Enhancing the Stability of Photogenerated Benzophenone Triplet Radical Pairs through Supramolecular Assembly

Article abstract of DOI:10.1021/jacs.8b08501

Supramolecular assembly of urea-tethered benzophenone molecules results in the formation of remarkably persistent triplet radical pairs upon UV irradiation at room temperature, whereas no radicals were observed in solution. The factors that lead to emergent organic radicals are correlated with the microenvironment around the benzophenone carbonyl, types of proximal hydrogens, and the rigid supramolecular network. The absorption spectra of the linear analogues were rationalized using time-dependent density functional theory calculations on the crystal structure and in dimethyl sulfoxide, employing an implicit solvation model to describe structural and electronic solvent effects. Inspection of the natural transition orbitals for the more important excitation bands of the absorption spectra indicates that crystallization of the benzophenone-containing molecules should present a stark contrast in photophysical properties versus that in solution, which was indeed reflected by their quantum efficiencies upon solid-state assembly. Persistent organic radicals have prospective applications ranging from organic light-emitting diode technology to NMR polarizing agents.

Full text of DOI:10.1021/jacs.8b08501

Subscriber access provided by University of South Dakota
Article
Enhancing the Stability of Photogenerated Benzophenone
Triplet Radical Pairs through Supramolecular Assembly
Baillie A DeHaven, Dustin W Goodlett, Ammon J. Sindt, Niklas Noll, Martina De
Vetta, Mark D. Smith, Corey R. Martin, Leticia Gonzalez, and Linda S. Shimizu
J. Am. Chem. Soc., Just Accepted Manuscript • DOI: 10.1021/jacs.8b08501 • Publication Date (Web): 13 Sep 2018
Downloaded from http://pubs.acs.org on September 14, 2018
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a service to the research community to expedite the dissemination
of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in
full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully
peer reviewed, but should not be considered the official version of record. They are citable by the
Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore,
the “Just Accepted” Web site may not include all articles that will be published in the journal. After
a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web
site and published as an ASAP article. Note that technical editing may introduce minor changes
to the manuscript text and/or graphics which could affect content, and all legal disclaimers and
ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or
consequences arising from the use of information contained in these “Just Accepted” manuscripts.
is published by the American Chemical Society. 1155 Sixteenth Street N.W.,
Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 8
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
Enhancing the Stability of Photogenerated Benzophenone Triplet
Radical Pairs through Supramolecular Assembly
†
†
†
†
‡
†
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Baillie A. DeHaven, Dustin W. Goodlett, Ammon J. Sindt, Niklas Noll, Martina De Vetta, Mark D. Smith, Corey R.
†
‡
†
Martin, Leticia González, and Linda S. Shimizu*
†
Department of Chemistry and Biochemistry, University of South Carolina, Columbia, USA
‡
Institute of Theoretical Chemistry, Faculty of Chemistry, University of Vienna, Währinger Strasse 17, 1090 Vienna, Austria
KEYWORDS: Supramolecular chemistry, triplet radical pairs, urea, benzophenone, and solid-state.
ABSTRACT: Supramolecular assembly of ureaꢀtethered benzophenone molecules results in the formation of remarkably persistent
triplet radical pairs upon UVꢀirradiation at room temperature, whereas no radicals were observed in solution. The factors that lead
to emergent organic radicals are correlated with the microenvironment around the benzophenone carbonyl, types of proximal hyꢀ
drogens, and the rigid supramolecular network. The absorption spectra of the linear analogs were rationalized using time dependent
density functional theory calculations on the crystal structure and in DMSO employing an implicit solvation model to describe
structural and electronic solvent effects. Inspection of the natural transition orbitals for the more important excitation bands of the
absorption spectra indicates that crystallization of the benzophenone containing molecules should present a stark contrast in photoꢀ
physical properties versus solution, which was indeed reflected by their quantum efficiencies upon solidꢀstate assembly. Persistent
organic radicals have prospective applications ranging from OLED technology to NMR polarizing agents.
INTRODUCTION
The supramolecular assembly of small molecules
through nonꢀcovalent interactions is proving to be a conꢀ
1
ꢀ3
venient approach in the design of hierarchical materials.
Controlled organization of discrete functional groups can
enhance chemical and physical properties. For example,
the solidꢀstate assembly of perylene bisimide dyes to form
4,5
transistors with nꢀtype charge transport properties and
πꢀconjugated materials that exhibit enhanced luminesꢀ
6,7
cence. Thus, further insight into how structure influꢀ
ences physical function is of great importance for the deꢀ
sign of synergistic materials with properties tailored to
specific applications. Here, we compare the impact solidꢀ
state assembly has on the photophysics of three benzoꢀ
phenone (BP) containing molecules. We report that orꢀ
ganization of BP units within distinct solidꢀstate environꢀ
ments quenches the lifetime and modulates the quantum
yield of phosphorescence. Moreover, remarkably persisꢀ
tent radicals are generated upon UVꢀirradiation at room
temperature. The quantity and stability of these radicals
vary with the chemical environment that surrounds the
key carbonyl unit (Figure 1). Thus, control over solidꢀ
state assembly of BP molecules can alter photophysical
properties and lead to the generation of persistent radical
pairs with potential applications ranging from OLED
Figure 1. Selfꢀassembly modulates the photophysics of BP derivꢀ
atives and gives rise to emergent organic radicals. (A) Structures
of ureaꢀbased BP containing macrocycles and linear analogs, 1ꢀ4.
