Nanoscale phase analysis of molecular cooperativity and thermal transitions in dendritic nonlinear optical glasses

Article abstract of DOI:10.1021/jp307370y

A broad nanoscopic study of a wide-range of dendritic organic nonlinear optical (NLO) self-assembly molecular glasses reveals an intermediate thermal phase regime responsible for both enhanced electric field poling properties and strong phase stabilizatio

Full text of DOI:10.1021/jp307370y

Article
pubs.acs.org/JPCB
Nanoscale Phase Analysis of Molecular Cooperativity and Thermal
Transitions in Dendritic Nonlinear Optical Glasses
Daniel B. Knorr, Jr.,†,‡ Stephanie J. Benight,§ Brad Krajina,† Cheng Zhang,∥ Larry R. Dalton,§
and Rene M. Overney*,†
́
†Department of Chemical Engineering, University of Washington, Seattle, Washington 98195-1750, United States
‡Weapons and Materials Research Directorate, United States Army Research Laboratory, Aberdeen Proving Ground, Maryland
21009, United States
§Department of Chemistry, University of Washington, Seattle, Washington 98195-1700, United States
∥Department of Chemistry and Biochemistry, South Dakota State University, Brookings, South Dakota 57007-0896, United States
ABSTRACT: A broad nanoscopic study of a wide-range of dendritic organic
nonlinear optical (NLO) self-assembly molecular glasses reveals an
intermediate thermal phase regime responsible for both enhanced electric
field poling properties and strong phase stabilization after poling. In this
paper, the focus is on dendritic NLO molecular glasses involving quadrupolar,
liquid crystal, and hydrogen bonding self-assembly mechanisms that, along
with chromophore dipole−dipole interactions, dictate phase stability.
Specifically, dendritic face-to-face interactions involving arene-perfluoroarene
are contrasted to coumarin-containing liquid crystal mesogen and cinnamic
ester hydrogen interactions. Both the strength of dendritic interactions and
the impact of dipole fields on the relaxation behavior have been analyzed by
nanoscale energetic probing and local thermal transition analysis. The
presence of dendritic groups was found to fundamentally alter transition
temperatures and the molecular relaxation behavior. Thermal transition analysis revealed that molecules with dendritic groups
possess an incipient transition (T1) preceding the glass transition temperature (T2) that provides increased stability and a well-
defined electric field poling regime (T1 < T < T2), in contrast to molecular groups lacking dendrons that exhibit only single
transitions. On the basis of enthalpic and entropic energetic analyses, thermally active modes below T1 were found to be
intimately connected to the dendron structure. Their corresponding activation energies, which are related to thermal stability,
increased moving from cinnamic ester groups to coumarin moieties to arene-perfluoroarene interacting groups. While dendritic
NLO materials were found to possess only enthalpic stabilization energies at temperatures relevant for device operation (T < T1),
the apparent molecular binding energies above T1 contain a substantial amount (up to ∼80%) of cooperative entropic energy.
The multiple interactions (from dipole−dipole interactions to local noncovalent dendritic interactions) are discussed and
summarized in a model that describes the thermal transitions and phases.
heating and poling process is to have sufficient spacing between
the high-dipole chromophores to inhibit undesired sponta-
neous antiparallel pairing of the dipoles, which results in
reduced EO activity.7
An apparent solution to reduce chromophore aggregation is
to lower the loading density of the chromophores, which in
turn, however, moderates the EO activity. A possible way to
achieve a high loading density with minimal chromophore
aggregation is the “site-isolation” approach,6 where bulky
constituents, e.g., dendrimers, are grafted to the chromophores.
The dendrimers act as chromophore cages, thereby screening
the electrostatic forces and shifting the overall shape of the
chromophores from an anisotropic rod-like shape to a more
INTRODUCTION
■
Organic second-order nonlinear optical (NLO) materials are
being actively pursued for applications in photonic devices such
as high-speed electro-optic (EO) modulators, optical switches,
and frequency converters.1−5 For practical applications, NLO
materials must have both high macroscopic EO activity,
quantified by the r33 value, and thermal stability within the
operating temperature range. High macroscopic EO activity can
be achieved by acentrically ordering a system containing a high
density of high-dipole chromophores via electric field poling at
elevated temperatures.1,6 In electric field poling, the initial state
of the molecular system is isotropically ordered at random, and
the system is heated, allowing for chromophore mobility.
During heating, an electric field is applied, and the high-dipole
chromophores align with respect to the electric field. Finally,
the system is cooled in the presence of the electric field to
obtain an acentrically ordered system. One challenge during the
Received: July 25, 2012
Revised: October 15, 2012
Published: October 25, 2012
© 2012 American Chemical Society
13793
dx.doi.org/10.1021/jp307370y | J. Phys. Chem. B 2012, 116, 13793−13805

The Journal of Physical Chemistry B
Article
Figure 1. Molecules of interest for material Class I (TED1, TED2, TED3, TED1A, TED1-CHO, HDFD), Class II (C1, C1-CHO), and Class III
(CE), as well as control molecules including a novel chromophore with a chain-containing bridge that has no dendritic groups (CC7) and a molecule
that has no chromophore and no dendritic groups (S1) that is of similar size. Incomplete chromophores indicate materials without electron acceptor
groups.
spherical one. This approach has been explored theoretically
and experimentally and shows promising results.2,6,8 Another
approach to reduce dipole antiparallel pairing with high
chromophore density involves self-assembled EO molecular
glasses through strong arene-perfluoroarene (Ar−ArF) π−π
interactions to improve poling efficiency and chromophore
alignment stability.9−11 Materials of this nature resulted in EO
coefficients exceeding those reported for NLO polymers.
For device operation, chromophore alignment imposed by
electric field poling must be stable over long periods of time,
i.e., alignment must exhibit thermal and temporal stability. As
poled organic NLO materials are inherently nonequilibrium
systems, the relaxation behavior and parameters that affect
them, such as thermal transition temperatures, molecular
energetics of thermally active modes, and molecular cooperative
phenomena, have to be carefully analyzed.10 Constraints must
13794
dx.doi.org/10.1021/jp307370y | J. Phys. Chem. B 2012, 116, 13793−13805

