temperature. Then ether (30 mL) was added
and the reaction mixture was washed with saturated aqueous NH4Cl (1 × 30 mL). The organic
phase was dried over MgSO4, the dried solution was filtered, and the filtrate concentrated to
dryness. Subsequent column chromatography on silica gel (hexane) afforded 11 as a colorless
oil (880 mg, 3.35 mmol, 90%).
1H NMR (400 MHz, CDCl3):  0.28 (s, 9H), 1.35 (s, 18H), 7.39 (m, 2H), 7.45 (m,
1H). 13C {1H} NMR (400 MHz, CDCl3):  0.8, 31.7, 35.0, 123.3, 127.4, 139.4, 149.7. IR
(KBr): 3046, 2963, 2905, 2869, 1589, 1477, 1462, 1408, 1392, 1362, 1262, 1249, 1203, 1151,
878, 825, 753, 713, 691, 624, 461 cm-1. MS, m/z (%): 263.20 (9, M + H), 262.20 (38, M),
249.2 (10), 248.2 (46), 247.2 (100, M - CH3). HRMS (EI) m/z: [M]+ Calcd for C17H30Si+
262.2117; Found 262.2121. Anal. Calcd. for C17H31Si: C, 77.78; H, 11.52. Found: C, 77.30;
H, 11.50.
Compound 13. To a Schlenk flask was added 3 (130 mg, 0.344 mmol, 1.0 equiv.), 12 (160
mg, 0.378 mmol, 1.1 equiv.), Pd(PPh3)4 (16 mg, 0.014 mmol, 4 mol %), and CuI (2 mg,
0.010 mmol, 3 mol %). The reaction vessel was thoroughly evacuated three times and filled
with argon. Subsequently, THF (4 mL) and triethylamine (2 mL) were added from a syringe
and the orange solution was stirred for 18 h at room temperature. A dense white solid
precipitated during this time. The crude reaction mixture was diluted with CH2Cl2 (80 mL),
and washed with saturated aqueous NH4Cl (2 × 20 mL). An orange organic phase was dried
over MgSO4. The dried solution was filtered and the filtrate concentrated to dryness. The
residue was purified by silica gel column chromatography (hexane/CH2Cl2 - 5:1) to provide
13 as an orange crystalline solid (102 mg, 0.152 mmol, 44%).
Mp > 285 °C (dec.). 1H NMR (400 MHz, CDCl3): δ 0.34 (s, 9H), 7.11 – 7.16 (m, 6H),
7.32 (m, 1H), 7.59 – 7.63 (m, 2H), 7.74 – 7.78 (m, 6H), 7.81 – 7.83 (m, 2H), 8.01 – 8.04 (m,
2H). 13C {1H} NMR (100 MHz, CDCl3): δ -0.4, 52.8, 53.1, 72.2, 76.4, 76.8, 77.3, 78.0, 78.7,
87.4, 87.7, 115.7 (d, J = 5 Hz), 122.3, 122.4, 123.7 (d, J = 4 Hz), 125.0, 126.0, 126.1, 127.3,
132.4, 133.71, 133.73, 134.4, 142.75, 142.85, 150.2 (d, J = 6 Hz), 152.3 (d, J = 2 Hz). IR
(KBr): 3070, 3017, 2958, 2927, 2896, 2854, 2238, 2111, 1607, 1595, 1564, 1497, 1453, 1429,
1416, 1305, 1251, 1225, 1200, 1179, 1149, 1134, 1052, 1032, 1010, 951, 922, 843, 784, 750,
727, 703, 639, 619, 525, 486 cm-1. MS, m/z (%): 672.1 (100, M + H). HRMS (APCI) m/z: [M
+ H]+ Calcd for C43H29Cl2N15NSi+ 672.1442; Found 672.1442. Anal. Calcd. for
C43H28Cl2N15NSi: C, 76.78; H, 4.20; N, 4.31. Found: C, 76.56; H, 4.32; N, 4.13.