(B) Monomers 2 and 3, presented as planar for simplicity, assemꢀ
ble through hydrogen bonding interactions. UVꢀirradiation gives
rise to persistent radicals as an emergent property. Reagents and
conditions: a. crystallization; b. UV-irradiation (360 nm, rt, under
N ). Inset: top down assembly motif of the BP sensitizer in each
2
crystal structure, 2 (left) and 3 (right).
8
ꢀ10
assembles into hexagonally packed columnar structures via
technology to NMR polarizing agents.
15
robust urea hydrogenꢀbonding interactions. Preorganization
Benzophenone, a prominent photosensitizer was first reꢀ
of the sensitizer impacted its photophysical properties by draꢀ
matically decreasing the quantum yield and lifetime. Most
1
1
ported to generate organic ketyl radicals in 1891. Although
inherent high reactivity makes electron paramagnetic resoꢀ
nance (EPR) characterizations difficult as they are known to
16
intriguingly, UVꢀirradiation of this crystalline solid gave rise
to organic radicals that persisted for weeks at room temperaꢀ
1
2
dimerize in solution forming benzopinacol. Radicals generꢀ
ated by BP in solution are unstable and are typically only obꢀ
10
ture when stored in the dark. Highꢀfield and variable temperꢀ
ature Xꢀband EPR studies accompanied by simulations sugꢀ
gested that UV irradiation of the crystals results in a resonance
stabilized radical pair through hydrogen abstraction. Our
served using EPR at low temperatures or through oneꢀelectron
13,14
reduction to form the radical anion.
Previously, we reportꢀ
10
ed a bisꢀurea macrocycle (1) that contains two BP units, which
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 2 of 8
hypothesis is that BP in the excited state abstracts a hydrogen
atom from a nearby molecule to form ketyl containing radical
pairs.
minimize the dipole moment. Bifurcated urea hydrogenꢀ
bond interactions guide the assembly of with NꢀHꢀꢀꢀO
distances ranging from 2.873(2)ꢀ2.968(2) Å (Figure 2B &
C). The BP sensitizer is ordered down the ab crystalloꢀ
graphic plane resulting in a lamellar packing motif with
aryl groups organized in an edgeꢀtoꢀface pattern and
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
2
Herein, we examine the chemical and photophysical properꢀ
ties of selfꢀorganized structures of BPꢀcontaining linear anaꢀ
logs and macrocycles (Figure 1A). Macrocycles 1 and 4 vary
the position of BP within the cyclic framework to probe how
orientation of the chromophore influences its crystalline packꢀ
ing. Linear analogs, 2 and 3, are comprised of two BP moleꢀ
cules covalently tethered through a single methylene urea
group and assemble through urea hydrogenꢀbonding interacꢀ
tions. The positions of the methyl substituents, meta or para
with respect to the BP carbonyl, were varied across two differꢀ
ent structures in order to explore their influence on crystal
packing as well as determine the types of Hꢀabstraction sites
near carbonyl oxygen. Our goal is to examine how orientation
of the BP sensitizer, as well as its relative position with respect
to Hꢀabstraction sites, impacts subsequent photophysical propꢀ
erties and if these assembled benzophenone also display the
ability to form persistent radicals upon UV irradiation or if this
emergent property is a function of the assembled macrocycles.
CgꢀꢀꢀCg distances (C = ring centroid) ranging from
g
4
.601(2)ꢀ4.825(2) Å. The CꢀHꢀꢀꢀC distances vary from
g
3.419(4)ꢀ3.637(3) Å with angles ranging from 127ꢀ135°.
The BP carbonyl oxygens reside in close proximity to aryl
protons on closely packed molecules of
distances as close as 2.60 Å.
2 with C=OꢀꢀꢀH
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Slow evaporation of
3 in dichloromethane (1 mg/ 1.6
mL) resulted in the formation of transparent needleꢀlike
crystals in the monoclinic system in the acentric P2 space
1
group. The profile of
3 is distinct, conforming to a Cꢀ
shape with both BP components oriented in close proximꢀ
ity and CgꢀꢀꢀC distances of 8.94 Å between alkyl substiꢀ
g
tuted aryl groups. The two BP carbonyls of
3 are oriented
in the same direction, although the carbonyls on neighborꢀ
ing molecules are opposing in direction. Predictable biꢀ
furcated ureaꢀurea hydrogen bonding interactions stack
the sensitizer down the a-axis with NꢀHꢀꢀꢀO distances
ranging from 2.800(6)ꢀ2.809(6) Å, Figure 3A. This asꢀ
sembly orients the BP units in a herringbone pattern along
the b-axis, while the aryl rings are parallel displaced down
the a-axis with a distance of 4.511(3) Å from centroid to
centroid. The carbonyl oxygens reside in close proximity
RESULTS AND DISCUSSION
The macrocycles and linear counterparts were syntheꢀ
sized in three to four steps using a simple alkylation of a
16
protected urea (triazinanone) as the key step. Protected
analog
see SI). Colorless solventꢀfree crystals of
obtained by recrystallization. Unfortunately, attempts to
crystallize through slow cooling, vapour diffusion, and
2
and macrocycle
4
were structurally characterized
(
2
and were
3
to benzyl and aryl protons on proximal molecules of
with C=OꢀꢀꢀH distances of 2.60 Å to methyl hydrogens,
.88 Å to methylene protons, and 2.64 Å to aryl hydroꢀ
gens.
In comparison, previously reported
transparent needleꢀlike crystals in the monoclinic system
in the P2 space group by slowꢀcooling a hot DMSO soluꢀ
3
4
2
microcrystallization techniques did not yield single crysꢀ
tals. We are currently screening a wide range of crystalliꢀ
zation techniques including conditions with potential coꢀ
crystal formers.