The Journal of Physical Chemistry B
Article
Figure 2. Synthesis of molecule S1.
be put in place on the molecular level to prolong the relaxation,
i.e., delay the loss of chromophore acentric ordering for as long
as possible.
In this study, correlations are established among the
molecular mobility, critical transition temperatures, temporal
(thermal) stability, and EO activity of electric field poled
dendritic NLO glasses. It involves a comprehensive analysis of
molecular systems considering an assortment of molecular
system specific parameters, from the strength of the
chromophore dipoles to a variety of dendritic interactions.
The molecular mobility is addressed in terms of its energetic
signature that possesses not only enthalpic activation barriers,
but also, more notable in these glass-forming materials, a strong
cooperative enthalpic signature. More specifically, arene-
perfluoroarene dendritic systems and coumarin-based chromo-
phore materials are discussed that both reveal three solid
phases, separated by two distinct transitions.10,11 It is the
intermediate phase that strongly depends on the interactions
involved, which provides the unique opportunity to engineer
rationally, from the molecular scale up, organic solid NLO
phases with distinct properties toward electric field poling.
These two NLO systems are contrasted to less interacting NLO
glasses, from systems with cinnamic dendritic groups,
chromophore systems without dendritic groups, to a system
that also lacks dipolar interactions. Beyond NLO materials, the
concepts of physical interaction between small molecules that
form amorphous glasses explored here is relevant for other
material systems, for example, those used for three-dimensional
lithography.28
Dendritic molecular glasses provide opportunities for
nanoscale architectural control and molecular and supra-
molecular engineering to simultaneously improve macroscopic
EO activity and thermal stability.8,12,13 Recent efforts have
focused on attached dendrons capable of self-assembly via
internal constraints imposed via noncovalent interactions.9,14
Along these lines, two separate material classes have been
explored, (i) arene-perfluoroarene (ArH-ArF) interactions and
(ii) liquid crystal (LC) forming moieties (e.g., coumarin
mesogens). Both have shown great potential in improving the
acentric order and therefore the macroscopic EO activity of
naturally amorphous NLO systems.9,15,16
In the first material class, face-to-face interactions between
perfluorinated and nonfluorinated aromatic rings are exploited
to provide internal constraints. Along these lines, as shown in
Figure 1, organic NLO molecular systems containing arene
dendritic moieties (i.e., phenyl, naphthyl, or anthryl groups)
and perfluoroarene (pentafluorophenyl groups) have shown
excellent EO activity (>300 pm/V) and thermal stability.14 This
class of self-assembling amorphous NLO materials was found to
exhibit very distinct temperature windows (bordered by
thermal transitions) for optimum field poling.9−11
The second material system involves another strategy for
providing temporal stability, wherein dendrons containing LC
mesogens are used to enhance poling-induced order (Figure 1).
In this strategy, specifically coumarin-based mesogens capable
of reducing the dimensionality of the macroscopic chromo-
phores system were employed.15 Coumarin derivatives which
form LC phases have been used extensively as alignment layers
in LC displays17 and in self-assembling monolayers.18 These
materials have the advantage that, apart from their ability to
form LCs, they can also undergo photoinitiated cross-linking
via the formation of cyclobutane dimers.19−21 Coumarin-based
mesogen systems generally form smectic A type LC phases in
various polymers.20,21 Along these lines, chromophores
containing coumarin mesogens (C1) were synthesized.22,23
A third materials system is represented by the molecule CE,
Figure 1, which, rather than the coumarin derivatives of C1,
contains cinnamic ester derivatives that are traditionally used as
photochromic dyes24 and are not mesogenic.25 Further,
cinnamic acids and cinnamic acid ester derivatives have
shown the tendency toward hydrogen bonding,26,27 which
provides the potential for self-assembly in this system.
EXPERIMENTAL SECTION
■
Materials and Synthesis. As pointed out, this study
involves three classes of NLO chromophore materials that are
distinguished by their dendritic groups. As illustrated in Figure
1, Class 1 materials are composed of arene-perfluoroarene
dendrons, with either heteroaromatic bridge chromophores,
YLD156, or polyene bridge chromophores, YLD124 or
YLD124-CHO. The entire dendritic chromophore molecules
are labeled in accordance with the literature,9,10,14 as HDFD,
TED1, TED2, TED3, or TED1-CHO. A selection of these
NLO molecules are illustrated in Figure 1. Class 2 materials
entail coumarin dendrons, and either complete (YLD 156) or
incomplete (YLD 156-CHO) heteroaromatic bridge chromo-
phores. The molecules are correspondingly labeled C1 and C1-
CHO. Class 3 materials contain cinnamic dendritic groups, and,
as with HDFD, only complete heteroaromatic bridge
chromophores (YLD 156). Class 1−3 materials were
synthesized according to the literature,10,14,22,23,29 and are
here contrasted with CC7, a NLO molecule that does not
13795
dx.doi.org/10.1021/jp307370y | J. Phys. Chem. B 2012, 116, 13793−13805

The Journal of Physical Chemistry B
Article
possess dendritic groups, and, S1, a nonpolar molecule that is
deprived of not only a dipole moment, but also dendritic
groups. The synthesis of CC7, introduced previously,22 will be
described in a future publication. Details about the synthesis of
S1 are provided below.
Instrumental and Analysis. Thermal transition and
relaxation temperatures were deduced from scanning-modu-
lation force microscopy (SM-FM) plots12,33,34 (Figure 3)
Thin films were produced by dissolving the sample materials
in suitable solvents (trichloroethane, dichloromethane, chloro-
form, or toluene) at a concentration of about 10 wt %, filtering
through a 0.2 μm PTFE-syringe filter, and then spin-coating
onto indium tin oxide (ITO) glass or silicon substrates. The
films with thicknesses of 0.5−1.0 μm were annealed overnight
in vacuum at 25−70 °C, depending on the boiling point of the
solvent and material glass transition temperature, to ensure the
removal of any residual solvent.
For the synthesis of S1 N,N-dibutyl aniline (97%, Aldrich), 1,
acetonitrile (HPLC grade, EMD), copper(II) bromide (99%,
Aldrich), 3-dimethylaminoacrolein (90%, Aldrich), tert-butyl
lithium (1.6 M in pentane, Acros Organics), tetrahydrofuran
(EMD), 1,4-phenylenediacetonitrile (99%, Aldrich), and
potassium tert-butoxide (1.0 M in THF, Aldrich) were used
as received. Absolute ethanol (Deacon Laboratories) was stored
with molecular sieve to ensure dryness. The reaction scheme is
provided in Figure 2. 4-Bromo-N,N-dibutylaniline (2) was
synthesized according to the literature.30 Synthesis steps of the
component (4) and (S1) are as follows.
3-[4-(Dibutylamino)phenyl]-2-propenal (4). The procedure
used to synthesize 4 is quite similar to that described in the
literature for an identical reaction wherein the only difference is
the functionality of the tertiary amine.31 In a flame-dried flask,
under argon, 4-bromo-N,N-dibutylaniline (2) (10.72 g, 37.7
mmol) was dissolved in THF (290 mL) and was cooled to −78
°C. tert-Butyl lithium (75.4 mmol, 47 mL) was added, and the
mixture was stirred for 5 min. 3-Dimethylaminoacrolein (189
mmol, 18.84 mL) was added and the mixture was stirred
overnight until room temperature was reached. Saturated
NH4Cl was added to the reaction mixture, followed by
extraction with ethyl acetate. Solvent was evaporated in vacuo
and the product was purified by silica column chromatography
using an ethyl acetate/hexanes (1:9) solvent affording 5.19 g
product (53% yield).
Figure 3. SM-FM setup and illustration of a typical SM-FM glass
transition signature. Temperature transition values are obtained from
“kinks” in contact stiffness versus temperature plots. The transient
regime that identifies the heterogeneity of structural phases is bordered
by the glass and rubber melt phase regimes, ϕ1 and ϕ2. G*(T)
represents the dynamic complex modulus of the material.
involving a standalone scanning force microscope (SFM)
(Explorer, Veeco, CA). Contact-mode silicon bar-shaped levers
(PPP-CONT, Nanosensors, normal and lateral spring constants
of ∼0.2 N/m and ∼80 N/m,35,36 respectively) and a custom
automated sample heating setup were used with a heating rate
of 1 °C/min between steady-state recordings of thermome-
chanical sample responses. All SM-FM measurements were
performed on spun-cast films at ambient pressure under a dry
nitrogen atmosphere with relative humidities below 10% and a
sinusoidal lateral force modulation frequency of ∼1 kHz. The
cantilevers were geometrically preconditioned to reveal smooth,
close to spherically shaped geometries.10
SM-FM is a nonscanning SFM method, with which variations
in the thermomechanical properties of materials are recorded as
a function of the temperature. The experimental physical
observable is the contact stiffness that includes the
thermomechanical properties of the sample, i.e., the temper-
ature dependent complex shear modulus G*(T), and the
contact geometry.34 The thermal gradient of the phase
response, dG*/dT, is geometry and rate dependent. Thermally
activated transitions are determined from abrupt changes
(“kinks”) in the shear mechanical response curve, as illustrated
in Figure 3 for a glass transition, which is one of several
relaxation processes that can be recorded through SM-FM.
Glass transitions are well-known to cause a significant drop in
the material’s compressibility. This abrupt change causes the
spherical SFM tip that is held at constant load to compress the
material further, and thus, through contact area augmentation,
results in a significant increase in the observed dynamic contact
stiffness, kc. In Figure 3, the regime identified as “transient
regime” between the two phases, ϕk with k = 1,2, reflects to
part the heterogeneity of two structural phases and the pressure
nonequilibrium resulting from the time-steps in SM-FM
experiments that are too short for the sluggish relaxation
behavior in the heterogeneous regime of glass formers. In the
case of polymers, the phase heterogeneity reflects a spatial
distribution of independent backbone and side-chain relaxa-
tions. It is important to note that a material relaxation other
than a glass or melting transition37 can exist without noticeable
4-[1-Cyano-4-[4-(dibutylamino)phenyl]-1,3-butadien-1-
yl]-a-[3-[4-(dibutylamino) phenyl]-2-propen-1-ylidene]-ben-
zeneacetonitrile (S1). The synthesis for this molecule (S1) is
very similar to that described in the literature.32 In a flame-dried
flask, under argon, 1,4-phenylenediacetonitrile (5) (61 mg, 0.39
mmol) was dissolved in ethanol (25 mL) via stirring. Then,
potassium tert-butoxide (3.12 mmol, 3.1 mL of 1 M solution in
THF) was added. The system was allowed to stir for several
minutes until the reaction mixture had a slightly bluish tint. 3-
[4-(Dibutylamino)phenyl]-2-propenal (4) (0.202 g, 0.78
mmol) dissolved in ethanol (5 mL) was added, and the
mixture was allowed to stir overnight at room temperature.
Solvent was evaporated in vacuo and the product was purified
by running silica column chromatography using an ethyl
acetate/hexanes (1:9) solvent affording 59 mg of a black solid
1
product (24% yield). H NMR (300 MHz, CDCl3) δ 7.52 (m,
4H), 7.40 (m, 4H), 7.25 (m, 4H), 6.98 (m, 2H), 6.64 (m, 4H),
3.32 (t, J = 6.0 Hz, 8H), 1.56 (m, 8H), 1.36 (m, 8H), 0.97 (t, J
= 6.0 Hz, 12H). 13C NMR (CDCl3) δ 149.6, 143.3, 143.1,
134.0, 129.9, 125.8, 123.0, 120.3, 118.0, 111.8, 108.0, 51.1, 29.7,
20.6, 14.3. LRMS (ESI) (M+, C44H54N4): calc’d, 638.4;
observed, 638.6 [M+Na].
13796
dx.doi.org/10.1021/jp307370y | J. Phys. Chem. B 2012, 116, 13793−13805

The Journal of Physical Chemistry B
Article
transient regime, as we will discuss below. Differential scanning
calorimetry (DSC) measurements were performed on a TA
Q20 DSC at a heating rate of 10 °C/min.
The energetics of thermal active modes (e.g., side-chain
rotations, molecular translations, or bonding/debonding
Table 1. SM-FM Transitions in Dendritic NLO Materials
Sorted by Increasing T2
T2 − T1
(°C)
type of dendritic
interaction
material
T1 (°C)
T2 (°C)
TED1-
41
2
46
2
5
Arene-perfluoroarene
CHO9,10
C1-CHO23
42
2
47
2
5
Coumarin-coumarin
(mesogenic)
Figure 5. SM-FM results for S1 and CC7 revealing only single
transitions for molecular glasses lacking dendritic interacting groups.
CE
50
59
3
2
2
2
2
2
59
70
2
2
2
2
2
2
9
Hydrogen bonding
Arene-perfluoroarene
Arene-perfluoroarene
Coumarin-coumarin
Arene-perfluoroarene
Arene-perfluoroarene
HDFD11,14
TED19,10
C123
10
15
15
11
8
59
74
61
76
signal between forward and reverse scans as a function of the
scan velocity at constant temperatures. These isothermal
friction-rate curves are superimposed to an arbitrary reference
isotherm by shifting them accordingly. From these shifts, the
energetic information can be deduced based on the time−
temperature superposition principle.41−43
TED29,10
TED39,10
75
86
118
126
Table 2. SM-FM Transitions of Non-Dendritic Materials
material
T2 (°C)
dendritic interactions
chromophores
All dipole moment calculations were conducted using
semiempirical VAMP geometry optimizations in Materials
Studio as described previously.10 Settings used were an NDDO
Hamiltonian44 with AM1* parameters,45−47 a restricted
CC7
S1
60
36
2
1
no
no
yes
no
́
events) were determined by intrinsic friction analysis (IFA),38
involving the SFM (Explorer, Veeco Inc.) system described
above, and operated under identical environmental conditions.
SM-FM data provided the temperature phase regimes within
which IFA was applied.
IFA is a spectroscopic rate method analogous to other
techniques that probe molecular relaxations, such as dielectric
spectroscopy, with the difference being that the velocity instead
of the frequency is swept at constant temperature. The method
has been applied successfully to polymers of wide complexity
and thin films down to organic monolayers.10,38−40 The
observable in IFA is the friction force that entails the coupling
phenomena between the one-dimensional sliding motion of the
SFM probe and the thermally active modes within the material.
If spectrally analyzed through IFA, the friction signal contains
the enthalpic and entropic energies of the thermal modes, as
discussed in detail elsewhere.38 Briefly, the friction force is
determined from the hysteresis in the torsional SFM deflection
Hartree−Fock formalism, a maximum step size of 1.0 Å,
automatic multiplicity, and medium convergence tolerance.
RESULTS AND DISCUSSION
■
Transition Temperature Measurements. The thermal
transition results of the SM-FM analyses for the dendritic NLO
materials, as introduced in Figure 1, are provided in Table 1.
Table 2 provides the transitions found for the nondendritic
materials in this study for comparison. All dendritic systems
possess two transitions, a finding that is in contrast to the
nondendritic materials that possess only one transition. A SM-
FM plot is provided for CE that is representative of all dendritic
NLO materials in this study, Figure 4a. Figure 4b provides a
contrasting sketch to Figure 3b. The second transitions at
temperatures T2 correspond to the transition temperatures
determined by DSC, as shown in the inset of Figure 4a, and
both DSC and SM-FM identify T2 as a glass-like transition.
Contrary to SM-FM glass transitions, the lower temperature
Figure 4. (a) Representative SM-FM plot of CE for dendritic NLO systems revealing two thermal relaxations in the dynamic contact stiffness (kc) at
T1 = 50 3 °C and T2 = 59 2 °C. Inset: DSC result for CE, with a transition at Tg of 62 °C, in good agreement with T2. (b) Illustration of SM-FM
plots for dendritic pendant group NLO systems highlighting the qualitative differences of the two transitions, and the three phase regimes.
13797
dx.doi.org/10.1021/jp307370y | J. Phys. Chem. B 2012, 116, 13793−13805

The Journal of Physical Chemistry B
Article
a
Table 3. IFA Results and Calculated Dipole Moments of Dendritic NLO Materials Sorted by Increasing Ea,1
material
C1-CHO23
Ea,1
Ea,2 (kcal/mol)
Ea,3
51
MW (g/mol)
ln vp @ Tref
Dcalc (D)
4
3
2
2
2
2
2
2
2
26
33
31
43
44
41
49
57
4
4
4
4
3
2
3
5
4
4
5
4
4
4
4
2
1102.1
1283.4
1212.1
1399.4
1427.2
1509.4
1609.5
1709.6
3.5
−0.1
2.1
6.8 2.5
11.0 2.5
9.7
CE
11
13
16
23
25
29
35
67
55
72
71
69
82
98
TED1-CHO9,10
C123
HDFD11,14
TED19,10
TED29,10
TED39,10
−4.5
−1.8
−2.8
−8.1
−9.5
11.1 2.7
--
15.2
15.3
15.7
a
Tref = 49.5 °C; D = Debye.
a
Table 4. IFA Results of Non-Dendritic Materials, and Compilation of Chromophore Dipole Moments
material
Ea,1 (kcal/mol)
Ea,3
45
MW (g/mol)
ln vp @ Tref
Dcalc (D)
S1
2
10
--
1
2
3
5
638.9
967.3
--
N/A
0.3
--
--
CC7
66
--
9.6 0.3
12.7
11.6
6.0
YLD124 Chromophore9,10
YLD156 Chromophore
YLD124-CHO9,10
--
--
--
--
--
--
--
--
YLD156-CHO
5.4
a
Tref = 49.5 °C; D = Debye.
transitions, T1, of dendritic materials do not show a transient
regime, identifying the transition as a local phenomenon, such
as an isolated side-chain rotation. The two transition temper-
atures T1 and T2 thereby reveal three thermal phase regimes,
Gϕκ*(T) with k = 1, 2, and 3.
spectroscopy master curve of the friction values reveals a
relaxation peak at vp ≈ 1 nm/s for CE at a reference
temperature of 49.5 °C, Figure 6b. It has to be noted that the
reference temperature chosen here is below T1 for CE, where
the material is in its glassy state. This is in agreement with the
other dendritic NLO systems with T1 values above 49.5 °C, as
verifiable by cross referencing Table 1 with Table 3. Dendritic
systems with T1 and even T2 transition values below the
reference temperature of 49.5 °C, such as C1-CHO and TED-
CHO (Table 1), reveal higher relaxation velocities than found
for glass phases, CE, HDFD, and TED1−3 (Table 3).
Returning to Figure 6b, the horizontal shifts, known as shift
factors ln aT, were plotted versus the reciprocal temperature, in
an Arrhenius-like fashion (inset of Figure 6b). The Arrhenius
plot provides the energetic values Ea,k (k = 1, 2, and 3) in
accordance with
The pendant group interactions responsible for causing a low
temperature transition, and thus phase ϕ2, are, as per Tables 1
and 2, perfluoroarene (ArH-ArF) interactions for HDFD and
the TED series,10,11,14 coumarin LC interactions for C1 and
C1-CHO,8 and hydrogen bonding, as found in this study for
CE (Figure 4a). Hydrogen bonding for the pendant groups in
CE is in accordance with studies involving cinnamic acids and
cinnamic acid ester derivatives.26,27 Specifically, hydrogen
bonding can occur between C−H(aromatic)···O or C−
H(olefinic)···O.27,48 The two molecular systems lacking
dendritic groups, i.e., S1 and CC7, exhibit only single
transitions (Figure 5). Thus, the second transition in NLO
dendritic systems can be attributed to the interacting pendant
groups of the NLO dendritic systems.
Energetic Relaxation Activation Analysis. A local
energetic analysis via IFA was conducted to elucidate the
three phases observed in dendritic NLO molecular systems, and
to assess the impact of intermolecular constraints originating
from dipole moments and pendant group interactions. The IFA
results of the three classes of NLO materials from Figure 1 are
compiled in Table 3. It provides phase specific energy values
Ea,k (k = 1, 2, 3) that correspond to the temperature regimes.
The energetic information of the two nondendritic materials is
listed separately in Table 4. In addition, both tables contain the
calculated dipole moments of the entire molecules and of the
chromophores alone.
⎡
⎢
⎤
⎥
∂ ln(aT)
⎣ ∂(1/T) ⎦P
E = −R
a
(1)
For instance, an energy value of Ea,1 of 11 2 kcal/mol is
deduced for CE below T1, which can be attributed to the
hydrogen bonding−debonding interaction between the cin-
namic ester pendant groups,26,27 distorted by the static dipolar
interacting field, emanating from the strong chromophore
dipoles, and the van der Waals (VdW) interactions of the
network. Energy values for the higher temperature phases of
CE or for other dendritic systems of this study are found in
Table 3. It is important to note that the energy values can either
be truly enthalpic or also possess a cooperative entropic
component that is introducing a length scale over which the
relaxation occurs, likely to exceed by orders of magnitude the
molecular length scale.40,49 As such, the energy Ea can be either
an enthalpic activation energy or an apparent energy containing
entropic and enthalpic components.
With CE, an IFA data set representative of all dendritic NLO
materials is provided and discussed. Figure 6a shows the
friction−velocity isotherms of CE below T1 = 50 3 °C (c.f.
Figure 4). A similar set of friction data exists between T1 and
T2, and above T2. Each set of friction isotherms was shifted
horizontally (and vertically if necessary) to an arbitrarily chosen
reference temperature, to produce a single overlapping master
curve, as illustrated for CE in Figure 6b. The velocity
In relaxation processes, cooperative phenomena can manifest
themselves in concerted molecular motions that involve
entropic energies that originate from configurational adjust-
ments. In polymers or strongly interacting molecular systems,
13798
dx.doi.org/10.1021/jp307370y | J. Phys. Chem. B 2012, 116, 13793−13805

The Journal of Physical Chemistry B
Article
debonding reaction is solely enthalpic. This applies also to the
two other classes of materials below T1. For Class 1 (HDFD
and TEDs) and Class 2 (C1 and C1-CHO) materials, the
interaction energies Ea,1 could be attributed to bonding−
debonding interactions between the arene-pentafluorophenyl
dendritic moieties and the coumarin pendant groups,
respectively.17,18,50−55
As can be seen from Tables 3 and 4, the energy values Ea,k
increase for subsequent temperature regimes (i.e., for k = 1 to
3). The increase can be attributed to entropic cooperative
energy contributions. As pointed out, the degree of
cooperativity is also attainable by IFA.49,40 Vertical force
shifting ΔFF in IFA is indicative of cooperativity in the system,
i.e.,49
TΔS*
ϕ′
ΔFF ≈ −
(2)
where ϕ′ is the contact area normalized stress activation
volume.49 On the basis of the nonzero shift values above T1 of
Figure 6c, which is representative of all dendritic materials
listed in Figure 1, energies above T1 carry both enthalpic and
entropic contributions. The origin for enthalpic interactions is
temperature invariant, so the difference between the phase
energies can be solely attributed to changes in the entropic
cooperativity. In the case of CE with Ea,2 and Ea,3 values of 33
4 and 67 4 kcal/mol, respectively, TΔS* can be determined
to be 22 4 and 56 4 kcal/mol for the two thermal phases
above T1 and T2, by deducting the enthalpic energy value of 11
2 kcal/mol from the two apparent energies. Similarly high
entropic contributions can be found for the other dendritic
systems. At temperatures far above T2 (i.e., at Tc ≈ T2 + 20 °C),
the energies return to be purely enthalpic as they are below T1
(Figure 6c), i.e., they have zero ΔFF shift values.
The energies of primary significance for NLO applications
are Ea,1 (T < T1) and Ea,2 (T1 < T < T2), as they reflect the
temporal stability under NLO operating conditions and the
activation energy relevant for the poling process, respectively.
The energy values Ea,3 provide information about the range of
interactions in NLO networks. These three energy values are
determined by the strength of the molecular dipole moments
and the interaction strength between the pendant groups.
Figure 7 reveals the impact of the dipole moment on Ea
values within the three thermal regimes. C1-CHO and TED1-
CHO, which possess incomplete chromophores with signifi-
cantly reduced dipole moments compared to their counterparts
C1 and TED1, reveal relatively low energies Ea,k (k = 1, 2, 3).
Thus, molecular NLO systems of chromophores with increased
dipole moments can be expected to possess improved temporal
stability, but are more difficult to pole. Also captured in Figure
7 is the impact of the dendritic interaction on the energetics.
Within the TED series, the activation energies increase with
increasing arene size, and more so for Ea,2 and Ea,3 than
expected from the enthalpic interaction captured by Ea,1, and
the slight increase in the dipole moment with larger pendant
groups. Thus, while there is an increase in enthalpic interaction
strength, given by Ea,1, increases due to entropic cooperative
effects in Ea,2 and Ea,3 have to be considered. On the basis of the
TED series, it can be concluded that the pendant-side group
interactions not only alter the enthalpic energy, but also impact
the molecular cooperativity within material phases to an even
higher degree. The stark change in the dependence of the
energetics on the dipole moments for the TED series illustrates
Figure 6. IFA results for CE: (a) Friction−velocity isotherms at
indicated temperatures. (b) Shifted isotherms forming a master
friction relaxation curve revealing the relaxation peak ln vp = −0.1 at a
reference temperature of 49.5 °C. (Inset) Arrhenius plots disclose the
apparent activation energies Ea,1 = 11 2 kcal/mol (T < T1), Ea.2 = 33
4 kcal/mol (T1 < T < T2), and Ea.3 = 67 4 kcal/mol (T2 < T). (c)
Vertical force shift ΔFF to collapse the isotherms on the master curve.
T1, T2, and Tc represent the two transitions 1 and 2, and the
temperature at which ΔFF vanishes above T2, respectively.
cooperative phenomena are observable during molecular
aligning processes, such as shear flow or electric field poling.
So, the energy Ea can be either an enthalpic energy or an
apparent energy containing entropic and enthalpic compo-
nents. It is the degree of vertical shifting (see Figure 6c)
necessary to generate the master curve that determines the
degree of cooperativity.40,49
From Figure 6c, with friction shift values close to zero below
T1, it can be inferred that for CE the hydrogen bonding−
13799
dx.doi.org/10.1021/jp307370y | J. Phys. Chem. B 2012, 116, 13793−13805

The Journal of Physical Chemistry B
Article
molecules that lack any pendant group interaction (squares
with diagonal lines) and are free to rotate or translate within the
system. These two examples are representative of expressing
the likelihood of finding mobile molecules, which increases with
higher lower-molecular-weight molecules. From this model, it
can be statistically expected that more “free” (unbound)
molecules will be found in dendritic systems of lower MW than
higher MW.
If we consider the bonds to be capable of bonding and
debonding, we anticipate an increase in open bonds with
increasing temperature. Consequently, a temperature can be
reached at which the less dense molecular structure (Figure 9b)
also exhibits mobile molecules. Considering that the transition
temperatures T1 and T2 are distinct manifestations of the effect
of the molecular mobility, we can expect low-molecular-weight
NLO material to reveal lower transition temperatures than
higher-molecular-weight materials. This is confirmed with the
transition data in Table 1, if cross-referenced with the molecular
weight information provided in Table 1. As the transition
temperatures are intimately connected to activation energies,
the presented simple model is in accordance with the finding
presented in Figure 8 that yields an increase in the activation
energy with increasing molecular size.
Figure 7. Dipole moment effect on the apparent activation energy for
arene-pentafluorophenyl (TED1, TED2, TED3, TED1-CHO), LC
(C1 and C1-CHO), and hydrogen bonding interacting pendant
groups (CE).
that changes in dendritic interactions are much more significant
than changes in dipole moment.
Figure 8 shows that, besides the dipole moments and
pendant group interactions, the molecular weight (MW) is also
The heuristic model (Figure 9) finds further experimental
support if we plot the logarithm of the molecular relaxation
velocity, vp, versus the corresponding activation energy Ea,3 at
equivalent reference temperatures for the molecules of Classes
1 to 3. The plot displays a linear relationship, Ea,3 (ln vp), for
the molecules with interacting dendritic groups, which
translates into a linear relationship of ln vp with the molecular
weight (Figure 10b). In contrast with the dendritic molecules
are CC7 and S1, i.e., molecules with missing pendant groups
(Figure 10b), which show fundamentally different behaviors.
Apparent Energy and Relaxation Frequency. As
pointed out above, the cooperation length in dendritic NLO
systems is significant, particularly around the poling temper-
ature necessary to achieve acentric ordering. For instance, the
entropic energy requirement (TdS* ≈ Ea,k − Ea,1; k = 1, 2) for
TED3 is about 22 and 63 kcal/mol, respectively (Table 3),
which is up to twice the enthalpic energy of the arene−
pentafluorophenyl interactions involved. In the case of CE, the
entropic cooperative contribution can be as much 84% of the
apparent energy. The apparent (total) energy Ea and the
cooperative entropic contribution TΔS* can be evaluated based
on the theory of absolute reaction rates applied to
relaxations56,57 via
Figure 8. Energy−molecular weight relationship below T1 and above
T2, i.e., Ea,1(MW), Ea,2(MW), Ea,3(MW), respectively. Linear fits reveal
the following fit parameters: Ea,1(MW) = 0.049MW − 50, Ea,2(MW) =
0.049MW − 28, Ea,3(MW) = 0.069MW − 26.
of importance in the energetics in NLO dendritic systems.
Within the MW range studied, approximate linear relationships
for Ea,k(MW) (k = 1, 2, 3) were found for all dendritic NLO
systems introduced in Figure 1. To elucidate this initially
unexpected behavior, we discuss the energy−molecular weight
relationship in terms of a simplified structural grid model, as
depicted in Figure 9.
In Figure 9a,b, we consider time snapshots of two dendritic
NLO molecular systems of maximum packing densities, and
identical system volume and temperature. The difference
between Figure 9a and b is found in the molecular weight, or
more specifically, the molecular number density, with low and
high MW in Figure 9a and b, respectively. While the
illustrations on the left in Figure 9a,b depict arbitrarily chosen
arrangements of close-packed dendritic NLO molecules, the
sketches on the right highlight the location of molecules (gray
filled squares), (ii) the location of molecular bonds formed
between pendant groups (filled dark circles), and (iii)
Ea = RT[1 + ln(kBT/2πhfP )] + TΔS* = ΔH* + TΔS*
(3)
where kB represents Boltzmann’s constant, h Planck’s constant,
R the universal gas constant, f P the peak frequency of the
relaxation,10,38,39,49 and ΔH* the enthalpic energy component
of relaxation. As IFA measurements involve velocities instead of
frequencies, a length scale ξ is required between the peak
velocity vp and peak frequency fp according to the tested
relationship40
vp
ξ =
fp
(4)
to benefit fully from eq 3.
Vanishing entropic contributions are found for dendritic
NLO materials below T1, as shown above with Figure 6b for
13800
dx.doi.org/10.1021/jp307370y | J. Phys. Chem. B 2012, 116, 13793−13805

The Journal of Physical Chemistry B
Article
Figure 9. Molecular size effect on the mobility of dendritic NLO molecules. (Left) Densely packed (a) low-molecular-weight system is compared to
(b) high-molecular-weight system. The molecules are composed of a core chromophore (arrow indicates dipole moment) and two pendant moieties.
(Right) Essence extracted representations of the molecular systems, highlighting the molecules with gray squares, the molecular bonds formed with
dark-filled circles, and free molecules with squares with diagonal lines.
CE, and below with Figure 11 for all three material classes,
represented by TED1, C1, and CE. Thus, for temperatures
below T1, eq 3 reduces to
with 10 < ΔT < 20 °C, (shown here only for CE) the
cooperative entropic contribution dies out.
Results from Control Systems CC7 and S1. CC7 and S1
were chosen as control systems, as neither possesses interacting
pendant groups. While CC7 contains a full chromophore that is
shielded from neighboring molecules by a “looping alkane
chain” to suppress dipole−dipole association, S1 represents a
molecule that partially resembles (at least in size and in
aromatic carbon content) the chromophores in this study, but
is deprived of a dipole moment. As discussed earlier, both CC7
and S1 exhibit only single temperature transitions (dubbed T2
for consistency), each revealing two thermal regimes of
energies, dubbed Ea,1 and Ea,3, respectively, for consistency
with the previous discussion. The transition results retrieved
from Figure 5 and Figure 12a,b, are provided in Table 4.
As expected, the lowest activation energies are found for S1.
Lacking any obvious means for self-assembly, its transition
value T2 has to be considered a melting transition. In its solid
state, below T2, the low energy value Ea,1 of 2 1 kcal/mol in
S1 originates from weak van der Waals interaction, i.e., dipolar
fluctuations that could be identified either as thermally active
modes within the material, or association−dissociation
processes between the sample and the probing SFM tip. A
mode relaxation frequency on the order of one tenth of a
picosecond can be inferred from eq 6 for Ea,1. Above the
melting temperature T2, a significant increase in energy is
ΔH* = RT[1 + ln(kBT/2πhfP )]
(5)
for dendritic NLO systems. This relationship yields for Ea ≫
RT a peak frequency of
⎛
⎞
⎟
⎠
Ea
RT
kT
2πh
⎜
fp ≈
exp −
⎝
(6)
which works for noncooperative relaxing systems, such as the γ-
relaxation (phenyl rotation) in polystyrene.43,58,59 In dendritic
NLO systems, however, this relationship would lead to too low
bonding/debonding frequencies that would not result in IFA
mode coupling events for the TED molecular systems.
It has recently been estimated that the enthalpic energetic
impact of the YLD124 chromophore (Figure 1) is around 20
kcal/mol for TED1.10 Thus, by subtracting the dipolar
constraining energies from the apparent energies Ea, pending
group bonding/debonding frequencies on the order of 0.1 to 1
ns can be anticipated for the dendritic NLO molecular glasses
studied here.
Figure 11 indicates that, above T1, relaxation processes in all
three classes of dendritic materials possess an additional
entropic cooperative component, as formally expressed in eq
3, reaching maximum cooperativity at T2. Above Tc = T2 + ΔT,
13801
dx.doi.org/10.1021/jp307370y | J. Phys. Chem. B 2012, 116, 13793−13805

The Journal of Physical Chemistry B
Article
Figure 10. Peak relaxation velocity ln vp relationships at the reference
temperature of 49.5 °C with (a) the apparent activation energy Ea,3
and (b) the molecular weight MW. Linear fits of molecules with
interacting pendant groups reveal for Ea,3 = 3.03 ln vp + 63 and ln vp
−0.022MW + 27.7. Contrasted to the linear fit in (b) are molecules
with missing pendant groups, i.e., CC7 and S1. For S1, * indicates the
maximum possible peak value.
Figure 12. (a) IFA Arrhenius results for S1 and CC7, disclosing the
apparent activation energies Ea,1 and Ea,3 below and above T2,
respectively, where the reference temperature (ln aT = 0) was chosen
separately for each data set to space the data for clarity. (b) Vertical
ΔFF shifting plots for S1 and CC7.
for the dendritic NLO systems (Figure 11), can only be
interpreted as a cantilever probe induced shear alignment effect.
Contrary to S1, CC7 is a dipolar chromophore and maintains
its solid phase above T2. The strength of the dipolar interaction
of 10 2 kcal/mol determined below T1 is comparable to the
apparent interaction strength observed for CE of 11 2 kcal/
mol with its hydrogen bond interacting cinnamic ester pendant
groups and the chromophore dipole interactions. Comparison
between S1 and CC7 is the key to understanding the
interaction in CC7, as S1 is a nonpolar molecule. The
difference in Ea,1 between these two molecular systems is 8
3
kcal/mol, which provides a measure of the dipole interaction
strength for CC7 below T1. This finding confirms earlier
results10,23 that were based on comparative findings involving
incomplete chromophoric systems, namely, C1-CHO and
TED1-CHO, and their full chromophore counterparts C1
and TED1 (Tables 3 and 4). As CC7 possesses a chromophore
with a dipole similar to that of CE (calculated as 9.6D and
11.6D, respectively, Tables 3 and 4), we can infer from the
energy difference above that about 3 kcal/mol (∼25%) of Ea,1
in CE is due to interacting cinnamic ester moiety pendant
groups (hydrogen bonding). In comparison, C1 with its highly
interacting mesogenic pendant groups shows an even larger
Figure 11. Compilation of the degree of entropic cooperativity TΔS*
≈ −ΔFF/ϕ′ expressed by IFA's vertical friction shifts ΔFF of
representative materials from the three dendritic classes, as defined
by Figure 1.
observed, yielding 45 3 kcal/mol. Such high-energy values in
the absence of strong interactions can only be explained by
entropic contributions, as confirmed by Figure 12b. We will
argue that this entropic effect is induced, which is different from
the intrinsic cooperativity discussed so far. The lack of a
molecular network response in S1 and an entropic signal ΔFF
that does not vanish at higher temperatures, as found above Tc
13802
dx.doi.org/10.1021/jp307370y | J. Phys. Chem. B 2012, 116, 13793−13805

The Journal of Physical Chemistry B
Article
Figure 13. General model of molecular motion in self-assembling dendritic chromophores where dotted lines indicate interacting dendritic groups:
(a) T < T1 isolated dendritic side bonding−debonding; (b) T1 < T < T2, rotational motion allowed by dendritic dissociation; (c) T2 < T translation
of molecules and interacting segments (region within the black line) due to widespread dendritic dissociation; (d) T2 ≪ T, uncorrelated translation
of molecules where dendritic association is rare.
side group contribution to the glass state below T1 of about 8
kcal/mol (∼50%) based on its Ea,1 value of 16 2 kcal/mol, as
per Table 3.15,23
separating the apparent energy in enthalpic and cooperative
entropic components.
Below T1, the apparent energies Ea,1 of the three classes of
materials reflect the spontaneous bonding−debonding reac-
tions between the dendritic side groups within the dipole field.
The process is enthalpic only, as per the IFA vertical shifts
(Figure 11). As illustrated in Figure 13a, only isolated
associations and dissociations of dendritic bonds (dotted
lines) are occurring spontaneously throughout the material
phase, leaving the molecular network intact. As the molecules
each possess two dendritic groups, molecular mobilities of
rotations and translations are statistically non-occurring.
At T1, a critical number of dendritic association and
dissociation events is reached that yield statistically (but only
locally available) dissociation of both dendritic bonds of
molecules providing them with rotational degrees of freedom
(Figure 13b). While the network up to T2 stays intact, i.e.,
molecular translation is still subdued, cooperative mobility
events are increasingly noticeable with temperature (Figure
11), mainly due to the rotating large dipoles involved. The
apparent energy Ea,2 is now expressed by eq 7, i.e., it contains
also a nonzero entropic component of cooperativity.
At T2, a critical temperature is reached at which the network
also accommodates molecular (or molecular chain) trans-
lations. This process is highly cooperative near T2 and gives rise
to a high degree of cooperativity expressed through Ea,3. T2 has
been identified as a glass transition, based on SM-FM and DSC
experiments (Figure 4a). With increasing temperature, thermal
noise is increasingly disrupting cooperative events, yielding at
about 20 °C above T2 (i.e., Tc) to vanishing cooperativity
(Figure 11).
Above T2, CC7 reveals an increase followed by a decrease in
cooperative entropy (Figure 12b). This qualitative behavior is
characteristic for the solid phase of a polymer rubber or melt
close to the glass transition. A comparison of the ratio between
the CC7 and CE peak values of the ΔFF shifts and energy
values Ea,3 is of interest. The CC7−CE comparison yields a
ΔFF peak value ratio of about 1:3, as per Figures 5c and 7b, and
an Ea,3 energy ratio of about 1:1 (based on values taken from
Tables 3 and 4). The ratios indicate that the cooperative
entropic contribution for NLO materials with interacting
pendant side groups is more significant than in merely dipolar
interacting systems, suggesting a higher order of mobility. This
aspect has been recently confirmed23 for C1 with its liquid
crystal pendant coumarin side groups, revealing matrix-assisted
poling properties, in which the molecular network exhibit a
mobility toward higher order.
It can be concluded from the comparison with the two
control systems, S1 and CC7, and earlier findings that dendritic
groups provide NLO organic glasses with (i) strong phase
stabilization, in particular below T1, (ii) substantial intrinsic
entropic cooperativity above T2, and (iii) structural alignment
with mobility toward higher order.
Molecular Phase Description Model. The results and
discussion above provide the basis for a molecular model that
describes the three phases and two transitions in dendritic NLO
molecular glasses (Figure 13). The overall description of the
energetics involved, considering also the constraining dipolar
fields that emanates from the chromophores (i.e., Edipole), can
be expressed as
SUMMARY AND CONCLUSION
■
Achieving both high temporal stability and high EO activity is
of paramount importance in molecular engineering of novel
organic NLO materials. Constraints and high EO active
Ea = (RT[1 + ln(kBT/2πhfP )] + Edipole) + TΔS*
(7)
= ΔH* + TΔS*
13803
dx.doi.org/10.1021/jp307370y | J. Phys. Chem. B 2012, 116, 13793−13805

The Journal of Physical Chemistry B
Article
components are built in to achieve this goal, as in this study by
dendritic side group interactions, and large heteroaromatic
chromophores or polyene bridge chromophores. With
cinnamic, coumarin, or arene-perfluoroarene dendritic side
groups, it has been shown that the thermal poling regime can
be separated neither from the NLO device operating
temperature regime (T < T1) nor from the dipole annihilation
regime at high temperatures (T > T2) that are known to lead to
antiparallel dipole pairing. As described in the model above
(Figure 13b), the origin for the existence of such a unique
intermediate phase regime can be attributed to the association−
dissociation mechanism of the dual bonds, through which the
dendritic molecules interact. Within the intermediate temper-
ature regime, T1 < T < T2, and close to T2, where molecular
(rotational) mobilities are entropically highly cooperative,
thermal electric field poling is most effective and beneficial
for achieving a lower dimensional organization, such as an
acentric chromophore order.
Considering the energetic information (Table 3 and Figure
7), we find the width of the intermediate thermal regime to
coincide with an increase in the interaction strength of the
pendant groups. In other words, while the interaction strength
is beneficial to the stability of the poled molecular system under
NLO operation by raising T1,10 it negatively impacts the
thermal window of opportunity to acentrically align the
chromophores during thermally activated poling. These two
competing conclusions summarize the findings above and are
captured in the plot of Figure 14 that relates the poling
molecules. Polyene bridge-type chromophores employed in the
TED series have been shown to produce higher r33 values than
their heteroaromatic counterparts,9 accounting for the higher
poling efficiency values for the TED series.
Thus, it has been shown that the presence of dendritic
groups fundamentally alters the observed transition temper-
atures and molecular relaxation behavior. Interacting dendritic
groups introduce an intermediate phase regime bordered by
two thermal transitions. The temperature gap between the two
transitions was found to be substantially widened by the
presence of chromophores. Molecules lacking dendrons
showed only one transition whether or not chromophores
were present. Low-temperature relaxation activation energies,
which correspond to long-term stability, followed the trend:
cinnamic ester groups, coumarin moieties, followed by arene-
perfluoroarene interacting groups, indicating the relative
strength of these interactions within the molecular glass.
Intermolecular interactions definitely improved the phase long
term stability and are expected to do so in the poled state as
well. Interestingly, dendritic glasses show a universal trend in
peak relaxation velocity as a function of both molecular weight
and activation energy, indicating that molecular size heavily
influences behavior due to the density of interacting dendritic
groups. It was also argued that dendritic groups provide NLO
organic glasses with structural alignment with mobility toward
higher order.
This study showed the importance of entropic cooperativity
in the design of novel materials that entail a multitude of
interactive constraints form dipole field interactions to local
dendritic bonding. To further our understanding on how to
effectively translate this information in future molecular designs
of self-assembling organic systems is part of our ongoing
studies.
AUTHOR INFORMATION
■
Corresponding Author
*E-mail: roverney@u.washington.edu.
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
This work was supported by a University Initiative Fund (UIF)
graduate fellowship through the Center for Nanotechnology
(CNT) at the University of Washington. D. B. Knorr thanks
ARCS for a graduate fellowship. Funding support from the
National Science Foundation (DMR-0905686) and the Air
Force Office of Scientific Research (FA9550-09-1-0589 MOD)
is gratefully acknowledged. C. Zhang is obliged to the support
from National Science Foundation (EEC 0812072) and the
NNIN Laboratory Experience for Faculty (LEF) Program at
the University of Washington.
Figure 14. Poling efficiencies9,14,23 as a function of Ea,1 for various
materials poled on ITO-coated glass, revealing the competition of EO
activity and stability.
efficiency (defined as the poling field, Ep, normalized EO
activity coefficient, r33) to the enthalpic activation energy Ea,1.
The latter is generally indicative of stability,10 and affects the
width of the intermediate temperature regime. Materials with
excellent stability and initial EO activity will be high on both
axes, i.e., in the top-right region of the plot. As discussed above
based on Tables 1 and 2, the width of the intermediate thermal
transition regime depends on (a) the size of the chromophore,
i.e., increasing for smaller chromophores, as shown with TED1-
CHO and C1-CHO, and (b) pendant group interaction
strength, decreasing with increasing strength, as found for with
CE and C1, and TED1, TED2 to TED3.
REFERENCES
■
(1) Kajzar, F.; Lee, K.-S.; Jen, A. K.-Y. Adv. Polym. Sci. 2003, 161
(Polymers for Photonics Applications II), 1−85.
(2) Ma, H.; Jen, A. K.-Y.; Dalton, L. R. Adv. Mater. 2002, 14, 1339−
1365.
(3) Dalton, L. R.; Steier, W. H.; Robinson, B. H.; Zhang, C.; Ren, A.;
Garner, S.; Chem, A.; Londergan, T.; Irwin, L.; Carlson, B.; et al. J.
Mater. Chem. 1999, 9, 1905−1920.
(4) Marder, S. R.; Kippelen, B.; Jen, A. K. Y.; Peyghammbarian, N.
Nature 1997, 388, 845−851.
As shown in Figure.14, the poling efficiency is also greatly
impacted by the type of chromophores incorporated into the
13804
dx.doi.org/10.1021/jp307370y | J. Phys. Chem. B 2012, 116, 13793−13805

The Journal of Physical Chemistry B
Article
(5) Shi, Y.; Zhang, C.; Zhang, H.; Bechtel, J. H.; Dalton, L. R.;
Robinson, B. H.; Steier, W. H. Science 2000, 288, 119−122.
(6) Robinson, B. H.; Dalton, L. R. J. Phys. Chem. A 2000, 104, 4785−
4795.
(7) Benight, S. J.; Robinson, B. H.; Dalton, L. R. Nano-engineering of
Molecular Interactions in Organic electro-Optic Materials. In
Molecular Interactions, Meghea, A., Ed.; InTech: Rijeka, Croatia, 2012.
(8) Dalton, L. R.; Benight, S. J.; Johnson, L. E.; Knorr, D. B.;
Kosilkin, I.; Eichinger, B. E.; Robinson, B. H.; Jen, A. K. Y.; Overney,
R. M. Chem. Mater. 2011, 23, 430−445.
(9) Zhou, X.-H.; Luo, J.; Huang, S.; Kim, T.-D.; Shi, Z.; Cheng, Y.-J.;
Jang, S.-H.; Knorr, D. B., Jr.; Overney, R. M.; Jen, A. K. Y. Adv. Mater.
2009, 21, 1976−1981.
(10) Knorr, D. B.; Zhou, X. H.; Shi, Z. W.; Luo, J. D.; Jang, S. H.; Jen,
A. K. Y.; Overney, R. M. J. Phys. Chem. B 2009, 113, 14180−14188.
(11) Gray, T.; Kim, T.-D.; Knorr, D. B., Jr.; Luo, J.; Jen, A. K. Y.;
Overney, R. M. Nano Lett. 2008, 8, 754−759.
(12) Gray, T.; Killgore, J. P.; Luo, J.; Jen, A. K.-Y.; Overney, R. M.
Nanotechnology 2007, 18, 044009.
(37) Gu, X.; Knorr, D. B., Jr.; Wang, G.; Overney, R. M. Thin Solid
Films 2011, 519, 5955−5961.
(38) Knorr, D. B. J.; Gray, T.; Overney, R. M. Ultramicroscopy 2009,
109, 991−1000.
(39) Knorr, D. B., Jr.; Widjaja, P.; Acton, O.; Overney, R. M. J. Chem.
Phys. 2011, 134, 104502.
(40) Sills, S.; Gray., T.; Overney, R. M. J. Chem. Phys. 2005, 123,
134902.
(41) Ferry, J. D. Dependence of viscoelastic behavior on temperature
and pressure. In Viscoelastic Properties of Polymers; John Wiley & Sons:
New York, 1980; Chapter 11.
(42) Ward, I. M. Mechanical Properties of Solid Polymers; Wiley-
Interscience: London, 1971.
(43) Sills, S.; Overney, R. M. Phys. Rev. Lett. 2003, 91, 095501.
(44) Pople, J. A.; Santry, D. P. J. Chem. Phys. 1965, 43, S129.
(45) Dewar, M. J. S.; Yuan, Y. C. Inorganic Chemistry 1990, 29,
3881−3890.
(46) Dewar, M. J. S.; Zoebisch, E. G. THEOCHEM 1988, 180, 1−21.
(47) Dewar, M. J. S.; Zoebisch, E. G.; Healy, E. F.; Stewart, J. J. P. J.
Am. Chem. Soc. 1985, 107, 3902−3909.
(13) Luo, J.; Liu, S.; Haller, M.; Liu, L.; Ma, H.; Jen, A. K.-Y. Adv.
Mater. 2002, 14, 1763−1768.
(14) Kim, T.-D.; Kang, J.-W.; Luo, J.; Jang, S.-H.; Ka, J.-W.; Tucker,
N.; Benedict, J. B.; Dalton, L. R.; Gray, T.; Overney, R. M.; et al. J. Am.
Chem. Soc. 2007, 129, 488−489.
́ ́ ́
(48) Kiss, J. T.; Felfoldi, K.; Kortvelyesi, T.; Palinko, I. Vibr. Spectrosc.
̈
̈
2000, 22, 63−73.
(49) Knorr, D. B. J.; Gray, T.; Overney, R. M. J. Chem. Phys. 2008,
129, 074504.
(50) Patrick, C. R.; Prosser, G. S. Nature 1960, 187, 1021−1021.
(51) Smith, V. B.; Massey, A. G. Tetrahedron 1969, 25, 5495−5501.
(52) Weck, M.; Dunn, A. R.; Matsumoto, K.; Coates, G. W.;
Lobkovsky, E. B.; Grubbs, R. H. Angew. Chem., Int. Ed. 1999, 38,
2741−2745.
(53) Collings, J. C.; Roscoe, K. P.; Robins, E. G.; Batsanov, A. S.;
Stimson, L. M.; Howard, J. A. K.; Clark, S. J.; Marder, T. B. New J.
Chem. 2002, 26, 1740−1746.
(54) Kishikawa, K.; Oda, K.; Aikyo, S.; Kohmoto, S. Angew. Chem.,
Int. Ed. 2007, 46, 764−768.
(55) Watt, S. W.; Dai, C.; Scott, A. J.; Burke, J. M.; Thomas, R. L.;
Collings, J. C.; Viney, C.; Clegg, W.; Marder, T. B. Angew. Chem., Int.
Ed. 2004, 43, 3061−3063.
(56) Starkweather, H. W., Jr. Macromolecules 1981, 14, 1277−1281.
(57) Starkweather, H. W., Jr. Macromolecules 1988, 21, 1798−1802.
(58) Arrese-Igor, S.; Arbe, A.; Frick, B.; Colmenero, J. Macromolecules
2011, 44, 3161−3168.
(59) Yano, O.; Wada, Y. J. Polym. Sci., Part A: Polym. Phys. 1971, 9,
669−686.
(15) Benight, S. J.; Johnson, L. E.; Barnes, R.; Olbricht, B. C.; Bale, D.
H.; Reid, P. J.; Eichinger, B. E.; Dalton, L. R.; Sullivan, P. A.; Robinson,
B. H. J. Phys. Chem. B 2010, 114, 11949−11956.
(16) Benight, S. J.; Bale, D. H.; Olbricht, B. C.; Dalton, L. R. J. Mater.
Chem. 2009, 19, 7466−7475.
(17) Hah, H.; Sung, S.-J.; Cho, K. Y.; Park, J.-K. Polym. Bull. 2008, 61,
383−390.
(18) Li, W. J.; Lynch, V.; Thompson, H.; Fox, M. A. J. Am. Chem. Soc.
1997, 119, 7211−7217.
(19) Yu, Y.-J.; Lee, S.-H.; Choi, D. H.; Jin, J.-I.; Tessler, N. Macromol.
Res. 2007, 15, 142−146.
(20) Tian, Y. Q.; Akiyama, E.; Nagase, Y.; Kanazawa, A.; Tsutsumi,
O.; Ikeda, T. J. Mater. Chem. 2004, 14, 3524−3531.
(21) Tian, Y. Q.; Kong, X. X.; Nagase, Y.; Iyoda, T. J. Polym. Sci., Part
A: Polym. Chem. 2003, 41, 2197−2206.
(22) Dalton, L.; Benight, S. Polymers 2011, 3, 1325−1351.
(23) Benight, S. J.; Knorr, D. B.; Johnson, L. E.; Sullivan, P. A.; Lao,
D.; Sun, J.; Kocherlakota, L. S.; Elangovan, A.; Robinson, B. H.;
Overney, R. M.; Dalton, L. R. Adv. Mater. 2012, 24, 3263−3268.
(24) Shen, K. H.; Kim, J. H.; Kim, G. W.; Cho, M. J.; Lee, S. K.; Choi,
D. H. Macromol. Res. 2005, 13, 180−186.
(25) Legge, C. H.; Whitcombe, M. J.; Gilbert, A.; Mitchell, G. R. J.
Mater. Chem. 1991, 1, 303−304.
(26) Palinko, I. Acta Crystallogr., Sect. B: Struct. Sci. 1999, 55, 216−
220.
(27) Kortvelyesi, T.; Palinko, I. Theochem 2000, 531, 99−103.
(28) Pires, D.; Hedrick, J. L.; De Silva, A.; Frommer, J.; Gotsmann,
B.; Wolf, H.; Despont, M.; Duerig, U.; Knoll, A. W. Science 2010, 328,
732−735.
(29) Luo, J.; Huang, S.; Cheng, Y.-J.; Kim, T.-D.; Shi, Z.; Zhou, X.-
H.; Jen, A. K.-Y. Org. Lett. 2007, 9, 4471−4474.
(30) Bhatt, S.; Nayak, S. K. Synth. Commun. 2007, 37, 1381−1388.
(31) He, M.; Leslie, T. M.; Sinicropi, J. A. Chem. Mater. 2002, 14,
4662−4668.
(32) Pond, S. J. K.; Rumi, M.; Levin, M. D.; Parker, T. C.; Beljonne,
D.; Day, M. W.; Bredas, J. L.; Marder, S. R.; Perry, J. W. J. Phys. Chem.
A 2002, 106, 11470−11480.
(33) Sills, S. E.; Overney, R. M.; Chau, W.; Lee, V. Y.; Miller, R. D.;
Frommer, J. J. Chem. Phys. 2004, 120, 5334.
(34) Overney, R. M.; Buenviaje, C.; Luginbuhl, R.; Dinelli, F. J.
Therm. Anal. Cal. 2000, 59, 205−225.
(35) Buenviaje, C. K.; Ge, S. R.; Rafailovich, M. H.; Overney, R. M.
Mater. Res. Soc. Symp. Proc. 1998, 552, 187−192.
(36) Wei, J. H.; He, M. Y.; Overney, R. M. J. Membr. Sci. 2006, 279,
608−614.
13805
dx.doi.org/10.1021/jp307370y | J. Phys. Chem. B 2012, 116, 13793−13805