Compound 14. To a Schlenk flask was added 4 (100 mg, 0.237 mmol, 1.0 equiv.), 12 (110
mg, 0.261 mmol, 1.1 equiv.), Pd(PPh3)4 (11 mg, 0.009 mmol, 4 mol %), and CuI (1 mg,
0.007 mmol, 3 mol %). The reaction vessel was thoroughly evacuated three times and filled
with argon. Subsequently, THF (4 mL) and triethylamine (2 mL) were added from a syringe
and the orange solution was stirred for 18 h at room temperature. A dense white solid
- 25 -
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 26 of 32
1
2
3
4
5
6
7
8
9
precipitated during this time. The crude reaction mixture was diluted with CH2Cl2 (80 mL),
and washed with saturated aqueous NH4Cl (2 × 20 mL). An orange organic phase was dried
over MgSO4. The dried solution was filtered and the filtrate concentrated to dryness. The
residue was purified by silica gel column chromatography (hexane/CH2Cl2 - 5:1) to yield 14
as an orange crystalline solid (101 mg, 0.141 mmol, 59%).
Mp > 240 °C (dec.). 1H NMR (400 MHz, CDCl3): δ 0.34 (s, 9H), 1.42 (s, 18H), 7.11 –
7.16 (m, 6H), 7.61 (m, 1H), 7.74 – 7.76 (m, 3H), 7.77 – 7.79 (m, 3H), 7.80 – 7.82 (m, 4H),
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
13
7.95 – 7.98 (m, 2H). C {1H} NMR (100 MHz, CDCl3): δ -0.4, 31.4, 35.1, 52.8, 53.1, 72.2,
75.6, 76.9, 77.2, 78.2, 78.3, 87.4, 87.7, 117.5 (d, J = 4 Hz), 122.3, 122.4, 122.9 (d, J = 4 Hz),
123.5, 125.9, 126.0, 133.6 (d, J = 2 Hz), 142.81, 142.82, 151.9, 152.6 (d, J = 5 Hz), 152.7 (d,
J = 2 Hz). IR (KBr): 3069, 3035, 3018, 2961, 2902, 2867, 2238, 2111, 1606, 1583, 1478,
1459, 1453, 1437, 1424, 1393, 1363, 1304, 1251, 1203, 1171, 1134, 1114, 1099, 1032, 1009,
974, 950, 900, 884, 843, 751, 744, 726, 699, 639, 616, 575, 534, 501, 483 cm-1. MS, m/z (%):
716.3 (100, M + H). HRMS (APCI) m/z: [M + H]+ Calcd for C51H47N15NSi+ 716.3473; Found
716.3478. Anal. Calcd. for C51H46N15NSi: C, 85.55; H, 6.48; N, 4.05. Found: C, 85.25; H,
6.42; N, 3.77.
Compound 15. A solution of TBAF in THF (1.0 M, 190 µL, 0.190 mmol, 1.5 equiv.) was
added to a stirred solution of 13 (85 mg, 0.126 mmol, 1.0 equiv) in wet THF (10 mL) at room
temperature. The reddish reaction mixture was stirred for 30 min. Then it was diluted with
CH2Cl2 (80 mL) and washed with water (2 × 20 mL). The clear orange organic phase was
dried over MgSO4. The dried solution was filtered and the filtrate concentrated to dryness.
The residue was purified by silica gel column chromatography (hexane/CH2Cl2 - 4:1) to give
15 as an orange crystalline solid (51 mg, 0.085 mmol, 67%).
Mp > 225 °C (dec.). 1H NMR (400 MHz, CDCl3): δ 2.44 (s, 1H), 7.13 – 7.18 (m, 6H),
7.31 (m, 1H), 7.59 – 7.64 (m, 2H), 7.73 – 7.75 (m, 3H), 7.77 – 7.80 (m, 3H), 7.81 – 7.84 (m,
13
2H), 8.01 – 8.04 (m, 2H). C {1H} NMR (100 MHz, CDCl3): δ 52.6, 53.1, 67.6, 68.6, 71.1,
76.4, 76.5, 76.8, 78.0, 78.6, 115.7 (d, J = 5 Hz), 122.3, 122.4, 123.7 (d, J = 4 Hz), 124.9,
126.09, 126.11, 127.3, 132.4, 133.7 (d, J = 2 Hz), 134.4, 142.6, 142.7, 150.2 (d, J = 6 Hz),
152.3 (d, J = 2 Hz). IR (KBr): 3290, 3070, 3017, 2924, 2853, 2239, 1607, 1593, 1563, 1496,
1454, 1443, 1429, 1418, 1301, 1284, 1225, 1217, 1202, 1179, 1150, 1099, 1053, 1032, 1011,
945, 925, 905, 858, 844, 790, 758, 754, 727, 707, 651, 639, 626, 619, 575, 526, 494, 479 cm-1.
MS, m/z (%): 600.1 (100, M + H). HRMS (APCI) m/z: [M + H]+ Calcd for C40H21Cl2N15N+
600.1047; Found 600.1046. Anal. Calcd. for C40H20Cl2N15N: C, 80.00; H, 3.36; N, 4.83.
Found: C, 79.83; H, 3.33; N, 4.58.
Compound 16. A solution of TBAF in THF (1.0 M, 190 µL, 0.190 mmol, 1.5 equiv.) was
added to a stirred solution of 14 (90 mg, 0.126 mmol, 1.0 equiv) in wet THF (10 mL) at room
temperature. The reddish reaction mixture was stirred for 30 min. Then it was diluted with
CH2Cl2 (50 mL) and washed with water (2 × 20 mL). The clear orange organic phase was
dried over MgSO4. The dried solution was filtered and the filtrate concentrated to dryness.
The residue was purified by silica gel column chromatography (hexane/CH2Cl2 - 4:1) to
provide 16 as an orange crystalline solid (81 mg, 0.126 mmol, 100%).
Mp > 220 °C (dec.). 1H NMR (400 MHz, CDCl3): δ 1.42 (s, 18H), 2.43 (s, 1H), 7.14 –
7.16 (m, 6H), 7.62 (s, 1H), 7.73 – 7.75 (m, 3H), 7.78 – 7.80 (m, 3H), 7.80 – 7.83 (m, 4H),
13
7.96 – 7.98 (m, 2H). C {1H} NMR (100 MHz, CDCl3): δ 31.4, 35.1, 52.6, 53.1, 67.7, 68.6,
71.2, 75.6, 76.5, 77.2, 78.1, 78.3, 117.6 (d, J = 2 Hz), 122.3, 122.4, 122.9 (d, J = 4 Hz), 123.4,
125.9, 126.06, 126.11, 133.6 (d, J = 2 Hz), 142.6, 142.8, 152.0, 152.5 (d, J = 5 Hz), 152.7 (d,
J = 2 Hz). IR (KBr): 3289, 3282, 3069, 3035, 2962, 2903, 2868, 2237, 1607, 1583, 1477,
1459, 1453, 1425, 1394, 1363, 1301, 1246, 1210, 1171, 1142, 1099, 1032, 1010, 974, 900,
- 26 -
ACS Paragon Plus Environment

Page 27 of 32
Journal of the American Chemical Society
1
2
3
4
5
884, 861, 842, 752, 727, 699, 639, 622, 575, 534, 479 cm-1. MS, m/z (%): 644.3 (100, M + H).
HRMS (APCI) m/z: [M + H]+ Calcd for C48H39N15N+ 644.3078; Found 644.3080. Anal.
Calcd. for C48H38N15N: C, 89.55; H, 5.95; N, 4.51. Found: C, 89.36; H, 6.01; N, 4.27.
6
7
8
ASSOCIATED CONTENT
9
Supporting Information. The Supporting Information is available free of charge on the ACS
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
Publications website.
1
Copies of H, 13C, 15NMR spectra of all new compounds, optimized geometries of
periodic cis- and trans- isomers of 1@TPP and 2@TPP (xyz), and an ORTEP view of a single
molecule and packing in 3, 4, 7, 8, S18, and S20 (PDF).
X-ray crystallographic data for compound 3 (CIF)
X-ray crystallographic data for compound 4 (CIF)
X-ray crystallographic data for compound 7 (CIF)
X-ray crystallographic data for compound 8 (CIF)
X-ray crystallographic data for compound S18 (CIF)
X-ray crystallographic data for compound S20 (CIF)
Optimized geometries of cis- and trans- isomers of 1@TPP and 2@TPP (xyz)
AUTHOR INFORMATION
Corresponding Author
* kaleta@uochb.cas.cz
ORCID
J. Kaleta: 0000-0002-5561-7580
G. Bastien: 0000-0001-6374-8620
M. Dračínský: 0000-0002-4707-8230
I. Císařová: 0000-0002-9612-9831
C. T. Rogers: 0000-0002-2278-6863
Notes
The authors declare no competing financial interest.
ACKNOWLEDGEMENTS
This work was supported by the Institute of Organic Chemistry and Biochemistry of
the Czech Academy of Sciences (RVO: 61388963), Czech Science Foundation (grant
number: 20-13745S, 20-01472S), and Ministry of Education, Youth and Sports (grant
number: LTAUSA19120). We thank Dr. Lucie Bednárová for help with IR spectra, Dr. Jan
Plutnar for acquiring PXRD, and Dr. Romana Hadravová for help with TEM imaging. Access
to computing and storage facilities owned by parties and projects contributing to the National
Grid Infrastructure MetaCentrum provided under the programme “Projects of Large Research,
Development, and Innovations Infrastructures” (CESNET LM2015042), is greatly
appreciated.
REFERENCES
1.
Rusch, T. R.; Hammerich, M.; Herges, R.; Magnussen, O. M. “Molecular Platforms as
Versatile Building Blocks for Multifunctional Photoswitchable Surfaces.” Chem.
Comm. 2019, 55, 9511-9514.
2.
Chu, Z.; Han, Y.; Bian, T.; De, S.; Král, P.; Klajn, R. „Supramolecular Control of
Azobenzene Switching on Nanoparticles.“ J. Am. Chem. Soc. 2019, 141, 1949-1960.
- 27 -
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 28 of 32
1
2
3
4
5
6
7
8
9
3.
4.
5.
Kaleta, J.; Wen, J.; Magnera, T. F.; Dron, P. I.; Zhu, C.; Michl, J. ”Structure of a
Monolayer of Molecular Rotors on Aqueous Subphase from Grazing-Incidence X-Ray
Diffraction.” PNAS 2018, 115, 9373-9378.
Chen. J.; Vachon, J.; Feringa, B. L. “Design, Synthesis, and Isomerization Studies of
Light-Driven Molecular Motors for Single Molecular Imaging.” J. Org. Chem. 2018,
83, 6025-6034.
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
Kaleta, J.; Chen, J.; Bastien, G.; Dračínský, M.; Mašát, M.; Rogers, C. T.; Feringa, B.
L.; Michl, J. “Surface Inclusion of Unidirectional Molecular Motors in Hexagonal
Tris(o-phenylene-dioxy)cyclotriphosphazene TPP.” J. Am. Chem. Soc. 2017, 139,
10486-10498.
6.
7.
Hammerich, M.; Herges, R. “Laterally Mounted Azobenzenes on Platforms.” J. Org.
Chem. 2015, 80, 11233-11236.
Kaleta, J.; Kaletová, E.; Císařová, I.; Teat, S. J.; Michl, J. “Synthesis of Triptycene-
Based Molecular Rotors for Langmuir-Blodgett Monolayers.” J. Org. Chem. 2015, 80,
10134-10150.
8.
9.
Kaleta, J.; Dron, P. I.; Zhao, K.; Shen, Y.; Císařová, I.; Rogers, C. T.; Michl, J.
“Arrays of Molecular Rotors with Triptycene Stoppers: Surface Inclusion in
Hexagonal Tris(o-phenylenedioxy)cyclotriphosphazene.” J. Org. Chem. 2015, 80,
6173-6192.
Kobr, L.; Zhao, K.; Shen, Y.; Comotti, A.; Bracco, S.; Shoemaker, R. K.; Sozzani, P.;
Clark, N. A.; Price, J. C.; Rogers, C. T.; Michl, J. “Inclusion Compound Based
Approach to Arrays of Artificial Dipolar Molecular Rotors. A surface Inclusion.” J.
Am. Chem. Soc. 2012, 134, 10122-10131.
10.
11.
12.
13.
Howee, M. E.; Garcia-Garibay, M. A. “Fluorescence and Rotational Dynamics of
Crystalline Molecular Rotor Featuring an Aggregation-Induced Emission
Fluorophore.” J. Org. Chem. 2019, 84, 9570-9576.
Howee, M. E.; Garcia-Garibay, M. A. “The Roles of Intrinsic Barriers and Crystal
Fluidity in Determining the Dynamics of Crystalline Molecular Rotors and Molecular
Machines.” J. Org. Chem. 2019, 84, 9835-9849.
Danowski, W.; van Leeuwen, T.; Abdolahzadeh, S.; Roke, D.; Brown, W. R.;
Wezenberg, S.; Feringa, B. L. “Unidirectionaly Rotary Motion in a Metal-Organic
Framework.” Nature Nanotechnology 2019, 14, 488-494.
Catalano, L.; Perez-Estrada, S.; Wang, H.-H.; Ayitou, A. J.-L.; Khan, S. I.; Terraneo,
G.; Metrangolo, P.; Brown, S.; Garcia-Garibay, M. A. „Rotational Dynamics of
Diazabicyclo[2.2.2]octane in Isomorphous Halogen-Bonded Co-crystals: Entropic and
Enthalpic Effects.“ J. Am. Chem. Soc. 2017, 139, 843-848.