1
crystallizes as
Crystals of
through slow cooling in a hot acetic acid solution (120 °C,
mg/mL). The sample crystallized in the triclinic system
in the acentric space group P1 (No. 1). The conformation
of is linear with the two BP units outstretched on both
2
were obtained as transparent plates
1
1
5
tion from 120 °C. The two BP carbonyl carbons of the
monomer are 10.2 Å apart and orient the carbonyl oxꢀ
ygens pointing outward towards the exterior of the macꢀ
rocycle. Ureaꢀurea hydrogenꢀbond interactions drive asꢀ
6
2
sembly stacking the BP molecules down the aꢀaxis alignꢀ
sides of the urea tether (Figure 2A). The two BP carbonyl
groups of the monomer are aligned antiꢀparallel, likely to
ing the aryl rings in an edgeꢀtoꢀface motif with CꢀHꢀꢀꢀCg
distances ranging from 3.559(6)ꢀ3.597(7) Å and angles
from 124ꢀ130° (Figure 3B). The columns encapsulate
disordered DMSO molecules. The macrocycles are hexꢀ
agonally packed and the BP units are staggered like
1
5
brickwork along the c-axis. The BP carbonyl oxygens
are in close proximity to neighbouring methylene and aryl
hydrogens with C=OꢀꢀꢀH distances of 2.41 Å and 2.68 Å,
respectively (Figure 3C).
To probe how crystal packing of BP units impacts the
overall photophysics, we measured the absorption, emisꢀ
sion, lifetime, and quantum yield for each sample in the
‡,16
solidꢀstate and in argonꢀpurged solutions of DMSO.
Table 1 compares these measurements with unsubstituted
BP and 1. For both linear analogs and macrocycles, the
Figure 2. Single crystal Xꢀray diffraction (SCꢀXRD) of linear
analog 2, which crystallizes through slow cooling in acetic acid in
the triclinic system as transparent platelets. (A) Thermal ellipsoid
plot. (B) View of the urea hydrogenꢀbond interactions that stack
BP units on top of each other and orient the aryl rings in an edgeꢀ
toꢀface motif down the urea tape. (C) Top down view of the urea
groups showing that the edgeꢀtoꢀface aryl packing pattern is mainꢀ
tained between neighboring BP units in two directions.
absorption spectra in solution maintained the major specꢀ
troscopic properties of BP, with a strong ππ* band rangꢀ
ing from 256ꢀ270 nm and a weak spin forbidden nπ* tranꢀ
sition from 335ꢀ345 nm. The molar absorptivity for these
ꢀ1
ꢀ1
compounds range from 297ꢀ622 M cm with the para
ꢀ
substituted compounds exhibiting values higher than unꢀ
substituted BP and the metaꢀsubstituted values being lowꢀ
ACS Paragon Plus Environment

Page 3 of 8
Journal of the American Chemical Society
tween the two rings of the BP unit shifted from 26.9° to
1
2
3
4
5
6
7
8
9
31.3° (Figure S35). Additionally, the benzenes directly
connected to the urea spacers move from being in plane
with each other and roughly perpendicular to the urea unit
to a more contorted structure. Nevertheless, the spectrum
calculated with implicit solvation in DMSO shows good
agreement with the experimental being only slightly blueꢀ
shifted (5 nm) with respect to the main absorption peak
raised by ππ* transitions. The computations also find the
dark nπ* transition as the lowest excited state. In comparꢀ
ison, the spectrum of
2 calculated for the crystal structure
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
in the gas phase is shifted by 119 nm to higher energies
with respect to the experimental solidꢀstate spectrum
(Figure 4A). Even though the excitation energies differ,
the shape of the experimental spectra is reproducible and
allows for the assignment of the lower energy absorption
peaks to their corresponding electronic excitations. Simiꢀ
Figure 3. SCꢀXRD analysis of analogue 3 and macrocycle 1. (A)
Slow evaporation of 3 forms transparent needleꢀlike crystals in the
monoclinic system with BP units assembled in a herringbone
pattern along the b-axis. The aryl rings are parallel displaced
down the a-axis. (B) Macrocycle 1 crystallizes in the monoclinic
system as needleꢀlike crystals with BP units stacked down the a-
axis resulting in edgeꢀtoꢀface aryl packing down the column. (C)
The columns pack hexagonally staggering BP units across the c-
axis.
larly, the computed absorption spectra for
3 using the
crystal structure geometry in the gas phase gave roughly
the same shape with two intense absorption bands raised
by the ππ* transitions (Figure 4B). Similar to 2, this specꢀ
trum was blueꢀshifted by 135 nm compared to the experꢀ
iment. As seen before, the nπ* transition was found as the
lowest excited state. The calculated spectrum for the
structure optimized in solution was again only slightly
shifted in comparison to the experiment (4 nm).
er (Figure S24ꢀ26). In comparison, solidꢀstate assembly
of 1ꢀ3 induces an overall bathochromic shift in the specꢀ
As seen in Figure 4C and D there is a stark contrast beꢀ
tween the occupied natural transition orbitals (NTOs) for
trum λmax = 355ꢀ382 nm, with broadening in the UV/vis
region. This red shift is similar to what is observed upon
2
and 3. In solution, the occupied NTO of 2 covers the
17
entire BP unit including both benzene rings, while the
corresponding occupied NTO in gas phase with the crysꢀ
tal structure geometry, shows contributions from only one
of the BP benzene rings. Moreover, many of the NTOs
contributing to the finer structure for the gas phase specꢀ
trum show the electron density localized on only one of
the two aromatic rings of the BP units (Figure S31). In
formation of Jꢀaggregates with dyes; although this is not
a perfect analogy as the BP chromophore is not planar.
