Comprehensive chirality sensing: Development of stereodynamic probes with a Dual (chir)optical response

Article abstract of DOI:10.1021/jo500959y

The attachment of a salicylaldehyde ring and a cofacial aryl or heteroaryl N-oxide chromophore onto a naphthalene scaffold affords stereodynamic probes designed to rapidly bind amines, amino alcohols, or amino acids and to translate this binding event via substrate-to-receptor chirality amplification into a dual (chir)optical response. 1-(3′-Formyl-4′-hydroxyphenyl)-8- (9′-anthryl)naphthalene (1) was prepared via two consecutive Suzuki cross-coupling reactions, and the three-dimensional structure and racemization kinetics were studied by crystallography and dynamic HPLC. This probe proved successful for chirality sensing of several compounds, but in situ IR monitoring of the condensation reaction between the salicylaldehyde moiety in 1 and phenylglycinol showed that the imine formation takes 2 h. Optimization of the substrate binding rate and the circular dichroism (CD) and fluorescence readouts led to the replacement of anthracene with smaller fluorophores capable of intramolecular hydrogen bonding. 1-(3′-Formyl-4′-methoxyphenyl)-8- (4′-isoquinolyl)naphthalene N-oxide (2) and its pyridyl analogue 3 combine fast substrate binding with distinctive chiral amplification. This asymmetric transformation of the first kind prompts CD and fluorescence responses that can be used for in situ determination of the absolute configuration, ee, and total concentration of many compounds. The general utility of the three chemosensors was successfully tested on 18 substrates.

Full text of DOI:10.1021/jo500959y

Article
pubs.acs.org/joc
Comprehensive Chirality Sensing: Development of Stereodynamic
Probes with a Dual (Chir)optical Response
Keith W. Bentley and Christian Wolf*
Department of Chemistry, Georgetown University, Washington, D.C. 20057, United States
S
* Supporting Information
ABSTRACT: The attachment of a salicylaldehyde ring and a cofacial aryl or heteroaryl N-oxide chromophore onto a
naphthalene scaffold affords stereodynamic probes designed to rapidly bind amines, amino alcohols, or amino acids and to
translate this binding event via substrate-to-receptor chirality amplification into a dual (chir)optical response. 1-(3′-Formyl-4′-
hydroxyphenyl)-8-(9′-anthryl)naphthalene (1) was prepared via two consecutive Suzuki cross-coupling reactions, and the three-
dimensional structure and racemization kinetics were studied by crystallography and dynamic HPLC. This probe proved
successful for chirality sensing of several compounds, but in situ IR monitoring of the condensation reaction between the
salicylaldehyde moiety in 1 and phenylglycinol showed that the imine formation takes 2 h. Optimization of the substrate binding
rate and the circular dichroism (CD) and fluorescence readouts led to the replacement of anthracene with smaller fluorophores
capable of intramolecular hydrogen bonding. 1-(3′-Formyl-4′-methoxyphenyl)-8-(4′-isoquinolyl)naphthalene N-oxide (2) and its
pyridyl analogue 3 combine fast substrate binding with distinctive chiral amplification. This asymmetric transformation of the first
kind prompts CD and fluorescence responses that can be used for in situ determination of the absolute configuration, ee, and
total concentration of many compounds. The general utility of the three chemosensors was successfully tested on 18 substrates.
porphyrin receptors exhibiting induced circular dichroism (CD)
can be used for the determination of the absolute configuration
of chiral compounds.2 Since then, a variety of stereodynamic
chirality sensors have been introduced by Rosini,3 Anslyn,4
Canary,4d,e,5 Borhan,6 Toniolo,7 Gawronski,8 our group,9 and
others.10
The simultaneous determination of the absolute configuration,
concentration, and enantiomeric composition of a chiral
compound with an optical sensor is a challenging task. Our
efforts to develop a probe that can accomplish such a complete
stereochemical analysis have directed us to a dual sensing mode
design that can provide an instant CD signal and either a UV or a
fluorescence readout upon detection of a chiral substrate.9i As is
shown in Scheme 1, a stereodynamic receptor moiety and a
chromophore are connected in close proximity to a rigid scaffold
to establish communication between the two units. The
underlying idea is to couple a fast substrate recognition event
INTRODUCTION
■
The amplification and sensing of molecular chirality play an
increasingly important role in the advancement of asymmetric
catalysis, biomedical treatments and diagnostics, material
sciences, and other fields. The synthesis and stereochemical
analysis of chiral compounds have developed into rapidly
growing, mutually dependent fields that are of vital importance
throughout the chemical sciences. The general availability of
numerous chiral catalysts that can be tested in an enantioselective
reaction together with the challenge to assess unpredictable
effects of solvents, counterions, temperature, additives, and other
reaction parameters on the chemical yield and stereoselectivity
has directed increasing attention to high-throughput screening
(HTS) efforts. As a result, progress in method development
often relies on automated parallel synthesis that quickly
generates hundreds of samples, while the analytical task, that is,
the determination of the reaction yield and product ee, typically
remains laborious and time-consuming. Chiroptical assays have
been considered particularly promising to address this bottle-
neck.1 Nakanishi and Berova showed more than 20 years ago that
Received: May 1, 2014
Published: June 17, 2014
© 2014 American Chemical Society
6517
dx.doi.org/10.1021/jo500959y | J. Org. Chem. 2014, 79, 6517−6531

The Journal of Organic Chemistry
Article
Scheme 1. Illustration of the Dual Sensing Mode Strategy
Figure 1. Top: Structures of chemosensors 1−3 exhibiting a salicylaldehyde unit for ee determination and a proximate chromophore for total
concentration analysis. Bottom: Imprinting of the chirality of amine substrates on the stereodynamic scaffold of sensor 1.
a
Scheme 2. Synthesis of 1 and Crystal Structures of Intermediates 4 and 5
a
Selected crystallographic measurements of 4: C′1−Br separation, 3.108 Å; twisting angle, 24.4°; splaying angle, 10.0°. Selected crystallographic
measurements of 5: phenylcentroid−anthrylcentroid, 3.311 Å; twisting angle, 23.6°; splaying angle, 8.0°.
with a chiral amplification process at the receptor site. In the
absence of a chiral bias, the probe exhibiting a stereolabile chiral
axis exists as a mixture of rapidly interconverting stereoisomers
and remains CD-silent. Binding of a chiral substrate at the
stereodynamic receptor unit and subsequent chirality imprinting
disturb this equilibrium and favor population of an axially chiral
conformation. This could be measured by CD analysis and thus
provide an entry to the determination of the absolute
configuration and ee of the substrate. Communication of the
binding event between the receptor part and the adjacent
chromophore would concomitantly affect the UV or fluorescence
properties of the chemosensor. This change would be
independent of the substrate chirality and allow nonstereose-
lective sensing of the total substrate amount. To become practical
for HTS purposes, the processes described above have to occur
within a few minutes and generate strong (chir)optical responses
that can be accurately quantified.
RESULTS AND DISCUSSION
■
We envisioned that a salicylaldehyde moiety connected via a
stereolabile chiral axis to a naphthalene scaffold could function
both as a receptor for a wide range of important chiral
compounds carrying an amino group (amines, amino alcohols,
and amino acids) and as the stereodynamic reporter unit (Figure
1). The acidic phenol function adjacent to the formyl group in
the salicaldehyde ring was expected to accelerate the substrate
binding process and allow for time-efficient chemosensing. The
formation of a rigid imine product would then induce an
asymmetric transformation of the first kind, generating a
stereochemical bias at the chiral aryl−aryl axis in the sensor
and a distinct CD signal. For example, the binding of an R-
configured amine would favor population of the R,P conformer
to avoid repulsion between the steric bulk of the bound substrate
and the proximate chromophore. This central-to-axial chirality
induction process thus disturbs the conformational equilibrium
of the stereodynamic sensor and affords a quantifiable CD
6518
dx.doi.org/10.1021/jo500959y | J. Org. Chem. 2014, 79, 6517−6531

The Journal of Organic Chemistry
Article
Figure 2. Top: Experimentally obtained HPLC chromatograms of compound 5 at four different temperatures. HPLC conditions: (S,S)-Whelk-O 1 (25
cm × 4.6 mm); mobile phase, CH2Cl2/hexanes (1:1); flow rate, 1 mL/min; UV detection, 254 nm. Bottom: Simulation of the elution profiles calculated
with Mimesis 3.1 and the Eyring plot, which yielded ΔS⧧ = −120.5 J mol−1 K−1, ΔH⧧ = 49.9 kJ/mol, and ΔG⧧° = 85.8 kJ/mol at 298.15 K.
response that is ultimately controlled by the bound substrate.
The consideration of a large chromophoric group as the second
reporter unit led to the structure of 1. The placement of
anthracene at the neighboring peri position in the naphthalene
ring was expected to enforce a cofacial orientation and π−π
interactions between the two reporter units. As a result, the
substrate recognition would be communicated from the
salicylidenimine to the anthracene moiety via steric, dipole,
and π-stacking interactions and thus alter the UV and
fluorescence output of the sensor. This change would be
independent of the chirality of the bound amino compound and
afford a means to determine the total substrate amount. We now
describe the development of dual mode sensors 1−3 and their
use for the simultaneous determination of the absolute
configuration, ee, and amount of a wide variety of unprotected
chiral amines, amino alcohols, and amino acids.
Synthesis, Racemization Kinetics, and Chirality Sens-
ing with Chemosensor 1. The synthesis of 1-(3′-formyl-4′-
hydroxyphenyl)-8-(9′-anthryl)naphthalene (1) required careful
optimization of two consecutive cross-coupling reactions
(Scheme 2). Initial attempts to prepare 1-(3′-formyl-4′-
methoxyphenyl)-8-bromonaphthalene (4) from 3-formyl-4-
methoxyphenylboronic acid and 1,8-dibromonaphthalene by
means of typical Suzuki coupling protocols gave low yields and
large amounts of the debrominated and disubstituted derivatives.
In an effort to improve the selectivity toward 4, we carefully
screened various palladium catalysts and optimized the solvent
composition, base, temperature, and reaction time. We finally
found that 4 can be obtained in 70% yield when the reaction is
performed with 7.5 mol % Pd(PPh3)4 and 5 equiv of K3PO4 in a
toluene/ethanol/water (3:2:1) mixture at 80 °C for 4 h. When
the subsequent reaction of 4 and anthracene-9-boronic acid was
carried out under the same conditions, only debromination was
observed. However, several changes, including an increase in the
catalyst loading, reaction time, and temperature, finally afforded
the sterically crowded triaryl 5 carrying two cofacial aryl rings
perpendicular to the naphthalene scaffold.11 To our surprise,
several attempts to remove the methoxy group following a range
of literature protocols with BBr3,9a MgCl2,12 CeCl3,13 and
AlCl314 were unsuccessful. We were pleased to find that the use of
LiCl in refluxing DMF produces 1 in 60% yield.15
The difficulties encountered during the synthesis of 4 and 5 are
a result of the steric strain and repulsion between the two peri
substituents. Crystallographic analysis of 4 reveals that the
salicylaldehyde unit and the bromine are forced out of the
naphthalene plane and spread apart from one another (Scheme
2). The twisting angle between the naphthalene−salicylaldehyde
bond and the naphthalene−bromine bond was determined to be
24.4°. In addition, the two bonds are splayed by 10.0°. While the
two aryl units in 5 are cofacial and perpendicular to the
naphthalene plane as expected, the crystal structure shows
twisting and a similar splaying as observed with 4. The proximate
anthracene ring also reduces the accessibility of the methoxy
group on the adjacent salicylaldehyde ring and therefore severely
hampers the demethylation reaction unless severe conditions are
applied.
Having established synthetic access to the sterically crowded
triaryl arrangement of 1 as well as crystallographic information
about the three-dimensional structure, we continued with the
determination of the activation free energy, ΔG⧧, for the rotation
of the salicylaldehyde unit about the chiral aryl−aryl axis. We
were unable to separate the rotational isomers of 1 by HPLC
using a variety of chiral columns even at very low temperatures.
However, the enantiomers of precursor 5 were resolved on the
(S,S)-Whelk-O 1 chiral stationary phase using dichloromethane/
hexanes (1:1) as the mobile phase (Figure 2). At −20 °C, the
enantiomers have retention times of 10.3 and 12.4 min, and a
small plateau between the peaks due to on-column racemization
was observed. As expected, the height of the plateau increased as
the temperature was raised stepwise to −5 °C. Simulation of the
elution profiles at four different temperatures provided the
corresponding enantiomerization rates obeying reversible first-
6519
dx.doi.org/10.1021/jo500959y | J. Org. Chem. 2014, 79, 6517−6531

The Journal of Organic Chemistry
Article
Figure 3. Top: Binding of amine substrates by sensor 1 and in situ IR analysis of the condensation reaction between 1 and 10. The disappearance of 1
was measured by the decrease in the absorbance intensity of the aldehyde stretching at 1668 cm−1 (red), and the generation of the imine was monitored
by the increase in the absorbance intensity of the imine stretching at 1635 cm−1 (blue). Bottom left: (a) CD spectra of the imines obtained from 1 and
(R)-10 (blue) and (S)-10 (red). (b) CD spectra of the imines obtained from 1 and (R)-18 (blue) and (S)-18 (red). (c) CD spectra of the imines
obtained from 1 and (R)-21 (blue) and (S)-21 (red). All of the spectra were collected at 7.5 × 10−5 M in MeOH (10 and 18) or CHCl3 (21). Bottom
right: Substrate scope of sensor 1. Only one enantiomer is shown.
order kinetics. Eyring plot analysis gave an activation enthalpy,
ΔH⧧, of 49.9 kJ/mol and an activation entropy, ΔS⧧, of −120.5 J
mol−1 K−1. The strikingly negative ΔS⧧ value corresponds well to
a highly ordered T-shaped transition state in which the
salicylaldehyde ring rotates through the naphthalene plane and
is perpendicular to the adjacent anthracene moiety.16 The
standard activation free energy, ΔG⧧°, was calculated as 85.8 kJ/
mol, which demonstrates that 5 and consequently the
demethylated analogue 1 undergo rapid racemization at room
temperature.
We envisioned that fixation of a chiral amine or amino alcohol
by the salicylaldehyde unit would generate a rigid salicylideni-
mine motif stabilized by intramolecular hydrogen bonding. On
the basis of the stereodynamic nature of 1, that is, the rapid
interconversion of the enantiomeric rotamers at room temper-
ature, this would result in a spontaneous asymmetric trans-
formation of the first kind favoring an axially chiral conformer
that carries the bulkiest group of the bound substrate pointing
away from the neighboring anthracene ring to occupy the least
sterically hindered space. The imprinting of the substrate
chirality on the conformational equilibrium of 1 was thus
expected to generate a predominant sense of axial chirality and a
characteristic CD response of the chemosensor. Despite the
proximity of the acidic phenol moiety, the condensation reaction
of 1 and phenylglycinol (10) at room temperature required 24 h
to come to completion according to 1H NMR analysis. However,
the addition of 10 mol % trifluoroacetic acid to an equimolar
mixture of 1 and 10 in chloroform significantly reduced the
reaction time (Figure 3).17 In-situ IR measurements showed that
the formyl stretching of the sensor (1668 cm−1) rapidly decreases
and almost completely disappears within 2 h while the imine
stretching of the condensation product (1635 cm−1) reaches a
maximum intensity. These results were confirmed by NMR and
MS analysis (see the Supporting Information).
exhibiting CD signals at low wavelengths. In accordance with a
previous study,9a we believe that the CD responses observed with
sensor 1 originate from the relative orientation of the two cofacial
chromophores, which are almost perfectly perpendicular to the
naphthalene plane. A variety of amino alcohols were then tested
to evaluate the general utility of 1 for enantioselective sensing. All
of the condensation reactions were conducted in chloroform in
the presence of catalytic amounts of trifluoroacetic acid and
stirred for 5 h prior to CD analysis. The CD spectra were
collected at micromolar concentrations, with alcoholic solvents
such as methanol, ethanol, and isopropanol giving the strongest
signals (see the Experimental Section). The imines derived from
amino alcohols 9−12 generally exhibit strong CD signals above
250 nm, and the absolute configurations can be correlated to the
CD spectra, with the R configuration generating a negative
maximum at 270 nm and the S configuration generating a
positive CD response at the same wavelength (Table 1 and the
Supporting Information). Encouraged by the success with
chemosensing of amino alcohols, we decided to pursue other
classes of compounds bearing an amino function, namely, amines
17−20 and amino acids 21−24 (Figure 3). While the amine
sensing was conducted as described above, the condensations
with amino acids were performed in DMSO in the presence of 1
equiv of tetrabutylammonium hydroxide (TBAOH) to improve
the solubility of the substrates. Again, characteristic Cotton
effects were observed using the stereodynamic probe 1 for
chirality sensing of a variety of aliphatic and aromatic amines and
amino acids. A close comparison of the CD spectra reveals that all
12 substrates representing important amines, amino alcohols,
and amino acids consistently generate a similar chiroptical
response of 1, namely, a strong positive CD signal at 270 nm
when a substrate with the S configuration is used and the
opposite (negative) Cotton effect when the corresponding R
enantiomer is used (entries 1−12 in Table 1).
The CD spectra of the imines obtained from 1 and the
enantiomers of 10 are shown in Figure 3. As expected, the
stereodynamic chemosensor shows a strong chiroptical signal
upon binding of the substrate, which is CD-silent in the absence
of 1 under otherwise the same conditions. It is known that
salicylidenimines can exist in the form of a quinoid-like structure
As mentioned above, the determination of the absolute
configuration, ee, and total concentration of chiral compounds
with a single chemosensor would afford a practical entry to a
complete stereochemical analysis and be most valuable for high-
throughput screening applications. Since the chiroptical studies
conducted with chemosensor 1 established its general suitability
6520
dx.doi.org/10.1021/jo500959y | J. Org. Chem. 2014, 79, 6517−6531

The Journal of Organic Chemistry
Article
Table 1. Summary of the CD Sensing Results with Chemosensors 1, 2, and 3
a
b
a
b
entry sensor substrate class
substrate
Δ
(mdeg) predicted sign
entry sensor substrate class
substrate
Δ
(mdeg) predicted sign
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
2
2
2
2
AA
AA
AA
AA
AA
AA
AA
AA
MA
MA
MA
MA
MA
MA
MA
MA
AC
AC
AC
AC
AC
AC
AC
AC
AA
AA
AA
AA
AA
AA
AA
AA
AA
AA
AA
AA
AA
AA
(1S,2R)-9
(1R,2S)-9
(R)-10
−49
+55
−48
+48
−7
−
+
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
AA
AA
AC
AC
AC
AC
AC
AC
AC
AC
AA
AA
AA
AA
AA
AA
AA
AA
AA
AA
AA
AA
AA
AA
AA
AA
AA
AA
AC
AC
AC
AC
AC
AC
AC
AC
(R)-16
+16
−15
+19
−15
+21
−27
+19
−16
+8
+
−
+
(S)-16
2
−
+
(R)-21
(S)-10
(S)-21
−
+
3
(1R,2R)-11
(1S,2S)-11
(1S,2R)-12
(1R,2S)-12
(R)-17
−
+
(R)-22
+7
(S)-22
−
+
4
−9
−
+
(R)-23
+12
−12
+11
−15
+15
−9
(S)-23
−
+
5
−
+
(R)-24
(S)-17
(S)-24
−11
+29
−33
+77
−78
+27
−27
+61
−63
+18
−17
+57
−47
+55
−53
+63
−66
+20
−21
+22
−22
+36
−42
+51
−42
+22
−22
−
+
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
6
(R)-18
−
+
(1S,2R)-9
(1R,2S)-9
(R)-10
(S)-18
−
+
7
(R)-19
−
+
(S)-19
+10
(S)-10
−
+
8
(R)-20
−5
+6
−
+
(1R,2R)-11
(1S,2S)-11
(1S,2R)-12
(1R,2S)-12
(R)-13
(S)-20
−
+
9
(R)-21
−19
+20
−10
+11
−24
+24
−10
+12
+13
−13
+13
−9
−
+
(S)-21
−
+
10
11
12
13
14
15
16
17
18
19
(R)-22
−
+
(S)-22
(S)-13
−
+
(R)-23
−
+
(R)-14
(S)-23
(S)-14
−
+
(R)-24
−
+
(R)-15
(S)-24
(S)-15
−
+
(1S,2R)-9
(1R,2S)-9
(R)-10
+
(R)-16
−
+
(S)-16
−
+
(1R,2R)-28
(1S,2S)-28
(R)-21
(S)-10
−
+
−
+
(1R,2R)-11
(1S,2S)-11
(1S,2R)-12
(1R,2S)-12
(R)-13
+8
−8
+3
−
+
(S)-21
−
+
(R)-22
−3
−
+
(S)-22
−
+
+19
−19
+12
−14
+13
−13
(R)-23
(S)-13
−
+
(S)-23
−
+
(R)-14
(R)-29
(S)-14
−
+
(S)-29
−
(R)-15
(S)-15
−
a
b
CD output at 270 nm for sensor 1 and 260 nm for sensors 2 and 3. The CD response to 28 was measured at 280 nm. Predicted CD sign at 270
nm for sensor 1, where R is negative and S is positive. Analysis of the CD readout at 260 nm for sensors 2 and 3 shows that R enantiomers give a
c
positive sign and S substrates give a negative CD response for all substrates. MA = monoamine, AA = amino alcohol, AC = amino acid. Reference 9i.
for enantioselective detection and possibly quantitative ee
analysis of amino alcohols, amines, and amino acids, we decided
to explore whether a change in the UV or fluorescence signal of 1
upon substrate binding would provide a means to determine the
substrate concentration independent of the enantiomeric
composition. Because the conversion of the salicylaldehyde to
a salicylidenimine moiety ultimately alters its steric bulk, dipole
moment, and π-stacking interactions, we expected that the
substrate binding event would affect not only the conformational
bias of the chemosensor but also the UV and fluorescence
properties of the neighboring anthracene ring.18 In addition to
the chiroptical sensing of the ee, this would then enable one to
determine the total concentration of the substrate directly from
the optical response of 1. Unfortunately, only a small increase in
the UV and fluorescence signals of 1 was observed upon imine
formation, and the changes were insufficient for quantitative
analysis (see the Supporting Information).
Although the enantioselective sensing of various chiral
compounds with 1 was quite encouraging, we concluded that
the incorporation of a fluorophore that (a) provides less steric
hindrance to the substrate fixation and (b) has a strong
interaction with a bound substrate that results in a distinct and
6521
dx.doi.org/10.1021/jo500959y | J. Org. Chem. 2014, 79, 6517−6531

The Journal of Organic Chemistry
Article
a
Scheme 3. Synthesis of Sensor 2 and Crystal Structure of Intermediate 6
a
Selected crystallographic measurements for 6: through-space C4′−Br separation, 3.073 Å; twisting angle, 9.5°; splaying angle, 8.6°.
Figure 4. Top: Amino alcohol detection with sensor 2. In-situ IR analysis of the condensation reaction between 2 and 10 and changes in the absorbance
of the aldehyde stretching at 1680 cm−1 (red) and the imine stretching at 1640 cm−1 (blue) are shown. Bottom left: (a) CD spectra of the imines
obtained from 2 and (1R,2R)-11 (blue) and (1S,2S)-11 (red). (b) CD spectra of the imines obtained from 2 and (R)-22 (blue) and (S)-22 (red).
Bottom right: Substrate scope of sensor 2. Only one enantiomer is shown.
quantifiable fluorescence response would afford a more useful
chemosensor. We assumed that a fluorophore smaller than
anthracene would not interfere with the imine formation step,
thus significantly reducing the reaction time. The incorporation
of a hydrogen-bond acceptor site in the fluorophore was
expected to affect both the asymmetric induction process and the
fluorescence response to the substrate binding event. These
considerations led to the structure of sensor 2 carrying a
salicylaldehyde ring for ee analysis and an isoquinoline N-oxide, a
moderately sized fluorophore and hydrogen-bond acceptor, for
total concentration analysis.
Determination of the Absolute Configuration, Enan-
tiomeric Excess, and Total Amount of Chiral Compounds
with Sensor 2. The experience obtained from the construction
of the sterically crowded scaffold of 1 greatly facilitated the
synthesis of 2. In this case, however, we realized that the
fluorophore had to be introduced first, followed by the
attachment of the O-methylated salicylaldehyde unit (Scheme
3). The Suzuki cross-coupling of 1,8-dibromonaphthalene and 4-
isoquinolineboronic acid largely left the second carbon−bromide
bond intact and gave 6 in 80% yield. Oxidation with m-CPBA in
dichloromethane afforded 7, and the second Suzuki coupling
with 3-formyl-4-methoxyphenylboronic acid provided 8 in 70%
yield. The yields of both coupling steps were significantly higher
than the aryl−aryl bond formations leading to compound 1,
which can be attributed to the reduced steric hindrance. The
crystal structure of 6 also indicates a significantly smaller twisting
angle between the bromine and the isoquinolyl ring. The
demethylation with BBr3 proceeded smoothly to give 2 under
relatively mild conditions.
In-situ IR analysis of the condensation reaction between 2 and
amino alcohol 10 showed that the imine formation proceeds
rapidly, as indicated by the disappearance of the formyl stretching
at 1680 cm−1 and the increase in the imine stretching at 1640
cm−1 (Figure 4). NMR and MS analysis confirmed complete
conversion of the starting material within 20 min even in the
absence of an acid catalyst. It is noteworthy that the rate of the
condensation reaction between 10 and 2 was found to be greater
than that with free salicylaldehyde under similar conditions,
which required 2 h for complete conversion to the imine (see the
Supporting Information). On the basis of the faster substrate
binding with the isoquinoline N-oxide-derived probe 2, we
concluded that it would be a much more practical chemosensor
than the sterically hindered prototype 1, warranting a closer
investigation of its (chir)optical sensing potential.
The CD spectra of the imines obtained from 2 and a wide
variety of unprotected amino alcohols 9−16 and amino acids
21−24 were collected (Figure 4). The condensation reactions
were carried out in chloroform, and the CD analyses were
performed with diluted samples in hexanes. In contrast to the
results obtained with sensor 1, the CD amplitudes were
diminished in alcoholic solvents such as methanol and ethanol.
This is in perfect agreement with the anticipated intramolecular
hydrogen-bonding motif between the alcohol or carboxylic acid
moiety of the substrate and the N-oxide of the isoquinolyl unit in
the chemosensor. The absolute configuration of the substrates
can be consistently correlated with the sign of the CD signal at
260 nm, with R enantiomers generating a positive CD maximum
and S enantiomers a negative one (entries 13−24 in Table 1).
6522
dx.doi.org/10.1021/jo500959y | J. Org. Chem. 2014, 79, 6517−6531

The Journal of Organic Chemistry
Article
Figure 5. (a) Fluorescence spectra of 2 obtained upon addition of up to 1 equiv of 10 (blue) and in the presence of an excess of the amino alcohol (red).
(b) Plot of the fluorescence intensity at 515 nm vs the [10]/[2] ratio. (c) Plots of the CD maxima of the phenylglycinol-derived imine of 2 at 275 nm
(blue), 295 nm (red), and 325 nm (green) vs the % ee of substrate 10. (d) Plots of the CD maxima of the tryptophan-derived imine at 275 nm (blue) and
295 nm (red) vs the % ee of 22.
a
Table 2. Quantitative Enantiomeric Excess and Concentration Sensing of Substrates 10 and 22
chemosensing results
phenylglycinol (10)
tryptophan (22)
sample composition
ee (%)
ee (%)
entry conc (mM) Ee (%) abs config conc (mM) 275 nm 295 nm 325 nm
avg
abs config entry 260 nm 290 nm
avg
abs config
1
2
3
4
0.63
1.34
2.67
3.23
87.0
76.0
68.0
89.0
R
R
S
0.63
1.30
2.62
3.38
92.0
62.8
65.4
85.6
96.6
77.1
66.3
82.0
87.1
77.7
59.5
92.7
91.9
72.5
63.7
86.8
R
R
S
5
6
7
8
84.1
71.4
67.8
90.4
86.4
73.5
67.7
89.4
85.3
72.5
67.8
89.9
R
R
S
S
S
S
a
See the Supporting Information for details.
a
Scheme 4. Asymmetric Transformation of the First Kind of the Salicylidenimine Obtained from Amino Alcohols and Sensor 2
a
The conformation that is stabilized by intramolecular hydrogen bonding and has the least steric repulsion is shown in blue.
In order to test the usefulness of 2 for ee determination of
amino alcohols, CD spectra of the imines obtained with
nonracemic samples of 10 covering a wide ee range were
collected. Calibration curves were then constructed by plotting
the intensity of the Cotton effects at three different wavelengths
versus the substrate ee (Figure 5). On the basis of the linear
regression equations derived from the calibration curves, the
enantiomeric compositions of four unknown samples were
determined (Table 2, entries 1−4). The average of the three
values obtained from the CD readouts at 275, 295, and 325 nm
gave the corresponding ee’s with an accuracy that compares well
with previously reported enantioselective chemosensing assays
6523
dx.doi.org/10.1021/jo500959y | J. Org. Chem. 2014, 79, 6517−6531

The Journal of Organic Chemistry
Article
and is sufficient for HTS purposes. Sensor 2 was also used
successfully for ee determination of tryptophan (22). In this case,
calibration curves were constructed for two different wave-
lengths, and the ee’s of the unknown samples were calculated
from the averages (Table 2, entries 5−8).
We were very pleased to find that 2 can also be used for the
determination of the concentration of amino alcohols, which is
another important advantage over chemosensor 1. The
fluorescence outputs of the imine obtained from 2 and 10
using varying amounts of 10 were measured. A steady
fluorescence enhancement was observed upon addition of up
to 1.0 molar equiv of the substrate, and the intensity remained
nearly constant in the presence of excess 10 (Figure 5). The 6-
fold increase in the fluorescence of 2 upon binding of 10 proved
to be independent of the substrate chirality and allowed accurate
determination of the concentration of the four nonracemic
samples tested (Table 2). For example, the chemosensing of a
0.63 mM solution of 10 having 87.0% ee gave virtually the same
concentration and 91.9% ee (entry 1 in Table 2). Similar results
were obtained with the other samples, that is, the concentrations
were generally determined with less than 5% error and the ee’s
were within a 6% margin. It is noteworthy that the accuracy of the
ee and concentration analyses suffices for HTS purposes and
compares well with results generally obtained with chemosensing
assays.
Figure 6. Front view (left) and side view (right) of the X-ray structure of
the (1S,2R,M,M)-configured imine obtained from 2 and (1S,2R)-9.
Selected crystallographic measurements: N1−H1, 1.773 Å; phe-
nylcentroid−isoquinolylcentroid, 3.455 Å; twisting angle, 25.7°; splaying
angle, 10.1°.
A comparison of the chiroptical responses of chemosensors 1
and 2 reveals that the latter generates a lower CD readout upon
detection of the amino alcohols 9 and 10 (compare entries 1 and
2 with entries 13 and 14 in Table 1). This may be attributed to
the coexistence of several diastereomeric isomers, each having
individual and possibly opposite CD effects. The free chemo-
sensor 2 exists as a mixture of two diastereomeric pairs of
enantiomers that undergo fast interconversion through rotation
of both the salicylaldehyde and isoquinolyl N-oxide moieties
about the chiral aryl−aryl bonds (Scheme 4). NMR analysis
revealed the presence of diastereomeric anti and syn isomers in a
7:3 ratio, each being a racemic mixture in the absence of a chiral
bias. Upon substrate binding, this equilibrium is disturbed, with
the intramolecular hydrogen-bonding motif between the N-oxide
and the alcohol of the substrate as well as steric effects favoring
one conformer. For example, chirality imprinting with an (R)-
amino alcohol is expected to favor population of the syn-(R,M,P)
diastereomer, placing the steric bulk of the substrate away from
the isoquinolyl N-oxide moiety and in the least sterically
hindered direction, although other stereoisomers may also be
present. Accordingly, condensation of 2 with an (S)-amino
alcohol would favor the enantiomeric syn-(S,P,M) conformation.
The coexistence of diastereomeric conformations with
individual chiroptical properties, however, ultimately limits the
overall CD intensity. In fact, we found that the intramolecular
hydrogen-bonding motif in the imine derived from 9 does not
prevail in the solid state. Single crystals of the condensation
product obtained from 2 and substrate (1S,2R)-9 were grown by
slow diffusion of hexanes into a concentrated chloroform
solution (Figure 6). Crystallographic analysis revealed an anti
orientation of the cofacial aryl rings, which places the steric bulk
of the substrate opposite to the isoquinolyl moiety and results in
a (1S,2R,M,M) configuration. The crystallization of this
particular conformer may be a result of an asymmetric
transformation of the second kind due to solid-state packing
forces, and the (1S,2R,M,P) conformer exhibiting a hydrogen
bond between the alcohol group of 9 and the N-oxide of 2 is still
likely to coexist and possibly predominate in solution.
Development of Stereodynamic Sensor 3. The fast
chirality sensing and accurate ee and concentration analysis of a
variety of chiral compounds with probe 2 led to the design of 1-
(3′-formyl-4′-methoxyphenyl)-8-(4′-pyridyl)naphthalene N-
oxide (3). Sensor 3 combines the stereodynamic and (chir)-
optical properties of 2 while exhibiting only one chiral axis, which
reduces the number of possible diastereomeric conformations of
the corresponding imine derivatives. We therefore anticipated
that 3 would be as useful as 2 in enantioselective chemosensing
applications and at the same time be more readily available and
simplify the analysis of the chiral induction process. The
synthesis of 3 in four steps and its use for sensing of amino
alcohols were recently reported by our laboratory (Scheme 5).9i
Chiroptical and crystallographic analysis of the condensation
products obtained from 3 and amino alcohols provided evidence
for a distinct asymmetric transformation of the first kind that is
consistent with the chiral amplification model described above
for sensor 2. We found that the imine formation locks the
stereodynamic scaffold of 3 into a single conformer that is
stabilized by intramolecular hydrogen bonding between the
substrate alcohol group and the pyridyl N-oxide moiety while the
steric bulk of the bound imino alcohol moiety is placed into the
least crowded direction (Scheme 5). This imprinting of the
substrate chirality on the sensor structure occurs with excellent
stereodivergence. Chiral amplification with an (R)-amino alcohol
produces the M conformation and a positive CD response at 260
nm, while the S antipode leads to the corresponding S,P
enantiomer and a negative CD maximum at the same wavelength
(entries 25−33 in Table 1).
We now report that the chirality sensing scope of 3 extends to
unprotected aliphatic and aromatic amino acids and that we
observed a characteristic CD response at high wavelength that
can be exploited for comprehensive chirality sensing (Figure
7).19 In analogy to the substrate recognition capabilities of 2, the
condensation reaction using 3 as the probe is significantly faster
than that with the bulky anthryl derivative 1. In-situ IR, NMR,
and MS analysis revealed that the substrate binding with 3 is
complete within 15 min even in the absence of an acid catalyst,
6524
dx.doi.org/10.1021/jo500959y | J. Org. Chem. 2014, 79, 6517−6531

The Journal of Organic Chemistry
Article
Scheme 5. Synthesis of 3 and Chiral Amplification with an (R)-Amino Alcohol9i
Figure 7. Top: Stereodivergence of the chirality imprinting of amino acids on sensor 3. Middle: In-situ IR analysis of the condensation reaction between
3 and 10. The disappearance of 3 was measured by the decrease in the absorbance intensity of the aldehyde stretching at 1685 cm−1 (red), and the
generation of the imine was monitored by the increase in the absorbance intensity of the imine stretching at 1645 cm−1 (blue). Bottom left: (a) CD
spectra of the imines obtained from 3 and (R)-23 (blue) and (S)-23 (red). (b) CD spectra of the imines obtained from 3 and (R)-29 (blue) and (S)-29
(red). All of the spectra were collected at 7.5 × 10−5 M in CHCl3. Bottom right: Substrates tested with sensor 3. Only one enantiomer is shown.
which proves its suitability for HTS purposes (see the Supporting
Information). In accordance with the chiroptical sensing of
amino alcohols, a positive CD response at 270 nm was
consistently observed for R-configured amino acids, whereas
the S antipodes yielded negative CD amplitudes at the same
wavelength (entries 34−37 in Table 1). These results suggest
that the binding of amino acids is followed by the same
instantaneous chiral induction process. The binding of an (R)-
amino acid by 3 thus leads to a predominant population of the
R,M conformation, which is stabilized by intramolecular
hydrogen bonding while steric repulsion between the amino
acid residue and the sensor scaffold is kept at a minimum. The
formation of the M conformer of the salicylidenimine derivative
of 3 coincides with a positive CD response at 270 nm. Similarly,
binding of an S-configured substrate favors the corresponding S,P
enantiomer, which affords a negative CD maximum at the same
wavelength (Figure 7).
wavelengths and the substrate ee (Figure 8). In addition, the
fluorescence properties of the sensor changed significantly upon
substrate binding. We observed that the intensity of the
fluorescence signal of 3 at 450 nm steadily decreases in the
presence of up to 1 molar equiv of tyrosine but remains constant
when additional amounts of 23 are added. Interestingly, the
fluorescence response of sensors 2 and 3 appears to depend on
the class of compounds detected. As shown in Figure 5, the
binding of phenylglycinol by 2 results in fluorescence enhance-
ment, and a similar fluorescence change was obtained when
sensor 3 was treated with an amino alcohol.9i By contrast, the
binding of an amino acid causes fluorescence quenching. The
imine derivatives of 2 and 3 formed with either phenylalanine or
tyrosine show a remarkable decrease in the fluorescence intensity
compared to the free sensor (Figures 8 and 19−21). Using the
dual CD/fluorescence output of 3, we were thus able to identify
the absolute configuration of the major tyrosine enantiomer and
to simultaneously determine the enantiomeric excess and the
total amount of five samples covering wide concentration and ee
ranges (Table 3).
We then applied 3 in the chirality, ee, and concentration
analysis of tyrosine (23). Calibration curves constructed from the
chiroptical response of 3 to nonracemic samples of 23 showed a
linear relationship between the CD amplitudes at three different
6525
dx.doi.org/10.1021/jo500959y | J. Org. Chem. 2014, 79, 6517−6531

