Tuning the Thermal Isomerization of Phenylazoindole Photoswitches from Days to Nanoseconds

Article abstract of DOI:10.1021/jacs.7b12871

The growing interest in light-driven molecular switches and optical oscillators led to the development of molecules that are able to interconvert from a stable to a metastable configuration upon photochemical triggering and to return to the thermodynamically stable form as soon as the light stimulus is removed. Controlling a wide range of back-isomerization lifetimes in the dark is a crucial goal for potential application of these compounds such as molecular machines. We herein present a novel class of easily synthesizable azo photoswitches based on the arylazoindole core. Most notably, minimal modifications of the core, such as methylation, dramatically change the Z-to-E thermal isomerization rate from days (Me in position 1) to the nanosecond range (Me in position 2). Moreover, fine-tuning of the Z-to-E lifetimes can be achieved choosing a proper dimethyl sulfoxide-water (or buffered water) solvent mixture. The photochemical and thermal mechanisms have been elucidated by a thorough computational and spectroscopic analysis. This allowed to detect three different pathways of thermal isomerization and to identify the hydrazone tautomer of the phenylazoindole as the major actor in the fast Z-E thermal isomerization of the NH-substituted switch in protic media.

Full text of DOI:10.1021/jacs.7b12871

Subscriber access provided by UNIVERSITY OF TOLEDO LIBRARIES
Article
Tuning the Thermal Isomerization of Phenylazoindole
Photoswitches from Days to Nanoseconds
Nadja A. Simeth, Stefano Crespi, Maurizio Fagnoni, and Burkhard König
J. Am. Chem. Soc., Just Accepted Manuscript • DOI: 10.1021/jacs.7b12871 • Publication Date (Web): 06 Feb 2018
Downloaded from http://pubs.acs.org on February 6, 2018
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a service to the research community to expedite the dissemination
of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in
full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully
peer reviewed, but should not be considered the official version of record. They are citable by the
Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore,
the “Just Accepted” Web site may not include all articles that will be published in the journal. After
a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web
site and published as an ASAP article. Note that technical editing may introduce minor changes
to the manuscript text and/or graphics which could affect content, and all legal disclaimers and
ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or
consequences arising from the use of information contained in these “Just Accepted” manuscripts.
Journal of the American Chemical Society 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 9
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
Tuning the Thermal Isomerization of Phenylazoindole Photoswitch-
es from Days to Nanoseconds
Nadja A. Simeth,†,‡ Stefano Crespi,§,‡ Maurizio Fagnoni§ and Burkhard König†,*
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
† Institut für Organische Chemie, Universität Regensburg, Universitätsstraße 31, 93040 Regensburg, Germany.
§ PhotoGreen Lab, Università di Pavia, Via Taramelli 12, 27100 Pavia, Italy.
ABSTRACT: The growing interest in lightꢀdriven molecular switches and optical oscillators led to the development of molecules
that are able to interconvert from a stable to a metastable configuration upon photochemical triggering and to return to the thermoꢀ
dynamically stable form as soon as the light stimulus is removed. Controlling a wide range of backꢀisomerization lifetimes in the
dark is a crucial goal for potential application of these compounds such as molecular machines. We herein present a novel class of
easily synthesizable azo photoswitches based on the arylazoindole core. Most notably, minimal modifications of the core such as
the methylation, dramatically change the ZꢀtoꢀE thermal isomerization rate from days (Me in position 1) to the nanosecond range
(Me in position 2). Moreover, fine tuning of the ZꢀtoꢀE lifetimes can be achieved choosing a proper dimethylsulfoxideꢀwater (or
buffered water) solvent mixture. The photochemical and thermal mechanisms have been elucidated by a thorough computational
and spectroscopic analysis. This allowed to detect three different pathways of thermal isomerization and to identify the hydrazone
tautomer of the phenylazoindole as the major actor in the fast ZꢀE thermal isomerization of the NHꢀsubstituted switch in protic meꢀ
dia.
been used to stabilize the thermal lifetime of azobenzenes,
whereas 4,4ʹ pushꢀpull or azophenolic systems (or a combinaꢀ
INTRODUCTION
Photoswitches have evolved into handy tools to control moꢀ
lecular properties, such as magnetization, fluorescence or bioꢀ
logical activity, or can be exploited as the centerpiece of moꢀ
lecular machines.1 Thereby, light is applied as an external
stimulant. It is easy to handle, unmatchable fast and precise in
practical usage, and orthogonal to biological systems and most
chemical transformations.1–3 Within the many classes of phoꢀ
toswitches investigated, azo dyes have emerged as a popular
example in the last decades including manifold applications,
e.g. in polymers,4,5 electronics,6 material science,7 protein
probes and in-vivo photoꢀcontrol of biomolecules,8–11 or as
lightꢀcontrollable catalysts.12 Among them, azobenzene (and
its derivatives) is undoubtedly the most prominent scaffold
investigated in both fundamental research and modern applicaꢀ
tions.13 The planar azobenzene E-isomer can be efficiently
converted into its metaꢀstable Zꢀisomer via UV irradiation,
with a consequent change in the electronic as well as geometꢀ
rical properties of the molecule.14,15 The Zꢀisomer is the main
component of the thus formed photostationary state (PSS) and
can interconvert to the more stable Eꢀisomer through visible
light irradiation or through thermal relaxation.15 Indeed, the
ideal thermal lifetime of the metaꢀstable Zꢀisomer strongly deꢀ
pends on the application of interest. Highly thermal stable
azoswitches are desired for applications as such as molecular
storage or logic devices, respectively.16 In contrast, fast relaxꢀ
ing azobenzenes are preferably employed, for instance, in realꢀ
time optical informationꢀtransmitting materials17 or in in-vivo
applications due to their temporally limited action.18 Thus, a
lot of effort has been put to modify the thermal lifetime of azꢀ
obenzenes. Most of these transformations have been achieved
through the introduction of different substituents onto the azoꢀ
benzene core. Orthoꢀfluoro19 or orthoꢀthiol substituents20 have
tion of both) can lead to extremely fastꢀrelaxing switches.17,21
More recently, also heteroaromatic scaffolds were considered
to broaden the range of azo dye properties, including their
thermal lifetime. Notably, azopyrimidine,21 azopyridines,22,23
azopyrroles,24 azoimidazole25–29 and azopyrazoles,30–32 or lateꢀ
ly the bicyclic azoindazole scaffold,33 were reported. They
were used, for example, to control the assembly of supramoꢀ
lecular systems or as photoꢀcontrollable ligands being, deꢀ
pending on the scaffold and its substitution pattern, highly staꢀ
ble (for a pyrazole derivative, the thermal lifetime is reported
to be up to 1000 days32) or fastꢀrelaxing systems (40 ns, for an
azopyrimidine21). However, a scaffold usually exhibits a cerꢀ
tain lifetime and the tunability is mostly limited to few orders
of magnitude, thus it belongs either to the thermally stable or
to the fast backꢀisomerizing family.
Herein, we present phenylazoindoles as photochromic scafꢀ
folds with extremely tunable thermal lifetimes. Indoles are
omnipresent in nature, substructures of drugs and various
dyes. Though previously employed as antibacterial agents,
foodꢀdyes,34 or metal complexing ligands,35 arylazoindoles
have not been considered as photochromic compounds so far.
We investigated both minimal chemical variations of the scafꢀ
fold (methylation in different positions) and solvent effects on
the thermal lifetime of the compounds. We elucidated the
mechanism of the isomerizations and take advantage of it to
tune the properties of the here presented photoswitches from
days to nanoseconds. Thereby, we focus on providing a deep
and general understanding of a new class of azo photoswitches
to pave the way to further rational developments, using this
novel core potentially for all kinds of applications.
ACS Paragon Plus Environment

Journal of the American Chemical Society
allowed than the one in Z-3a (cf. Section 5.3 in the Supporting
Page 2 of 9
RESULTS AND DISCUSSION
1
2
3
4
5
6
7
8
Information).
Synthesis. The phenylazoindoles 3aꢀc were synthesized
through azocoupling of diazonium salt 2 and the appropriate
indoles 1aꢀc (see Scheme 1 and Section 2 in the Supporting
Information).
Moving from apolar to polarꢀaprotic solvents in 3a led to τ
shifting from the millisecond to the second range (Entries 1ꢀ5,
12ꢀ13, Table 1). This trend becomes less discernible with
Z-3b, which only in dioxane has a τ > 1 s (entry 2, Table 1). In
3a,b it seems that the increased speed of the thermal ZꢀtoꢀE
relaxation goes hand in hand with the protonꢀavailability
(compare entries 7 and 13 for Z-3a). A striking example is 3a
in mixtures of MeOH or DMSO with water. While the addiꢀ
tion of water in methanol does not drastically modify the lifeꢀ
time (cf. Table 1, entry 7, 9, 10 for Z-3a), the increased protic
environment in aprotic DMSO lowers τ from seconds to miꢀ
croseconds (cf. Table 1, entry 13ꢀ15). This allowed tuning the
lifetime only by changing the waterꢀcontent in DMSO. Similar
observations can be made for Z-3b. Here, 1:1 mixtures of waꢀ
ter and MeOH or DMSO afford lifetimes of 41.4 µs or
32.1 µs, respectively. Moreover, in a mixture of acidic water
and DMSO, the relaxation time decreases to 2.2 ꢀs for Z-3a
and even reaches the nanoꢀsecond range (743 ns) for Z-3b (enꢀ
try 16). A similar result was obtained for both 3a and 3b when
DBU was used as additive (entry 17), suggesting a protonꢀ
transfer mechanism contributes to the speed of the thermal
isomerization. Reaching the subꢀmicrosecond scale is crucial
for applications requiring realꢀtime information transmission,
and up to now, this goal was achieved in azobenzene cores
only in one case by using phenolic pushꢀpull systems.21 The
dependency on protic environment becomes more evident
comparing the lifetimes of Z-3a in MeOH and MeOD, which
show a significant kinetic isotope effect (entries 7 and 8 in Taꢀ
ble 1). A KIE=5.5 indicates that a proton transfer is involved
in the rateꢀdetermining step in this solvent. Increasing the conꢀ
centration of 3a in MeOH leads to faster isomerization kinetꢀ
ics (entry 11). In anhydrous toluene no concentration effect on
τ was found, see Table S2. The presence of an intermolecular
interaction in 3a can be followed by NMR (see Section 4.3 in
the Supporting Information). In contrast to the other experiꢀ
ments, Z-3b in methanol relaxes with a biexponential kinetic.
This suggests, compared with the previous findings, that the
presence of a protic environment modifies the mechanism of
the dark reaction for Z-3a and Z-3b. Indeed, for the specific
case of 3b in MeOH, two mechanisms of thermal relaxation
are in competition (cf. entry 7 for 3b). A detailed kinetic analꢀ
ysis of the thermal isomerization reaction is present in Section
7 of the Supporting Information. In particular, both a dependꢀ
ency on concentration of the photoswitch and on water was
found for the NꢀH containing photoswitch 3b. The kinetic law
that was derived is similar to the one obtained for the arylaꢀ
zoimidazoles, compounds that are supposed to isomerize
through an intermolecular proton transfer leading to the forꢀ
mation of a hydrazone intermediate (Chart 1).17,26,32
Scheme 1. Synthesis of 3a-c.
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
Spectroscopy. The absorption and isomerization properties
of phenylazoindoles 3aꢀc were investigated in various solꢀ
vents, from nonꢀpolar (aromatic and nonꢀaromatic) to polar
aprotic and polarꢀprotic ones. Most prominent is the broad abꢀ
sorption band between 359 and 380 nm. We could assign the
aforementioned band to the π→π* transition of the azo group
(see Table S1; for the recorded UV/Vis spectra, see Section 3
in the Supporting) owing to the similarities with azobenzene
and other azoheteroarenes, and TDꢀDFT simulations (see Secꢀ
tion 5.3 in the Supporting).15,33 By trend, methylation (3b and
3c) induces a slight bathochromic shift of the band compared
to 3a (between 4 to 11 nm). The weaker n→π* band is present
between 400 to 500 nm as a tailing of the π→π* transition.
Due to the high variety of lifetimes (τ) of the Zꢀisomer enꢀ
countered, we decided to study the different kinetics of therꢀ
mal relaxation processes of Z-3a-c using three different experꢀ
imental setups. In particular, for the slower kinetics, the EꢀtoꢀZ
isomerization was induced through irradiation with LEDs of
400 nm or 365 nm and the back isomerization from the PSS in
the dark was recorded through an UV/Vis spectrometer (seꢀ
conds to hours range, see for example Figure 1A) or multiplate
UV/Vis reader (hours to days). The faster kinetics were studꢀ
ied via nanosecond laser flash photolysis with excitation proꢀ
vided by a Nd:YAG laser at 355 nm (full details and pictures
of the various experimental setups are included in the Supportꢀ
ing Information, Section 1). Multiple isomerization cycles can
be achieved by pulsed irradiation of a single wavelength (see
for example Figure 1B, concerning the pulsed irradiation with
400 nm of 3a in DMSO).
The results are summarized in Table 1 (see Table S2 for a
wider number of entries, spectra and isomerization kinetics).
Compounds Z-3a and Z-3b exhibited thermal isomerization
lifetimes comparable with or far lower than the second range.
Laser flash photolysis allowed us to describe the transient abꢀ
sorption signal belonging to the Zꢀform of 3a and 3b. In parꢀ
ticular, in 3b a broad absorption band appeared around 450 nm
immediately after the laser shot, attributable to the n→π* tranꢀ
sition of its Zꢀisomer (Figure 1C). On the other hand, only the
bleaching of the π→π* of the Eꢀisomer can be observed in 3a.
TDꢀDFT UV/Vis absorption simulations of the most stable Zꢀ
isomers of 3a,b allowed us to conclude that the geometry of
Zꢀ3b is causing the n→π* transition to be more symmetryꢀ
Chart 1. Formula of the s-trans Hydrazone Tautomer of
3a,b
N
N
H
N
R1
R1 = H, Me
ACS Paragon Plus Environment

Page 3 of 9
Journal of the American Chemical Society
Table 1. Z-to-E Isomerization Lifetimes for 3a-c.
1
Lifetime τ
3b
2
3
4
5
6
7
8
9
Entry
Solventsa
3cc
1.1 h
17.6 h
1.13 d
6.0 min
17.1 min
3a
cyclohexane
dioxane
mesitylene
benzene
42.5 msb
84.9 msb
12.3 msb
78.5 msb
47.5 msb
116.8 msb
1.9 sc
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
3.3 msb
167 msb
60.2 msb
1.4 msb
toluene
aqueous toluened
MeOH
6.8 msb
37.7 msb
14.8 msb
2.9 msb
4.7 msb
4.3 sc
2.4 h
454 ꢀs (51%), 78 ꢀs (49%) e
MeOD
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
41.9 ꢀsb
41.4 ꢀsb
MeOH water 9:1
MeOH water 1:1
MeOHf
1.9 ꢀsb
MeCN (dry)
DMSO (dry)
1.2 d
2.6 d
6.5 sc
1.3 msb
111.2 ꢀsb
266 ꢀsb
DMSO water 9:1
DMSO water 1:1
DMSO ꢀ buffered water (pH = 4) f 1:1
DMSO water 1:1 + 25 ꢀL DBU
187.9 ꢀsb 32.1 ꢀsb
743 nsb
46 ns
2.2 ꢀsb
31 ns
a The concentration was adjusted to 50 ꢀM for all the measurements. b Determined via nanosecond laser flash photolysis. c Determined
d
via UV/Vis spectroscopy. Toluene saturated with water (15 ꢀL of water were added to 2 mL of a 50 ꢀM solution of 3b in toluene. The
e
emulsion thus formed was sonicated for 1h.) The decay is biexponential. In parentheses is shown the contribution of the single lifetimes
to the whole decay. f Concentration of 3a = 75 ꢀM g pH = 4.00 ± 0.02 (20 °C) (Citric acid = 0.056 M; NaOH = 0.11 M; HCl = 0.044 M)
Figure 1. A: Thermal relaxation from the PSS (τ=4.3 s) of 3a in DMSO after irradiation with λirrad=400 nm. B: Multiple isomerization
cycles of 3a in DMSO (black bars: dark, blue bars: irradiation with λirrad=400 nm; the observed wavelength corresponds to λobs=375 nm)
indicating a good fatigue resistance. C: Transient signal registered for 3b in DMSO. The disappearing of the n→π* band belonging to Zꢀ3b
is accompanied by the restoration of π →π* band of Eꢀ3b. D: Transient signal registered for 3b in a DMSO/water 1:2 mixture (detailed
spectrometric data and kinetics in Section 3.3 in the Supporting Information). The decay of the n→π* band of Zꢀ3b is followed by the forꢀ
mation of a hypsochromically shifted transient signal. Then, the π →π* band of Eꢀ3b reappears.
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 4 of 9
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 2. A: Thermal Z-E isomerization of compound 3b in toluene. An isosbestic point can be observed, indicating the monomolecular
decay of the n→π* band of Z-3b in favour of the formation of the π →π* of E-3b. B: The addition of water to toluene induces a change in
the transient spectra. As like as in DMSO (cf. Figure 2d), a new species, blueꢀshifted compared to the n→π* band of Z-3b, can be detected
by laser flash photolysis. This intermediate contributes to the restoration of the π →π* band of E-3b. C: Transient signal of the restoration
of the π →π* band of E-3b (λobs =370 nm) in toluene and water saturated toluene. A forced protic environment inside toluene increases the
overall Zꢀto-E isomerization rate.
The water content dependency led to an interesting saturaꢀ
tion kinetic, that is another evidence of a mechanism characꢀ
terized by an intermolecular proton transfer that is mediated
by the presence of protic medium.
Moreover, we observed notable differences in the transient
spectra of 3a,b. In the absence of protic media, the transient
signal of the Zꢀisomer directly interconverts to the Eꢀform
(that exhibits an isosbestic point in 3b, where the n→π* abꢀ
sorption band of the Z form can be clearly discerned, see Figꢀ
ure 1C). However, in protic environment, the signal of Zꢀ3b
shifts hypsochromically, giving rise to a second band, which
belongs to an intermediate of the ZꢀtoꢀE thermal conversion
that eventually collapses with the restoration of the Eꢀisomer
(cf. Figure 1D). In 3a a new band arises at a similar waveꢀ
length of the comparable one in 3b, before the Eꢀ3a π→π* is
restored). To prove our theory and to generalize the effect of a
protic medium on the isomerization reaction of 3a,b, we satuꢀ
rated toluene with water (cf. Figure 2 and Table 1, entry 6 for
3b). Consequently, the isomerization rate increased sixtyfold
and a transient signal, similar to the intermediate signal obꢀ
served above, appears. Furthermore, the importance of the NH
moiety in the thermal ZꢀtoꢀE relaxation was proven studying
Figure 3. Thermal relaxation from the PSS (85:15 Z:E ratio,
τ=2.6 d) of 3c in DMSO after irradiation with λirrad = 400 nm.
ACS Paragon Plus Environment

Page 5 of 9
Journal of the American Chemical Society
hand (ꢁH‡=14 kcal/mol at the (CꢀPCM)ꢀB3LYP/6ꢀ31G(d) in
MeCN) are comparable with the experimental EyringꢀPolanyi
plot obtained for 3a in MeCN (ꢁH‡=12 kcal/mol; Figure 4D)
compound 3c. Indeed, the methylation of the indoleꢀnitrogen
1
2
3
4
5
6
7
8
(in 3c) causes significant changes of the relaxation properties
with respect to the N–H containing derivatives (3a,b). Comꢀ
pound 3c exhibits longer lifetimes (minutes to days, see Table
1) compared to 3a,b. Moreover, in polarꢀaprotic solvents
(DMSO and MeCN) as well as in mesitylene, 3c shows τ largꢀ
er than one day. Owing to their stable photostationary state,
we were able to determine the PSS population of 3c in DMSO
Most interestingly, using DFT calculations we proved the
participation of a hydrazone intermediate in a third relaxation
pathway. We reasoned that the hydrazone formation in
azoindoles is only feasible intermolecularly, due to the imposꢀ
sibility for the indolic proton to reach the azo group intramoꢀ
lecularly (see Scheme 2; Section 5.4 in the Supporting Inforꢀ
mation; for a more detailed representation of the mechanism
see Section 8 of the Supporting Information).
1
(Figure 3) and MeCN applying HꢀNMRꢀspectroscopy (85:15
and 61:39 of Z:E ratio, respectively; see Section 4.4 in the
Supporting Information).
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
However, to the best of our knowledge, spectroscopic obserꢀ
vations similar to the ones we witnessed have not been made
so far. Phenylazoindoles themselves are known to give rise to
a hydrazone tautomerism.36 Thus, we aimed to explain the inꢀ
volvement of an azoꢀhydrazone and why minimal modificaꢀ
tions of the indolic core lead to a surprisingly wide span of
thermal relaxation rates through a detailed computational
study.
Scheme 2. Simplified Intermolecular Hydrazone Mecha-
nism
Computational studies. To seek for similarities with the
azobenzene core, we endeavoured to analyze both the photoꢀ
chemical E-to-Z and thermal Z-to-E interconversion pathways.
Representing the most general scaffold, 3a was selected for
the study. Hence, we performed an MSꢀCASPT2/CASSCF
calculation of the E-toꢀZ pathway, both from the first (S1,
n→π* in character; cf. Figure 4A) and the second excited state
(S2, π→π*). The analysis of the molecular structure, its elecꢀ
tronic distribution and the g and h vectors defining the branchꢀ
ing space of the conical intersections (CoIns) (cf. Figure 4A
and B), helped us to assign the photochemical decay pathways
(a complete analysis of the CoIns is present in Section 5.2.2 in
the Supporting Information).37 Similarly to azobenzene, popuꢀ
lation of the n→π* singlet excited state leads to a rotational
conical intersection (S0←S1, CoInA), which generates the
Z-isomer.14 This process is barrierless. Excitation to the S2
level is followed by a deactivation to S1 via a conical intersecꢀ
tion topologically similar to the S0←S1 one (CoInC). The molꢀ
ecule has now enough potential energy either to go through
CoInA relaxing to Zꢀ3a or to funnel through another conical
intersection to the sole Eꢀ3a isomer (CoInB). CoInB is only acꢀ
cessible in this way, because it is located energetically uphill
compared to the FranckꢀCondon geometry of the S1.
However, without the assistance of water (representing the
simplest proton donor) the interaction of two molecules of 3a
is characterized by high energy barriers, which make this
pathway not competitive with the previously analyzed monoꢀ
molecular thermal isomerizations. On the other hand, two exꢀ
plicit molecules of water bridging two Z-conformers of 3a
lower the energetical barriers for the hydrazone formation,
thus explaining the increase of rate observed with protic solꢀ
vents in NꢀH containing azoindoles (Section 7 in the Supportꢀ
ing Information). We propose the following mechanism. The
first step involves a waterꢀmediated proton transfer from the
indole nitrogen of one molecule of Z-3a onto the azoꢀmoiety
of the second one (βꢀnitrogen atom with respect to the indole,
see Figure 4C). The barrier for the formation of this complex
is only 17.3 kcal/mol due to the stabilizing action of the water
molecules. The ensuing generation of the two s-cis hydrazones
is defined by a low activation energy (ꢁE‡=3.5 kcal/mol) and
results to be slightly exothermal (ca. 1 kcal/mol). After the roꢀ
tation around the newly formed NꢀNꢀsingle bonds (characterꢀ
ized by (ꢁE‡= 8.6 and 10.2 kcal/mol), the sꢀtrans hydrazone
dimer is formed, which is located inside a potential well (ca.
30 kcal/mol more stable than the initial waterꢀbridged Z-3a
dimer; see figure 4C). Due to this fact, the transient signal of
the intermediate present in protic media can be assigned to the
sꢀtrans hydrazone (compare Figure 4E, showing the simulated
TDꢀDFT UV/Vis spectra of the azohydrazone of 3b with the
experimentally observed band, figure 2D).
In accordance with our experiments, the simulated photoꢀ
chemical behaviour of 3a, and in general of phenylazoindole
cores, do not differ drastically compared to the parent azobenꢀ
zene photoswitch.14 In contrast, the thermal Z-to-E pathway
was studied via broken symmetry (BS)ꢀDFT taking into acꢀ
count three different mechanisms, viz. the inversion, the rotaꢀ
tion (both of them typical of azobenzene)38 and the intermoꢀ
lecular hydrazoneꢀformation (with and without water assisting
molecules). We found that the indole moiety lowers the rotaꢀ
tional activation barrier with respect to the inversion in 3aꢀb
(ꢁE‡=20.3 and 30.7 kcal/mol for the rotation and inversion
mechanism, respectively, at the ωB97XD/6ꢀ31G(d) level of
theory; Figure 3C; comparable findings were recently reported
on arylazoindazoles33), while NꢀMe indole (3c) increases it
(ꢁE‡=33.8 and 30.7 kcal/mol for the rotation and inversion
mechanism, respectively, at the ωB97XD/6ꢀ31G(d) level of
theory). That could be an explanation for the higher lifetimes
encountered in aprotic solvents for 3c compared to 3a and 3b,
where other pathways, apart monomolecular relaxation, were
not detected. The computed thermochemical values we have in
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 6 of 9
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 4. A: Photochemical E-toꢀZ isomerization pathways calculated at the CASSCF(10,8)/6ꢀ31G(d) level of theory, showing the differꢀ
ent pathways 3a can cover after excitation to the first or second excited state. B: Branching space of Conical Intersection A (CoInA) repreꢀ
sented in figure A. The derivative coupling (g) and gradient difference vectors (h), removing the degeneracy between the S1 and S0, are
highlighted with green arrows on the geometry at which the conical intersection occurs. C: Thermal reaction coordinates (RC) restoring E-
3a from Z-3a. In particular, the rotation, inversion and hydrazone formation mechanisms are compared. In the box, the spectroscopically
detectable hydrazone of 3a is represented. D: Thermal relaxation kinetics of 3a in MeCN at different temperatures. The related Eyringꢀ
Polanyi plot is depicted in the inset E: (TD)ꢀPBE0/6ꢀ311+G(2d,p) simulated UV/Vis spectrum of the sꢀtrans conformer of the azoꢀ
hydrazone of 3b in DMSO. In the inset, the optimized structure of the same species is shown.
ACS Paragon Plus Environment

Page 7 of 9
Journal of the American Chemical Society
Then, the system undergoes two consecutive waterꢀmediated
ACKNOWLEDGMENT
1
2
3
4
5
6
7
8
protonꢀtransfers leading to the E-isomer of 3a. The hydrazone
pathway possesses a slightly lower activation energy in protic
media than the rotation or the inversion ones. This explains the
surprisingly fast isomerization kinetics found with increased
proton availability of the medium for 3a,b.With this mechaꢀ
nism in hand, we also expect 3a and 3b to be able to switch
faster at higher concentrations, nicely explaining the concenꢀ
trationꢀdependent behavior we found in our experiments (entry
11, Table 1; Section 7 in Supporting Information).
We are grateful to Regina Hoheisel and Julia Zach for technical
support and to Prof. S. M. Bonesi (University of Buenos Aires)
for his help with the kinetic analysis. This work was supported by
CINECA SCAI, with computer time granted by ISCRA projects
(project code: HP10CBEIAU and HP10C8U1NY). N.A.S. thanks
the Studienstiftung des Deutschen Volkes for a doctoral scholarꢀ
ship. S.C. gratefully thanks P. Sbazzeguti, Prof. S. Protti and Dr.
D. Ravelli for their help and fruitful discussions.
9
REFERENCES
Moreover, the competition between the hydrazone and the
rotation pathway can be recognized for 3b in MeOH for its
biexponential decay. In contrast, due to its methylated nitrogen
atom, the azoꢀhydrazone formation is precluded to 3c.
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)
Molecular Switches; Feringa, B. L., Browne, W. R., Eds.;
WileyꢀVCH Verlag GmbH & Co. KGaA: Weinheim,
Germany, 2011.
(2)
Velema, W. A.; Szymanski, W.; Feringa, B. L. J. Am. Chem.
Soc. 2014, 136, 2178–2191.
CONCLUSIONS
(3)
(4)
Withers, N. Nat. Chem. 2010, 2, 11–11.
In summary, the photochromic properties of phenylazoinꢀ
doles have been investigated for the first time. We synthesized
compounds 3aꢀc representing three very basic derivatives of
this class of azo dyes. All compounds photoꢀisomerize applyꢀ
ing light of 355, 365 and 400 nm wavelength. Thereby, Nꢀ
methylated 3c showed the typical behaviour of an azobenzene
(τ between several minutes to two days). In contrast to this
findings, NꢀH azoindoles exhibit shorter lifetimes, strongly
dependent on the protic character of the solvent. We took furꢀ
ther advantage of this relationship and tuned the reaction kiꢀ
netics, increasing the proton content of the solvents. Best reꢀ
sults were obtained in mixtures of DMSO and water (1:1) usꢀ
ing DBU as an additive, giving rise to lifetimes between 33.9
and 46 ns for 3a or 3b, respectively. The surprisingly short
lifetimes are explained by a thorough mechanistic study, asꢀ
suming the intermolecular formation of a hydrazone. Such a
mechanism was studied in detail for the first time, spectroꢀ
scopically, kinetically and computationally. Compared to the
canonical inversion and rotation pathways, this conversion is
energetically favoured in protic media. Thus, faster Z-to-E reꢀ
laxation rates can be achieved for 3a,b in contrast to 3c.
Hence, only slight changes of the substituent position or the
solvent composition can drastically tune the properties of pheꢀ
nylazoindoles. Such extremely versatile properties render this
core structure a suitable candidate for the wide range of appliꢀ
cations of photoswitches.
Gelebart, A. H.; Jan Mulder, D.; Varga, M.; Konya, A.; Vanꢀ
tomme, G.; Meijer, E. W.; Selinger, R. L. B.; Broer, D. J.
Nature 2017, 546, 632–636.
(5)
(6)
(7)
CamachoꢀLopez, M.; Finkelmann, H.; PalffyꢀMuhoray, P.;
Shelley, M. Nat. Mater. 2004, 3, 307–310.
del Valle, M.; Gutiérrez, R.; Tejedor, C.; Cuniberti, G. Nat.
Nanotechnol. 2007, 2, 176–179.
Baroncini, M.; d’Agostino, S.; Bergamini, G.; Ceroni, P.;
Comotti, A.; Sozzani, P.; Bassanetti, I.; Grepioni, F.;
Hernandez, T. M.; Silvi, S.; Venturi, M.; Credi, A. Nat.
Chem. 2015, 7, 634–640.
(8)
Banghart, M.; Borges, K.; Isacoff, E.; Trauner, D.; Kramer,
R. H. Nat. Neurosci. 2004, 7, 1381–1386.
(9)
Gautier, A.; Gauron, C.; Volovitch, M.; Bensimon, D.;
Jullien, L.; Vriz, S. Nat. Chem. Biol. 2014, 10, 533–541.
Vomasta, D.; Högner, C.; Branda, N. R.; König, B. Angew.
Chem. Int. Ed. 2008, 47, 7644–7647.
Simeth, N. A.; Kneuttinger, A. C.; Sterner, R.; König, B.
Chem Sci 2017, 8, 6474–6483.
Neilson, B. M.; Bielawski, C. W. ACS Catal. 2013, 3, 1874–
1885.
Photochromism: molecules and systems, Rev. ed.; Dürr, H.,
BouasꢀLaurent, H., Eds.; Elsevier: Amsterdamꢁ; Boston,
2003.
(10)
(11)
(12)
(13)
(14)
(15)
(16)
Conti, I.; Garavelli, M.; Orlandi, G. J. Am. Chem. Soc. 2008,
130, 5216–5230.
Bandara, H. M. D.; Burdette, S. C. Chem Soc Rev 2012, 41,
1809–1825.
Andréasson, J.; Pischel, U.; Straight, S. D.; Moore, T. A.;
Moore, A. L.; Gust, D. J. Am. Chem. Soc. 2011, 133, 11641–
11648.
ASSOCIATED CONTENT
(17)
(18)
GarcíaꢀAmorós, J.; Velasco, D. Beilstein J. Org. Chem.
2012, 8, 1003–1017.
Kienzler, M. A.; Reiner, A.; Trautman, E.; Yoo, S.; Trauner,
D.; Isacoff, E. Y. J. Am. Chem. Soc. 2013, 135, 17683–
17686.
Supporting Information. General procedures, synthesis of comꢀ
pounds 3a-c, spectroscopic experiments, NMR experiments, comꢀ
putational analysis. This material is available free of charge via
the Internet at http://pubs.acs.org.
(19)
(20)
(21)
(22)
(23)
(24)
Bléger, D.; Schwarz, J.; Brouwer, A. M.; Hecht, S. J. Am.
Chem. Soc. 2012, 134, 20597–20600.
AUTHOR INFORMATION
Samanta, S.; McCormick, T. M.; Schmidt, S. K.; Seferos, D.
S.; Woolley, G. A. Chem. Commun. 2013, 49, 10314.
GarciaꢀAmorós, J.; DíazꢀLobo, M.; Nonell, S.; Velasco, D.
Angew. Chem. Int. Ed. 2012, 51, 12820–12823.
Venkataramani, S.; Jana, U.; Dommaschk, M.; Sonnichsen,
F. D.; Tuczek, F.; Herges, R. Science 2011, 331, 445–448.
Suwa, K.; Otsuki, J.; Goto, K. Tetrahedron Lett. 2009, 50,
2106–2108.
Corresponding Author
* Eꢀmail: burkhard.koenig@chemie.uniꢀregensburg.de
ORCID
Nadja A. Simeth: 0000ꢀ0001ꢀ8130ꢀ883X
Stefano Crespi: 0000ꢀ0002ꢀ0279ꢀ4903
Maurizio Fagnoni: 0000ꢀ0003ꢀ0247ꢀ7585
Burkhard König: 0000ꢀ0002ꢀ6131ꢀ4850
Calbo, J.; Weston, C. E.; White, A. J. P.; Rzepa, H. S.; Conꢀ
trerasꢀGarcía, J.; Fuchter, M. J. J. Am. Chem. Soc. 2017, 139,
1261–1274.
Author Contributions
(25)
Zhao, L.; Liu, J.; Zhou, P. J. Phys. Chem. A 2017, 121, 141–
150.
‡These authors contributed equally.
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 8 of 9
(26)
(27)
(28)
(29)
(30)
(31)
(32)
Otsuki, J.; Suwa, K.; Sarker, K. K.; Sinha, C. J. Phys. Chem.
A 2007, 111, 1403–1409.
Heitmann, G.; Schütt, C.; Herges, R. Eur. J. Org. Chem.
2016, 2016, 3817–3823.
Weston, C. E.; Richardson, R. D.; Fuchter, M. J. Chem
Commun 2016, 52, 4521–4524.
(33)
(34)
TraviesoꢀPuente, R.; Budzak, S.; Chen, J.; Stacko, P.; Jaꢀ
strzebski, J. T. B. H.; Jacquemin, D.; Otten, E. J. Am. Chem.
Soc. 2017, 139, 3328–3331.
Barden, T. C. In Heterocyclic Scaffolds II; Gribble, G. W.,
Ed.; Springer Berlin Heidelberg: Berlin, Heidelberg, 2010;
Vol. 26, pp 31–46.
Ding, H.; Peng, Z.; Wang, J.; Lu, P.; Wang, Y. Org Biomol
Chem 2016, 14, 7114–7118.
Babür, B.; Seferoğlu, N.; Aktan, E.; Hökelek, T.; ꢂahin, E.;
Seferoğlu, Z. J. Mol. Struct. 2015, 1081, 175–181.
Olivucci, M.; Bernardi, F.; Celani, P.; Ragazos, I.; Robb, M.
A. J. Am. Chem. Soc. 1994, 116, 1077–1085.
Yu, L.; Xu, C.; Zhu, C. Phys Chem Chem Phys 2015, 17,
17646–17660.
1
2
3
4
5
6
7
8
Wendler, T.; Schütt, C.; Näther, C.; Herges, R. J. Org.
Chem. 2012, 77, 3284–3287.
(35)
(36)
(37)
(38)
Wang, Y.ꢀT.; Liu, X.ꢀY.; Cui, G.; Fang, W.ꢀH.; Thiel, W.
Angew. Chem. Int. Ed. 2016, 55, 14009–14013.
Stricker, L.; Fritz, E.ꢀC.; Peterlechner, M.; Doltsinis, N. L.;
Ravoo, B. J. J. Am. Chem. Soc. 2016, 138, 4547–4554.
Weston, C. E.; Richardson, R. D.; Haycock, P. R.; White, A.
J. P.; Fuchter, M. J. J. Am. Chem. Soc. 2014, 136, 11878–
11881.
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
ACS Paragon Plus Environment

Page 9 of 9
Journal of the American Chemical Society
Insert Table of Contents artwork here
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
9
ACS Paragon Plus Environment