14.
Kaleta, J.; Bastien, G.; Wen, J.; Dračínský, M.; Tortorici, E.; Císařová, I.; Beale, P. D.;
Rogers, C. T.; Michl, J. “Bulk Inclusions of Double Pyridazine Molecular Rotors in
Hexagonal Tris(o-phenylene)cyclotriphosphazene (TPP).” J. Org. Chem. 2019, 84,
- 28 -
ACS Paragon Plus Environment

Page 29 of 32
Journal of the American Chemical Society
1
2
3
4
5
8449-8467.
6
7
8
9
15.
16.
Bracco, S.; Castiglioni, F.; Comotti, A.; Galli, S.; Negroni, M.; Maspero, A.; Sozzani,
P. “Ultrafast Molecular Rotors and Their CO2 Tuning in MOFs with Rod Like
Ligands.” Chem.: Eur. J. 2017, 5, 10328-10337.
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
Rodríguez-Fortea, A.; Kaleta, J.; Méziècre, C.; Allain, M.; Canadell, E.; Wzietek, P.;
Michl, J.; Batail, P. “Asymmetric Choreography in Pairs of Orthogonal Rotors.” ACS
Omega 2018, 3, 1293-1297.
17.
18.
Comotti, A.; Bracco, S.; Sozzani, P. “Molecular Rotors Built in Porous Materials.”
Acc. Chem. Res. 2016, 49, 1701-1710.
Dron, P. I.; Zhao, K.; Kaleta, J.; Shen, Y.; Wen, J.; Shoemaker, R. K.; Rogers, C. T.;
Michl, J. “Bulk Inclusions of Pyridazine-Based Molecular Rotors in Tris(o-
phenylenedioxy)cyclotriphosphazene (TPP).” Adv. Funct. Mater. 2016, 26, 5718-
5732.
19.
20.
Kumar, S.; Merelli, M.; Danowski, W.; Rudolf, P.; Feringa, B. L.; Chiechi, R. C.
“Chemical Locking in Molecular Tunneling Junction Enables Nonvolatile Memory
with Large On-Off Ratios.” Adv. Mater. 2019, 31, 1807831.
Green, J. E.; Choi, J. W.; Boukai, A.; Bunimovich, Y.; Johnston-Halperin, E.;
DeIonno, E.; Luo, Y.; Sheriff, B. A.; Xu, K.; Shin, Y. S.; Tseng, H. R.; Stoddart, J. F.;
Heath, J. R. „A 160-Kilobit Molecular Electronic Memory Patterned at 10(11) Bits per
Square Centimeter.“ Nature 2007, 445, 414.
21.
Allcock, H. R.; Siegel, L. A. „Phosphonitrilic Compounds. III. Molecular Inclusion
Compounds of Tris(o-phenylenedioxy)phosphonitrile Trimetr.“ J. Am. Chem. Soc.
1964, 86, 5140-5144.
22.
23.
Allcock H. R. (American Cyanamid Co): US3356768 (A) - 1967-12-05.
Sozzani, P.; Bracco, S.; Comotti, A.; Ferretti, L.; Simonutti, R. “Methane and Carbon
Dioxide Storage in a Porous van der Waals Crystal.” Angew. Chem. 2005, 117, 1850-
1854.
24.
25.
26.
27.
Bracco, S.; Comotti, A.; Ferretti, L.; Sozzani, P. “Supramolecular Aggregation of
Block Copolymers in the Solid State As Assisted by the Selective Formation of
Inclusion Crystals.” J. Am. Chem. Soc. 2011, 133, 8982-8994.
Sozzani, P.; Comotti, A.; Simonutti, R.; Meersmann, T.; Logan, J. W.; Pines, A. “A
Porous Crystalline Molecular Solid Explored by Hyperpolarized Xenon.” Angew.
Chem. Int. Ed. 2000, 39, 2695-2699.
Zhao, K.; Dron, P. I.; Kaleta, J.; Rogers, C. T.; Michl, J. ”Arrays of Dipolar Molecular
Rotors in Tris(o-phenylenedioxy)cyclotriphosphazene.” Top. Curr. Chem. 2014, 354¸
163-212.
Allcock H. R. „Phosphonitrilic Compounds. II. Reactions of Phosphonitrilic Chlorides
- 29 -
ACS Paragon Plus Environment