The Journal of Organic Chemistry
Article
Figure 9. (a) CD spectra of the imines obtained from 1 and (R)-24
(solid blue) and (S)-24 (solid red) in the presence of TBAOH and the
CD response upon addition of HCl (dashed lines). (b) CD spectra of
the imines obtained from 2 and (R)-22 (solid blue) and (S)-22 (solid
red) in the presence of TBAOH and the CD response upon addition of
HCl (dashed lines). (c) CD spectra of the imines obtained from 3 and
(R)-29 (solid blue) and (S)-29 (solid red) in the presence of TBAOH
and the CD response upon addition of HCl (dashed lines).
Figure 8. (a) Fluorescence spectra of 3 obtained upon addition of up to
1 equiv of 23 (blue) and in the presence of excess of the amino acid
(red). (b) Plot of the fluorescence intensity at 450 nm vs the [23]/[3]
ratio. (c) CD spectra of the imine obtained with nonracemic samples of
23. (d) Plots of the CD maxima at 274 nm (blue), 305 nm (green), and
335 nm (red) vs the % ee of tyrosine.
The CD sensing of amino acids with the triaryl probes 2 and 3
proved to be pH-sensitive, which is in accordance with the
general importance of intramolecular hydrogen bonding
between the bound substrate and the heteroaryl N-oxide units
(Figure 9 and the Experimental Section). While the CD response
of the imine obtained from 1 and 24 did not change upon
addition of 1 equiv of HCl, the intensities of the Cotton effects
measured for amino acid-derived condensation products of 2 and
3 increased significantly. In the presence of base, the free
carboxylate group of the bound amino acid cannot participate in
intramolecular hydrogen bonding with the neighboring hetero-
aryl N-oxide group. However, the generation of the free
carboxylic acid function upon acidification enables intra-
molecular hydrogen bonding with the N-oxide moiety of 2 and
3. As a result, a more distinct asymmetric transformation of the
first kind favors population of the conformer showing the
hydrogen-bonding motif, which ultimately leads to a stronger
CD effect (Figure 7).
process that favors population of an axially chiral sensor
conformation. This asymmetric transformation of the first kind
turns on a characteristic chiroptical response from the sensor
moiety that is correlated to the chirality and ee of the bound
substrate. The covalent fixation of the substrate at the
salicylaldehyde moiety is also recognized by the adjacent
chromophore, thus inducing a strong change in the fluorescence
output of the sensor. Because this second response is not
stereoselective, it can be used to quantify the total amount of the
substrate present in solution. This dual response sensor design
has several attractive features: The substrate binding is fast and
can be accelerated with catalytic amounts of acid if needed; the
(chir)optical readouts occur at high wavelengths, which avoids
interference with (chiral) impurities; micromolar sensor and
substrate concentrations suffice for accurate determination of ee
and concentration; the sensing method is operationally simple
and suitable to automation and high-throughput screening
applications; time-consuming product purification is not
necessary, which allows convenient in situ analysis; and solvent
waste production is reduced compared with chromatographic
methods.
CONCLUSION
■
The incorporation of a salicylaldehyde ring and a neighboring
aryl or heteroaryl chromophore at the peri positions of
naphthalene affords stereodynamic circular dichroism/fluores-
cence sensors that can be used for rapid determination of the
absolute configuration, enantiomeric excess, and total concen-
tration of a variety of amines, amino alcohols, and amino acids at
micromolar concentrations. The imine formation at the
salicylaldehyde unit initiates a spontaneous chiral amplification
EXPERIMENTAL SECTION
■
1. General Information. All of the reagents and solvents were
commercially available and used without further purification. Reactions
were carried out under an inert atmosphere and anhydrous conditions.
Flash chromatography was performed on silica gel (particle size 40−63
μm). NMR spectra were obtained at 400 MHz (1H NMR) and 100 MHz
a
Table 3. Quantitative Enantiomeric Excess and Concentration Sensing of Tyrosine (23)
chemosensing results
ee (%)
sample composition
entry
conc (mM)
ee (%)
abs config
conc (mM)
274 nm
303 nm
335 nm
avg
abs config
1
2
3
4
5
0.56
1.01
1.73
2.36
2.93
87.0
76.0
12.0
26.0
68.0
R
R
R
S
0.59
1.08
1.73
2.32
2.97
87.4
74.2
11.8
27.1
66.3
88.3
73.8
13.2
28.3
65.4
90.1
76.7
12.9
25.3
66.8
88.6
74.9
12.6
26.9
66.2
R
R
R
S
S
S
a
See the Supporting Information for details.
6526
dx.doi.org/10.1021/jo500959y | J. Org. Chem. 2014, 79, 6517−6531

The Journal of Organic Chemistry
Article
(13C NMR) using CDCl3 as the solvent and TMS as the reference.
Sensor 3 was prepared as previously reported.9i The quantum yields of
1−3 were determined in chloroform as 0.2%, 0.4%, and 0.05%,
respectively, following literature procedures.20
(s, 1H), 9.33 (s, 1H). 13C NMR: δ 119.7, 125.2, 125.5, 126.4, 127.0,
127.5, 127.8, 129.2, 130.1, 130.3, 130.6, 132.1, 133.9, 134.0, 134.2, 136.0,
136.6, 143.2, 151.7. Mp: 170−171 °C. Anal. Calcd for C19H12BrN: C,
68.28; H, 3.62; N, 4.19. Found: C, 68.07; H, 3.62; N, 4.13.
2. Syntheses of 1-(3′-Formyl-4′-hydroxyphenyl)-8-(9′-
anthryl)naphthalene (1) and 1-(4′-Isoquinolyl)-8-(3′-formyl-4′-
hydroxyphenyl)naphthalene N-Oxide (2). 1-(3′-Formyl-4′-me-
thoxyphenyl)-8-bromonaphthalene (4). A solution of 1,8-dibromo-
naphthalene (500 mg, 1.7 mmol), 3-formyl-4-methoxyphenylboronic
acid (472.0 mg, 2.6 mmol), Pd(PPh3)4 (151.5 mg, 0.13 mmol), and
K3PO4 (927.7 mg, 4.4 mmol) in 18 mL of toluene/ethanol/water (3:2:1
v/v) was stirred at 80 °C for 4 h. The resulting mixture was allowed to
cool to room temperature, quenched with water, and extracted with
CH2Cl2. The combined organic layers were dried over MgSO4 and
concentrated in vacuo. Purification by flash chromatography on silica gel
(CH2Cl2/hexanes 4:1) afforded 417 mg (1.2 mmol, 70% yield) of a
yellow solid.
1-(4′-Isoquinolyl)-8-bromonaphthalene N-Oxide (7). A solution of
5 (470 mg, 1.4 mmol) and m-CBPA (728 mg, 4.2 mmol) in 15 mL of
CH2Cl2 was stirred at room temperature for 12 h. The mixture was
washed with 2 M NaOH, dried over MgSO4, and concentrated in vacuo.
Purification by flash chromatography on silica gel (CH2Cl2/MeOH
20:1) afforded 441 mg (1.26 mmol, 90% yield) of a light-brown solid.
1H NMR: δ 7.14 (d, J = 8.5 Hz, 1H), 7.34−7.42 (m, 2H), 7.48 (d, J =
7.1 Hz, 1H), 7.56−7.62 (m, 2H), 7.75 (d, J = 7.6 Hz, 1H), 7.77 (d, J = 8.4
Hz, 1H), 7.96 (d, J = 8.6 Hz, 1H), 8.05 (d, J = 8.9 Hz, 1H), 8.16 (s, 1H),
8.85 (s, 1H). 13C NMR: δ 119.2, 125.0, 125.5, 126.8, 128.9, 128.9, 129.2,
129.3, 130.1, 130.9, 131.1, 131.1, 131.9, 134.2, 135.3, 135.9, 136.9, 138.8.
Mp: 99−100 °C. Anal. Calcd for C19H12BrNO: C, 65.16; H, 3.45; N,
4.00. Found: C, 65.14; H, 3.75; N, 3.84.
1-(4′-Isoquinolyl)-8-(3′-formyl-4′-methoxyphenyl)naphthalene
N-Oxide (8). A solution of 6 (440 mg, 1.3 mmol), 3-formyl-4-
methoxyphenylboronic acid (339.2 mg, 1.9 mmol), Pd(PPh3)4 (109 mg,
0.1 mmol), and K3PO4 (669 mg, 3.2 mmol) in 18 mL of toluene/
ethanol/water (3:2:1 v/v) was stirred at 100 °C for 12 h. The resulting
mixture was allowed to cool to room temperature, quenched with water,
and extracted with CH2Cl2. The combined organic layers were dried
over MgSO4 and concentrated in vacuo. Purification by flash
chromatography on silica gel (CH2Cl2/MeOH 20:1) afforded 357.6
mg (0.88 mmol, 70% yield) of a light-brown solid. NMR analysis
showed a mixture of syn and anti isomers with a ratio of 80:20.
1H NMR: δ 3.71 (s, 0.6H), 3.89 (s, 2.4H), 6.01 (d, J = 8.8 Hz, 0.2H),
6.64 (d, J = 7.7 Hz, 0.2H), 6.70 (d, J = 8.8 Hz, 0.8H), 6.90 (s, 0.8H), 6.99
(d, J = 8.0 Hz, 1H), 7.10 (d, J = 7.7 Hz, 0.8H), 7.27−7.31 (m, 2H), 7.39−
7.48 (m, 3.2H), 7.56−7.65 (m, 2H), 7.94 (s, 1H), 8.02 (d, J = 7.7 Hz,
1H), 8.11 (d, J = 7.7 Hz, 1H), 8.37 (s, 1H), 9.92 (s, 0.8H), 10.36 (s,
0.2H). 13C NMR: δ 56.0, 109.2, 114.2, 122.2, 124.2, 125.3, 125.7, 125.8,
128.3, 128.8, 129.1, 129.2, 129.7, 130.2, 130.9, 130.9, 131.0, 133.5, 134.1,
134.7, 134.8, 136.5, 137.4, 138.1, 139.5, 159.9, 188.2. Mp: 194−195 °C.
Anal. Calcd for C27H19NO3: C, 79.98; H, 4.72; N, 3.45. Found: C, 79.85;
H, 4.92; N, 3.45.
1H NMR: δ 4.02 (s, 3H), 7.12 (d, J = 8.5 Hz, 1H), 7.39−7.50 (m, 3H),
7.68 (d, J = 8.2 Hz, 1H), 7.81 (dd, J = 8.4 Hz, 8.2 Hz, 2H), 7.90 (d, J = 8.0
Hz, 1H), 7.98 (s, 1H), 10.56 (s, 1H). 13C NMR: δ 55.8, 110.7, 119.8,
124.0, 125.3, 126.1, 129.0, 129.1, 129.5, 129.8, 131.4, 133.8, 135.3, 136.1,
137.4, 138.6, 161.1, 190.0. Mp: 194−195 °C. Anal. Calcd for
C18H13BrO2: C, 63.36; H 3.84. Found: C, 63.18; H, 4.06.
1-(3′-Formyl-4′-methoxyphenyl)-8-(9′-anthryl)naphthalene (5). A
solution of 3 (400 mg, 1.2 mmol), anthracene-9-boronic acid (390 mg,
1.8 mmol), Pd(PPh3)4 (208 mg, 0.2 mmol), and K3PO4 (636.8 mg, 3.0
mmol) in 15 mL of toluene was stirred at 120 °C for 18 h. The resulting
mixture was allowed to cool to room temperature, quenched with water,
and extracted with CH2Cl2. The combined organic layers were dried
over MgSO4 and concentrated in vacuo. Purification by flash
chromatography on silica gel (CH2Cl2/hexanes 4:1) afforded 206 mg
(0.5 mmol, 40% yield) of a yellow solid.
1H NMR: δ 3.61 (s, 3H), 5.53 (d, J = 8.5 Hz, 1H), 6.30 (dd, J = 8.5 Hz,
2.2 Hz, 1H), 6.63 (d, J = 2.1 Hz, 1H), 7.08 (d, J = 7.0 Hz, 1H), 7.20−7.41
(m, 7H), 7.49 (dd, J = 7.7 Hz, 7.5 Hz, 1H), 7.63 (dd, J = 7.8 Hz, 7.4 Hz,
1H), 7.78 (d, J = 8.5 Hz, 1H), 7.84 (d, J = 8.5 Hz, 1H), 8.05−8.08 (m,
2H), 8.11 (d, J = 8.2 Hz, 1H), 9.63 (s, 1H). 13C NMR: δ 55.3, 108.0,
121.4, 124.7, 125.0, 125.1, 125.3, 125.4, 125.7, 125.9, 126.9, 127.1, 127.7,
127.8, 127.9, 129.2, 129.4, 130.0, 130.7, 130.9, 131.2, 131.3, 131.8, 132.0,
133.8, 134.7, 135.0, 135.6, 137.5, 139.1, 158.8, 188.4. Mp: 188−189 °C.
Anal. Calcd for C32H22O2: C, 87.65; H, 5.06. Found: C, 87.85; H, 5.27.
1-(3′-Formyl-4′-hydroxyphenyl)-8-(9′-anthryl)naphthalene (1). A
solution of 4 (200 mg, 0.46 mmol) and LiCl (193 mg, 4.6 mmol) in 5
mL of DMF was stirred at 150 °C for 12 h. The resulting mixture was
allowed to cool to room temperature, quenched with water, and
extracted with CH2Cl2. The combined organic layers were washed with
brine, dried over MgSO4, and concentrated in vacuo. Purification by
flash chromatography on silica gel (CH2Cl2/hexane 2:1) afforded 87.1
mg (0.2 mmol, 60% yield) of a yellow solid.
1-(4′-Isoquinolyl)-8-(3′-formyl-4′-hydroxyphenyl)naphthalene N-
Oxide (2). A solution of 7 (350 mg, 0.86 mmol) and BBr3 (1 M in
CH2Cl2, 2.6 mL, 2.6 mmol) in 10 mL of CH2Cl2 was stirred at room
temperature for 2 h. The resulting mixture was quenched with 2-
propanol, washed with water, dried over MgSO4, and concentrated in
vacuo. Purification by flash chromatography on silica gel (CH2Cl2/
MeOH 20:1) afforded 253 mg (0.65 mmol, 75% yield) of a white solid.
NMR analysis showed a mixture of syn and anti isomers with a ratio of
70:30.
1H NMR: δ 5.96 (d, J = 8.4 Hz, 0.3H), 6.59 (s, 1H), 6.69 (d, J = 8.4
Hz, 0.7H), 7.06−7.17 (m, 2H), 7.27−7.48 (m, 2H), 7.66−7.87 (m, 5H),
7.89−7.92 (m, 1H), 8.04 (d, J = 8.10 Hz, 1H), 8.12 (d, J = 8.1 Hz, 1H),
8.34 (s, 0.3H), 8.44 (s, 0.7H), 9.08 (s, 0.7H), 9.82 (s, 0.3H), 10.60 (s,
0.7H), 10.98 (s, 0.3H). 13C NMR: δ 114.7, 115.2, 117.9, 124.9, 125.3,
125.6, 125.7, 126.0, 128.3, 128.7, 128.9, 129.0, 129.3, 129.3, 129.4, 129.4,
130.7, 130.8, 130.9, 131.0, 131.0, 132.9, 133.0, 134.2, 134.7, 134.9, 136.4,
136.8, 137.0, 137.8, 137.8, 139.2, 139.2, 159.2, 159.8, 195.2, 196.0. Mp:
232−233 °C. Anal. Calcd for C26H17NO3: C, 79.78; H, 4.38; N, 3.58.
Found: C, 79.81; H, 4.72; N, 3.40.
3. Enantioselective CD Sensing Experiments. 3.1. General
Procedure for Chemosensing of Amines and Amino Alcohols. A stock
solution of sensor 1 or 2 (0.00375 M) in CHCl3 was prepared, and 350
μL portions were transferred to 4 mL vials. Solutions of the substrates
(0.026 M in CHCl3) were prepared. To each vial containing 350 μL of
stock solution was added 1 equiv (50 μL, 0.0013 mmol) of the substrate.
The reaction mixtures were stirred overnight for sensor 1 and 15 min for
sensor 2. The reaction times could be reduced to 5 h for sensor 1 by
addition of 10 mol % trifluoroacetic acid or p-toluenesulfonic acid. The
CD analysis was conducted with sample concentrations of 7.50 × 10−5
M in MeOH for sensor 1 and in hexanes for sensor 2. CD spectra were
collected with a standard sensitivity of 100 mdeg, a data pitch of 0.5 nm,
a bandwidth of 1 nm, a scanning speed of 500 nm s−1, and a response of
1H NMR: δ 5.65 (d, J = 8.4 Hz, 1H), 6.26 (d, J = 2.2 Hz, 1H), 6.30
(dd, J = 8.4 Hz, 2.3 Hz, 1H), 7.07 (d, J = 7.0 Hz, 1H), 7.19−7.38 (m,
5H), 7.44 (d, J = 8.3 Hz, 1H), 7.48−7.53 (m, 2H), 7.66 (dd, J = 8.1 Hz,
8.0 Hz, 1H), 7.81 (ddd, J = 8.6 Hz, 8.5 Hz, 2.9 Hz, 2H), 8.06 (d, J = 8.3
Hz, 1H), 8.11−8.14 (m, 2H), 8.56 (s, 1H), 10.42 (s, 1H). 13C NMR: δ
113.7, 117.2, 124.8, 124.9, 125.0, 125.2, 125.4, 125.5, 126.0, 126.7, 127.1,
128.4, 128.5, 129.4, 129.5, 130.2, 130.8, 130.9, 131.2, 131.3, 131.8, 132.1,
132,3, 132.9, 135.0, 135.1, 135.5, 137.4, 138.8, 158.9, 195.6. Mp: 199−
200 °C. Anal. Calcd for C31H20O2: C, 87.71; H, 4.75. Found: C, 87.97;
H, 5.08.
1-Isoquinolyl-8-bromonaphthalene (6). A solution of 1,8-dibro-
monaphthalene (500 mg, 1.7 mmol), 4-isoquinolineboronic acid (453.7
mg, 2.6 mmol), Pd(PPh3)4 (151.5 mg, 0.13 mmol), and K3PO4 (927.7
mg, 4.4 mmol) in 18 mL of toluene/ethanol/water (3:2:1 v/v) was
stirred at 80 °C for 4 h. The resulting mixture was allowed to cool to
room temperature, quenched with water, and extracted with CH2Cl2.
The combined organic layers were dried over MgSO4 and concentrated
in vacuo. Purification by flash chromatography on silica gel (CH2Cl2/
EtOAc 2:1) afforded 470 mg (1.4 mmol, 80% yield) of a yellow solid.
1H NMR: δ 7.26−7.36 (m, 2H), 7.51−7.61 (m, 4H), 7.71 (d, J = 7.4
Hz, 1H), 7.95 (d, J = 8.0 Hz, 1H), 8.01 (dd, J = 9.3 Hz, 9.3 Hz, 2H), 8.48
6527
dx.doi.org/10.1021/jo500959y | J. Org. Chem. 2014, 79, 6517−6531

The Journal of Organic Chemistry
Article
0.5 s using a quartz cuvette (1 cm path length). The data were baseline-
corrected and smoothed using a binomial equation. Control experi-
ments with free substrates showed no CD signal in the region of interest.
3.2. General Procedure for Chemosensing of Amino Acids. A stock
solution of sensor 1, 2, or 3 (0.005 M) in DMSO was prepared, and 1 mL
portions were transferred to 5 mL vials containing substrate 21−24 and
29 (0.005 mmol). Tetrabutylammonium hydroxide (1 M in methanol,
0.005 mmol, 5 μL) was then added. The reaction mixtures were stirred
overnight for sensor 1 and 15 min for sensors 2 and 3. CD analysis was
conducted with sample concentrations of 7.50 × 10−5 M in CHCl3 for all
sensors, and the instrument settings were the same as described above.
3.3. Analysis of Solvent Effects. The imines obtained with sensor 1
and amino alcohols or amines showed solvent-dependent CD readouts.
The imines derived from 9, 10, and 18 were formed as described above,
and CD spectra were collected in CHCl3 and methanol (Figures
10−12).
Figure 12. CD spectra of the imines obtained from 1 and (R)-18 in
CHCl3 (dashed blue) or MeOH (solid blue) and (S)-18 in CHCl3
(dashed red) or MeOH (solid red).
Figure 10. CD spectra of the imines obtained from 1 and (1R,2R)-9 in
CHCl3 (dashed blue) or MeOH (solid blue) and (1S,2S)-9 in CHCl3
(dashed red) or MeOH (solid red).
Figure 13. CD spectra of the imines obtained from 3, TBAOH, and (R)-
21 (solid blue) and (S)-21 (solid red) and CD responses of the imines
obtained from 3 and (R)-21 (dashed blue) and (S)-21 (dashed red)
upon addition of HCl.
Figure 11. CD spectra of the imines obtained from 1 and (R)-10 in
CHCl3 (dashed blue) or MeOH (solid blue) and (S)-10 in CHCl3
(dashed red) or MeOH (solid red).
3.4. Comparison of CD Outputs Obtained with the Protonated
and Deprotonated Forms of the Imines Obtained from Sensors 1, 2,
and 3 and Amino Acids. Imine formation with sensor 3 and amino
acids 21, 22, 23, and 29 was conducted in the presence of TBAOH as
described above. After the condensation was complete, 1 equiv of HCl
(1.25 M in EtOH, 4 μL) was added, changing the color from dark to light
yellow. CD spectra were collected as described above (Figures 13−16).
The imine formation with sensor 2 and amino acid 22 was conducted
in the presence of TBAOH as described above. After the condensation
was complete, 1 equiv of HCl (1.25 M in EtOH, 4 μL) was added. and a
CD spectrum was collected as described above (Figure 17).
Figure 14. CD spectra of the imines obtained from 3, TBAOH, and (R)-
22 (solid blue) and (S)-22 (solid red) and CD responses of the imines
obtained from 3 and (R)-22 (dashed blue) and (S)-22 (dashed red)
upon addition of HCl.
The imine formation with sensor 1 and amino acid 24 was conducted
in the presence of TBAOH as described above. After the condensation
was complete, 1 equiv of HCl (1.25 M in EtOH, 4 μL) was added, and a
CD spectrum was collected as described above (Figure 18).
6528
dx.doi.org/10.1021/jo500959y | J. Org. Chem. 2014, 79, 6517−6531

The Journal of Organic Chemistry
Article
Figure 18. CD spectra of the imines obtained from 1, TBAOH, and (R)-
24 (solid blue) and (S)-24 (solid red) and CD responses of the imines
obtained from 1 and (R)-24 (dashed blue) and (S)-24 (dashed red)
upon addition of HCl.
Figure 15. CD spectra of the imines obtained from 3, TBAOH, and (R)-
23 (solid blue) and (S)-23 (solid red) and CD responses of the imines
obtained from 3 and (R)-23 (dashed blue) and (S)-23 (dashed red)
upon addition of HCl.
obtained from 2 and phenylalanine, 2 and tyrosine, and 3 and
phenylalanine. The results are shown in Figures 19−21.
Figure 16. CD spectra of the imines obtained from 3, TBAOH, and (R)-
29 (solid blue) and (S)-29 (solid red) and CD responses of the imines
obtained from 3 and (R)-29 (dashed blue) and (S)-29 (dashed red)
upon addition of HCl.
Figure 19. Fluorescence spectra obtained for sensor 2 (blue) and the
imine obtained from 2 and 1 equiv of phenylalanine (red). Both spectra
were collected in CHCl3 (4.0 × 10−4 M) at an excitation wavelength of
340 nm.
Figure 17. CD spectra of the imines obtained from 2, TBAOH, and (R)-
22 (solid blue) and (S)-22 (solid red) and CD responses of the imines
obtained from 2 and (R)-22 (dashed blue) and (S)-22 (dashed red)
upon addition of HCl.
Figure 20. Fluorescence spectra obtained for sensor 2 (blue) and the
imine obtained from 2 and 1 equiv of tyrosine (red). Spectra were
collected in CHCl3 (4.0 × 10−4 M) at an excitation wavelength of 340
nm.
4. Fluorescence Changes upon Imine Formation. To determine
the impact of imine formation on the fluorescence of sensors 2 and 3,
fluorescence spectra were obtained for sensors 2 and 3 and the imines
6529
dx.doi.org/10.1021/jo500959y | J. Org. Chem. 2014, 79, 6517−6531

The Journal of Organic Chemistry
Article
(c) Tartaglia, S.; Padula, D.; Scafato, P.; Chiummiento, L.; Rosini, C. J.
Org. Chem. 2008, 73, 4865.
(4) (a) Nieto, S.; Lynch, V.; Anslyn, E.; Kim, H.; Chin, J. J. Am. Chem.
Soc. 2008, 130, 9232. (b) Nieto, S.; Dragna, J.; Anslyn, E. Chem.Eur. J.
2010, 16, 227. (c) Leung, D.; Anslyn, E. Org. Lett. 2011, 13, 2298.
(d) Joyce, L. A.; Maynor, M. S.; Dragna, J. M.; da Cruz, G. M.; Lynch, V.
M.; Canary, J. W.; Anslyn, E. V. J. Am. Chem. Soc. 2011, 133, 13746.
(e) Joyce, L.; Canary, J.; Anslyn, E. Chem.Eur. J. 2012, 18, 8064.
(f) You, L.; Pescitelli, G.; Anslyn, E. V.; Di Bari, L. J. Am. Chem. Soc.
2012, 134, 7117. (g) You, L.; Berman, J. S.; Lucksanawichien, A.;
Anslyn, E. V. J. Am. Chem. Soc. 2012, 134, 7126. (h) You, L.; Berman, J.
S.; Anslyn, E. V. Nat. Chem. 2011, 3, 943.
(5) (a) Holmes, A.; Zahn, S.; Canary, J. Chirality 2002, 14, 471.
(b) Holmes, A. E.; Das, D.; Canary, J. W. J. Am. Chem. Soc. 2007, 129,
1506. (c) Canary, J. W.; Mortezaei, S.; Liang, J. Chem. Commun. 2010,
46, 5850.
(6) (a) Yang, Q.; Olmsted, C.; Borhan, B. Org. Lett. 2002, 4, 3423.
(b) Li, X.; Tanasova, M.; Vasileiou, C.; Borhan, B. J. Am. Chem. Soc.
2008, 130, 1885. (c) Li, X.; Borhan, B. J. Am. Chem. Soc. 2008, 130,
16126. (d) Li, X.; Burrell, C. E.; Staples, R. J.; Borhan, B. J. Am. Chem.
Soc. 2012, 134, 9026. (e) Anyika, M.; Gholami, H.; Ashtekar, K. D.;
Acho, R.; Borhan, B. J. Am. Chem. Soc. 2014, 136, 550.
Figure 21. Fluorescence spectra obtained for sensor 3 (blue) and the
imine obtained from 3 and 1 equiv of phenylalanine (red). Spectra were
collected in CHCl3 (4.0 × 10−4 M) at an excitation wavelength of 350
nm.
(7) (a) Mazaleyrat, J.-P.; Wright, K.; Gaucher, A.; Toulemonde, N.;
Wakselman, M.; Oancea, S.; Peggion, C.; Formaggio, F.; Setnicka, V.;
Keiderling, T. A.; Toniolo, C. J. Am. Chem. Soc. 2004, 126, 12874.
(b) Dutot, L.; Wright, K.; Gaucher, A.; Wakselman, M.; Mazaleyrat, J.-
P.; De Zotti, M.; Peggion, C.; Formaggio, F.; Toniolo, C. J. Am. Chem.
Soc. 2008, 130, 5986.
(8) (a) Sciebura, J.; Skowronek, P.; Gawronski, J. Angew. Chem., Int. Ed.
2009, 48, 7069. (b) Sciebura, J.; Gawronski, J. Chem.Eur. J. 2011, 17,
13138.
5. Determination of the Rotational Energy Barrier of Sensor 1
by Dynamic HPLC. The free energy barrier to rotation, ΔG⧧, of the 3′-
formyl-4′-methoxyphenyl unit of sensor precursor 5 was determined by
DHPLC. Compound 5 (1 mg, 0.002 mmol) was dissolved in 1.5 mL of
CH2Cl2/hexanes (1:1 v/v). The HPLC analysis was performed using
1:1 CH2Cl2/hexanes as the mobile phase, a flow rate of 1 mL/min, an
injection volume of 20 μL, and the (S,S)-Whelk-O 1 column as the chiral
stationary phase at various temperatures. The rotational barrier was
determined by simulation of the elution profiles obtained at −4.8, −10.4,
−14.9, and −20.2 °C with the computer program Mimesis 3.1. The rate
constant of enantiomerization was optimized until the simulated and
experimentally obtained elution profiles were superimposable.
(9) (a) Ghosn, M. W.; Wolf, C. J. Am. Chem. Soc. 2009, 131, 16360.
(b) Ghosn, M. W.; Wolf, C. J. Org. Chem. 2011, 76, 3888. (c) Ghosn, M.
W.; Wolf, C. Tetrahedron 2011, 67, 6799. (d) Iwaniuk, D. P.; Wolf, C. J.
Am. Chem. Soc. 2011, 133, 2414. (e) Iwaniuk, D. P.; Wolf, C. Org. Lett.
2011, 13, 2602. (f) Iwaniuk, D. P.; Bentley, K. W.; Wolf, C. Chirality
2012, 24, 584. (g) Iwaniuk, D. P.; Wolf, C. Chem. Commun. 2012, 48,
11226. (h) Zhang, P.; Wolf, C. Chem. Commun. 2013, 49, 7010.
(i) Bentley, K. W.; Wolf, C. J. Am. Chem. Soc. 2013, 135, 12200.
(j) Bentley, K. W.; Wolf, C. J. Am. Chem. Soc. 2013, 135, 18052.
(10) (a) Kim, H.; So, S. M.; Yen, C. P.-H.; Vinhato, E.; Lough, A. J.;
Hong, J.-I.; Kim, H.-J.; Chin, J. Angew. Chem., Int. Ed. 2008, 47, 8657.
(b) Waki, M.; Abe, H.; Inouye, M. Angew. Chem., Int. Ed. 2007, 46, 3059.
(c) Katoono, R.; Kawai, H.; Fujiwara, K.; Suzuki, T. J. Am. Chem. Soc.
2009, 131, 16896. (d) Wezenberg, S. J.; Salassa, G.; Escudero-Adan, E.
C.; Benet-Buchholz, J.; Kleij, A. W. Angew. Chem., Int. Ed. 2011, 50, 713.
(e) Kuwahara, S.; Nakamura, M.; Yamaguchi, A.; Ikeda, M.; Habata, Y.
Org. Lett. 2013, 15, 5738. (f) Kim, M. J.; Choi, Y. R.; Jeon, H.-G.; Kang,
P.; Choi, M.-G.; Jeong, K.-S. Chem. Commun. 2013, 49, 11412. For a
review, see: (g) Wolf, C.; Bentley, K. W. Chem. Soc. Rev. 2013, 42, 5408.
For chirality sensing based on CPL analysis, see: (h) Song, F.; Wei, G.;
Jiang, X.; Li, F.; Cheng, Y.; Zhu, C. Chem. Commun. 2013, 49, 5772. For
chirality sensing with helical microfibers, see: (i) Zou, W.; Yan, Y.; Fang,
J.; Yang, Y.; Liang, J.; Deng, K.; Yao, J.; Wei, Z. J. Am. Chem. Soc. 2014,
136, 578.
(11) Attempts to reverse the coupling sequence and to install the
anthracene unit prior to the salicylaldehyde moiety were unsuccessful
because of prevailing debromination.
(12) Bao, K.; Fan, A.; Dai, Y.; Zhang, L.; Zhang, W.; Cheng, M.; Yao, X.
Org. Biomol. Chem. 2009, 7, 5084.
(13) Yadav, J.; Reddy, B.; Madan, C.; Hashim, R. Chem. Lett. 2000, 29,
738.
(14) Du, Z. T.; Lu, J.; Yu, H. R.; Xu, Y.; Li, A. P. J. Chem. Res. 2010, 34,
222.
(15) Bernard, A.; Ghiani, R.; Piras, P.; Rivoldini, A. Synthesis 1989, 287.
(16) In Dynamic Stereochemistry of Chiral Compounds; Wolf, C., Ed.;
RSC Publishing: Cambridge, U.K., 2008; pp 84−94.
(17) The reaction time can be reduced to 90 min in the absence of an
acid catalyst by microwave irradiation (chloroform at 100 °C, 100 W).
ASSOCIATED CONTENT
* Supporting Information
■
S
Quantitative ee and concentration analysis; MS, CD, fluo-
rescence, and NMR spectra; and crystallographic details (CIF).
This material is available free of charge via the Internet at http://
pubs.acs.org.
AUTHOR INFORMATION
Corresponding Author
*E-mail: cw27@georgetown.edu.
■
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
This material is based upon work supported by the National
Science Foundation under CHE-1213019.
■
REFERENCES
■
(1) Leung, D.; Kang, S. O.; Anslyn, E. V. Chem. Soc. Rev. 2012, 41, 448.
(2) (a) Matile, S.; Berova, N.; Nakanishi, K.; Novkova, S.; Philpova, I.;
Blagoev, B. J. Am. Chem. Soc. 1995, 117, 7021. (b) Kurtan, T.; Nesnas,
N.; Koehn, F. E.; Li, Y.-Q.; Nakanishi, K.; Berova, N. J. Am. Chem. Soc.
2001, 123, 5974. (c) Huang, X.; Fujioka, N.; Pescitelli, G.; Koehn, F. E.;
Williamson, R. T.; Nakanishi, K.; Berova, N. J. Am. Chem. Soc. 2002, 124,
10320. (d) Balaz, M.; De Napoli, M.; Holmes, A. E.; Mammana, A.;
Nakanishi, K.; Berova, N.; Purrello, R. Angew. Chem., Int. Ed. 2005, 44,
4006. (e) Berova, N.; Pescitelli, G.; Petrovic, A. G.; Proni, G. Chem.
Commun. 2009, 5958.
(3) (a) Superchi, S.; Casarini, D.; Laurita, A.; Bavoso, A.; Rosini, C.
Angew. Chem., Int. Ed. 2001, 40, 451. (b) Superchi, S.; Bisaccia, R.;
Casarini, D.; Laurita, A.; Rosini, C. J. Am. Chem. Soc. 2006, 128, 6893.
6530
dx.doi.org/10.1021/jo500959y | J. Org. Chem. 2014, 79, 6517−6531

The Journal of Organic Chemistry
Article
(18) (a) Cao, H.; Heagy, M. D. J. Fluoresc. 2004, 14, 569. (b) Martínez-
Manez, R.; Sancenon, F. Chem. Rev. 2003, 103, 4419. (c) Mei, X.; Wolf,
́
́
̃
C. Chem. Commun. 2004, 2078. (d) Mei, X.; Wolf, C. J. Am. Chem. Soc.
2004, 126, 14736. (e) Mei, X.; Wolf, C. J. Org. Chem. 2005, 70, 2299.
(19) In contrast to sensor 1 carrying a bulky anthracene unit adjacent
to the salicylaldehyde ring, the sterically less crowded probes 2 and 3 do
not show a CD response upon binding of chiral amines. We believe that
in the case of 2 and 3 the absence of the intramolecular hydrogen-
bonding motif results in conformationally flexible imine derivatives that
are CD-inactive because of insufficient central-to-axial chirality
induction.
(20) Jones, G., II; Jackson, W. R.; Choi, C.-Y. J. Phys. Chem. 1985, 89,
294−300.
6531
dx.doi.org/10.1021/jo500959y | J. Org. Chem. 2014, 79, 6517−6531