Optically Reconfigurable Monolayer of Azobenzene Donor Molecules on Oxide Surfaces

Article abstract of DOI:10.1021/acs.langmuir.6b04585

The structural configuration of molecules assembled at organic-inorganic interfaces within electronic materials strongly influences the functional electronic and vibrational properties relevant to applications ranging from energy storage to photovoltaics.

Full text of DOI:10.1021/acs.langmuir.6b04585

Subscriber access provided by UB + Fachbibliothek Chemie | (FU-Bibliothekssystem)
Article
Optically reconfigurable monolayer of
azobenzene-donor molecules on oxide surfaces
Kyle M McElhinny, Peishen Huang, Yongho Joo, Catherine Kanimozhi,
Arunee Lakkham, Kenji Sakurai, Paul G. Evans, and Padma Gopalan
Langmuir, Just Accepted Manuscript • DOI: 10.1021/acs.langmuir.6b04585 • Publication Date (Web): 07 Feb 2017
Downloaded from http://pubs.acs.org on February 12, 2017
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a free service to the research community to expedite the
dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts
appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been
fully peer reviewed, but should not be considered the official version of record. They are accessible to all
readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered
to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published
in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just
Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor
changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers
and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors
or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
Langmuir 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 45
Langmuir
1
2
3
4
Optically reconfigurable monolayer of azobenzene-donor molecules on oxide surfaces
5
6
7
8
Kyle M. McElhinny1*, Peishen Huang1, Yongho Joo1, Catherine Kanimozhi1, Arunee Lakkham1,
Kenji Sakurai2,3, Paul G. Evans1b*, Padma Gopalan1a*,
9
1 Department of Materials Science and Engineering, University of Wisconsin-Madison, Madison,
Wisconsin 53706, USA.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
2 National Institute for Materials Science, Tsukuba, Ibaraki 305-4007, Japan
3 University of Tsukuba, Tsukuba, Ibaraki 305-8577, Japan
a* Electronic mail: pgopalan@wisc.edu
b* Electronic mail: pgevans@wisc.edu
Abstract
The structural configuration of molecules assembled at organic-inorganic interfaces within
electronic materials strongly influences functional electronic and vibrational properties relevant to
applications ranging from energy storage to photovoltaics. Controlling and characterizing the
structural state of an interface and its evolution under external stimuli is crucial both for
fundamental understanding of the factors influenced by molecular structure and to the
development of methods for materials synthesis. It has been challenging to create complete
molecular monolayers that exhibit external reversible control of the structure and electronic
configuration. We report a monolayer/inorganic interface consisting of an organic monolayer
assembled on an oxide surface exhibiting structural and electronic rearrangement under ultraviolet
illumination. The molecular monolayer is linked to the surface through carboxylate link, with the
backbone bearing an azobenzene functional and the head group consisting of a Re-bipyridine
group. Optical spectroscopy, x-ray photoelectron spectroscopy, atomic force microscopy, and x-
ray reflectivity show that closely packed monolayers are formed from these molecules via the
Langmuir-Blodgett technique. Reversible photoisomerization is observed in solution and in
monolayers assembled on Si and quartz substrates. The reconfiguration of these monolayers
1
ACS Paragon Plus Environment

Langmuir
Page 2 of 45
1
2
3
4
5
6
provides additional means to control excitation and charge transfer processes important in
applications in catalysis, molecular electronics, and solar energy conversion.
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
2
ACS Paragon Plus Environment

Page 3 of 45
Langmuir
1
2
3
4
Introduction
5
6
7
8
9
The creation of materials incorporating optically reconfigurable interfaces between organic
and inorganic components provides access to a wide range of new electronic and optical
phenomena with both fundamental and practical importance. Dynamically reconfiguring the
structure of an interfacial molecular monolayer through the use of external stimuli has the potential
to provide precise control over charge transfer dynamics, the energy offsets of interfacial electronic
states, dipole moments or polarizations, and vibrational phenomena. In the past, various structural
configurations have been created by varying the conditions under which the monolayers are
assembled. Such studies indicate that changes in the structural state can include change in the
molecular height and conformation1–4 and an increase or decrease in the dipole moment and net
polarization.5–7
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
In addition to these modifications of the fundamental properties of the interfaces,
computational studies proposed that it could be possible to create high-energy-density solar fuels
through a mechanism in which energy is stored in the metastable molecular configuration, which
is released through isomerization into the stable trans isomeric state via optical absorption.8–10
This technology was recently realized through the creation of a high energy density
azobenzene/graphene hybrid as a solar fuel exhibiting a high energy density of up to 112 W h kg-
1 with a half-life of 33 days.11 Additionally, a photoresponsive cross-linked azobenzene/polymer
assembly has demonstrated its feasibility as a high performance light driven actuator.12 Similarly,
the structural cues provided by azobenzene monolayers can drive photoinduced alignment in liquid
crystal films and create surfaces with photoswitchable wettability. 1,13–15
A further range of opportunities to achieve reconfigurable interfacial control of materials
properties arises in organic donor/inorganic semiconductor materials through the incorporation of
3
ACS Paragon Plus Environment

Langmuir
Page 4 of 45
1
2
3
4
a photoexcitable molecular monolayer at the interface. Molecular donors deposited on
5
6
16–18
semiconducting oxides such as TiO2
ZnO19, and SnO220 , can inject electrons into the oxide
,
7
8
9
from their metal-to-ligand charge transfer excited state on the picosecond timescale.16–19 Molecular
monolayers and thin films exhibit variation in molecular orientation, conformation, and binding
motif, as well as macroscopic parameters including thickness, roughness, packing density, and
aggregation, all of which can affect the charge transfer rates. Likewise, the rates of charge transfer,
through either conjugated links or through-space tunneling, depend on factors such as the
composition and structure of the chromophore and the anchoring group, the extent of conjugation
of the molecular bridge group, and macroscopic structural factors such as the packing density and
the aggregation state.16,18,21,22 In this context, Re-bipyridine complexes have been extensively
studied as photoabsorbers and electron donors.18,23,24
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Donor molecule/semiconductor interfaces are electronically and structurally intriguing
because of the strong connection between charge transfer rates and the spatial separation of the
donating group and the substrate. This donor-surface distance can be manipulated through the
incorporation of a reconfigurable group which changes the distance between donor and substrate
in response to an external stimulus.18,21,22 The challenge has been to fabricate reconfigurable
interfaces with a specific structure and to correlate the corresponding electronic kinetics
systematically. Additionally, it is challenging to directly evaluate the structure and structural
changes quantitatively during and after external stimulus in these reconfigurable systems.
Reconfigurability can be introduced into organic/inorganic interfaces by assembling
molecules containing azo-groups at interfaces.1–3,25 The azo group switches from trans to cis
configurations following electronic excitation by a photon resonant with the π to π* transition at a
wavelength of 320 to 350 nm. The cis-to-trans transformation can be induced by exciting the n to
4
ACS Paragon Plus Environment

Page 5 of 45
Langmuir
1
2
3
4
5
6
7
8
9
π* transition via the absorption of blue light at a wavelength of 400-450 nm.4 In solution, the
azobenzene molecule reconfigures on a 10 ps timescale.26,27 Thermal isomerization back to the
thermodynamically stable trans state occurs much more slowly, on the order of milliseconds to
days depending on the stability and steric bulk of the other constituents of the azobenzene-
containing molecule.4 The timescale on which the trans to cis ratio in an ensemble changes is often
much longer due to the different probabilities of the elementary isomerization processes. This
change in the ensemble ratio of trans and cis isomers has commonly been observed to be on the
timescale of seconds.5,28 The photoisomerization leads to changes in the molecular conformation,
from the planar, elongated trans isomer to the compact three-dimensional cis isomer, which is
accompanied by a change in the dipole moment, a decrease in the intensity of π to π*peak and an
increase in the intensity of the n to π* peak in the absorption spectra.4
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Steric hindrance and the restricted free volume available due to dense molecular packing
limit the photoisomerization rates of azobenzene monolayers, prolonging the timescale of the
initial reconfiguration to the order of seconds.2,28–30 In self-assembled azobenzene monolayers with
100% coverage, vibrational sum-frequency generation measurements have observed no evidence
for photoisomerization, despite the observation of photoisomerization at sub-monolayer
coverages.29 Additional insight into the role of mesoscopic effects is observed in scanning
tunneling microscopy (STM) experiments, which show that chains with widths of a few molecules
and lengths of 20 to 300 Å photoisomerize more slowly under UV radiation than their isolated
counterparts and exhibited concerted switching.2 The concerted switching phenomenon observed
using STM was described as arising from intermolecular electronic coupling due to enhanced -
orbital overlap of the neighboring azobenzene groups.2 These previous experimental results offer
5
ACS Paragon Plus Environment

Langmuir
Page 6 of 45
1
2
3
4
5
6
insight into the behavior of localized and isolated sets of molecules, but do not describe the
behavior of complete monolayers.
7
8
9
In this work we report the synthesis of a novel photoisomerizable donor (ReAzoC)
molecule (Figure 1) and its assembly on oxide surfaces. The molecules were organized into an
ordered monolayer by Langmuir-Blodgett (LB) deposition, an approach we have previously
developed for a series of other rhenium-bipyridine complexes.31 Structural characterization shows
that LB assembly yields a monolayer structure with the rhenium bipyridine complex as the head
group and with carboxylic acid anchor to the substrate, as in Figure 1(b). The molecules show
spectroscopic evidence of photoisomerization in solution and as deposited in a monolayer. The
structure is thus further tunable using the trans to cis photoisomerization of the azobenzene group.
Additionally, LB assembly can be reproducibly used to create monolayers with a series of different
molecular coverage by varying surface pressure and an optimized single monolayer when
deposited at the correct conditions.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
The order resulting from this simple deposition process permits more precise structural
characterization of the molecular reconfiguration than has been possible in the past. The
incorporation of the Re metal complex in the monolayer increases the electron density contrast,
enabling precise x-ray reflectivity (XRR) studies of the mechanisms of structural reconfiguration.
XRR quantitatively characterizes the structure and structural changes in large areas of complete
monolayers with sensitivity to in-plane structural roughness at the 100 nm lateral length scale and
the molecular configuration at the Angstrom scale. The XRR studies reported here characterize
the degree of order within the monolayer, the orientation of molecules on the surface, the
isomerization state, and the evolution of the structure under optical exposure.
6
ACS Paragon Plus Environment

Page 7 of 45
Langmuir
1
2
3
4
5
6
7
8
9
The combination of structural and spectroscopic characterization studies described below
shows that ReAzoC can be organized into a closely-packed complete monolayer. Optical
absorption measurements indicate that 13-15% of the molecules on the surface can be reversibly
optically driven between structural states by illumination with alternating UV and blue
illumination and that larger fractions of up to 47-52% can be irreversibly transformed at high
optical power. X-ray and optical spectroscopy provide independent and consistent estimates of the
fraction of isomerized molecules and the resulting monolayer structure. At low optical power, a
small fraction of molecules isomerize in a spatially random distribution. At higher optical powers,
where optical absorption indicates that UV light isomerizes approximately 50% of molecules there
is an inhibition of the cis-to-trans reverse photoisomerization. XRR measurements show that the
mean height of the donor groups within the monolayer is reproducibly changed as a result of the
trans-to-cis isomerization.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Experimental Section
ReAzoC Synthesis. Synthesis of 4’-bromomethyl-4-methyl-2,2’-bipyridine (3) (Modified
synthetic conditions from ref. 32)
A solution of 4,4’-Dimethyl-2,2’-dipyridine (1, 1.00 g, 5.43 mmol) in anhydrous
tetrahydrofuran (THF, 30 mL) was prepared under protection of argon (Ar) and pre-cooled at -78
C. Lithium diisopropylamide (LDA) stock solution (2.85 mL, 2.0 M, in THF) was then added
dropwise to the reaction mixture. A dark brown solution was generated and stirred at room
temperature for an additional 2 hours. The solution was cooled to -78 °C before adding
chlorotrimethylsilane (1.5 mL, 1.18 mmol). After 5 min, the reaction was quenched by ethanol (3
mL) and stirred for 10 min. The final reaction mixture was isolated and purified using saturated
sodium bicarbonate (NaHCO3) aqueous solution (50 mL), deionized (DI) water and ethyl acetate
7
ACS Paragon Plus Environment

Langmuir
Page 8 of 45
1
2
3
4
5
6
7
8
9
(3 × 200 mL). The organic phase was combined and dried by magnesium sulfate (Mg2SO4),
filtered, concentrated and purified in a silica plug to yield a white solid (2, 1.35 g, 96%). 1H-NMR
(500 MHz, CDCl3, TMS) δ (ppm): 0.04 (s, 9H), 2.21 (s, 2H), 2.43 (s, 3H), 6.95 (dd, 1H), 7.12 (d,
1H), 8.05 (s, 1H), 8.22 (s, 1H), 8.47 (d, 1H), 8.52 (d, 1H).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
CsF (2.24 g, 14.7 mmol) was added to a mixture of 2 (1.40 g, 5.4 mmol) and
dibromotetrafluoroethane (1.36 mL, 11.4 mmol) in 50 mL of dimethylformamide (DMF). The
reactants formed a suspension and were stirred at room temperature for 3 h. The final reaction
mixture was isolated and purified using water/ethyl acetate, washed by brine (3 × 200 mL), dried
by Mg2SO4, filtered, concentrated, and purified in a silica plug to yield a colorless oil, which turned
into a white solid under vacuum (3, 1.4 g, 97%). 1H-NMR (500 MHz, CDCl3, TMS) δ (ppm): 2.45
(s, 3), 4.48 (s, 2), 7.16 (d, 1H), 7.34 (d, 1H), 8.24 (s, 1H), 8.42 (s, 1H), 8.55 (d, 1H), 8.68 (d, 1H).
Synthesis of 4-[2-(4-hydroxyphenyl)diazenyl]-benzoic acid methylester (5) (Optimized
synthetic route according to ref. 33)
4-amino-benzoic acid methyl ester (1.00 g, 6.62 mmol) was dissolved in 100 mL DI water
with the minimum of HCl aqueous solution (1 M). A freshly prepared sodium nitrite (NaNO2)
aqueous solution (480 mg, 50 mL) was then slowly added under ice bath. The mixture was stirred
at room temperature for 20 min to generate 4-(methoxycarbonyl)-benzenediazonium chloride (4),
which was used in situ without any further isolation or purification.
Phenol (685 mg, 7.28 mmol) was dissolved in a component solvent of acetic acid and water
(1:2, v/v), and then slowly added to the prepared solution of compound 4 at room temperature.
The reaction was quenched by satiated NaHCO3 aqueous solution carefully under ice bath until
the formation of gas bubbles ceased. A light yellow solid was precipitated, filtered and washed by
large amount of water as the final product (5, 1.70 g, 98%). 1H-NMR (500 MHz, DMSO-D6, TMS)
8
ACS Paragon Plus Environment

Page 9 of 45
Langmuir
1
2
3
4
5
6
7
8
9
δ (ppm): 3.88 (s, 3H), 6.93 (d, 2H), 7.83 (d, 1H), 7.88 (d, 1H), 8.11 (d, 1H). 13C-NMR (126 MHz,
DMSO-D6, TMS) δ (ppm): 166.19, 163.62, 155.35, 145.28, 130.90, 130.74, 125.99, 122.59,
116.84, 52.78.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Synthesis of 4-{2-[4-(4’-methoxyl-4-methyl-2,2’-bipyridyl)-phenyl]diazenyl}-benzoic acid
methylester (6) (Optimized synthetic conditions on different reactants according to ref. 34)
A solution of compound 3 (1.43 g, 5.43mmol) in DMF (50 mL) was combined with
compound 5 (1.43 mg, 5.58 mmol), K2CO3 (1.2 g, 8.68 mmol), and 18-crown-6 (75 mg, 0.28
mmol) while stirring at 65 C overnight. The resulting reddish solution was concentrated using a
rotary evaporator and recrystallized in chloroform/methanol three times to yield a yellow solid
(1.85 g, 78%). 1H-NMR (500 MHz, CDCl3, TMS) δ (ppm): 2.45 (s, 3H), 3.95 (s, 3H), 5.27 (s, 2H),
7.11 (d, 2H), 7.16 (d, 1H), 7.42 (d, 1H), 7.91 (d, 2H), 7.96 (d,2H), 8.17 (d,2H), 8.27 (s,1H), 8.48
(s, 1H), 8.56, (d, 1H), 8.70 (d, 1H). 13C-NMR (126 MHz, CDCl3, TMS) δ (ppm): 166.60, 161.15,
156.70, 155.52, 155.27, 149.56, 149.02, 148.29, 147.38, 146.54, 131.30, 130.59, 125.24, 124.97,
122.41, 122.10, 121.40, 118.94, 115.20, 68.67, 52.29, 21.22. MS (C26H22N4O3) m/z: [M+H]+
calculated: 439.1765 found: 439.1761.
Synthesis of 4-{2-[4-(4’-methoxyl-4-methyl-2,2’-bipyridyl)-phenyl]diazenyl}-benzoic acid
(7)
DMF (30 mL) was used to dissolve compound 6 (300 mg, 0.684 mmol). Sodium hydroxide
(NaOH, 80 mg, 2 mmol) was first dissolved in water (1 mL) and then added to the DMF solution
of 6. The reaction was kept under 90 C for 18 hours, then the DMF was removed through the use
of a rotary evaporator. The crude yellow slush was first washed by water/methanol (10:1 v/v), then
recrystallized by ethyl acetate/hexane to obtain a yellow powder as the final product (240 mg,
83%). 1H-NMR (500 MHz, DMSO-D, TMS) δ (ppm): 2.48 (s, 3H), 5.44 (s, 2H), 7.29 (d, 2H), 7.41
9
ACS Paragon Plus Environment

Langmuir
Page 10 of 45
1
2
3
4
5
6
7
8
9
(d, 1H), 7.60 (d, 1H), 7.90 (d, 2H), 7.95 (d,2H), 8.12 (d,2H), 8.35 (s,1H), 8.53 (s, 1H), 8.59 (d,
1H), 8.74 (d, 1H). 13C-NMR (126 MHz, DMSO-D, TMS) δ (ppm): 166.61, 166.46, 161.08, 157.48,
154.21, 149.32, 148.92, 147.46, 147.29, 146.43, 132.04, 130.17, 125.26, 124.61, 122.21, 121.83,
121.78, 118.75, 115.37, 67.96, 20.54. MS (C25H20N4O3) m/z: [M+H]+ calculated: 425.1608,
found: 425.1602.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Synthesis of 4-{2-[4-(4’-methoxyl-4-methyl-2,2’-bipyridyl)-phenyl]diazenyl}-benzoic acid
Re complex (8)
Compound 7 (300 mg, 0.236 mmol) and pentacarbonylchlororhenium(I) (93.7 mg, 0.260
mmol) was dissolved in degassed DMF (15 mL) under protection of Ar. The solution was stirred
at 70 C for 15 hours. The DMF was pumped off under vacuum, and the solid crude product was
recrystallized by a compound solvent of diethyl ether and hexane. A bright orange powder was
obtained as the final product (8, 141 mg, 82%). 1H-NMR (500 MHz, DMSO-D, TMS) δ (ppm):
2.58 (s, 3H), 5.54 (s, 2H), 7.35 (d, 2H), 7.61 (d, 1H), 7.82 (d, 2H), 7.94 (d,2H), 8.01 (d,2H), 8.13
13
(s,2H), 8.69 (s, 1H), 8.85 (d, 1H), 8.86 (d, 1H), 9.04 (d, 1H). C-NMR (126 MHz, DMSO-D,
TMS) δ (ppm): 197.73, 197.65, 190.04, 166.61, 166.47, 162.17, 160.79, 155.38, 154.44, 154.28,
152.95, 152.33, 152.20, 150.28, 146.63, 132.21, 130.50, 128.49, 125.43, 124.96, 122.20, 121.99,
115.61, 67.59, 20.85. MS (C28H20ClN4O6Re) m/z: [M+H]+ calculated: 731.0702, found: 731.0693.
Substrate Materials and Preparation. Quartz substrates were obtained from Structure
Probe, Inc. under the trade name SPI SuperSmooth Quartz Slides. The manufacturer’s
specification for the root-mean-square (rms) roughness of these quartz slides is less than 0.5 nm.
Minimizing substrate roughness maximizes the quality of XRR data can be achieved through both
improvements in the vertical alignment of electron rich portion of molecule and a reduction in
interfacial roughness. Si/SiO2 substrates were purchased from Montco Silicon Technologies. The
10
ACS Paragon Plus Environment

Page 11 of 45
Langmuir
1
2
3
4
Si/SiO2 and quartz substrates were cleaned by the piranha process (7:3 H2SO4:H2O2) at 90C for
5
6
7
30 min and rinsed with DI water.
8
9
LB Deposition. LB characterization and deposition experiments were performed using a
LB trough (KSV KIMA Medium size KN2002, Helsinki, Finland). The surface pressure-area (π-
A) isotherm of donor molecules molecule was measured with a Wilhelmy balance (Pt plate). The
water subphase was treated to have a resistivity of 18.2 MΩ cm by a Milli-Q water filtration station
(Millipore, Inc.). The molecular layers were transferred to the solid substrates using the LB dipping
technique at a set surface pressure. A ReAzoC solution with a concentration of 200 L of (1
mg/mL CHCl3:THF=9:1) was used. The solvent was allowed to evaporate for 30 min before
compression of the monolayer. A drop cast film was prepared for comparison by dropping ReAzoC
solution (1mg/mL CHCl3:THF=9:1) onto a Si substrate prepared via the piranha treatment and
allowed to evaporate for 24 h in vacuum.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Atomic Force Microscopy. The surface morphology of the monolayers of ReAzoC
molecules was imaged using a Nanoscope III Multimode atomic force microscope (Digital
Instruments). A non-contact imaging mode was used for the AFM measurement. A triangular
cantilever with an integral pyramidal Si3N4 tip was used. The typical imaging force was on the
order of 1 nN. The thickness of the monolayer was determined using line profiles through pinholes
in the film.
Ultraviolet-visible (UV-Vis) Spectroscopy. UV-Vis absorption measurements were
performed in a transmission geometry in a wavelength range of 200 to 800 nm using a Shimadzu
PC-2401 spectrometer.
11
ACS Paragon Plus Environment

Langmuir
Page 12 of 45
1
2
3
4
5
6
7
8
9
Light Sources. Two different light sources were used to photoisomerize the molecules.
Prizmatix FC8-LEDs with wavelengths of 365 nm and 465 nm were used for the UV-Vis studies.
Light emitting diode (LED) sources (Engin LZ1-00DB00 and LZ1-00UV00, with peak
wavelengths of 460 nm and 365 nm, respectively) were employed for XRR studies and studies of
sequential XRR measurements and optical exposures.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
X-Ray Photoelectron Spectroscopy (XPS). A Thermo Scientific K-alpha XPS was used
for compositional analysis of films deposited on Si/SiO2 substrates. The XPS instrument used a
microfocused monochromated Al Kα X-ray source with a spot size of 400 µm. A pass energy of
50 eV was used to collect survey spectra. Spectra were analyzed using the manufacturer’s
Avantage software and peak fitting with Gaussian/Lorentzian peak shape and a Shirley/Smart
baseline. The surface coverage, defined as the number of molecules per unit area on the substrate,
was determined using:
ꢎꢒꢛꢓ
ꢅꢆꢘꢚꢐ
ꢄ,ꢔꢕꢖꢗꢘꢆꢙ
ꢈ
ꢋ
ꢇꢌꢍꢎꢏꢐꢑ
ꢅꢆ,ꢅꢆꢉ ꢅꢆ,ꢅꢆꢉ
ꢊ
ꢃ
ꢄ
ꢇ
ꢅꢆ
ꢊ
ꢀꢁ ꢂ ꢃ
.
(1)
ꢎꢒꢛꢓ
ꢅꢆꢘꢚꢐ
ꢇ
ꢄ
ꢄ,ꢔꢕꢖꢗꢘꢆꢙ
ꢅꢆ
ꢑ
⁄
⁄
Here, ꢀꢁ is number of atoms per unit area for element x, and ꢜꢁ ꢜꢝand ꢞꢟ ꢞꢁ are the ratio of
integrated peak areas for element x, and the ratio of the sensitivity factors of element x on the
substrate Si, respectively. The factor ꢠꢇꢌ,ꢇꢌꢡ is the number of Si atoms per unit volume in SiO2,
ꢊ
and ꢢꢇꢌ,ꢇꢌꢡ is inelastic mean free path (IMFP) of Si photoelectrons in SiO2. The layer thickness is
ꢊ
t, and ꢢꢁ,ꢣꢤꢥꢦꢍꢌꢧ is the IMFP for the electrons emitted from element x in the molecular monolayer.
X-Ray Reflectivity. XRR measurements were performed at the University of Wisconsin-
Madison and at the National Institute for Materials Science (NIMS). Measurements at NIMS used
a custom-built x-ray reflectometer with an x-ray source operating at 1.6 kW producing Cu Kα
radiation. The incident beam width measured using a knife edge was 42.6 μm. Measurements
12
ACS Paragon Plus Environment

Page 13 of 45
Langmuir
1
2
3
4
5
6
7
8
9
conducted at UW-Madison used a Panaltytical X’PERT MRD using Cu Kα radiation at a source
power of 1.8 kW, for which the incident beam width was 220 μm. The analysis of the
experimentally acquired XRR intensity distributions was conducted using the GenX software
package.35 The GenX package employs a curve fitting algorithm that adjusts values of the interface
roughness, layer thickness, and density to optimize a figure of merit Log(R1):
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
∑
ꢬꢭlog ꢰ ꢲꢳ ꢴ log ꢰ ꢞ ꢳꢭꢵ
ꢱ
ꢱ
ꢌ
ꢮꢯ
ꢌ
ꢮꢯ
ꢌ
ꢎ
ꢐ
ꢨꢩꢪ ꢫ1 ꢂ
ꢎꢶꢐ
∑
ꢬlog ꢰ ꢲ ꢳꢵ
ꢱ
ꢌ
ꢮꢯ
ꢌ
where Yi is the measured reflected intensity, Si is the corresponding simulated value, and the sum
spans the series of data points.
Results and Discussion
Design and Synthesis of ReAzoC. We designed a unique azo-based chromophore
ReAzoC that contains three moieties relevant to the XRR studies. The head group containing the
bipyridine group can complex with transition metal ions to increase the electron density contrast
and hence the x-ray scattering cross section. The middle unit of the molecule contains the photo-
switchable azo group which is used to reconfigure the monolayer by photo-isomerization from
trans to the cis form. The tail group of the molecule is a polar carboxylic acid group which anchors
the molecule at the water sub phase during assembly and subsequently to the oxide substrate.
Our initial attempts to connect the bipyridine head group to the azo-benzene unit via an
36,37
ester link were unsuccessful because the bipyridine is labile either under low pH
or in the
presence of acyl halides38 in traditional esterification reactions. Mitsunobu esterification requires
a organophosphate catalysts, which interferes with the azo-benzene in the chromophore and thus
cannot be used in this case.39,40 The Steglich esterification showed good compatibility with azo-
13
ACS Paragon Plus Environment

Langmuir
Page 14 of 45
1
2
3
4
5
6
7
8
9
benzene in our previous work41,42, but it led to low yields when hydroxy-azobenzenes were reacted
with aromatic acid.43 Thus, we choose a ether linker via SN2 substitution44 of 4’-bromomethyl-4-
methyl-2,2’-bipyridine (3) by activated methyl-4’-hydroxy-4-azobenzylcarboxylate (5) under mild
reaction conditions. These two precursors were obtained with over 90% yields under optimized
conditions. As a final step the carboxylic acid was deprotected by basic hydrolysis. It was essential
to conduct the deprotection as the very last step because the carboxylic acid of (5) has low
solubility in most organic solvents leading to low yields of methyl-4’-(4-methyl-2,2’-bipyridyl-4’-
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
methoxyl)-4-azobenzylcarboxylate
(6)
and
4’-(4-methyl-2,2’-bipyridyl-4’-methoxyl)-4-
azobenzylcarboxylic acid (7). In order to form the carboxylic acid from sodium carboxylate salt,
the pH was adjusted to 4.5 as the bipyridine group is acid sensitive.45 Based on these observations
an eight-step (Scheme 1) synthesis was carried to achieve the final material with an overall yield
1
of 48%, and the purity was confirmed by H/13C-NMR and high resolution mass spectrometry
(MS).
Langmuir-Blodgett (LB) Deposition. The LB method is typically used to assemble
amphiphilic molecules such as lipids at the air-water interface and transferring the ordered
monolayer on to a substrate. Although not truly amphiphilic, the ReAzoC molecule has a more
polar carboxylic acid tail group which makes it a good candidate for stabilization at air/water
interface and to preferentially interact with oxide substrates.31 For LB monolayer deposition,
ReAzoC was spread on the water subphase. After complete solvent evaporation, a surface
pressure-mean molecular area isotherm was recorded at 20 °C with a compression speed of 10
mm/min (Figure 2). The isotherm in Figure 2 exhibits three distinct phases: gas, liquid, and solid,
each separated by discontinuities in the slope of the pressure-area curve. For the isotherm of
ReAzoC donor molecules, the transition from a two-dimensional gas to liquid phase and from
14
ACS Paragon Plus Environment

Page 15 of 45
Langmuir
1
2
3
4
5
6
7
8
9
liquid to solid phase occurs at a mean molecular area of 125 Å2 and 50 Å2, respectively, as shown
in Figure 3. Based on bond angles and bond lengths given by Dominey et al.,46 the trans state of
the molecule theoretically occupies an area of 37.44 Å2. The 50 Å2 mean molecular area observed
at a surface pressure of 25 mN/m is thus close to the theoretical cross-sectional area of ReAzoC in
the thermodynamically favored trans conformation. The pressure-area curve suggests that
transferring a monolayer of ReAzoC to a solid substrate at the transition point between the solid
and liquid phases in the isotherm produces the most densely packed single monolayer which can
be deposited. This hypothesis is supported below by detailed XPS characterization.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
AFM images of monolayers transferred from the LB trough to a substrate at a surface
pressure of 25 mN/m are shown in Figure 3. AFM images in Figure 3(a) exhibit low rms surface
roughness of 0.218 nm, far less than the total thickness of the molecule. The holes evident in Figure
3(a) and (b) allow the thickness of the film to be measured. The height profile inset in Figure 3(b)
shows a height difference of 2.3 nm between the top and bottom of the monolayer, which
corresponds to the height of the ReAzoC molecule, This height is predicted to be 2.4 nm based on
the bonding configurations given in ref. 46.
XPS Characterization of Molecular Coverage. The surface concentration of ReAzoC
molecules within layers deposited at different surface pressures (5, 15, 25, and 40 mN/m) and via
drop casting was measured by XPS. The drop cast film which is expected to be disordered was
used as a comparison. The XPS data for N(1s), Cl(2p), and Re(4f) transitions for a monolayer
transferred at 25 mN/m and a drop cast film are shown in Figure 4(a) and (b), respectively. The
N(1s) peak at 399-404 eV, Cl(2p) peak at 198-204 eV, and Re(4f) peaks at 41-49 eV together
confirm that the ReAzoC molecules have been transferred to the substrate. The surface coverages
calculated from the Cl(2p), Re(4f), and N(1s) peaks in the XPS spectrum are shown in Table 1.
15
ACS Paragon Plus Environment

Langmuir
Page 16 of 45
1
2
3
4
5
6
7
8
9
The dependence of the molecular coverage on surface pressure is shown in Figure 5. The
surface coverage of the ReAzoC molecules in LB films transferred at the optimum surface pressure
of 25 mN/m was 1.84 molecules/nm2. This is consistent with the estimate of 2 molecules/nm2 from
the LB isotherm (Figure 2). The theoretical maximum coverage of a single monolayer of ReAzoC
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
is 2.7 molecules/nm2 calculated from the molecular dimensions estimated based on Dominey et
,
al.46 When taken together the XPS spectra, AFM profile, and LB isotherm confirm the formation
of an optimized ReAzoC monolayer at a surface pressure of 25 mN/m.
UV-Vis Spectroscopy. The trans-cis photoisomerization of ReAzoC thin films deposited
on a transparent quartz substrate was monitored using UV-Vis transmission spectroscopy. The
absorption spectrum before irradiation (black curve, Figure 6(a)) exhibits a strong peak at 360 nm
arising from the π-π* electronic transition of trans-azobenzene. Absorption spectra for ReAzoC at
a concentration of 2.74  10-6 M in a chloroform-THF (9:1) solution are shown in Figure 6(b).
Absorption spectra were acquired under ambient room light and the initial configuration thus
consisted of a mixture of trans and cis states. The two absorption peaks at 244 nm and 285 nm
arise from the π-π* transition of the rhenium bipyridine group.47,48
Illumination of the transferred LB film at a wavelength of 365 nm LB film of ReAzoC
leads to a decrease in the absorbance of the 360 nm absorption band, reaching a photostationary
state after a 10 s exposure at 16.9 mW/cm2 (Figure 6(a)). The absorbance of the 360 nm peak
recovered to its initial value after a 10 s exposure to 465 nm wavelength illumination at 9.5 mW/
cm2. The photostationary state observed in the spectra of the thin film is consistent with
observations in solution, as in Figure 6(b). The change in the optical spectrum provides direct
evidence for the trans-cis photoisomerization of both monolayer and solution forms of ReAzoC.
16
ACS Paragon Plus Environment

Page 17 of 45
Langmuir
1
2
3
4
5
6
7
8
9
The absorption cross section was estimated using the UV-Vis spectra shown in Figure 6.
In solution, the absorption cross section at 352 nm was 6.8  10-16 cm2 for the trans isomer and
4.0  10-16 cm2 for the cis isomer. The absorption cross section of the molecular monolayer was
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
1.8  10-16 cm2 for the trans isomer and 1.6  10-16 cm2 for the cis isomer. The difference between
the absorption cross section of ReAzoC in solution and the ReAzoC monolayer can be in part
attributed to the anisotropic distribution of the molecular orientation in the monolayer. The
transition dipole moment of the azobenzene π-π* transition is oriented approximately parallel to
the long axis of the azobenzene group and is thus oriented nearly along the monolayer surface
normal. The electric-field vector of the incident radiation is in a plane perpendicular to the surface
normal under the normal-incidence conditions used for the UV-Vis measurements. The absorbance
under these conditions is proportional to 3/2 Sin2(θ), where θ is the angle between the transition
dipole moment of the azobenzene and the surface normal.28 With the molecules oriented
approximately normal to the surface as is the case in the ReAzoC monolayer, this factor serves to
reduce the absorption in comparison to the solution.
The absorption cross section of the π-π* transition in an azobenzene-containing monolayer
can be estimated from values given in ref. 30. With a correction based on the approximately 20%
quantum yield of the reorientation transtition,49 the total absorption cross section from ref. 30 is on
the order of 3.5  10-17 cm2. The peak position of π-π* transition in the LB monolayer was red-
shifted by 8 nm relative to that in solution. The origin of this red-shift is unclear at this point.
However multiple factors such as the tilt of the molecular orientation with respect to the surface
normal, contributions to the maximum at 360 nm from the S2 π-π* transition in Re-bipyridine
group,18 and possible solvatochromic effects influencing the S2 transition may play a role.
17
ACS Paragon Plus Environment

Langmuir
Page 18 of 45
1
2
3
4
5
6
7
8
9
The optical absorption spectra were used to estimate the fraction of molecules that undergo
the trans-to-cis isomerization. The 360 nm absorption band has contributions from the absorbance
of the azobenzene group in both the cis and trans isomer as well as a small contribution from the
Re-bipyridine head group.18 However, if this small contribution is neglected, an estimation of the
lower limit for the fraction of molecules isomerized can be estimated by an extension of Beer’s
Law:
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
ꢜ ꢂ ꢷꢸꢤꢦꢍꢹꢀꢸꢤꢦꢍꢹ ꢺ ꢷꢻꢌꢹꢀꢻꢌꢹ ꢎ3ꢐ
Here εn is the absorption constant for the isomer and Nn is the number of molecules of each isomer
in the optical beam path, and NTrans + NCis = NTotal. The absorption constants of the trans and cis
isomer are related by a scalar n, the wavelength-dependent ratio of the trans-cis isomer absorption
constants. Thus the absorbance of the peak can be rewritten as a function of a single isomer
independent absorption constant.
ꢜ ꢂ ꢼꢷꢀꢸꢤꢦꢍꢹ ꢺ ꢷꢀꢻꢌꢹꢎ4ꢐ
The fractional change in the absorbance can be related to the fraction of molecules
ꢾ
isomerized by computing the fractional change ꢃꢽꢃ , where A is the absorbance of the initial state
ꢃ
and A’ is the absorbance of the final state. With equation (4) we find the following expression for
the fractional change in absorbance:
ꢜ ꢴ ꢜꢿ ꢼꢷ∆ꢀꢸꢤꢦꢍꢹ ꢺ ꢷ∆ꢀꢻꢌꢹ
ꢂ
ꢎ5ꢐ
ꢜ
ꢼꢷꢀꢸꢤꢦꢍꢹ ꢺ ꢷꢀꢻꢌꢹ
Here ΔNTrans and ΔNCis are the differences between the number of molecules of the trans and cis
state present in the initial and final states, respectively. Necessarily, ΔNTrans = - ΔNCis = ΔN, thus:
18
ACS Paragon Plus Environment

Page 19 of 45
Langmuir
1
2
3
4
5
6
ꢜ ꢴ ꢜꢿ
ꢜ
ꢎꢼ ꢴ 1ꢐ∆ꢀ
ꢂ
ꢎ6ꢐ
ꢼꢀꢸꢤꢦꢍꢹ ꢺ ꢀꢻꢌꢹ
7
8
9
We assume that the initial state, before UV exposure can be reasonably approximated as
consisting only of trans isomers, which is applicable because the trans state of the azobenzene
group is the thermodynamically stable state. In this approximation, NTrans = N and NCis = 0 and
equation (6) becomes:
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
∆ꢀ
ꢀ
ꢼ
ꣀꢜ
ꢂ
ꢎ7ꢐ
ꢼ ꢴ 1 ꢜ
Literature values of n are between 5 and 10.4,50 The absorption spectra for ReAzoC molecules in
solution (Figure 6(b)) have ꣁꢃ ꢂ 40%, and thus between 44 and 50% of the molecules in solution
ꢃ
are photoisomerized. When deposited as a monolayer (Figure 6 (a)), ꣁꢃ ꢂ 12%, and thus only 13
ꢃ
to 15% of the molecules on the surface are photoisomerized by a UV intensity of 16.9 mW/cm2.
This limited photoisomerization is commonly observed in para-substituted azobenzenes as a result
of very rapid thermal relaxation which makes the creation of large fractions of cis isomers
difficult.51
The reversibility of the trans-cis photoisomerization of ReAzoC was characterized by
alternating 10 s periods of illumination at wavelengths of 365 nm and 465 nm wavelength at optical
intensities of 16.9 mW/cm2 and 9.5 mW/cm2, respectively. A decrease and recovery of the
absorbance of the monolayer at 352 nm in solution and at 360 nm is evident in the summaries of
the absorption peak heights after repeated exposures shown in Figure 6(c) and (d). At low optical
intensities the molecules in solution deposited in a monolayer undergo reversible
photoisomerization.
19
ACS Paragon Plus Environment

Langmuir
Page 20 of 45
1
2
3
4
X-Ray Reflectivity Structural Characterization.
5
6
7
8
9
The fractional intensity of an incident x-ray beam reflected by a thin film on a substrate
depends on the electron-density profile of the interface along the surface normal direction.52–57
A
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
model consisting of the substrate, the monolayer, and the ambient environment was used to
interpret the XRR measurement. The model parameters are ρSub, the electron density of the
substrate, which is held constant, σSub the roughness of the substrate-monolayer interface, and
LMonolayer, ρMonolayer, σMonolayer, which are the thickness, electron density and roughness of the
monolayer-air interface, respectively. An initial point for the refinement of the reflectivity model
was obtained by predicting the height using the distance along the long axis of the ReAzoC
molecule from literature values of bond angles and bond lengths.46 The initial electron density
estimate was obtained by counting the number of electrons within the theoretical volume occupied
by the molecule. Based on the model for the ReAzoC molecule, an ideally packed monolayer of
trans conformation molecules is expected to have an electron density of 0.396 e-/Å3, thickness 24
Å, and a substrate roughness near that of bare silicon. A corresponding monolayer of cis
conformation molecules is expected to have an electron density of 0.521 e-/Å3 and a thickness of
8.45 Å. Additional geometric and experimental parameters such as sample width, beam width, and
direct beam intensity are measured to allow the fit to match the measured intensity with no overall
scaling parameter. Further discussion of the geometric model is given in ref. 31
An XRR measurement of a ReAzoC monolayer deposited at a surface pressure of 25 mN/m
is shown in Figure 7(a). The reflectivity curve for the 25 mN/m samples exhibits total external
reflection at incident angles lower than 2θ = 0.4° and a single broad reflectivity fringe, including
an intensity minimum at 2θ = 1.4° and a maximum at 2θ = 1.8°. Only one fringe is observed as a
result of the small thickness of the monolayer and the limited dynamic range of the laboratory-
20
ACS Paragon Plus Environment

Page 21 of 45
Langmuir
1
2
3
4
5
6
7
8
9
based x-ray instrumentation. The reflectivity from a monolayer deposited at 15 mN/m, Figure 7(b),
shows a weak reflectivity fringe as seen by the changes in the curvature of reflected intensity near
2θ = 1.5° and 2θ = 2.25°. The reflectivity signal from the drop-cast film, Figure 7(c), and bare
silicon, Figure 7(d), do not exhibit interference fringes. The extracted electron density profiles for
each sample are shown in the inset of the reflectivity curves. Fitting parameters for the XRR curves
are summarized Table 2. For a monolayer deposited at a surface pressure of 25 mN/m, the
substrate-monolayer roughness and electron density of the single layer agree with the theoretical
model. This indicates that the monolayer is closely packed, with the molecules orientated and
ordered with their long axis more or less normal to the surface.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
A systematic XRR study of several samples shows small variations in the apparent
monolayer thickness of films deposited under nominally identical deposition conditions.
Thicknesses are consistent for samples in a single batch, but are different in different deposition
events and substrates. The mean molecular thickness obtained by averaging the thicknesses
measured for each sample is 29.6 Å with a standard deviation of 3.2 Å. A further difference
between the predicted and measured heights can arise from the roughness of the substrate, which
is nearly 1/3 the height of the molecule. The physical interpretation of the parameters of the XRR
model is thus complicated by the relative magnitudes of the surface roughness and of the
molecules.
An alternative to the single layer model is to conceptually split the molecule into two layers
of different density with a boundary between the two layers placed at the ether group connecting
the head and the bridge groups. The result is a two-layer model with an electron rich head group
with a density of 0.718 e-/Å3, 8.3 Å in thickness and an electron poor bridge and attachment group
with a density of 0.226 e-/Å3 15.7 Å in thickness. The substrate-air interface roughness of the
,
21
ACS Paragon Plus Environment

Langmuir
Page 22 of 45
1
2
3
4
5
6
7
8
9
silicon substrate, is fit to be 7.6 Å rms. The interface can be modeled as a monolayer of molecules,
assuming this two box model, conformally coating a rough silicon substrate. In this case, the out-
of-plane electron density profile created by integrating the density along the in-plane direction
appears to form a single layer with average density and increased thickness as a result of the
comparable magnitudes of the molecular length and surface roughness and the comparable
densities of the molecular head group and the Si substrate.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
The XRR results exhibit a clear difference in the electron density profiles of samples
deposited with different surface densities. The model fit of the monolayer deposited using a surface
pressure of 15 mN/m deposited shows a reduction in magnitude of the thickness and electron
density parameters. This reduction in the magnitude of the thickness and electron density is
indicative of a reduced monolayer coverage with the fit parameters being a weighted average
between the densities of the molecular monolayer and air. This weighted average interpretation is
used to account for samples with sub monolayer coverage to account for the lack of lateral
homogeneity. The reflectivity pattern is indicative of a surface with the molecules organized into
closely packed islands due to the spatial constraint imposed by the lateral coherence length of x-
rays. The weighted average model estimates the surface coverage for the 15 mN/m film to be 75%
the value of the complete single monolayer, as compared to 68% as observed via XPS. This
discrepancy arises from uncertainty in accounting for the interfacial roughness for the bare silicon
and complete monolayer films used in the weighted average model. Further reduction of those
parameters in the case of the drop cast films is consistent with a randomly distributed array of
molecules or the formation of three-dimensional aggregates.
X-Ray Reflectivity Characterization of Isomerization.
22
ACS Paragon Plus Environment

Page 23 of 45
Langmuir
1
2
3
4
5
6
7
8
9
XRR provides insight into the structural changes, including changes in the thickness of the
molecular layer, resulting from optical illumination. XRR studies of the reconfiguration were
conducted using monolayers deposited at a surface pressure of 25 mN/m. The experimental
procedure used for the structural studies of the reconfiguration process consisted of a series of
alternating exposures to UV and blue light at wavelengths of 365 nm and 460 nm, respectively.
The UV and blue optical intensities for this structural study were 94 mW/cm2 and 100 mW/cm2,
respectively, far higher than the intensities used to obtain the optical spectra in Figure 6.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
XRR measurements of the as-deposited structure are shown in Figure 8(a). Following this
initial configuration, the monolayer was exposed to blue light for 30 s and the XRR curve was
remeasured. No change in the structure was observed (Figure 8(a), solid blue line), as expected
because the initial state is composed of molecules in the trans isomeric state. The same sample
was then exposed to first 30 s and then 360 s of UV light. After the UV exposure there is an
increase in intensity near the local minimum (Figure 8(a), solid violet line). This increase in
intensity near the local minima is consistent with a change in the structure due to the
photoisomerization of the monolayer from the trans state to the cis state.
The XRR data can be used to estimate the fraction of the monolayer isomerized as a result
of UV exposure. Using the simple structural model described above, the reflected intensity curve
can be modeled as the weighted average of the reflectivity from the starting structure, which is
assumed to be dominated by the trans isomer, and the reflectivity from a theoretical 100% cis
monolayer. Adding the intensities in this way is appropriate for sample configurations in which
the isomeric domains have lateral extents on the order of the x-ray transverse coherence length,
which is 350 nm for the experimental geometry used here. An analysis of the XRR data shown in
Figure 7 using this two-domain model gives an isomerization of 112% of the area of the
23
ACS Paragon Plus Environment

Langmuir
Page 24 of 45
1
2
3
4
5
6
7
8
9
monolayer. An analysis of UV-Vis studies under these optical conditions using equation (7) gives
an isomerized fraction of 47 to 52% depending on the value of the constant n. The difference
between these fractions may arise because the x-ray measurements are insensitive to isomerized
areas that are smaller than the lateral x-ray coherence length on the order of 100 nm.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
When the XRR experiment was performed with the optical conditions as were used for the
UV-Vis spectroscopy studies, no statistically significant change in the reflected intensity was
observed. The combination of x-ray and spectral data thus suggests that for experiments conducted
with low-intensity light, the isomerized molecules are randomly distributed, rather than grouped
into domains, because a change in the XRR curve is not observed under low intensity illumination.
The return to the trans state was probed by studying the structure under blue illumination
following UV illumination. The initial UV exposure was followed by exposures with durations of
360 s (Figure 8(d), solid blue line), 720 s (Figure 8, dashed red line) and 1440 s (Figure 8, dotted
black line) to light with 460 nm wavelength at an optical intensity of 100 mW/cm2. An XRR
measurement was performed after each of these exposures, but no evidence of structural change
in the monolayer was observed in the XRR signal.
To further investigate the monolayer structural changes, the molecular structure was
probed by a series of alternating XRR and UV-Vis spectroscopy measurements performed with
same power density and exposure times as were used for the measurements shown in Fig. 8. A 25
mN/m LB sample was prepared on a quartz substrate, its initial structure was probed using XRR,
and a subsequent series of measurements were conducted using alternating optical illumination
and XRR. The 365 nm illumination was carried out at a power density of 94 mW/cm2 for 360 s.
The 460 nm illumination was performed at a power density of 100 mW/cm2 for 720 s. Optical
24
ACS Paragon Plus Environment

Page 25 of 45
Langmuir
1
2
3
4
spectroscopy measurements were performed before and after each XRR measurement. The
5
6
resulting XRR measurements are shown in Error! Reference source not found. (a).
7
8
9
The reflectivity curve from the starting structure in Error! Reference source not found.
(a) (black curve) shows a local minimum at 2θ=2°. Upon exposure to UV light (violet) the intensity
at the local minimum increases and the reflectivity fringe becomes less pronounced. Upon
subsequent exposure to blue light, the reflectivity curve remains unchanged. This progression in
the reflectivity curves is consistent with the results observed in Figure 8.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
The UV-Vis spectra resulting from the high power and duration optical exposure are shown
in Error! Reference source not found. (b). The starting UV-Vis spectrum (black, solid) remains
unchanged after the x-ray measurement (black, dashed). After UV exposure (violet, solid) the
absorption peak at 360 nm is reduced from 0.031 to 0.018, a decrease of 42%. After the XRR
measurement, the magnitude of the absorbance increases to 0.021, a recovery of 23% of the
original transformation. After blue light exposure, the absorbance magnitude remains unchanged
in comparison to the absorbance before blue exposure. This progression of UV-Vis spectra and
with the x-ray results together indicate that a fraction of the area of monolayer undergoes trans-
to-cis photoisomerization, followed by a small fraction of the isomerized portion of the monolayer
recovering to the trans state, and no additional recovery occurring as a result of blue optical
exposure. At this higher power and duration exposure, XRR and UV-Vis spectra exhibit an
isomerization from trans to cis states, but also indicate a suppression of the reverse cis to trans
isomerization. The limited cis to trans photoisomerization is consistent with steric effects from
either the bulkiness of the molecule or monolayer close-packing. Damage to the monolayer from
UV/Ozone is unlikely, as the wavelength of UV light used in the experiment is reported as
generating extremely low concentrations of ozone.58
25
ACS Paragon Plus Environment

Langmuir
Page 26 of 45
1
2
3
4
Conclusions
5
6
7
8
9
Molecular monolayers of a new molecule with electron donating characteristics and
externally reconfigurable structure have been fabricated. The physical and electronic structure of
the molecules and monolayers have been characterized and show near full coverage when
deposited via the LB technique. The molecules and monolayers showed evidence of structural
reconfiguration in response to optical exposure both in solution and when confined to a monolayer.
A structural model for the photoisomerization and reconfiguration in the monolayer has been
developed, in which a fraction of the monolayer is isomerized, depending on the intensity and
duration of optical exposure. We also present evidence of limited reversibility of the
photoisomerization in monolayers, hypothesized to be steric hindrance or other molecule-surface
interactions.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
26
ACS Paragon Plus Environment

Page 27 of 45
Langmuir
1
2
3
4
Acknowledgements
5
6
7
8
9
P.G., P.H., and C.K acknowledge support by the Division of Materials Sciences and Engineering,
Office of Basic Energy Science, U.S. Department of Energy under Award No. ER46590 for the
synthesis, characterization, and self-assembly studies. P.E, K.M., Y.H, and A.L. acknowledge
support from the National Science Foundation through the University of Wisconsin Materials
Research Science and Engineering Center (MRSEC) (NSF grant no. DMR-1001288) for the x-ray
studies. K.M. also acknowledges support from 3M Corporation through 3M Science and
Technology Fellowship Program and from the Japanese Society for the Promotion of Science as a
partner of the National Science Foundation EAPSI program. This material is based upon work
supported by the National Science Foundation under Grant No. 1515273.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
27
ACS Paragon Plus Environment

Langmuir
Page 28 of 45
1
2
3
4
References
5
6
7
8
(1) Delorme, N.; Bardeau, J.-F.; Bulou, A.; Poncin-Epaillard, F. Azobenzene-Containing
Monolayer with Photoswitchable Wettability. Langmuir 2005, 21, 12278–12282.
9
(2) Zheng, Y. B.; Pathem, B. K.; Hohman, J. N.; Thomas, J. C.; Kim, M.; Weiss, P. S.
Photoresponsive Molecules in Well-Defined Nanoscale Environments. Adv. Mater. 2013,
25, 302–312.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
(3) Han, M.; Honda, T.; Ishikawa, D.; Ito, E.; Hara, M.; Norikane, Y. Realization of Highly
Photoresponsive Azobenzene-Functionalized Monolayers. J. Mater. Chem. 2011, 21,
4696–4702.
(4) Merino, E.; Ribagorda, M. Control over Molecular Motion Using the Cis-Trans
Photoisomerization of the Azo Group. Beilstein J. Org. Chem. 2012, 8, 1071–1090.
(5) Kim, M.; Safron, N. S.; Huang, C.; Arnold, M. S.; Gopalan, P. Light-Driven Reversible
Modulation of Doping in Graphene. Nano Lett. 2012, 12, 182–187.
(6) Weber, R.; Winter, B.; Hertel, I. V.; Stiller, B.; Schrader, S.; Brehmer, L.; Koch, N.
Photoemission from Azobenzene Alkanethiol Self-Assembled Monolayers. J. Phys. Chem.
B 2003, 107, 7768–7775.
(7) Hsu, C.-C.; Wang, Y.-T.; Yabushita, A.; Luo, C.-W.; Hsiao, Y.-N.; Lin, S.-H.; Kobayashi,
T. Environment-Dependent Ultrafast Photoisomerization Dynamics in Azo Dye. J. Phys.
Chem. A 2011, 115, 11508–11514.
(8) Kolpak, A. M.; Grossman, J. C. Azobenzene-Functionalized Carbon Nanotubes As High-
Energy Density Solar Thermal Fuels. Nano Lett. 2011, 11, 3156–3162.
(9) Zhitomirsky, D.; Cho, E.; Grossman, J. C. Solid-State Solar Thermal Fuels for Heat
Release Applications. Adv. Energy Mater. 2016, 6, 1502006.
(10) Feng, W.; Luo, W.; Feng, Y. Photo-Responsive Carbon Nanomaterials Functionalized by
Azobenzene Moieties: Structures, Properties and Application. Nanoscale 2012, 4, 6118–
6134.
(11) Luo, W.; Feng, Y.; Cao, C.; Li, M.; Liu, E.; Li, S.; Qin, C.; Hu, W.; Feng, W. A High
Energy Density Azobenzene/Graphene Hybrid: A Nano-Templated Platform for Solar
Thermal Storage. J. Mater. Chem. A 2015, 3, 11787–11795.
(12) Qin, C.; Feng, Y.; Luo, W.; Cao, C.; Hu, W.; Feng, W. A Supramolecular Assembly of
Cross-Linked Azobenzene/Polymers for a High-Performance Light-Driven Actuator. J.
Mater. Chem. A 2015, 3, 16453–16460.
(13) Seki, T. New Strategies and Implications for the Photoalignment of Liquid Crystalline
Polymers. Polym. J. 2014, 46, 751–768
28
ACS Paragon Plus Environment

Page 29 of 45
Langmuir
1
2
3
4
5
6
(14) Zong, Y.; Tawa, K.; Menges, B.; Rühe, J.; Knoll, W. Azobenzene-Containing Polyamic
Acid with Excellent Langmuir−Blodgett−Kuhn Film Formation Behavior Suitable for All-
Optical Switching. Langmuir 2005, 21, 7036–7043.
7
8
9
(15) Cristofolini, L.; Berzina, T.; Fontana, M. P.; Konovalov, O. Structure and Stability of
Molecular Layers of a Photosensitive Azo-Polyacrylate by X-Rays Reflectivity and GID.
Mol. Cryst. Liq. Cryst. Sci. Technol. Sect. Mol. Cryst. Liq. Cryst. 2002, 375, 689–699.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
(16) Asbury, J. B.; Hao, E.; Wang, Y.; Lian, T. Bridge Length-Dependent Ultrafast Electron
Transfer from Re Polypyridyl Complexes to Nanocrystalline TiO2 Thin Films Studied by
Femtosecond Infrared Spectroscopy. J. Phys. Chem. B 2000, 104, 11957–11964.
(17) Xiong, W.; Laaser, J. E.; Paoprasert, P.; Franking, R. A.; Hamers, R. J.; Gopalan, P.;
Zanni, M. T. Transient 2D IR Spectroscopy of Charge Injection in Dye-Sensitized
Nanocrystalline Thin Films. J. Am. Chem. Soc. 2009, 131, 18040–18041.
(18) Paoprasert, P.; Laaser, J. E.; Xiong, W.; Franking, R. A.; Hamers, R. J.; Zanni, M. T.;
Schmidt, J. R.; Gopalan, P. Bridge-Dependent Interfacial Electron Transfer from
Rhenium−Bipyridine Complexes to TiO2 Nanocrystalline Thin Films. J. Phys. Chem. C
2010, 114, 9898–9907.
(19) Bauer, C.; Boschloo, G.; Mukhtar, E.; Hagfeldt, A. Electron Injection and Recombination
in Ru(dcbpy)2(NCS)2 Sensitized Nanostructured ZnO. J. Phys. Chem. B 2001, 105, 5585–
5588.
(20) Basu, K.; Benetti, D.; Zhao, H.; Jin, L.; Vetrone, F.; Vomiero, A.; Rosei, F. Enhanced
Photovoltaic Properties in Dye Sensitized Solar Cells by Surface Treatment of SnO2
Photoanodes. Sci. Rep. 2016, 6, 23312.
(21) Christianson, J. R.; Schmidt, J. R. Structural Heterogeneity and Dynamics of Dyes on
TiO2: Implications for Charge Transfer across Organic–inorganic Interfaces. Phys. Chem.
Chem. Phys. 2015, 17, 3731–3740.
(22) Ruther, R. E.; Cui, Q.; Hamers, R. J. Conformational Disorder Enhances Electron Transfer
Through Alkyl Monolayers: Ferrocene on Conductive Diamond. J. Am. Chem. Soc. 2013,
135, 5751–5761.
(23) Walters, K. A.; Premvardhan, L. L.; Liu, Y.; Peteanu, L. A.; Schanze, K. S. Metal-to-
Ligand Charge Transfer Absorption in a rhenium(I) Complex That Contains a π-
Conjugated Bipyridine Acceptor Ligand. Chem. Phys. Lett. 2001, 339, 255–262.
(24) Busby, M.; Matousek, P.; Towrie, M.; Clark, I. P.; Motevalli, M.; Hartl, F.; Vlček, A.
Rhenium-to-Benzoylpyridine and Rhenium-to-Bipyridine MLCT Excited States of Fac-
[Re(Cl)(4-benzoylpyridine)2(CO)3] and Fac-[Re(4-benzoylpyridine)(CO)3(bpy)]+:ꢀ A
Time-Resolved Spectroscopic and Spectroelectrochemical Study. Inorg. Chem. 2004, 43,
4523–4530.
(25) Ferri, V.; Elbing, M.; Pace, G.; Dickey, M. D.; Zharnikov, M.; Samorì, P.; Mayor, M.;
Rampi, M. A. Light-Powered Electrical Switch Based on Cargo-Lifting Azobenzene
Monolayers. Angew. Chem. Int. Ed. 2008, 47, 3407–3409.
29
ACS Paragon Plus Environment