The absorption properties of
2 and 3 were examined
through time dependent density functional theory (TDꢀ
DFT) calculations to characterize the excited states of
these molecules. This was done by calculating the absorpꢀ
tion spectra of
2 and 3 using the crystal structures in the
comparison,
sity distribution moving from gas phase to solution. In
both cases for , the occupied NTO covers the entire BP
unit. These calculations suggest that should present a
stark contrast in its photophysical properties when in the
solidꢀstate versus in solution, while we
3 shows little difference in the electron denꢀ
gas phase and an optimized geometry with the polarized
1
8
continuum model (PCM) in DMSO. The excited states
3
1
9
20
were calculated at the ωB97XD /6ꢀ31+G** level of
theory. More computational details are given in the SI.
2
During optimization of 2, the average dihedral angle beꢀ
Table 1. Measured photophysical properties of BP and the BPꢀurea molecules in DMSO solution compared to the solidꢀ
state.
PHOTOPHYSICAL
1
2
3
4
BP
PROPERTIES
ꢀ
1
ꢀ1
a
a
ε (M cm )
6
22
449
317
297
342
a
a
a
a
ππ*, 270
nπ*, 345
ππ*, 260
nπ*, 335
ππ*, 256
nπ*, 340
ππ*, 265
nπ*, 340
ππ*, 270
nπ*, 345
λmax,
Solution
Abs (nm)
a
Crystals
Solution
Crystals
Solution
Crystals
Solution
Crystals
ππ*, 355
ππ*, 382
474
528
ππ*, 374
465
526
ꢀꢀ
502
ꢀꢀ
ππ*, 381
a
a
λmax,
Em (nm)
435
489
435
a
a
450
ꢀꢀ
0.32
1.5
2.0
1.3
1.5
ꢀꢀ
ꢀꢀ
τ (ns)
a
a
0.94
<0.3
5.0
23,000
ꢀꢀ
<0.1
<0.3
<0.3
<0.3
ꢀꢀ
ꢀꢀ
0.5
φ (%)
a
a
a
Values obtained from reference 16.
ACS Paragon Plus Environment

Journal of the American Chemical Society
talline lattice. Literature reports also correlate suppressed moꢀ
Page 4 of 8
23ꢀ25
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
bility with increased quantum yields.
Recent studies have
shown that haloꢀsubstituted BP units exhibit enhanced phosꢀ
23
phorescence when organized in the solidꢀstate. In solution,
the quantum yields were sufficiently diminished but they beꢀ
2
3
came highly emissive when frozen with liquid nitrogen.
Here, we show that restricting molecular motion of BP enꢀ
riched a radiative decay pathway of the triplet excited state
when only one of BP’s benzene ring participates in the excitaꢀ
tion process.
Typically, upon FranckꢀCondon excitation, BP undergoes
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
rapid intersystem crossing (ISC) from S ꢁT ꢁT excited
1
2
1
26
states, which can abstract nearby hydrogens to form ketyl
containing radical pairs as well as undergo other excited state
ES) or thermal processes. Scheffer proposed that intramoꢀ
27
(
lecular photochemical Hꢀabstraction is preferred when the
C=OꢀꢀꢀH distance is below the sum of the van der Waals radii
28
of the oxygen and hydrogen atoms (2.72 Å), while others
have observed intermolecular Hꢀabstraction by BP with
Figure 4. The excited states of 2 and 3 were characterized using
TDꢀDFT calculations. The normalized experimental solidꢀstate
absorption spectra of (A) 2 and (B) 3 are compared to their calcuꢀ
lated spectra in the gas phase including the corresponding spectral
lines; the numbers indicate the electronic excited state. NTOs for
the main transitions of (C) 2 and (D) 3 in gas phase compared to
solution, where red/blue = occupied orbital and yellow/green =
virtual orbital.
29,30
C=OꢀꢀꢀH distances as far as 3.13 Å.
Figure 5 compares the
microenvironment around the BP groups in the three systems
and shows that there are, indeed, closely preorganized hydroꢀ
gens (<2.72 Å to BP oxygen). The BP carbonyl in 1 is orgaꢀ
nized more closely to neighboring benzyl protons (2.41 Å)
versus the harder to abstract aryl protons (2.68 Å). In compariꢀ
son, in 2 the carbonyl oxygen is in close proximity to only the
aryl protons (2.60 Å), which have a higher bond dissociation
energy (BDE). BP has been found to abstract hydrogen atoms
do not expect strong media effects for 3.
31
from benzene rings, albeit slowly. Our hypothesis is that UV
irradiation of 2 crystals may produce a triplet radical pair in
low efficiency versus 1. Finally, for 3 there are proximal benꢀ
The emission spectra recorded in solution (DMSO, 0.9
ꢀ 1.0 mM) exhibited transitions ranging from 435ꢀ502 nm
and displayed an overall red shift upon solidꢀstate assemꢀ
zylic CH (2.60 Å), benzylic methylenes (2.88 Å), and aryl
bly, λem = 450ꢀ528 nm. The phosphorescence lifetime of
and in solution (DMSO, 0.9 mM) were the shortest at
.5 ns, while exhibited a slightly longer lifetime at 2.0
ns. Upon solidꢀstate assembly the lifetimes of and
2
3
protons (2.64 Å), which suggests that several different triplet
radical pairs could be formed. Simple BDE arguments predict
the ketyl radicals may be formed more easily in compounds 1
and 3 as compared with 2, which only contains close aryl hyꢀ
drogens; as homolytic BDEs are lower for benzyl protons verꢀ
4
1
3
2
3
were slightly quenched to 0.94 ns and 1.3 ns, respectively.
Such deactivation in lifetime suggests that these comꢀ
pounds are prone to intermolecular selfꢀquenching similar
2
1,22
to other BP compounds.
A comprehensive study on
this phenomenon by the GarciaꢀGaribay group demonꢀ
strated that the lifetime of BP nanocrystals with electron
donating substituents are dramatically shorter than in soꢀ
lution, varying over 9 orders of magnitude depending on
2
2
the electron donating ability of the substituents. This is
attributed to intermolecular selfꢀquenching via a charge
2
2
transfer mechanism. The shorter observed lifetimes for
and in the crystals are consistent with these prior reꢀ
ports, as the alkyl groups are mildly donating.
2
3
The phosphorescent quantum yields of 2ꢀ4 in DMSO soluꢀ
tion (25 ꢁM and 1 mM) displayed efficiencies of less than
0
.3% in all cases. The low quantum efficiency is attributed to
unrestricted rotation and vibrations of the sensitizer when alꢀ
lowed to move freely in solution. Interestingly, crystallization
of 2 dramatically increased its quantum yield to 5.0% but did
not influence 3 as predicted by computation. The calculations
suggest that the solidꢀstate geometry of 2 forces each of BPs
benzene rings to act independently, whereas in solution the
linear analog is able to orient itself so that both benzenes parꢀ
ticipate in the excitation of the π bands resulting in a loss of
independent chromophores. This demonstrates that the higher
quantum yield observed for 2 upon solidꢀstate assembly is
likely due to suppressed mobility when locked within the crysꢀ
Figure 5. Comparison of the microenvironments around the BP
carbonyl obtained from the SCꢀXRD of compounds 1ꢀ3. (A) 1 has
neighboring aryl and benzyl CH protons. (B) The carbonyl of 2
2
resides in close proximity to only neighboring aryl protons. (C)
The more complex structure of 3 is oriented close in space to two
types of benzyls protons (CH and CH ) as well as (D) aryl proꢀ
3
2
tons.
ACS Paragon Plus Environment

Page 5 of 8
Journal of the American Chemical Society
32
sus aryl (88 kcal/mol vs 111 kcal/mol, respectively). Despite
this, BP has been known to abstract all three types of proꢀ
tons.
1
2
3
4
5
6
7
8
9
29ꢀ31
In an effort to correlate structure with the formation and
stability of the UVꢀgenerated triplet radical pairs, we
turned to Xꢀband EPR spectroscopy. First, EPR spectra
were recorded on solutions of
2 and 3 in dichloromethane
§
(1 mM) pre and post UVꢀirradiation (1 h). As expected,
no EPR signal was observed pre or post UV indicating
that any ketyl radical formed is quickly terminated in soꢀ
lution (Figure S38). Upon UV irradiation the linear anaꢀ
log solutions yellowed and showed only minor spectroꢀ
scopic changes by absorption spectroscopy (Figure S43).
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Next, solidꢀstate EPR spectra were recorded on triply
recrystallized samples of
2
and
UV. After one hour of UV irradiation, the transparent
crystals of turned reddishꢀbrown in color (Figure 6A),
while the needleꢀlike crystals of became opaque upon
3 (~10 mg) pre and post
§§
2
3
removal from the mother liquor and showed a slight yelꢀ
lowing in color upon UV irradiation (Figure 6B). UV
irradiation of the crystals resulted in the formation of radꢀ
icals in both
respectively (Figure 6 and S39). Irradiation of
gave rise to an isotropic EPR signal. The EPR line width
of was similar, but a weak second transition was obꢀ
2
and
3
with gꢀvalues of 2.005 and 2.007,
2
crystals
Figure 6. Photophysical properties of the triply recrystallized
samples of 2 and 3 pre and post UVꢀirradiation. (A) The transparꢀ
ent crystals of 2 exhibit green fluorescence under UV light and
become brownꢀred upon UVꢀirradiation. (B) The needleꢀlike crysꢀ
tals of 3 show quenched emission and only slightly yellow in
color after UVꢀirradiation. (C) EPR of 2 post UV and subsequent
dark decay study demonstrating that the radicals are persistent for
several days at rt. (D) EPR of 3 exhibits persistent radicals after
irradiation with a significant change in EPR line shape within 2 h
post UV. Comparison of absorption spectra of (E) 2 and (F) 3 in
solution and their recrystallized solids pre and post UV irradiation
for one hour, the new absorbance band in 2 at λ = 557 nm is laꢀ
belled.
3
served at g = 2.003. A gꢀvalue of 2.003 has previously
33
been attributed to the BP ketyl.
The concentration of radical pairs generated after one
hour of UVꢀirradiation was approximated using a calibraꢀ
tion of standard solutions of TEMPO in benzene (Figure
1
0,34
S40).
the overall area of the spectra, which were then compared
to the TEMPO calibration. One hour UVꢀirradiation of
Double integration of the EPR signals provides
2
generated the same amount of radical as a 0.053 mM soꢀ
lution of TEMPO in benzene, suggesting that approxiꢀ
mately 1 in 5,000 molecules of
parison, after similar UVꢀirradiation host
2
have a radical. In comꢀ
showed ~1 in
0,000 molecules have a radical, whereas the linear anaꢀ
log shows radicals in ~1 in 25,000 molecules (similar to
1
between 2 and 3 hours after irradiation with an increase in
population of radicals at g = 2.003. The overall line width
of the spectra remained similar, but the area of the signal
was decreased by half just 45 hours after UV irradiation
1
0
3
3
a 0.009 mM TEMPO solution). The amount of UVꢀ
generated radical formed increases with longer irradiation
(Figure S42). In contrast, macrocycle
1 displays a persisꢀ
times. This result shows that
five times more radical than
2
generates approximately
tent radical, which exhibits a modest amount of radical 26
days after irradiation. In accordance with the spin selecꢀ
tion rule, recombination reactions of triplet geminate radiꢀ
cal pairs are forbidden and must first undergo ISC to yield
3
after one hour of UV irraꢀ
diation and demonstrates that radical formation is not deꢀ
terred by higher homolytic BDEs. Though, it is possible
that the persistence of the radical pair is playing a role in
the observed concentration difference.
2
7,33
a singlet radical pair in order to form products.
In soꢀ
lution, Hꢀabstraction by BP generally occurs in 10ꢀ100 ns
while recombination is considered the rateꢀlimiting step
Therefore, the persistence of the radicals was probed
using dark decay studies where the samples were stored at
room temperature in the dark after irradiation and EPR
spectra were recorded over time. The dark decay study of
2
7
(
>1 ꢁs). The enhanced stability of these emergent radiꢀ
cals in the solidꢀstate post UVꢀirradiation is attributed to
delocalization of the radical pairs, which is further stabiꢀ
lized by the rigidity of the BP units upon assembly. Studꢀ
ies have shown that selfꢀrecombination reactions of the
BP ketyl have a rate constant that is an order of magniꢀ
2
shows that there was little change in line shape and gꢀ
value (2.005) 140 days post UVꢀirradiation (Figure 6C).
Sixty days after UVꢀirradiation, the area of the EPR signal
retained half its initial amount, demonstrating the remarkꢀ
2
7
tude lower than crossꢀreactions. This seems to be reꢀ
flected in the stability of as the ketyl radical signal (g =
3
able persistence of the radicals of
parison, dark decay studies on recrystallized
2
(Figure S41). In comꢀ
showed a
2.003) became more prominent the longer the sample
remained at room temperature.
3
faster decay and exhibited dramatic changes in the EPR
line shape (Figure 6D). Post UV the broad EPR line exꢀ
hibited a gꢀvalue of 2.007 with a weak transition at g =
.003. Two hours after irradiation the EPR signal retained
a similar line shape although a stark change was observed
Figure 6EꢀF compares the absorption spectra of triply
recrystallized samples (2 and 3) before and after 1h irraꢀ
diation. The absorption spectra of both crystalline samples
post UV retained their major spectroscopic properties,
2
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 6 of 8
although both signals broadened into the visible region.
Most intriguingly, irradiation of afforded a new absorbꢀ
ance band at 557 nm which is consistent with where both
after UVꢀirradiation (1 h) varied sixꢀfold with
ingly showing the largest amount even though only aryl
protons with higher BDE are close in proximity (2.60 Å)
for abstraction, while macrocycle exhibited the least
amount of radicals. Radicals of also displayed the greatꢀ
est persistence, exhibiting approximately half the EPR
signal after 140 days. In each case, the persistence of the
UVꢀgenerated radicals was attributed to resonance stabiliꢀ
zation about the rigid crystalline framework and may shed
light on the impact solidꢀstate assembly has on the reꢀ
combination of ketyl containing radicalꢀpairs. Future
work will be focused on elucidating the factors that govꢀ
ern the formation, stability, and applications of the radiꢀ
cals. We are also exploring the efficiency of these BP
sensitizers to undergo tripletꢀtriplet annihilation pathways
with molecular oxygen to generate reactive oxygen speꢀ
cies.
2 surprisꢀ
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
2
1
2,35,36
the triplet and resulting ketyl absorb.
It should be
1
noted that the triplet state of BP’s absorption peak signifiꢀ
cantly overlaps with its corresponding ketyl, however the
triplet is known to absorb out to wavelengths >600
2
3
5,36
nm.
served in the spectra of
UVꢀirradiation of affords increased amounts of radicals
versus the other derivatives. Samples of and were anaꢀ
This long wavelength absorption was not obꢀ
10
3
or
1
,
again suggesting that
2
2
3
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
lysed by SCꢀXRD after UV irradiation and revealed no
1
significant structural changes. Similarly, H NMR spectra
were obtained on irradiated samples showing no spectral
changes, which is consistent with the estimated concentraꢀ
tion of the radicals (Figure S44ꢀS45). Finally, the emisꢀ
sion behaviour of the UV irradiated crystals was also inꢀ
vestigated with no major changes observed upon excitaꢀ
tion at 355 nm (Figure S27).
ASSOCIATED CONTENT
We have demonstrated that UVꢀirradiation of selfꢀ
assembled BP containing molecules can give rise to perꢀ
sistent organic radicals in marked contrast to their behavꢀ
iour in solution. The concentration of the radicals is low
but is influenced by structure and assembly, as is their
persistence. Paraꢀsubstitution of BP containing radical
Supporting Information. Supporting data including synthetic
procedures, product characterization, Xꢀray, photophysical charꢀ
acterization, and EPR spectra are available free of charge via the
Internet at http://pubs.acs.org.
AUTHOR INFORMATION
Corresponding Author
pairs resulted in longerꢀlived radical species while meta
ꢀ
substituted radical pairs displayed decreased stability. A
comprehensive study on a library of BP containing crysꢀ
tals with varying substituent patterns may be fruitful to
further elucidate the rules that govern ketyl radical pair
formation and their subsequent stability.
*shimizls@mailbox.sc.edu
ORCID
Linda Shimizu: 0000ꢀ0001ꢀ5599ꢀ4960
Leticia González: 0000ꢀ0001ꢀ5112ꢀ794X
CONCLUSIONS
Author Contributions
In summary, three new BPꢀcontaining molecules were
synthesized and two afforded single crystals that assemꢀ
bled the photosensitizer through ureaꢀurea hydrogen
bonding interactions. We investigated the impact solidꢀ
state assembly has on their photophysics and explored
their ability to form persistent radicals as a result of UVꢀ
irradiation. Solidꢀstate assembly of the materials resulted
in a bathochromic shift in both their absorption and emisꢀ
sion spectra and quenched their phosphorescent lifetime,
which is attributed to BPs selfꢀquenching character. The
The manuscript was written through contributions of all authors.
Notes
‡
DMSO, an aggressive solvent, was selected because macrocyꢀ
cle
1 and 4 are poorly soluble. Absorption and emission specꢀ
tra are strongly influenced by solvent polarity so samples were
kept consistent when possible.
§
EPR experiments were performed in N degassed dichloroꢀ
2
methane, DMSO was not selected as the water impurity abꢀ
sorbed the microwaves.
§§
The samples were triply recrystallized to ensure that any
quantum efficiency of
although crystallization influenced their quantum yield
differently. Crystallization of enhanced its quantum
efficiency by an order of magnitude but did not influence
2 and 3 was <0.3% in solution,
radical generated is a result of BP photochemistry.
2
ACKNOWLEDGMENTS
This work was supported in part by the National Science Foundaꢀ
tion (CHEꢀ1608874 and OIAꢀ1655740), a Fulbright US Scholar
award, and the MarieꢀCurie Action (TCCMꢀINTꢀEJDꢀ642294).
We thank Dr. Tom Makris and Dr. Jose Amaya for thrilling EPR
discussions.
3. TDꢀDFT calculations on the crystal structures of
2 and
3
in the gas phase and in solution were consistent with
these experimental observations. The computations sugꢀ
gested that crystallization of and would influence their
photophysical properties differently, predicting a dramatic
change in photophysics for and little or no difference
was expected for
Selfꢀassembly of compounds 1ꢀ3 resulted in three disꢀ
2
3
2
ABBREVIATIONS
3.
BP, Benzophenone; DMSO, dimethylsulfoxide; EPR, electron
paramagnetic resonance; DNP, dynamic nuclear polarization;
NMR, nuclear magnetic resonance; OLED, organic light emitꢀ
ting diode; SCꢀXRD, single crystal Xꢀray diffraction; TDꢀ
DFT, time dependent density functional theory; NTO, natural
transition orbital; PCM, polarized continuum model; ES, exꢀ
cited state; ISC, intersystem crossing; BDE, bond dissociation
energy; TEMPO, 2,2,6,6ꢀTetramethylpiperidine 1ꢀoxyl.
tinct crystal structures that vary the microenvironment
around the BP carbonyl. Remarkably, all of the crystalline
compounds exhibit persistent radicals upon UVꢀ
irradiation even though no radicals were observed in soluꢀ
tion. The radical formation is attributed to BP carbonyls’
close proximity to neighboring Hꢀabstractions sites within
the crystal structures. The amount of radicals generated
ACS Paragon Plus Environment

Page 7 of 8
Journal of the American Chemical Society
REFERENCES
20. Ditchfield, R.; Hehre, R. W.; Pople, J. A. J. Chem.
Phys. 1971 54, 724ꢀ728.
21. Simoncelli, S.; Kuzmanich, G.; Gard, M. N.; Garciaꢀ
1
2
3
4
5
6
7
8
9
,
1
2
3
.
.
.
Lehn, J. ꢀM. Proc. Natl. Acad. Sci. USA 2002
763ꢀ4768.
Stupp, S. I.; Palmer, L. C. Chem. Mater. 2014
07ꢀ518.
Adhikari, B.; Lin, X.; Yamauchi, M.; Ouchi, H.;
Aratsu, K.; Yagai, S. Chem. Commun. 2017 53, 9663ꢀ
683.
Würthner, F.; Kaiser, T. E.; SahaꢀMöller, C. R. Angew.
Chem. Int. Ed. 2011 50, 3376ꢀ3410.
Würthner, F.; SahaꢀMöller, C. R.; Fimmel, B.; Ogi, S.;
Leowanawat, P.; Schmidt, D. Chem. Rev. 2016 116
62ꢀ1052.
Mei, J.; Leung, N. L. C.; Kwok, R. T. K.; Lam, J. W.
Y.; Tang, B. Z. Chem. Rev. 2015 115, 11718ꢀ11940.
,
99
,
4
Garibay, M. A. J. Phys. Org. Chem. 2010, 23, 376ꢀ
381.
,
26,
5
22. Kuzmanich, G.; Simoncelli, S.; Gard, M. N.; Spänig,
F.; Henderson, B. L.; Guldi, D. M.; GarciaꢀGaribay,
,
M. A. J. Am. Chem. Soc. 2011, 133, 17296ꢀ17306.
9
2
3. Yuan, W. Z.; Shen, X. Y.; Zhao, H.; Lam, J. W. Y.;
Tang, L.; Lu, P.; Wang, C.; Liu, Y.; Wang, Z.; Zheng,
Q.; Sun, J. Z.; Ma, Y.; Tang, B. Z. J. Phys. Chem. C
4
.
.
,
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
5
2010
24. Yang, J.; Zhen, X.; Gao, X.; Ren, Z.; Wang, J.; Xie,
Y.; Li, J.; Peng, Q.; Pu, K.; Li, Z. Nat. Commun. 2018
, 1ꢀ10.
25. Gong, Y.; Tan, Y.; Li, H.; Zhang, Y.; Yuan, W.;
, 114, 6090ꢀ6099.
,
,
9
,
6
.
.
9
,
7
Shi, J.; Aguilar Suarez, L. E.; Yoon, S. –J.; Varghese,
S.; Serpa, C.; Park, S. Y.; Lüer, L.; RocaꢀSanjuán, D.;
MiliánꢀMedina, B.; Gierschner, J. J. Phys. Chem. C
Zhang, Y.; Sun, J.; Tang, B. Z. Sci. China Chem. 2013
,
56, 1183ꢀ1186.
26. Marazzi, M.; Mai, S.; RocaꢀSanjuán, D.; Delcey, M.
2
017
Lee, S. Y.; Yasuda, T.; Yang, Y. S.; Zhang, Q.;
Adachi, C. Angew. Chem. Int. Ed. 2014 53, 6402ꢀ
406.
,
121, 23166ꢀ23183.
G.; Lindh, R.; González, L.; Monari, A. J. Phys. Chem.
Lett. 2016, 7, 622ꢀ626.
8
.
.
27. Dormán, G.; Nakamura, H.; Pulsipher, A.; Prestwich,
G. D. Chem. Rev. 2016 116, 15284ꢀ153998.
8. Scheffer, J. R. Acc. Chem. Res. 1980 13, 290ꢀ296.
29. Ito, Y.; Matsuura, T.; Tabata, K.; JiꢀBen, M.; Fukuyaꢀ
ma, K.; Sasaki, M.; Okada, S. Tetrahedron 1987 43
307ꢀ1312.
,
,
6
2
,
9
Cao, W.; Wang, W. D.; Xu, H. ꢀS.; Sergeryev, I. V.;
Struppe, J.; Wang, X.; MentinkꢀVigier, F.; Gan, Z.;
Xiao, M. ꢀX.; Wang, L.ꢀY.; Ding, S. ꢀY.; Bai, S.;
,
,
1
30. Koshima, H.; Hessler Bittl, D. P.; Miyoshi, F.; Yang,
Wang, W. J. Am. Chem. Soc. 2018, 140, 6969ꢀ6977.
Y.; Matsuura, T. J. Photochem. Photobiol. A: Chem.
1
0. DeHaven, B. A.; Tokarski III, J. T.; Korous, A. A.;
MentinkꢀVigier, F.; Brugh, A. M.; Forbes, M. D. E.;
Makris, T. M.; van Tol, J.; Bowers, C. R.; Shimizu, L.
1
995, 86, 171ꢀ176.
3
1. Buettner, A. V.; Dedinas, J. J. Phys. Chem. 1971, 75,
187ꢀ191.
S. Chem. Eur. J. 2017
,
23, 8315ꢀ8319.
32. McMillen, D. F.; Golden, D. M. Ann. Rev. Phys.
Chem. 1982
3. Woodward, J. R. Prog. React. Kinet. Mech. 2002
65ꢀ207.
, 33, 493ꢀ532.
1
1. Scott, T. A.; Ooro, B. A.; Collins, D. J.; Shatruk, M.;
Yakovenko, A.; Dunbar, K. R.; Zhou, H. ꢀC. Chem.
Commun. 2009, 65ꢀ67.
3
3
, 27,
1
4. Danielsson, R.; Albertsson, P.ꢀA.; Mamedov, F. Styrꢀ
1
1
1
1
2. Iwakura, Y.; Takeda, K.; Nakazawa, T.; Watarai, O. J.
ing, S. Biocimica. Et Biophysica. Acta. 2004, 1608, 53ꢀ
61.
Polymer Sci. C: Polymer Lett. 1968
3. Murai, H.; Imamura, T.; Obi, K. Chem. Phys. Lett.
982 87, 295ꢀ298
4. Fischer, P. H. H.; Zimmermann, H. Z. Naturforsch A
968 23, 1339ꢀ1342.
5. Dewal, M. B.; Xu, Y.; Yang, J.; Mohammed, F.;
Smith, M. D.; Shimizu, L. S. Chem. Commun. 2008
909ꢀ3911.
6. Geer, M. F.; Walla, M. D.; Solntsev, K. M.; Strassert,
C. A.; Shimizu, L. S. J. Org. Chem. 2013 78, 5568ꢀ
578.
7. Cai, K.; Xie, J.; Zhang, D.; Shi, W.; Yan, Q.; Zhao, D.
J. Am. Chem. Soc. 2018 140, 5764ꢀ5773.
8. Mennucci, B. Wiley Interdiscip. Rev.: Comput. Mol.
Sci. 2012 , 386ꢀ404.
9. Chai J. D.; HeadꢀGordon, M. Phys. Chem. Chem. Phys.
008 10, 6615ꢀ6620.
,
6
, 115ꢀ118.
35. Turro, N. J.; Ramamurthy, V.; Scaiano, J. C. Modern
Molecular Photochemistry of Organic Molecules; Uniꢀ
versity Science Books: Sausalito, 2010; pp 483 – 624.
1
,
36. Encinas, M. V.; Scaiano, J. C. J. Am. Chem. Soc. 1981
,
1
,
103, 6393ꢀ6397.
,
3
1
,
5
1
1
1
,
,
2
2
,
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 8 of 8
Table of Contents (TOC).
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
8
ACS Paragon Plus Environment