Photoalignment layers for liquid crystals from the di-π-methane rearrangement

Article abstract of DOI:10.1021/ja312090p

Photoalignment of nematic liquid crystals is demonstrated using a di-π-methane rearrangement of a designed polymer. The alignment mechanism makes use of the strong coupling of the liquid crystal directors to dibenzobarrelene groups. The large structural c

Full text of DOI:10.1021/ja312090p

Communication
pubs.acs.org/JACS
Photoalignment Layers for Liquid Crystals from the Di-π-methane
Rearrangement
Jason R. Cox, Jeffrey H. Simpson, and Timothy M. Swager*
Department of Chemistry, Massachusetts Institute of Technology, Cambridge, Massachusetts 02139, United States
S
* Supporting Information
ABSTRACT: Photoalignment of nematic liquid crystals is
demonstrated using a di-π-methane rearrangement of a
designed polymer. The alignment mechanism makes use of
the strong coupling of the liquid crystal directors to
dibenzobarrelene groups. The large structural changes that
accompany photoisomerization effectively passivate seg-
ments of the polymer, allowing the remaining dibenzo-
barrelene groups to dominate the director alignment.
Photoisomerization requires triplet sensitization, and the
polymer was designed to have a uniaxially fixed rigid
structure and rapid triplet energy transfer from the
proximate benzophenone units to the dibenzobarrelene
groups. The isomerization was observed to be regiospe-
cific, and thin films showed alignment.
evices that exploit the unique opto-electronic and self-
Dassembling characteristics of liquid crystalline (LC)
materials are a cornerstone of modern technology.1 A majority
Figure 1. Photochemical behavior of dibenzobarrelene substrates6
under direct and sensitized irradiation conditions (top). Polymeric di-
π-methane rearrangement of P1 facilitated by intrapolymer triplet
sensitization (bottom).
of technologies require a uniform alignment of the LC
phasesa challenge that traditionally has been met through
the unidirectional “rubbing” of polymeric substrates.2 Despite
the crude nature of this approach, it continues to be a dominant
method to induce LC alignment. The rubbing approach,
however, is not without disadvantagesspecifically, it can
serve as a starting point for the design of a polymer capable of
undergoing a regioselective di-π-methane rearrangement.
generate dust3 and static charge at the interface of the substrate,
is not compatible with surfaces with complex topographies, and
can damage sensitive electronic components that are part of
integrated devices.4 To overcome this problem, noncontact
alignment methods have been developed, the majority of which
employ axis-selective photochemical transformations, to gen-
erate highly aligned LC phases.4,5 Many of these approaches
exploit well-known photochemical reactions such as the
isomerization of unsaturated centers and photochemically
allowed cycloaddition reactions. A conspicuously undeveloped
photochemical reaction for this application is the well-known
di-π-methane rearrangement, which is a robust photochemical
transformation that generates large shape changes in the
molecular structure of the reactant (see Figure 1, 1→2).6 To
address the need for new, efficient transformations that can
create alignment layers, we report here the application of the di-
π-methane rearrangement to create films capable of alignment
of nematic liquid crystals.
Drawing upon the ability of triptycene structures to enhance
the alignment of LC molecules through the minimization of
internal free volume (IFV),8 we envisioned that substituted
dibenzobarrelene structures could provide similarly strong
director field interactions with the LC phases (Figure 2).
Upon photolysis, the expected semibullvalene-like products
would exhibit diminished IFV, thus eliminating the alignment
preferences of the dibenzobarrelenes.
As shown in Figure 1 (top), the di-π-methane rearrangement
of rigid structures such as dibenzobarrelene 2 proceeds through
a triplet manifold, whereas direct irradiation primarily yields 3.9
Noting this behavior, we incorporated a triplet sensitizer in the
polymer backbone directly adjacent to the di-π-methane
substrate (Figure 1, P1). Such a design motif is attractive for
two reasons: (1) the proximity of sensitizer and substrate
enforces rapid energy transfer,10 and (2) once sensitization of
the di-π-methane substrate is complete, newly formed
benzophenone diradicals11 can cross-link the film, thereby
Zimmerman and co-workers described the di-π-methane
rearrangement in full detail, yielding predictive measures for the
determination of regioselectivity and the role of excited-state
multiplicity in the reaction mechanism.7 These seminal works
Received: December 11, 2012
© XXXX American Chemical Society
A
dx.doi.org/10.1021/ja312090p | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Communication
Scheme 1. Synthesis of Monomers (4, 5), Polymer (P1), and
Model Compound (6)
Figure 2. Schematic depiction of proposed photoalignment mecha-
nism: red/green barbed structures, P1/P1a; blue ellipsoids, LC
molecules.
substituted barrelene scaffoldssuch substrates could poten-
yielding an insoluble polymeric layer between the LC layer and
substrate. The first point is of critical importance, given that the
initial directional information imparted by photolysis with
plane-polarized light can be lost through energy transfer to
non-nearest neighbors. Hence, it is key that the molecular
locations that absorb the light undergo photoisomerization.
Our proposed photoalignment is schematically depicted in
Figure 2 and operates as follows: (1) selective excitation of
benzophenone units by polarized light with transition dipole
moments oriented parallel to the electric vector of the exciting
light, (2) rapid intersystem crossing (ISC) to the first triplet
state, (3) fast energy transfer to adjacent di-π-methane units,
and (4) subsequent rearrangement. In thin films, such a cascade
would result in a new surface landscape where molecules
residing in the plane perpendicular to the exciting light would
be left unreacted and still capable of strongly influencing the
LC director though strong interactions with the dibenzobarre-
lene groups. The isomerized units lack the clefts that created
the IFV and the strong interactions with the LC directors,
thereby allowing the remaining P1 to dictate the direction of
LC alignment.
The syntheses of monomers 4 and 5, as well as the synthesis
of P1, are outlined in Scheme 1. The synthetic approaches to
monomers 4 and 5 are straightforward, affording both
components in reasonable yields via short synthetic sequences.
The Sonogashira−Hagihara reaction was chosen to generate
the target polymer P1 because it is a robust cross-coupling
polymerization method for the synthesis of poly(phenylene
ethynylenes) (PPEs).12 A critical element of the design is that
the polymer backbone rigidly fixes the dibenzobarrelene
orienting groups along a given direction, and similarly the
passivated photoisomerized groups are locked in position. This
is accomplished by a two-fold connection through ethynyl
linkages that uniaxially fixes the direction of monomer units in
P1.
tially yield two racemic photoproducts (Figure 3).
Figure 3. Postulated photochemical pathways available to model
compound 6 after sensitized excitation.
We surmised that the extended conjugation available to
Diradical 1 would favor Pathway 1, thus yielding rac-DP1 as
the major product. Upon sensitized photolysis of 6, only a
1
single photoproduct was observed by H NMR, as shown in
Figure 4. During the photolysis, we monitored the emergence
of the two new methine resonances (Figure 4, shown as red and
green), which was concomitant with a reduction of the original
bridgehead methine (normalized in Figure 4, shown as blue).
The new signals belong to the benzylic and doubly benzylic
cyclopropyl methines that are part of the newly assembled
semibullvalene-like structure. Gratifyingly, further 2D NMR
analysis (see Supporting Information) revealed that the newly
formed product was rac-DP1 as initially predicted.
Having established that the reaction proceeds efficiently and
selectively at the small molecule level, we turned our attention
to the photochemical behavior of P1. For these reactions, the
photochemistry was performed under ambient conditions,
without the exclusion of oxygen from the reaction vessel.
Conducting photoalignment operations under strictly inert
conditions is not optimal for industrial processes. We
Prior to engaging directly into photochemical studies of P1,
we conducted small-molecule studies employing model
compound 6 in order to gain insight into the photochemical
behavior of this class of substrates. We were particularly
interested in the implications of using unsymmetrically
B
dx.doi.org/10.1021/ja312090p | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Communication
notion that the reaction proceeds with a high level of
regioselectivity.
With a firm understanding of the photochemical behavior of
P1, we directed our efforts toward the fabrication of LC
alignment layers. In these experiments, P1 was spun-coated
onto glass slides and irradiated with polarized UV light (see
Supporting Information for full details). LC cells were
fabricated and loaded with a small quantity of MLC-6884.13
To evaluate the alignment, the test cells were heated to the
clearing point, slowly cooled to the nematic phase, and then
analyzed using a polarizing microscope. As shown in Figure 6,
Figure 4. Di-π-methane rearrangement of model compound 6 (top).
Time-elapsed 1H NMR spectra recorded during the sensitized
photolysis of 6 (bottom). The intensity is normalized to the resonance
at 6.1 ppm, and the signals in the blue, green, and red boxes are
assigned to the protons colored similarly in the structures.
hypothesized that the proximity of sensitizer and substrate
would facilitate the transformation, despite the fact that such
reactions typically require rigorous exclusion of oxygen to
prevent quenching of the triplet state by oxygen. In addition,
we were interested in verifying that the observed product
distribution at the small-molecule level translated to P1. In
order to generate robust photoalignment layers based on the
mechanism depicted in Figure 2, it is necessary to ensure the
formation of rac-DP1. The molecular structure of this product
contains two methyl ester moieties, projecting in a fashion that
can block interactions between liquid crystal molecules and the
plane created by the rigid phenylethynyl linkages, effectively
eliminating strong director field interactions. The alternative
photoproduct, rac-DP2, projects the methyl ester groups away
from the rigid PPE backbone and therefore could possibly
generate competitive alignment sites.
Figure 6. Polarized optical micrographs of LC cells containing MLC-
6884 and thin films of P1. The upper images correspond to the non-
irradiated control cell, and the bottom images correspond to cells that
have been irradiated by polarized UV light.
the non-irradiated cell does not exhibit any alignment. This is
evidenced by the fact that there is no apparent alignment in the
schlieren textures, and rotation by 45° does not elicit any
changes in intensity in the transmission of light through the
cell. In stark contrast, the irradiated cell viewed under cross-
polarization displays a uniform alignment, without a schlieren
texture, that displays extinction when one of the polarizers is
aligned with the direction of the polarization of the photolysis.
The cell becomes bright when the cell direction is turned so
that its alignment axis is oriented at 45° to the crossed
polarizers of the microscope.
As shown in Figure 5, the rearrangement of P1 takes place
smoothly and cleanly in the presence of oxygen. Similar to the
regioselectivity observed with model compound 6, only a single
set of product methines are evident, lending credence to the
In summary, we have shown that the di-π-methane
rearrangement is an effective approach toward the photoalign-
ment of LC phases. The reaction is clean and highly
regioselective at both the small-molecule and polymer scales.
We postulate that the alignment results from the minimization
of IFV via the interaction of unreacted P1 dibenzobarrelene
moieties and rigid rod nematogens. Further work is currently
underway to explore this concept in greater detail.
ASSOCIATED CONTENT
* Supporting Information
Details of synthesis, photolysis, and device fabrication. This
material is available free of charge via the Internet at http://
pubs.acs.org.
■
S
1
Figure 5. Di-π-methane rearrangement of P1 (top). Time-elapsed H
AUTHOR INFORMATION
Corresponding Author
tswager@mit.edu
■
NMR spectra of the reaction (bottom). The signals in the blue, green,
and red boxes are assigned to the protons colored similarly in the
structures.
C
dx.doi.org/10.1021/ja312090p | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Communication
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
This work was supported by the National Science Foundation
DMR-1005810 and Army Research Office through the Institute
for Soldier Nanotechnologies (W911NF-07-D-004).
REFERENCES
■
(1) (a) Kim, K.-H.; Song, J.-K. NPG Asia Mater. 2009, 1, 29.
(b) Semenza, P. Inf. Disp. 2009, 25, 22. (c) Lee, K. H.; Kim, H. Y.;
Park, K. H.; Jang, S. J.; Park, I. C.; Lee, J. Y. Dig. Tech. Pap.-Soc. Inf.
Disp. Int. Symp. 2006, 37, 1079.
(2) Schadt, M.; Helfrich, W. Appl. Phys. Lett. 1971, 18, 127.
(3) Kawatsuki, N. Chem. Lett. 2011, 40, 548.
(4) Yaroshchuk, O.; Reznikov, Y. J. Mater. Chem. 2012, 22, 286.
(5) The examples are too numerous to list. For representative
examples on the topic, please see the following and references cited
therein: (a) Chigrinov, V. G.; Kozenkov, V. M.; Kwok, H. S.
Photoalignment of Liquid Crystalline Materials: Physics and Applications;
John Wiley & Sons: England, 2008. (b) Ichimura, K. Chem. Rev. 2000,
100, 1847.
(6) Zimmerman, H. E.; Grunewald, G. L. J. Am. Chem. Soc. 1966, 88,
183.
(7) For reviews on the topic, see the following: (a) Hixson, S. S.;
Mariano, P. S.; Zimmerman, H. E. Chem. Rev. 1973, 73, 531.
(b) Zimmerman, H. E.; Armesto, D. Chem. Rev. 1996, 96, 3065.
(8) (a) Long, T. M.; Swager, T. M. Adv. Mater. 2001, 13, 601.
(b) Long, T. M.; Swager, T. M. J. Am. Chem. Soc. 2002, 14, 3826.
(c) Zhu, Z.; Swager, T. M. J. Am. Chem. Soc. 2002, 124, 9670.
(9) (a) Zimmerman, H. E.; Givens, R. S.; Pagni, R. M. J. Am. Chem.
Soc. 1968, 90, 6096. (b) Zimmerman, H. E.; Givens, R. S.; Pagni, R. M.
J. Am. Chem. Soc. 1968, 90, 4191.
(10) Turro, N. J. Modern Molecular Photochemistry; University
Science Books: Sausolito, CA, 1991.
(11) (a) Dorman, G.; Prestwich, G. D. Biochemistry 1994, 33, 5661.
(b) Christensen, S. K.; Chiappelli, M. C.; Hayward, R. C.
Macromolecules 2012, 45, 5237.
(12) Bunz, U. H. F. Chem. Rev. 2000, 100, 1605.
(13) MLC-6884 is negative dielectric anisotropic nematic LC mixture
produced by Merck KGaA.
D
dx.doi.org/10.1021/ja312090p | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX