Sensitive Assays by Nucleophile-Induced Rearrangement of Photoactivated Diarylethenes

Article abstract of DOI:10.1021/jacs.8b02982

Upon light-induced isomerization, diarylethenes (DAEs) equipped with reactive aldehyde moieties rearrange selectively in the presence of amines, accompanied by decoloration. In a comprehensive study, the probe structure was optimized with regard to its inherent reactivity in the nucleophile-triggered rearrangement reaction. Detailed structure-reactivity relationships could be derived, in particular with regard to the type of integrated (het)aryl moieties as well as the location of the formyl residue, and the probes' intrinsic reactivity with primary and secondary amines was optimized. Utilizing an ancillary base, the initially formed rearrangement product can engage in a subsequent catalytic cycle, leading to an amplified decoloration process. This additional catalytic pathway allows us to enhance the sensitivity of our method and successfully discriminate between amines and thiols. Moreover, probes that exhibit strong analyte-induced fluorescence modulation have been designed to further decrease the detection limit by using a more sensitive read-out. The optimized DAE probes are promising molecular components for future programmable sensing materials and devices.

Full text of DOI:10.1021/jacs.8b02982

Article
Cite This: J. Am. Chem. Soc. XXXX, XXX, XXX−XXX
pubs.acs.org/JACS
Sensitive Assays by Nucleophile-Induced Rearrangement of
Photoactivated Diarylethenes
Sebastian Fredrich, Aurelio Bonasera, Virginia Valderrey, and Stefan Hecht*
̈
Department of Chemistry & IRIS Adlershof, Humboldt-Universitat zu Berlin, Brook-Taylor-Str. 2, 12489 Berlin, Germany
*
S Supporting Information
ABSTRACT: Upon light-induced isomerization, diaryle-
thenes (DAEs) equipped with reactive aldehyde moieties
rearrange selectively in the presence of amines, accompanied
by decoloration. In a comprehensive study, the probe structure
was optimized with regard to its inherent reactivity in the
nucleophile-triggered rearrangement reaction. Detailed struc-
ture−reactivity relationships could be derived, in particular
with regard to the type of integrated (het)aryl moieties as well
as the location of the formyl residue, and the probes’ intrinsic
reactivity with primary and secondary amines was optimized.
Utilizing an ancillary base, the initially formed rearrangement
product can engage in a subsequent catalytic cycle, leading to
an amplified decoloration process. This additional catalytic pathway allows us to enhance the sensitivity of our method and
successfully discriminate between amines and thiols. Moreover, probes that exhibit strong analyte-induced fluorescence
modulation have been designed to further decrease the detection limit by using a more sensitive read-out. The optimized DAE
probes are promising molecular components for future programmable sensing materials and devices.
INTRODUCTION
arise from common sulfur-containing precursors, in particular
methionine.
Routine laboratory techniques of quantifying biogenic
■
14
The qualitative discrimination and quantification of specific
molecular entities plays a crucial role in understanding and
monitoring physiological processes. Amines are, for example,
amines and thiols are mainly based on chromatography
15−17
coupled with mass spectrometry.
Because of their high-
involved in important biological regulation mechanisms, such a1s
costs and complexity, simpler tools for analyte evaluation in
food matrices were recently established. Although most rely on
dopamine in the stimulation/inhibition of potassium channels.
Similarly, thiols occupy a prominent role in molecular biology,
18−23
ordinary colorimetric assays,
the limit of color perception
ranging from the formation of stabilizing disulfide cross-links in
2
is only challenged in a few examples where catalytic/enzymatic
α-keratin to the prevention of oxidative stress and the
24,25
3
,4
mechanisms further decrease the detection limit.
Among colorimetric assays, approaches using fluorescence
coenzymatic role in several reactions with glutathione.
Nevertheless, various amines and thiols are connected to
pathological conditions, for instance dopamine present during
different stages of Parkinson’s disease.
In addition to the specific nature of the (bio)analytes, their
26−28
detection exhibit significant advantages.
Complex matrices
5
often comprise a high background color that needs to be
considered, whereas fewer samples show autofluorescence or
quenching. This latter effect can be used to determine the
concentration of certain analytes via Stern−Volmer analysis
6
concentration is another key parameter to survey. Industrial
effluvia in potable water sources or food spoilage can be
29
using fluorescent dyes that respond to amino groups. Other
techniques exploit turn-on fluorescence by reacting the analyte
with a sensor molecule to generate a highly emissive product or
harmful when the concentration of biogenic amines overcomes
7
a certain limit. Consequently, the presence of these amino
8
compounds is indicative of insufficient hygiene conditions and
potential toxicological effects. Biogenic amines, found in a
wide range of foods, most notably meat and fish, consist of low-
volatile amines formed via thermal or enzymatic decarbox-
27,28,30−32
9
to shift its emission maximum.
In order to externally control the operation of the sensor,
approaches have been developed using photochromic com-
pounds to regulate the detection of amino acids, cyanide
anions, thiols, organophosphorus compounds, and other
10−12
ylation of amino acids by bacteria.
Examples include
ethanolamine, spermidine, spermine, and the most widely
studied cadaverine, histamine, and putrescine, produced by
decarboxylation of lysine, histidine, and ornithine, respec-
18,19,33−37
analytes.
In most of these “smart” assays, the added
optical switching feature plays a major role due to the existence
1
1
tively. Volatile sulfur compounds (VSCs) are another
important class of analytes in food chemistry and increasing
Received: March 17, 2018
13
attention has been paid to their organoleptic relevance. They
©
XXXX American Chemical Society
A
DOI: 10.1021/jacs.8b02982
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
of one inactive state while the other is able to probe the desired
analyte. Recently, we reported diarylethene (DAE) single-
component probes, which undergo facile decoloration in the
Scheme 1. Overview of the Investigated DAE Probes
3
8
presence of amines after light-induced activation. The
remote-control over reactivity is remarkable and allows to
store the dormant probe for long time, even in an environment
contaminated with the analyte. Despite these advantages, we
wanted to lower the detection limit and clarify some details of
the mechanism. This motivated us to lead our new studies into
four main directions: (i) investigation of different structural
motifs, in order to understand key structural parameters that
could improve sensitivity and speed (Figure 1); (ii) detailed
insight into the sensing mechanism, including the different
reactivities of the involved species; (iii) application of the thus
optimized molecular systems to selectively and efficiently
recognize various classes of biomolecules, providing a unique
tool for multiple-targets trace sensing; (iv) utilization of the
specific spectral features of the inactive, active, and neutralized
forms to create turn-on sensors with high optical sensitivity.
groups using metalation by n-BuLi followed by reaction with
dimethylformamide (DMF). The aldehyde function was
masked as an acetal before further functionalization by Suzuki
coupling to give rise to a large variety of precursors (Schemes
S1 and S2). Assembly of the majority of the target DAEs was
achieved by sequential reaction of the metalated heteroaromatic
building blocks to perfluorocyclopentene followed by depro-
tection under acidic workup conditions (see Schemes S3−
39
S7). An analogous synthesis was used to prepare phenyl-
substituted DAEs 7 and 8 (see Scheme S6). Note that synthesis
of indole-substituted DAEs 10/11 as well as 19 required
N‑protection by a tert-butyloxycarbonyl (BOC) group.
Preparation of the benzothiophene-bridged DAEs 16−18 was
realized using two separate cross-coupling reactions (see
Scheme S8). Further details about the synthesis and character-
ization data can be found in the Supporting Information.
Structure−Reactivity Relationships. In order to have a
fair comparison, the compound showing the best performance
Figure 1. Light-induced reactivity modulation of diarylethene aldehyde
probes toward nucleophiles.
38
in our initial work, DAE-O, was used as benchmark. As typical
for DAEs, the open isomer Oo absorbs only in the UV-region
(
λmax = 285 nm) and undergoes 6π-electrocyclization upon UV-
In this article, we report a comprehensive investigation of the
structure−reactivity relationships of aldehyde-substituted DAE
probes deciphering the influence of the position of the reactive
carbonyl moiety as well as the nature of substituents on the
response to amines and thiols. On the basis of our detailed
mechanistic understanding, we describe two approaches,
namely catalytic amplification and fluorescence read-out, to
discriminate between amine and thiol analytes and increase the
overall sensitivity of our remote-controlled detection systems.
light irradiation (λ = 365 nm), indicated by purple coloration
of the solution and the rise of a new spectral band (λmax
580 nm) in the visible range (Figure 2, red line). According to
chromatographic analysis of the final irradiated solution, the
photostationary state (PSS) contains about 85% of the closed
irr
=
isomer Oc. Irradiation with visible light (λ > 500 nm) leads to
irr
full recovery of the open isomer Oo. This switching process can
be repeated over several cycles. The addition of n-octylamine
(OA) to the solution of pure Oo does not have any relevant
effect for several weeks, when some amount of the
corresponding imine Ooi is being formed, which can be
switched reversibly to its closed form Oci. On the contrary, the
irradiated mixture containing Oc immediately undergoes
decoloration when OA is added, resulting in a yellowish
solution (Figure 2, blue line). We thus concluded that the
closed isomer is involved in an irreversible rearrangement
RESULTS AND DISCUSSION
■
Synthesis. In order to identify the most sensitive probe, i.e.,
the molecular structure with the highest photoinduced
reactivity toward the analyte, several structural variations were
synthesized and subsequently explored (Scheme 1). For
synthesis, different heterocycles were equipped with formyl
B
DOI: 10.1021/jacs.8b02982
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
fully condensed into the corresponding imine 1ci in around 20
h (see Figure S56).
As expected, the imine can be switched back and forward by
means of visible and UV-light irradiation, respectively, without
inducing irreversible changes of the molecular scaffold (Figure
3
). We rationalize these findings by considering that upon ring-
closure the aromatic aldehyde group experiences only a slight
modulation of its electronic nature and thus reactivity toward
amines. More importantly, the intermediate hemiaminal is
inactive toward rearrangement, in strong contrast to the one
formed in case of Oc. Similarly, the reactivity of the
symmetrical compound 2 with two aldehyde moieties at the
periphery was found analogous.
Figure 2. Top: Proposed mechanism involving initial photoactivation
of Oo and subsequent reaction of formed Oc with OA to yield
rearrangement product Or. Bottom: Evolution of UV/vis absorption
spectra after reaching the PSS at λ = 365 nm (red line, [Oc] = 3.6 ×
irr
−
5
−2
1
0
M in CH CN) upon addition of OA (1.0 × 10 M). The inset
shows the decay of absorbance at λmax = 580 nm corresponding to the
3
decreasing concentration of Oc.
reaction, producing the photoinactive compound Or. A
reasonable mechanism for the observed rearrangement reaction
evokes the initial nucleophilic attack of the amine to the highly
reactive, aliphatic aldehyde of the closed DAE, resulting in the
corresponding hemiaminal intermediate. Subsequently, this
intermediate rearranges to the more stable compound Or
accompanied by formation of N-octylformamide (Figure 2,
top).
Figure 3. Top: Proposed mechanism for reversible formation of imine-
based DAEs 1ci and 1oi. Bottom: Evolution of UV/vis absorption
spectra during reversible photoisomerization of 1ci (red line, 5.5 ×
On the basis of these previous findings, the first new design
we decided to investigate concerned DAE compounds with
aldehyde groups in the outer part of the molecular skeleton
−5
1
0
M in CH CN) to 1oi (blue line) and vice versa upon irradiation
3
40
(
see series I in Figure 1). For an easy comparison of the
(λirr = 617 nm) and (λirr = 302 nm), respectively.
general behavior and sensing capacities, these molecules have a
similar design, only exchanging the phenyl substituents in the
parent DAE-O by aldehydes in the periphery and bearing two
methyl groups at the ring-closing inner α-position of the
molecule (see DAEs 1 and 2 in Scheme 1). These DAEs
preserve the general features described by our group in a
Once we confirmed the necessity of positioning the aldehyde
functionality in the core position to enable an effective
rearrangement, we introduced two formyl groups in the two
inner α-thienyl positions of the DAE structure. The resulting
symmetrical derivative 3 should in principle exhibit a higher
probability of reacting with amines. However, upon irradiation
of 3o with UV-light the generation of two new species, the
expected closed isomer 3c, but also another compound with
inferior mass, i.e., 3r, was observed (see Figure S25). Isolation
40
different context for compounds with nonfluorinated bridge.
While in the presence of an excess of amine (OA,
2
4
.0 × 10− M) Oc is completely transformed into the
rearrangement product in less than 3 min, its positional
analogue 1c does not undergo any rearrangement but instead is
C
DOI: 10.1021/jacs.8b02982
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
and characterization by UV/vis spectroscopy (see Figure S54
(Table 1) we have refrained from providing general sensitivity
factors for the reaction with nonspecified amines.
1
19
and Table S7), HRMS, H and F NMR revealed that the
formed 3r is no longer switchable and resembles the structure
of the other rearrangement products. In a control experiment,
an irradiated mixture containing 3o, 3c, and 3r, was monitored
in the darkness, but even after 2 h, no (further) transformation
of 3c to 3r was observed (see Table S3). Thus, our hypothesis
to explain these findings assumes that continued irradiation of
Though introducing EWGs into the DAE-scaffold clearly
enhances the reactivity of the ring-closed aldehydes, this effect
is probably weakened by the rather electron-rich thiophene
moieties and the substantial distance of the electron-accepting
substituents from the reactive center.
To further optimize the reactivity, we were considering less
electron-rich aromatic rings attached directly to the central
olefinic bridge and therefore replaced either thiophene moiety
by a phenyl ring (see DAEs 7 and 8 in Scheme 1). On the one
hand, 7o bearing a benzaldehyde moiety, upon irradiation gave
a variety of products, which we attribute to nonselective
photocyclization with regard to the position of ring-closure. On
the other hand, changing the opposite thiophene to a phenyl
moiety was expected to block the rearrangement due to
replacing the weak C−S bond, to be broken, by a stronger
C−C bond. To our surprise, addition of amines to solutions
containing photogenerated 8c also led to rapid bleaching of the
visible band over a time range comparable with other switches
in our library. Puzzled by this result, we were able to isolate the
formed rearrangement product, which turned out to be difficult
due to a low ratio of 8c in the PSS caused by its thermal
instability at room temperature (t1/2 = 4.5 h, see Table S3). A
(
photogenerated) 3c with UV-light triggers the rearrangement
event, even in the absence of amine.
After establishing one formyl group in the core of the DAE
structure to be optimal, we evaluated the effect of different
substituents in the outer phenyl pendants (see series II in
Figure 1). As long as no competing aldehyde is located there,
the mechanism of the rearrangement should not differ from the
one as shown in Figure 2. The rearrangement is presumably
driven by formation of the aromatic central 6-membered ring as
well as the resonance stabilized formamide and should be
accelerated by the presence of electron-withdrawing groups
(
EWGs). Thus, a trifluoromethyl (−CF ) group was
3
introduced in para-position of the phenyl moiety adjacent to
either the aldehyde function to render it more reactive or the
opposite thiophene moiety to facilitate its breakdown (see
DAEs 4 and 5 in Scheme 1). Indeed, photoactivated compound
1
19
5
c was found to be more reactive toward amines with rate
constants up to 1.6 times larger in the case of spermidine
SPER) when compared to reference Oc (Table 1). Placing the
combination of techniques (1D- and 2D- H and F NMR
experiments, HRMS, see Supporting Information, Figures
S20−S22, S55, Table S7) allowed us to determine the chemical
structure of 8r, which is consistent with the assumption that the
rearrangement does not involve breaking of a C−C bond
(
Table 1. Rate Constants for the Amine-Induced
41
Rearrangement Reaction Representing the Sensitivity of
Improved DAE-Probes 4c−6c in Comparison to Reference
within the phenyl residue as described by Ho et al. It seems
that despite limited stabilization by the formed cross-
conjugated π-system, elimination of HF drives the rearrange-
ment (Figure 4). Computational results support that formation
−5
Oc (all 2 × 10 M) in the Presence of Various Amines
3
(
5 × 10− M)
EWG closer to the reactive site further accelerated the
rearrangement as the reaction with SPER proceeds at twice
the rate of the benchmark switch. In an effort to combine both
of these effects, DAE-6 was synthesized and its closed isomer
6
c rearranges 2.7 times faster than Oc in the presence of SPER.
Figure 4. Comparison between the proposed mechanisms for the
rearrangement of DAE-Oc and DAE-8c to Or and 8r, respectively.
Experimentally determined rate constants for Oc and 8c in the
presence of different amines.
A similar tendency was observed for other amines (Table 1)
and thus the derived structure−reactivity relationship appears
to be general. However, in view of exceptions such as for BA
D
DOI: 10.1021/jacs.8b02982
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
of this unexpected rearrangement product is indeed thermo-
dynamically favored (see Table S8). Interestingly, the
optimized geometry of 8c exhibits different C−F bond lengths
(see Figures S76 and S77 and Table S9), suggesting an
enhanced and biased reactivity of the closed isomer (see section
in the Supporting Information). The observed rearrangement
is related to previous reports involving fluoride elimination of
6
42−44
45,46
the closed isomer of diarylethenes
or other pathways,
most notably the irreversible thermally bleaching of some
47−51
oxidized dithienylethenes due to release of SO2.
All the above results are in agreement with our proposed
mechanism (see Figure 2, top). Clearly, the introduction of
EWGs in key positions both enhances the reactivity of the
formyl group toward initial nucleophilic attack of the amine
analyte and accelerates the rearrangement of the formed
intermediate. It appears that resonance stabilization of the
aromatic fused benzenes as well as formamide products (or
alternatively HF elimination in the case of 8c) is the main
driving force behind the observed rearrangement.
Catalytic Amplification. In order to further enhance the
sensitivity of our system, we thought of engaging the initially
produced rearrangement product in a subsequent catalytic
cycle, which would involve nucleophilic attack of its own
precursor, thereby accelerating decoloration of the closed DAE
probe. However, under neutral conditions explored thus far the
formed thiol Or appears not to be reactive enough to induce
the rearrangement. Thus, we were considering to add a tertiary
amine base to deprotonate the formed thiol, thereby greatly
enhancing its nucleophilicity. Note that though primary and
secondary amines trigger the rearrangement, tertiary amines
cannot and therefore should be able to serve solely as the base.
On the basis of these considerations, we were exploring basic
conditions for the rearrangement, which should initially be
triggered by substoichiometric amounts of primary or
secondary amine analytes and subsequently be amplified in a
catalytic self-decoloration process.
Figure 5. Proposed mechanism of the analyte-induced self-catalyzed
rearrangement under basic conditions.
sensitivity of our probes. First, we designed an advanced
molecular probe carrying an integrated tertiary amine group
with a pK similar to DMBA and unable to react by itself with
a
the aldehyde giving a false positive response. However, this new
DAE probe 9, containing a N,N-dimethylamino group at the
periphery, exhibited intense fluorescence under irradiation at
365 nm leading to formation of only a negligible amount of
ring-closed, activated probe at the PSS. Though a PSS rich in 9c
can be achieved by quenching fluorescence through addition of
Initial experiments reacting Oc with substoichiometric
amounts of OA led to partial bleaching of the visible band,
corresponding to the amount of amine added. In addition, OA
was completely consumed and converted to N-octylformamide.
Then, a tertiary amine base, i.e., N,N-dimethylbenzylamine
+
(DMBA) shown to be unreactive by itself, was added in excess,
acid (the corresponding ammonium salt 9-H does not emit,
and suddenly the bleaching of the visible band takes place, with
a comparable rate yet until complete consumption of Oc (see
Figure S58), increasing the sensitivity of the probe. This finding
strongly supports our reaction mechanism (Figure 5). It is
important to note that the overall concentration of nucleophile
in solution remains constant, reflecting the level of initially
added amine.
In view of this modified, base-catalyzed mechanism, we can
now explain why after the addition of amine to the closed DAE
probe, a small band appears in the absorption spectrum
between 400 and 500 nm, which persists until the end of the
reaction (see Figure 2, bottom). Upon addition of acid, the
band disappears and therefore we attribute it to the thiolate
rearrangement product formed due to the basic environment
resulting from the nonreacted amine (see Figure S57). This
suggest that the catalytic pathway takes place also in the
presence of excess of amine, but its contribution to the
decoloration velocity is negligible because the concentration of
nucleophile is quasi constant under these conditions.
see Figure S30 and Tables S1−S4), it contradicts the idea of an
internal base. These findings led to the design of the two DAE
probes 10 and 11, which contain an indole fragment integrated
either in regular or inverse fashion, i.e., connected via its 2- and
3-position, respectively. To our surprise, and disappointment,
neither 10 nor 11 underwent photocyclization even after
prolonged irradiation times with UV-light in acetonitrile
solution. Once again, this problem presumably arising from
formation of a twisted intramolecular charge transfer (TICT)
state can be circumvented by changing to a nonpolar
5
2,53
solvent,
such as n-hexane in which efficient photo-
cyclization is observed (see Figure S31 and Tables S2−S4).
Because of the strong solvent dependency of the pK of amines
a
and thiols, however, the formation of thiolates in such a
nonpolar medium is more difficult to achieve and the presence
of one equivalent of the tertiary amine base would presumably
not have a significant effect.
After our unsuccessful attempts to design an internally base-
assisted probe, we then were aiming to decrease the pK of the
a
On the basis of the new, more detailed mechanism, we
followed three strategies to exploit it in order to enhance the
generated thiol to achieve deprotonation already under neutral
conditions. With all previously discussed DAE probes, the
E
DOI: 10.1021/jacs.8b02982
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
rearrangement resembles an ene-thiol and thus we designed 12
carrying a benzothiophene moiety, which after rearrangement
gives rise to formation of a more acidic thiophenol. Though the
rate of the decoloration reaction is comparable with Oc, upon
addition of substoichiometric amounts of amine, however, no
evidence for a self-induced decoloration of 12c could be
observed (see Figure S71).
Finally, tertiary amine was added right at the start of the
decoloration reaction. The setup was slightly changed,
employing a cuvette with shorter path length (1 mm) to
allow for 1 order of magnitude higher concentrations according
to Lambert−Beer Law. Thus, a solution containing our generic
DAE probe Oc is reacted with substoichiometric amount (less
than 0.5 equiv) of OA in absence and presence of DMBA,
respectively (Figure 6). While in the absence of the base, the
DAE probe is gradually consumed resulting in a partial
decoloration of the sample, the presence of the ancillary base
results in a more pronounced bleaching in accordance with the
catalytic mechanism. In the absence of primary or secondary
amine analyte, the base causes only negligible decoloration.
visible absorption band and a degradation of 6c, similarly to
what was observed for the catalytic experiments (see Figures
S65−67,74,75). This allows us to selectively distinguish
between amine and thiol analytes. In other experiments,
different quantities of DMBA were used while keeping the
concentration of the thiol constant and a linear correlation
between basicity and the apparent rate of the decoloration was
found (see Figure S68). Cleavage of the thioester formed in the
catalytic process (see DAE-t in Figure 5) and thus a
contribution of subsequently potentially generated
thiol/thiolate could be ruled out (see Figure S60). In strong
contrast to thiols, the presence of alcohols did not lead to any
detectable decoloration and thus rearrangement as the basicity
of DMBA is too low to generate alkoxides. Stronger bases
capable of deprotonating alcohols, such as sodium hydride and
tert-butoxide, led to immediate decoloration of the DAE
solution, even in the absence of alcohols. Similar negative
results were obtained testing for phenol as the prototype of
aromatic alcohols. In view of the observed chemoselectivity of
decoloration we developed a simple flowchart for discriminat-
ing primary/secondary amines from thiols (Figure 7), even in
the presence of alcohols.
−
4
Figure 6. Decoloration rate of Oc (6.5 × 10 M in CH CN) with
3
−4
substoichiometric amounts of OA (3.0 × 10 M) in the absence (red)
and presence (blue) of 1 vol % DMBA monitored by the absorbance
decay at 580 nm. The control experiment contains a solution of Oc
and DMBA only (black).
Figure 7. Sequential sensing scheme (flowchart) for the visual
discrimination of primary/secondary amines and thiols employing the
same DAE probe 6.
Discrimination of Analytes. In addition to amines, we
were interested in expanding the scope of our method to the
potential detection of other nucleophiles, in particular alcohols
and thiols that represent typical constituents of food matrices as
well. On the basis of our above finding that deprotonation
appears to be essential to induce sufficient reactivity in the case
Fluorescence Read-Out. After increasing the reactivity of
the DAE probe by optimizing its substitution pattern and
implementing an additional base-catalyzed amplification
mechanism, we sought to further lower the detection limit by
developing a more sensitive read-out. Following the absorbance
spectral change of a colored solution toward bleaching is
effective but lacks sensitivity, especially for the naked eye, when
compared to the evolution of color from a noncolored negative
sample. As a result, a slight coloration can more readily be
observed than a slight decoloration event, giving rise to a lower
threshold. When considering a DAE, which is colorless in both,
the open and closed isomer but is supposed to be colored in the
rearranged form, the problem is that the latter typically exhibits
a hypsochromically shifted absorption maximum as compared
to the closed isomer due to its less extended π-system. The
solution for this problem is the evolution of a band upon
of thiols and in view of the much higher pK values of alcohols
a
as compared to thiols, we were speculating that their
discrimination should be possible simply by choosing an
appropriate ancillary base.
For these experiments, we decided to explore the more
reactive DAE probe 6c and use DMBA as the base. In two
blank experiments, solutions of 6c were treated with 4-tert-
butylbenzyl mercaptan and ethanol, respectively. The model
thiol was selected as it resembles an aliphatic thiol yet exhibits
low volatility and thus an acceptable smell, whereas ethanol was
chosen because it is omnipresent in foods and readily available
in high purity. In both cases, in the absence of DMBA no
decoloration was observed (see Figure S59). In the presence of
DMBA, however, addition of the thiol induced bleaching of the
F
DOI: 10.1021/jacs.8b02982
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
nucleophile addition that is not present in the absorption
spectrum.
benzothiophene shows slight emission (Φ = 0.019) compared
f
to a barely higher intensity of dibenzothiophene (Φ =
f
54,55
The DAE probe 14 in both of its switching states as well as
its rearranged structure meets these requirements. It comprises
on the reactive side a benzothiophene moiety that is converted
via nucleophile-induced rearrangement of 14c into a
dibenzothiophene scaffold in 14r (Figure 8, top). Such
structures are known to show a considerable fluorescence,
0.08).
The more promising related example is benzofuran
with negligible fluorescence and dibenzofuran with a much
higher fluorescence quantum yield (Φ = 0.53) of its
Thus, DAE probe 15 bearing a
f
55,57
unsubstituted form.
benzofuran moiety was synthesized and, as expected, the
colorless acetonitrile solution of the 15o does not show
fluorescence at any excitation wavelength. Upon UV-induced
ring-closure, only very weak emission signals appear at around
54,55
whereas a simple benzothiophene is not emissive.
Indeed,
1
4o exhibits a very weak and broad emission band at around
3
00 nm when exciting the sample between 200 and 250 nm.
580 nm. This emission band disappears completely upon UV-
Addition of an excess of OA to this sample led to the evolution
of an intense emission band centered at 475 nm upon
excitation with 400 nm (see Figure S49). Although this example
illustrates our design concept and the fluorescence quantum
induced ring-closure. Note that residual emission resulting from
incomplete ring-closure in the PSS is mostly reabsorbed by the
intense visible band of 14c (λmax = 542 nm). Addition of an
excess of OA induced rearrangement to give 14r, accompanied
by the rise of a sharp emission band at 365 nm upon excitation
at 250 nm (see Figure S48). Thus, one should be able to detect
amines with a high sensitivity by directly measuring the amount
of rearranged form 14r via its fluorescence since emission
intensity can be considered linear dependent to the
yield (Φ = 0.017) is higher than the one of compound 14, it is
f
below the maximum that can be reached using this approach.
To boost the sensitivity of our system further, clearly higher
fluorescence quantum yields are needed. The structural
difference between the open and the rearranged forms of the
diarylethenes is the annulated benzene to the aryl ring bearing
the aldehyde in the open isomer. In addition, one could
imagine replacing the bridge by a unit, which is not fluorescent
but after benzene annulation it is. For this purpose, we decided
56
concentration at low optical densities (OD < 0.1).
Though the concept of fluorescence detection can be nicely
illustrated with DAE probe 14, the sensitivity of the system is
low due to the relatively small fluorescence quantum yield (Φ
45
f
to use a benzothiophene moiety as the bridging unit as well
=
0.007) and the significant emission of the open form. For this
and prepared DAE probe 16. Indeed, the rearranged form 16r
reason, alternative structures were considered. Unsubstituted
shows rather intense fluorescence (Figure 8, bottom, Φ =
f
0
.024). The reaction of the closed form 16c with the amine is,
however, significantly slower than for the other investigated
DAE probes. This is primarily attributed to the electron-rich
character of the benzothiophene bridge when compared to the
hexafluorocyclopentene bridging unit, rendering the aldehyde
less electrophilic. For this reason, we decided to further modify
the connectivity between the reactive and the bridging moieties
(
see Scheme 1). DAE probe 17, having an inverted benzo-
thiophene bridge when compared to 16, shows a slightly slower
reaction with OA yet the fluorescence of 17r is significantly red-
shifted (λem = 490 nm, λexc = 370 nm, Φ = 0.025, see Figure
f
S51), which we ascribe to the more planar structure of its bent
tribenzodithiophene core. With regard to fluorescence, a
combination of a benzothiophene bridge and a benzofuran as
the reactive formylated aryl unit appeared attractive. However,
DAE probe 18 shows a lower fluorescence quantum yield (Φ =
f
0
.008, see Figure S52) and due to the electron-rich nature of
the benzofuran its reactivity is so low that even upon addition
of an excess of OA no immediate decoloration occurs.
From the above findings, it is clear that fused heterocycles
can induce fluorescence mainly in the rearranged form.
Excitation and emission wavelengths are dependent on size
and planarity of the fluorophore. While introduction of one
heterocycle in the bridge increases the probe’s fluorescence
efficiency it concomitantly decreases its reactivity. Moreover,
breaking overall three aromatic rings upon photocyclization
destabilizes the closed isomer leading to deterioration of the
photoactivated probe. To overcome this problem, a switch
exhibiting the desired analyte-induced fluorescence modulation
but constructed using rather electron-poor building blocks was
needed. In this context, we reinvestigated DAE probe 10,
carrying an indole and a thiophene moiety connected via a
perfluorocyclopentene bridge. As expected, the open isomer
shows a rather intense emission at 390 nm (λexc = 325 nm),
which upon ring-closure is bleached until a certain point,
reflecting the ratio of open and closed isomers at the PSS.
Figure 8. Top: Different forms of DAE probes 14 and 16. Bottom:
Absorption (solid lines) and emission spectra (dashed lines) of 16 (4
−5
×
10 M in CH CN) in the open form (black), PSS mixture at 365
nm (red), and rearranged form originating from the PSS mixture
blue). Emission spectra were excited at 335 nm.
3
(
G
DOI: 10.1021/jacs.8b02982
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
However, upon addition of OA, a new emission band appears,
which is associated with the rearranged form and strongly red-
shifted when compared to the one of the open isomer. More
importantly, at the excitation maximum of 380 nm the 10o
shows no emission and 10r can be excited without interference
groups and inner reactive aldehydes. Moreover, (ii) unravelling
of an additional base-catalyzed pathway allowed us to lower the
detection below the stoichiometric limit and (iii) to successfully
discriminate between primary/secondary amines and thiols.
Furthermore, (iv) probes with significant analyte-induced
fluorescence modulation were designed, allowing us to further
enhance the sensitivity of our method. Future work will be
concerned primarily with implementation of our DAE probes
into actual sensing materials and devices.
(
Figure 9). Moreover, the fluorescence intensity (Φ = 0.03) is
f
ASSOCIATED CONTENT
Supporting Information
The Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/jacs.8b02982.
■
*
S
Synthetic procedures, compound characterization and
photochemical spectral data, experimental methods,
kinetic measurements data, and further details concern-
ing computational models (PDF)
AUTHOR INFORMATION
Corresponding Author
sh@chemie.hu-berlin.de
■
*
ORCID
Stefan Hecht: 0000-0002-6124-0222
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
V.V. is indebted to the Alexander von Humboldt Foundation
for providing a postdoctoral fellowship. Generous support by
the European Research Council (ERC via ERC-2012-
STG_308117 “Light4Function”), the European Commission
Figure 9. Top: Different forms of DAE probe 10. Bottom: Emission
−5
spectra of 10 (4 × 10 M in n-hexane) in the open (black) and
rearranged form (blue) upon excitation with 325 nm (dashed lines)
and 380 nm (solid lines).
(
via MSCA-ITN “iSwitch” GA No. 642196), and the German
Research Foundation (DFG via SFB 658, project B8) is
gratefully acknowledged.
the highest among all investigated rearrangement products and
the reaction with amines is rather fast, resulting in a high
sensitivity of this probe. The only disadvantage is the necessity
of using n-hexane as solvent to enable UV-induced ring-closure
since no photocyclization was observed in acetonitrile.
REFERENCES
■
(
(
1) Sasaki, K.; Sato, M. Nature 1987, 325, 259−262.
2) Fraser, R. D. B.; MacRae, T. P.; Sparrow, L. G.; Parry, D. A. D.
Int. J. Biol. Macromol. 1988, 10, 106−112.
3) Meister, A.; Anderson, M. E. Annu. Rev. Biochem. 1983, 52, 711−
760.
(4) Noctor, G.; Foyer, C. H. Annu. Rev. Plant Physiol. Plant Mol. Biol.
1998, 49, 249−279.
5) Gingrich, J. A.; Caron, M. G. Annu. Rev. Neurosci. 1993, 16, 299−
21.
6) Gao, T.; Tillman, E. S.; Lewis, N. S. Chem. Mater. 2005, 17,
(
In order to keep the structure as similar to 10 as possible but
to improve its switching behavior in acetonitrile, its BOC-
protected precursor molecule 19 was investigated. Gratifyingly,
DAE probe 19 is switching in acetonitrile and shows a similar
reactivity toward OA as described for 10. Though the
rearrangement product 19r emits with a slightly lower
fluorescence quantum yield (Φ = 0.022, λ = 320 nm, λem
(
3
(
2
(
(
904−2911.
f
exc
7) Bardocz, S. Trends Food Sci. Technol. 1995, 6, 341−346.
=
440 nm), open and closed isomers do not exhibit any
emission (see Figure S53). Therefore, with compounds 10 and
9 we were able to identify appealing probes that allow for
8) Naila, A.; Flint, S.; Fletcher, G.; Bremer, P.; Meerdink, G. J. Food
Sci. 2010, 75, R139−R150.
9) Ladero, V.; Calles-Enriquez, M.; Fernandez, M.; Alvarez, M. A.
Curr. Nutr. Food Sci. 2010, 6, 145−156.
(10) Ruiz-Capillas, C.; Jimenez-Colmenero, F. Crit. Rev. Food Sci.
Nutr. 2005, 44, 489−599.
(11) Bulushi, I. A.; Poole, S.; Deeth, H. C.; Dykes, G. A. Crit. Rev.
Food Sci. Nutr. 2009, 49, 369−377.
́ ̀ ́
12) Roig-Sagues, A. X.; Hernandez-Herrero, M. M.; Lopez-Sabater,
E. I.; Rodríguez-Jerez, J. J.; Mora-Ventura, M. T. Lett. Appl. Microbiol.
997, 25, 309−312.
13) Vermeulen, C.; Gijs, L.; Collin, S. Food Rev. Int. 2005, 21, 69−
37.
(14) Landaud, S.; Helinck, S.; Bonnarme, P. Appl. Microbiol.
Biotechnol. 2008, 77, 1191−1205.
1
(
fluorescence modulation, either by shifting or turning-on the
emission, as an alternative read-out to absorption.
́
CONCLUSION AND OUTLOOK
■
We were able to identify smart sensor molecules for
nucleophile detection as potent probes that can be remotely
controlled by light irradiation. Exploring (i) the relationships
between DAE structure and photoswitching properties as well
as reactivity of the ring-closed form toward an analyte-induced
rearrangement, we could ascertain the most effective
substitution pattern comprising outer electron-withdrawing
(
1
(
1
H
DOI: 10.1021/jacs.8b02982
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
(
15) Johnson, D. C.; Dobberpuhl, D.; Roberts, R.; Vandeberg, P. J.
Chromatogr. A 1993, 640, 79−96.
16) Chesnov, S.; Bigler, L.; Hesse, M. Eur. J. Mass Spectrom. 2002, 8,
−16.
17) Latorre-Moratalla, M. L.; Bosch-Fuste,
Cid, S.; Veciana-Nogues, M. T.; Vidal-Carou, M. C. J. Chromatogr. A
009, 1216, 7715−7720.
18) Lin, J.-F.; Kukkola, J.; Sipola, T.; Raut, D.; Samikannu, A.;
(51) Kodama, R.; Sumaru, K.; Morishita, K.; Kanamori, T.; Hyodo,
K.; Kamitanaka, T.; Morimoto, M.; Yokojima, S.; Nakamura, S.;
Uchida, K. Chem. Commun. 2015, 51, 1736−1738.
(
1
(
(52) Irie, M.; Sayo, K. J. Phys. Chem. 1992, 96, 7671−7674.
́
J.; Lavizzari, T.; Bover-
(53) Herder, M.; Pat
011, 47, 460−462.
54) Seixas de Melo, J.; Rodrigues, L. M.; Serpa, C.; Arnaut, L. G.;
Ferreira, I. C. F. R.; Queiroz, M. J. R. P. Photochem. Photobiol. 2003,
7, 121−128.
55) Nijegorodov, N.; Luhanga, P. V. C.; Nkoma, J. S.; Winkoun, D.
P. Spectrochim. Acta, Part A 2006, 64, 1−5.
56) Miller, J. N. Standards in fluorescence spectrometry; Chapman and
Hall, 1981.
57) Tero, T.-R.; Salorinne, K.; Lehtivuori, H.; Ihalainen, J. A.;
Nissinen, M. Chem. - Asian J. 2014, 9, 1860−1867.
̈
zel, M.; Grubert, L.; Hecht, S. Chem. Commun.
2
́
(
2
(
7
(
Mikkola, J.-P.; Mohl, M.; Toth, G.; Su, W.-F.; Laurila, T.; Kordas, K. J.
Mater. Chem. A 2015, 3, 4687−4694.
(
19) Leng, P. Q.; Zhao, F. L.; Yin, B. C.; Ye, B. C. Chem. Commun.
(
2
(
015, 51, 8712−8714.
20) Rakow, N. A.; Sen, A.; Janzen, M. C.; Ponder, J. B.; Suslick, K. S.
(
Angew. Chem., Int. Ed. 2005, 44, 4528−4532.
(
(
21) Kumpf, J.; Bunz, U. H. F. Chem. - Eur. J. 2012, 18, 8921−8924.
22) Askim, J. R.; Mahmoudi, M.; Suslick, K. S. Chem. Soc. Rev. 2013,
4
(
(
2, 8649−8682.
23) Ellman, G. L. Arch. Biochem. Biophys. 1959, 82, 70−77.
24) Punakivi, K.; Smolander, M.; Niku-Paavola, M. L.; Mattinen, J.;
Buchert, J. Talanta 2006, 68, 1040−1045.
(25) Wang, Z.; Liu, F.; Lu, C. Biosens. Bioelectron. 2014, 60, 237−243.
(26) Ueno, T.; Nagano, T. Nat. Methods 2011, 8, 642−645.
(27) Weigele, M.; DeBernardo, S. L.; Tengi, J. P.; Leimgruber, W. J.
Am. Chem. Soc. 1972, 94, 5927−5928.
28) Udenfriend, S.; Stein, S.; Bohlen, P.; Dairman, W.; Leimgruber,
W.; Weigele, M. Science 1972, 178, 871.
29) Rochat, S.; Swager, T. M. Angew. Chem., Int. Ed. 2014, 53,
792−9796.
30) Venkateswarulu, M.; Gaur, P.; Koner, R. R. Sens. Actuators, B
015, 210, 144−148.
31) Mani, P.; Ojha, A. A.; Reddy, V. S.; Mandal, S. Inorg. Chem.
(
̈
(
9
(
2
(
2
(
(
(
017, 56, 6772−6775.
32) Secor, K. E.; Glass, T. E. Org. Lett. 2004, 6, 3727−3730.
33) Mohr, G. J. Dyes Pigm. 2004, 62, 77−81.
34) Mertz, E.; Beil, J. B.; Zimmerman, S. C. Org. Lett. 2003, 5,
3
127−3130.
35) Shao, N.; Jin, J. Y.; Cheung, S. M.; Yang, R. H.; Chan, W. H.;
Mo, T. Angew. Chem., Int. Ed. 2006, 45, 4944−4948.
36) Shiraishi, Y.; Adachi, K.; Itoh, M.; Hirai, T. Org. Lett. 2009, 11,
482−3485.
37) Shiraishi, Y.; Yamamoto, K.; Sumiya, S.; Hirai, T. Phys. Chem.
Chem. Phys. 2014, 16, 12137−12142.
38) Valderrey, V.; Bonasera, A.; Fredrich, S.; Hecht, S. Angew.
Chem., Int. Ed. 2017, 56, 1914−1918.
39) Hanazawa, M.; Sumiya, R.; Horikawa, Y.; Irie, M. J. Chem. Soc.,
Chem. Commun. 1992, 206−207.
40) Kathan, M.; Kovarícek, P.; Jurissek, C.; Senf, A.; Dallmann, A.;
Thunemann, A. F.; Hecht, S. Angew. Chem., Int. Ed. 2016, 55, 13882−
3886.
41) Ho, T. I.; Wu, J. Y.; Wang, S. L. Angew. Chem., Int. Ed. 1999, 38,
558−2560.
42) Kobatake, S.; Imagawa, H.; Nakatani, H.; Nakashima, S. New J.
Chem. 2009, 33, 1362−1367.
43) Lvov, A. G.; Shirinian, V. Z.; Kachala, V. V.; Kavun, A. M.;
Zavarzin, I. V.; Krayushkin, M. M. Org. Lett. 2014, 16, 4532−4535.
44) Lvov, A. G.; Shirinian, V. Z.; Zakharov, A. V.; Krayushkin, M.
M.; Kachala, V. V.; Zavarzin, I. V. J. Org. Chem. 2015, 80, 11491−
1500.
45) Galangau, O.; Nakashima, T.; Maurel, F.; Kawai, T. Chem. - Eur.
J. 2015, 21, 8471−8482.
46) Lvov, A. G.; Shirinyan, V. Z. Chem. Heterocycl. Compd. 2016, 52,
(
(
3
(
(
(
(
̌
̌
̈
1
(
2
(
(
(
1
(
(
6
(
(
58−665.
47) Shoji, H.; Kobatake, S. Chem. Commun. 2013, 49, 2362−2364.
48) Shoji, H.; Kitagawa, D.; Kobatake, S. New J. Chem. 2014, 38,
9
(
(
33−941.
49) Kitagawa, D.; Kobatake, S. Chem. Lett. 2011, 40, 93−95.
50) Delbaere, S.; Berthet, J.; Shiozawa, T.; Yokoyama, Y. J. Org.
Chem. 2012, 77, 1853−1859.
I
DOI: 10.1021/jacs.8b02982
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX