Tomasz Mądry1, Agnieszka Czapik1, Marcin Kwit1,2. 1. Department of Chemistry, Adam Mickiewicz University, Uniwersytetu Poznanskiego 8, 61 614 Poznan, Poland. 2. Center for Advanced Technologies, Adam Mickiewicz University, Uniwersytetu Poznanskiego 10, 61 614 Poznan, Poland.
Abstract
A readily available stereodynamic and the electronic circular dichroism (ECD)-silent 2,5-di(1-naphthyl)-terephthalaldehyde-based probe has been applied for chirality sensing of primary amines. The chiral amine (the inductor) forces a change in the structure of the chromophore system through the point-to-axial chirality transmission mechanism. As a result, efficient induction of optical activity in the chromophoric system is observed. The butterflylike structure of the probe, with the terminal aryl groups acting as changeable "wings", allowed for the generation of exciton Cotton effects in the region of 1Bb electronic transition in the naphthalene chromophores. The sign of the exciton couplets observed for inductor-reporter systems might be correlated with an absolute configuration of the inductor, whereas the linear relationship between amplitudes of the specific Cotton effect and enantiomeric excess of the parent amine gives potentiality for quantitative chirality sensing. Despite the structural simplicity, the probe turned out to be unprecedentedly highly sensitive to even subtle differences in the inductor structure (i.e., O vs CH2).
A readily available stereodynamic and the electronic circular dichroism (ECD)-silent 2,5-di(1-naphthyl)-terephthalaldehyde-based probe has been applied for chirality sensing of primary n class="Chemical">amines. The chiral amine (the inductor) forces a change in the structure of the chromophore system through the point-to-axial chirality transmission mechanism. As a result, efficient induction of optical activity in the chromophoric system is observed. The butterflylike structure of the probe, with the terminal aryl groups acting as changeable "wings", allowed for the generation of exciton Cotton effects in the region of 1Bb electronic transition in the naphthalene chromophores. The sign of the exciton couplets observed for inductor-reporter systems might be correlated with an absolute configuration of the inductor, whereas the linear relationship between amplitudes of the specific Cotton effect and enantiomeric excess of the parent amine gives potentiality for quantitative chirality sensing. Despite the structural simplicity, the probe turned out to be unprecedentedly highly sensitive to even subtle differences in the inductor structure (i.e., O vs CH2).
The stereochemical
analysis of chiral molecules, understood as
determination of a structure and/or composition of stereoisomers,
is crucial in many areas of the chemical synthesis and, particularly,
in the pharmaceutical industry.[1−7] The chiroptical methods that are based onnondestructive intern class="Chemical">action
of randomly oriented molecules with linearly or circularly polarized
light constitute convenient and, very often, the only research tools
for determining the structure of chiral entities in a solution. Due
to the straightforward response and sensitivity, electronic circular
dichroism (ECD) spectroscopy has been widely used in stereochemical
studies.[8] To date, a lot of general, or
operating in a very narrow area, correlation rules that bind the main
quality of the ECD spectrum, namely, the Cotton effect (CE) at a given
wavelength, with the three-dimensional (3D) structure of the chiral
entity have been proposed.[8] Although it
has been repeatedly demonstrated by us and others that most of the
traditional empirical correlation rules are inadequate and often incorrect,[9] the exceptional exciton chirality method is still
highly useful in stereochemical studies. In principle, the geometrical
relationship between the interacting electric dipole transition moments
(EDTMs) of allowed (usually) π–π* electronic transitions,
generated in (at least) two isolated chromophores, directly reflects
the chirality of the whole system.[10−13] As long as the coupling between
electric and non-negligible intrinsic magnetic transition dipole moments
(MDTMs), associated with given EDTMs, does not overwhelm the coupling
between electric dipoles, the exciton chirality method works perfectly.
Fortunately, these exceptions of exciton chirality rule are rare and
limited to the compounds of the bis-phenanthrene type or in which
the circular dichroism (CD) spectra are originated from electric forbidden
transitions.[14−18]
The use of the methods, especially the density functional
theory
(DFT)-based, offered by theoretical chemistry, is an alternative to
the empirical rules. For the ECD, it comes down to solving the Rosenfeld
equation of optically active electron transitions.[19−23]While the structure of the majority of optically
active compounds
allows direct ECD measurements, there are some compounds that do not
have either a suitable chromophore or their ECD readouts are difficult
to be unequivocal interpreted, even when supported by theoretical
methods. In such circumstances, the structure of a given compound
(usually called an “inducer” or “inductor”)
can be modified by an attn class="Chemical">achment of the chromophoric chirality probe
(commonly called a “reporter” or “chirality sensor”).
The main feature that should characterize the suitable sensor is the
ability to adapt the structure of the probe to the chiral environment.
The probe alone is CD-silent but stereodynamic, and as a result of
the dynamic inductor–reporter structural matching, the appearance
of nonzero CEs in the region of the UV absorption of the sensor is
observed.[24] The advances in the field of
the optical analysis of chiral compounds have been summarized in the
recent reviews.[25−31]
Among the pool of naturally occurring or artificial chiral
compounds,
amines seem to be one of the most important ones as they play the
leading role in the biochemical processes, in developing new synthetic
methods and in designing/synthesizing the leading substances, drugs,
and materials.[1,32]Even a cursory review of
the available literature precedents led
to the conclusion that the efficiency of the chirality sensing of
n class="Chemical">amines is a function of the mode of action of the probe and the method
of binding the inductor.[33−64] The metalloporphyrin tweezers introduced to a stereochemical analysis
by Berova and Nakanishi (Chart a) are considered sensitive probes of the chirality of noncovalently
bound species. However, for efficient binding to the probe, an analyte
requires the presence of two functional groups in a molecular skeleton.
This prerequisite limited the number of potential analytes to those
having already two functional moieties or to those monofunctional
ones, into which the second, artificial functional group can easily
be introduced.[43,65] Recently, Borhan and co-workers
have proposed a porphyrin-containing flexible host molecule (called
MAPOL, Chart a) that
binds chiral monoamines via hydrogen-bonding formation. The binding
of amine is associated with adapting a core-specific P- or M-helical conformation by biphenyl, which is
revealed by the appearance of exciton couplets (ECs) in the region
of the Soret band.[39] Although the chiroptical
response of porphyrin-based sensors is impressive, their wider applications
may be hampered by multistep synthesis and/or the need to use a large
excess of guests.
Chart 1
(a) Porphyrin-Based Noncovalent Probes for Chirality
Sensing and
(b) Exemplary Sterodynamic Probes for the Covalent Binding and Chirality
Sensing of Amines
An alternative approach
is based on the formation of a covalent
bond between the inducer and chirality sensor. The most common methods
for covalent binding of the n class="Chemical">amine rely on N-functionalization including
N-heterocycle formation, imination, or amidation reactions. Some representative
examples of the sterodynamic probes for chirality sensing of amines
are presented in Chart b.
The operation simplicity, on the one hand, and versatility
of the
sensors, on the other hand, make the imine-based probes particularly
useful for chirality sensing.[63,64,66] Widely utilized in the asymmetric synthesis, i.e., n class="Chemical">allenes, point-to-axial
chirality transfer[67−69] recently has gained popularity in chirality sensing.
Now, point-to-axial chirality transfer is efficiently applied as the
modus operandi of some chiroptical probes for the stereochemical characterization
of optically active amines and other compounds.[39,70] The common denominator of these probes is the presence of at least
two aromatic fragments capable of twisting relatively to each other
upon binding the inducer. The nonplanarity of aromatic parts, in principle,
caused the appearance of intense CEs, mostly of the exciton character.
Very recently, we have shown that the dinaphthylmethane-based stereodynamic
probes were turned out to be highly sensitive chirality sensors for
optically n class="Chemical">active secondary alcohols.[71] However,
some difficulties in the introduction of dinaphthylmethane on the
polar functional groups preclude the wider applicability of these
probes.
Feeling that the problem of chirality sensing still
has not lost
its topicality, we decided to put some efforts into designing, synthesizing,
and providing evidence for the efficiency of new sensitive stereodynamic
reporters for primary n class="Chemical">amines. We assumed that the probe should be
readily available, simple, and have a modular structure. The probe
core would consist of two or three directly bound aryl parts. One
of them would serve as the amine binder, whereas the other(s) as supporting
chromophore systems responsible for generating the CE exciton through
the point-to-axial chirality transfer mechanism. Thus, the energetic
barrier of flipping the chromophore(s) conformation should be low
enough not to interfere with the sensing process. The amine would
be bound to the receptor covalently by forming the C=N imine
double bond. Last but not least, the solubility of the probe-analyte
systems in hydrocarbons eliminates the possibility of specific interactions
with polar solvents, which, in some cases, might complicate the analysis.
The intention behind this idea is illustrated in Figure .
Figure 1
Schematic representation
of the designed stereodynamic triaryl
probe. Red arrows indicate polarizations of the electronic transition
of the highest oscillator strengths within anthracene and naphthalene
chromophores.
Schematic representation
of the designed stereodynamic triaryl
probe. Red arrows indicate polarizations of the electronic transition
of the highest oscillator strengths within anthracene and n class="Chemical">naphthalene
chromophores.
It is worth emphasizing that,
in the case of the “flipping”
trichromophoric systems, the operational efficiency may go hand in
hand with the complexity of the ECD readout. Looking at this issue
from a different angle and abstrn class="Chemical">acting from the efficiency in chirality
sensing, these chromophores themselves will constitute interesting
objects for the ECD study.
Results and Discussion
As the substrates
for the synthesis of probes, we have chosen 2-bromobenzaldehyde
and symmetrical n class="Chemical">2,5-dibromoterephthalaldehyde, which in turn can be
transformed into respective bi- and triaryl systems through the Suzuki
reaction.[72−74] Due to the well-known spectroscopic characteristics,
we limited the number of substrates to 1-naphthyl- and 9-anthrylboronic
acids. The reactions of 2-bromobenzaldehyde and 2,5-dibromoterephthalic
aldehyde with the naphthyl boronic acid smoothly provided respective
probes 1a and 2 (Chart a).
Chart 2
(a) Structures of Compounds under
Study, (b) Oak Ridge Thermal Ellipsoid
Plot (ORTEP) Drawing Showing an X-ray-Determined Structure of Dialdehyde 2 with Atom Numbering, and (c) Torsion Angles That Characterize
a Molecular Conformation
Despite many attempts, the use of 9-anthrylboronic acid and n class="Chemical">2,5-dibromoterephthalaldehyde
for Suzuki coupling did not provide the desired bis-arylated product.
However, 2-(9-anthryl)benzaldehyde (1b) was obtained
under the standard conditions with a reasonable yield (64%). Probe 1b turned out to be slightly light-sensitive; therefore, separation,
storage, and further operations on this compound had to be conducted
with limited light.
As the X-ray diffraction study revealed
that in the crystal structure
of 2 (Chart b) both n class="Chemical">naphthyl groups (denoted here A and C) are twisted relative to the central aromatic moiety (denoted B). The values of both twist α1 and α2 angles (defined here as C2–C1–C2′–C1′
and C4′–C5′–C1″–C2″,
see Chart c) are equal
to 121.7°. This in turn excludes the possibility for electron
delocalization within the whole chromophoric system; therefore, each
unit might act independently. One should keep in mind that the sign
and magnitude of exciton CEs are primarily the function of the dihedral
angle and the distance between the interacting EDTMs. On the other
hand, the perpendicular or parallel mutual orientations of chromophores
may cause disappearance of the exciton Cotton’s effects and,
thus, inefficiency of a given chirality probe.
For preliminary
tests on the efficiency of the probes in chirality
sensing, we have chosenn class="Chemical">(R)-3,3-dimethylbutan-2-amine.
The amine is characterized by a large spatial diversity and, therefore,
may constitute the point of reference in stereochemical studies. The
ECD measurements run in cyclohexane clearly showed the advantage of
probe 2 over the remaining two. As one can see, the tested
imine systems 3a and 3b gave ECD output
that might provide structural information if some efforts are put
into extracting them from the spectral data (see Figure and Table ). The ECD spectra of 3a and 3b are characterized by the presence of one well-distinguished
ECD band, appearing at 224 and 258, respectively, and two bands of
smaller intensity in the higher-energy region. None of the ECD spectra
exhibit the CE pattern typical of exciton coupling. Having taken into
account the fact that the UV pattern for anthracene chromophore seemed
to be the most promising in the context of probe designing, the ECD
readout for 3b can be considered disappointing.
Figure 2
UV (upper panels)
and ECD (lower panels) spectra of (a) 3a (solid black
lines) and 3b (dashed black lines), (b) 4a, (c) 4c, and (d) 4f (blue lines),
measured in cyclohexane and calculated at the TD-CAM-B3LYP/6-311++G(d,p)
level and ΔΔG-based Boltzmann-averaged
UV and ECD spectra of 4a, 4c, and 4f (red lines). The calculated spectra were wavelength-corrected
to match the UV maximum. Only the 185–350 nm region is shown.
Experimental and calculated ECD spectra of 4f were multiplied
by a factor of 4 to increase their visibility.
Table 1
ECD Data Measured in Cyclohexane and
Sensitivity Factors G = |A|/ε
× 10–4 Estimated for Imines 3a, 3b, and 4a–j and
Amines 6a–j
In parentheses, the absolute configuration
at the nitrogen-containing stereogenic center is indicated.
Short-wavelength CEs are undistinguishable
due to the high noise level.
UV (upper panels)
and ECD (lower panels) spectra of (a) 3a (solid bln class="Chemical">ack
lines) and 3b (dashed black lines), (b) 4a, (c) 4c, and (d) 4f (blue lines),
measured in cyclohexane and calculated at the TD-CAM-B3LYP/6-311++G(d,p)
level and ΔΔG-based Boltzmann-averaged
UV and ECD spectra of 4a, 4c, and 4f (red lines). The calculated spectra were wavelength-corrected
to match the UV maximum. Only the 185–350 nm region is shown.
Experimental and calculated ECD spectra of 4f were multiplied
by a factor of 4 to increase their visibility.
In parentheses, the absolute configuration
at the nitrogen-containing stereogenic center is indicated.Short-wavelength CEs are undistinguishable
due to the high noise level.On the contrary, the ECD spectrum measured in cyclohexane for 4a is charn class="Chemical">acterized by the high amplitudes of CEs, especially
in the higher-energy region, when compared to the ECD spectra of 3a and 3b (see Figure and Table ). In particular, the ECD spectrum of 4a exhibits long-wavelength negative CE at 290 nm and a well-distinguished
positive/negative/positive couplet appearing between 230 and 188 nm.
While the long-wavelength CE might be attributed to electronic transitions
involving the central chromophoric unit (B), we assumed
that the shape of the ECD spectrum in the higher-energy region is
a result of exciton interactions between both extreme naphthyl chromophores
(AC interactions) as well as interactions of each of
the naphthyl groups with an aromatic diimino linker (AB and BC interactions, vide infra).
The most promising
probe 2 was subsequently used for
further studies. Among the available primary amines, we have chosen
arbitrarily those that seemed representative of this class of compounds.
In general, condensations between the n class="Chemical">amines and 2 proceed
smoothly in toluene, provided almost quantitatively the respective
imines (4b–j). Although most of the
compounds thus obtained were crystalline, in fact, no additional purification
of the crude reaction mixtures was necessary. Note that compound 5 constitutes a model used for theoretical studies.
Taking advantage of good solubility of these compounds in nonpolar
solvents, we have measured their UV and ECD spectra in cyclohexane
(see the exemplary UV and ECD spectra in Figure c,d). For easier comparison of the given
chromophoric system with others, and to measure sensitivity in chirogenesis,
we used the so-called sensitivity fn class="Chemical">actor G defined
as G = |A|/ε.[71] For the purpose of this work, the sensitivity factor combines
the absolute value of the amplitude |A| (A = Δεlong – Δεshort) of the couplet observed in the region of the high-intensity
electronic transition (1Bb) in the naphthalene
chromophore (ca. 220 nm), which may be a result of exciton interactions,
with the extinction coefficient (ε at ca. 220 nm). The experimental
ECD data along with estimates for sensitivity factors G of imines 4b–j are juxtaposed in Table .
The process
of point-to-axial chirality induction in imines is
revealed by generating CEs of amplitudes ranging from small but visible
ones to strong ones—in dependence on the structure of the inductor.
As it can be clearly seen, the probe is sensitive to all inductor
systems, even to that charn class="Chemical">acterized by very little structural diversity.
Additionally, we have noticed some straightforward correlation between
the size of the substituents flanking the stereogenic center and the
magnitude of CEs. For example, a stepwise deterioration of steric
congestion in imines 4a–c manifests
itself in a similar systematic reduction of the amplitude of the respective
CEs, appearing in the spectral range 230–180 nm, as well as
a reduction of the G factor values. The observed
relationship between G values and the size of the
substituent at the stereogenic center is not linear, however. The
increase of bulkiness of cyclohexyl substituent, as in the case of
imine 4d, resulted in an increase of magnitude of CEs.
On the other hand, more sterically congested bornyl skeleton affects
the resulted ECD spectrum even less than ethyl and methyl groups in 4c.
Imine 4f constitutes the most challenging
task for
probe 2. The parent compound n class="Chemical">(3R)-aminotetrahydrofurane
is characterized by a very small steric diversity. For the efficient
work, probe 2 must distinguish the oxygen atom from the
methylene group, both involved in a five-membered ring. Although the
observed, easily detectable CEs are not larger than those registered
ones, e.g., for 4c, their appearance clearly confirms
the point-to-axial chirality transfer occurring in this and in the
other compounds under study. Consequently, the sensitivity factor,
which equals 0.7 × 10–5, is the smallest of
the all estimated.
For 4a–f is seen a direct correlation
between the shape of the ECD spectrum and absolute configuration at
the stereogenic center. In the 230–180 nm spectral region,
the positive/negative/positive (+/–/+) sequence of CEs corresponds
to the R absolute configuration, whereas the −/+/–
sequence of CEs is characteristic of the S absolute
configuration. However, when one of the aliphatic substituents at
the stereogenic center is repln class="Chemical">aced by an aromatic group, as in the
case of imines 4g and 4h, the found relationship
between the sequence of CEs and absolute configuration at the stereogenic
center is reversed.
The ECD spectra of 4i and 4j are difficult
to interpret. Due to the observed induction of optical activity, there
is no doubt that the presence of the dynamically adapting probe fragment
is crucial for the appearance of intense Cotton effects. However,
the experimental spectra of 4i and 4j are
superpositions of contributions originating from the intern class="Chemical">actions
within the chromophores that make up the probe as well as between
the probe, taken as a whole, and the naphthyl or phthalimide chromophores.
Imines 4 could be smoothly transferred into corresponding
n class="Chemical">amines 6. A lack of imine chromophores made the central
unit CD-silent in the analytical higher-energy region. On the other
hand, after the reduction of imine bonds, one can expect increased
flexibility of the whole system, at least within the inductor moiety.
In fact, the ECD spectra of amines 6 are less complex
than those measured for imines 4. At the same time, we
noticed decreasing magnitudes of CEs and thus decreasing sensitivity
factors values.
When the ECD spectra measured for 4a–h and their reduced counterparts were compared,
it turned
out that in the diagnostic region of 1Bb electronic
transitions in the naphthalene chromophore are almost the mirror images
(neglecting details of the shape of the curves and the amplitude of
the effects). As expected, the ECD spectra of n class="Chemical">amine systems having
additional chromophoric systems, although deprived of some elements,
are still difficult to interpret. A more serious disadvantage of the
amine derivatives is their lower solubility in nonpolar solvents.
For some derivatives, for which the ECD spectra are measured in acetonitrile,
the solvent effect is clearly visible. This is particularly seen in
the cases of inductors characterized by either not a very diverse
structure of substituents at the stereogenic center or by the presence
of aryl substituents.
The above studies only allow for the statement
of the fact of the
point-to-axial chirality transfer and the qualitative associations
of this phenomenon with the structure of the inducer. Therefore, we
have decided to expand these studies to include solid-state structural
analysis of some derivatives, with a particular emphasis onn class="Chemical">imine 4a. Using the propensity to form crystals suitable for X-ray
diffraction analysis, we have determined solid-state structures of
imines 4a–d and 4f and
amines 6a–c and 6f (details
are provided in the Supporting Information (SI)). We expected that X-ray analysis would allow for an easy and explicit
correlation of the observed CEs with the structure of the molecule,
especially with the conformation of the chromophore.
A special
opportunity to examine the effect of a substituent on
the chromophore structure is given by a direct comparison of the structures
of the respective pairs: 4a–6a, 4b–6b, 4c–6c, and 4f–6f (Figure ). The molecular structure of compounds in
the crystal structure is unchangeable and, in our opinion, closely
related to the molecular pn class="Chemical">acking and weak intermolecular interactions.
Surprisingly, the general tendency observed in these compounds, regardless
of their chemical character, is the pursuit of pseudocentrosymmetry
in the solid state, when neglecting the presence of chiral substituents
at the nitrogen atoms. This structural feature is associated with
the opposite signs of α1 and α2 angles
and, in this way, with the (almost) antiparallel orientation of naphthalene
rings. Hence, for such conformations, we cannot expect any optical
activity originated from naphthyl–naphthyl (AC) interactions.
Figure 3
Overlays of X-ray diffraction-determined solid-state structures
of (a) 4a (green) and 6a (orange), (b) 4b (aquamarine) and 6b (red), (c) 4c (deep yellow) and 6c (blue), and (d) 4f (deep yellow) and 6f (deep blue). For the disordered
molecules, the fragments with lower occupancy factors are shown as
a thin line. C-bound hydrogen atoms have been omitted for clarity,
and the nitrogen and oxygen atoms are shown as balls.
Overlays of X-ray diffraction-determined solid-state structures
of n class="Gene">(a) 4a (green) and 6a (orange), (b) 4b (aquamarine) and 6b (red), (c) 4c (deep yellow) and 6c (blue), and (d) 4f (deep yellow) and 6f (deep blue). For the disordered
molecules, the fragments with lower occupancy factors are shown as
a thin line. C-bound hydrogen atoms have been omitted for clarity,
and the nitrogen and oxygen atoms are shown as balls.
The crystallographic studies, although interesting from the
crystal
engineering point of view, have not provided any ultimate answer regarding
the origin of the induced optical activity observed for the studied
compounds. Therefore, to cast some light on the problem of the dynamic
chirality induction, we have carried out theoretical studies on the
structure–chiroptical property relationships for all of the
n class="Chemical">imines 4a–j, with a special emphasis
on the arbitrary chosen representative example 4a (discussed
later) as well as on the model compound 5.[75−78] In the latter case, the chiral substituents at iminenitrogen atoms
were replaced with methyl groups, which significantly facilitate calculations.
In particular, we were interested whether the observed ECD spectra,
in the higher-energy region, result from the long-range naphthyl–naphthyl
(AC) or short-range naphthyl–imine (AB and BC) interactions. In other words, we would answer
the question to what extent the terminal naphthalenes flip is responsible
for the ECD response.
First, we have established the conformational
freedom of the model
compounds 5. Calculated at the B3LYP/6-311++G(d,p) level,
the potential energy surfn class="Chemical">ace (shown in the Figure a) for the systematic change of angles α1 and α2 clearly indicates that the low-energy
conformers are characterized by the values of α angles ranging
from ±70 to ±120°. A significant increase of energy
is noticed for conformers in which at least one of the aryl–aryl
moiety is pointed toward planarity. By analogy to the B–A–B-type triads, the low-energy
conformers of imine 5 and its real congeners can be considered
as C or S (see Scheme ).[79−81]
Figure 4
(a) Molecular energy
of 5 as a function of angles
α1 and α2. (b) Computed at the TD-CAM-B3LYP/6-311++G(d,p)
level, long- (Rlong) and short-wavelength
(Rshort) rotatory strengths corresponding
to experimental exciton couplets of the 1Bb electronic
transition as a function of angles α1 and α2. (c) Deconvolution of the ECD spectrum of the low-energy
conformer of 5. The ECD spectrum calculated for the whole
conformer is in black, the spectrum calculated for the AB part is represented by the blue solid line, the spectrum calculated
for the AC part of the molecules is shown as the red
line, and the effect of summation (S = 2 ×
AB + AC) is shown as a brown dashed line. Wavelengths were
not corrected. Insets indicate the position of the low-energy electronic
transition in 5. (d) Main molecular orbitals involved
in the low-energy electronic transitions in the low-energy conformer
of 5.
Scheme 1
Low-Energy C- and S-Type Conformers
of Imine 5 and the Relationship between the Twist of
the α Angles and Symmetry of the Molecule
(a) Molecular energy
of 5 as a function of angles
α1 and α2. (b) Computed at the TD-CAM-B3LYP/6-311++G(d,p)
level, long- (Rlong) and short-wavelength
(Rshort) rotatory strengths corresponding
to experimental exciton couplets of the 1Bb electronic
transition as a function of angles α1 and α2. (c) Deconvolution of the ECD spectrum of the low-energy
conformer of 5. The ECD spectrum calculated for the whole
conformer is in bln class="Chemical">ack, the spectrum calculated for the AB part is represented by the blue solid line, the spectrum calculated
for the AC part of the molecules is shown as the red
line, and the effect of summation (S = 2 ×
AB + AC) is shown as a brown dashed line. Wavelengths were
not corrected. Insets indicate the position of the low-energy electronic
transition in 5. (d) Main molecular orbitals involved
in the low-energy electronic transitions in the low-energy conformer
of 5.
The lowest-energy, chiral conformers of 5 have C2 symmetry and both naphthyl groups
deviate
by 10° from perpendicularity (α1 = α2 = 100 or −100°). A little bit higher in energy
conformers of parallel or antiparallel orientation of n class="Chemical">naphthyl rings
are achiral and characterized either by the opposite values of α1 and α2 angles or one α angle twisted
by ±100° and the second amounting to ±80°. However,
the calculated energy differences between C2-, C-, and C1-symmetrical minimum-energy conformers are negligible. Thus, given
the method error, all of the conformers that are characterized by
twist angles ranging from ±100 to ±110° are equally
probable.
In the next step, for each low-energy conformer of 5, the ECD spectrum was calculated at the TD-CAM-B3n class="Gene">LYP/6-311++G(d,p)
level. This allowed for the calculations of the three-dimensional
surfaces that connect the predicted long- and short-wavelength rotatory
strengths (respectively for CEs experimentally observed at around
230 and 215 nm) with the conformation of the probe. In this particular
spectral region between 230 and 200 nm is expected to see exciton
couplets originated from interactions between 1Bb electronic transitions within naphthalene chromophores.
The
exemplary surfaces estimated for long- and short-wavelength
rotatory strengths shown in Figure b (see the remaining results in the Supporting Information) approximately remain in relation:
an object to its mirror image. As it has been supposed, the C-symmetrical conformers of 5 are
En class="Chemical">CD-silent in the spectral region of 1Bb electronic
transitions in naphthalene. These studies suggest, but still do not
constitute compelling evidence, the dominant contribution of interactions
between terminal naphthyl chromophores to the overall rotatory strength.
To further confirm this hypothesis, we have chosen the low-energy C2-symmetrical conformer of 5, characterized
by the values of α1 and α2 angles
equal to −100° and we have divided it into two moieties.
The first of them, mapping AB and BC intern class="Chemical">actions
between nearest chromophores, consisted of naphthalene and the imine
part of the molecule (the remaining naphthalene fragment was replaced
by the hydrogen atom). The second one reflected the AC interactions between terminal naphthyl groups. For each of the substructures,
the ECD spectrum was calculated at the TD-CAM-B3LYP/6-311++G(d,p)
level (see Figure c). Finally, the ECD spectrum calculated for the “whole”
low-energy conformer of 5 was compared with the one being
the result of summation S = 2 × AB + AC.
It is clearly seen that both the calculated spectrum
for the isolated AC system and the compiled (S) spectrum are almost
undistinguishable. The only differences are slightly higher magnitudes
of the respective CEs found in the compiled spectrum S. This analysis confirms that in the 230–200 nm region, the
observed CEs are mainly due to the n class="Chemical">AC exciton-type interactions.
The contribution of AB and BC interactions
to the overall rotatory strength in this particular spectral region
seemed negligible. However, these AB and BC interactions are gaining importance in the higher-energy region
of the ECD spectrum, i.e., below 200 nm. The comparison of the compiled
spectrum S with that calculated for the low-energy conformer
of 5 shows a significant degree of the overall similarity.
However, in the latter case (the ECD spectrum of 5),
a red shift (ca. 10 nm) of the respective rotatory strengths is noted.
To establish the physical reasons behind this small discrepancy, we
have taken a look into orbitals involved in the electronic transitions
in the low-energy conformer of 5. As expected, the lowest-energy
electronic transitions engaged orbitals coming from the central imine
unit B. Going to higher energies, the rotatory strengths
between 230 and 200 nm originated mainly from the orbitals centered
in the naphthalene units (Figure d). It should be noted that some contribution from
electronic transitions involving highest occupied molecular orbital
(HOMO) – 1 and HOMO – 3 from naphthalene and the lowest
unoccupied molecular orbital (LUMO) orbital centered at the central
unit B to the overall rotatory strengths was also established.
As an effect of approximation and then summation of Gaussian functions
for each of the individual rotatory strengths, the red shift of the
absorption bands is observed. Finally, the higher-energy region of
the ECD spectrum of the low-energy conformer of 5 is
again dominated by transitions involving molecular orbitals from the
central unit B. In the case of the AC representing
a degenerate coupled-oscillator system, the calculated rotatory strengths
originate solely from transitions involving orbitals centered in the
naphthalene. Hence, the deconvolution approach does not allow to capture
all of the details of the spectrum; at least, it leads to a qualitative
estimation of the factors affecting the observed effects.
It
is worth noting that the direct correlation of the sign of the
exciton couplet with the twist of the α angles is not straightforward.
However, keeping in mind that the highest intensity of the En class="Chemical">DTM in
naphthalene is polarized along the long axis of the chromophore, one
may feel entitled to use the ω angle, defined as ω = C8–C1–C1″–C8″,
as the angle between interacting EDTMs. To be precise, the sign of
the given exciton CE is a function of the product of the sine of a
torsional angle between EDTMs as well as the sine and cosine (the
latter for nondegenerated systems) functions of in-plane angles between
the given electric dipole and the line connected the midpoints of
interacting electric dipoles.[11−13] Thus, assuming the dominant contribution
of A/C interactions to the ECD spectrum,
and polarization of EDTM’s close to (or even parallel to) the
long axis of the naphthalene chromophore, the sign of exciton CE will
depend on angle ω. Note that even if the orientation of the
EDTMs is tilted off the long axis of the naphthalene chromophore,
this does not affect the sign but only the magnitude of the excitonic
rotatory strength.
For the above-discussed low-energy conformer
of 5,
the negative sign of the exciton couplet remains in agreement with
the negative sign of the ω angle (ω = −36°).[11−13]To establish the mechanism of the optical activity in the
real
system, we applied a similar procedure (see Calculation details in
the Supporting Information) to n class="Chemical">imines 4a–j. Since the detailed elaboration of
each calculated structure may obscure the problem, we have decided
to limit the in-depth discussion to the representative example 4a. The remaining results will be briefly commented on, and
all of the theoretical results are skipped to the SI.
Four low-energy conformers of 4a are
found by calculations
at the B3LYP/6-311++G(d,p) level of the theory (Table and Figure a). The conformers are charn class="Chemical">acterized by the deviation
of α angles from perpendicularity by ±11 to ±20°,
with a reasonable agreement with the X-ray data obtained for the real
compound. The conformation of the imine groups, described by β1 and β2 angles (see Chart c), in all of the cases is synperiplanar.
Table 2
Relative Gibbs Free Energies (ΔΔG, in kcal mol–1); ΔΔG-Based Percentage Populations (Pop.); Values of α1, α2, β1, β2, and ω Angles (in deg) and Predicted Sign of the Exciton Couplet
(EC), Calculated at the B3LYP/6-311++G(d,p) Level of Theory for Individual
Low-Energy Conformers of 4a
conf.a
ΔΔG
Pop.
α1b
α2b
β1c
β2c
ωd
CE
1
1.77
3
105
105
178
178
29
(+)
2
1.12
10
101
101
179
179
21
(+)
4
0.00
65
104
–104
176
–178
180
0
6
0.64
22
–111
–111
–171
–171
–41
(−)
Conformers are
numbered according
to their appearance during the conformational search.
α1 = C2–C1–C2′–C1′;
α2 = C2″–C1″–C5′–C4′.
β1 = C2′–C1′–C=N;
β2 = C5′–C4′–C=N.
ω = C8–C1–C1″–C8″.
Figure 5
(a) Overlay
of low-energy conformers of 4a calculated
at the B3LYP/6-311++G(d,p) level of individual conformers of 4a. (b) ECD spectra calculated at the TD-CAM-B3LYP/6-311++G(d,p)
level for the individual low-energy conformers of 4a.
Wavelengths were not corrected. Structural and spectral data for a
given conformer are shown in the same color: conf. no 1, green; conf.
no. 2, blue; conf. no. 4, black; and conf. no. 6, red.
(a) Overlay
of low-energy conformers of 4a calculated
at the B3LYP/6-311++G(d,p) level of individual conformers of 4a. (b) ECD spectra calculated at the TD-CAM-B3n class="Gene">LYP/6-311++G(d,p)
level for the individual low-energy conformers of 4a.
Wavelengths were not corrected. Structural and spectral data for a
given conformer are shown in the same color: conf. no 1, green; conf.
no. 2, blue; conf. no. 4, black; and conf. no. 6, red.
Conformers are
numbered according
to their appearance during the conformational search.α1 = C2–C1–C2′–C1′;
α2 = C2″–C1″–C5′–C4′.β1 = C2′–C1′–C=N;
β2 = C5′–C4′–C=N.ω = C8–C1–C1″–C8″.While three of these structures
are characterized by the same signs
of both α angles and by C2 symmetry,
one of the lowest energy (conf. no. 4) closely resembles the X-ray-determined
crystal structure of 4a, at least in the context of the
preferred chromophore conformation. Namely, the C1-symmetrical conformer no. 4 is charn class="Chemical">acterized by the
values of α1 and α2 angles equal
to −104 and 104°, respectively. The population of this
conformer, estimated on the basis of relative Gibbs energies, amounts
to 65%.
Since the interacting n class="Chemical">EDTMs in naphthyl chromophores
are orientated
antiparallelly (ω ≈ 180°), there is no possibility
for optically active AC exciton interactions. Thus, the
ECD spectrum, calculated for the conformer no. 4, is almost flat (see Figure b) in the spectral
region dominated by naphthalene–naphthalene interactions. Hence,
the contribution of this conformer does not affect the overall ECD
spectrum, in accordance with the empirical estimation. The second
low-energy conformer no. 6 is much less populated (22%), but, in fact,
this particular conformer determines the sign and the pattern of the
averaged ECD spectrum. Due to the small contribution to the conformational
equilibrium, the effect of the remaining, higher-energy, conformer
nos. 1 and 2 on the overall ECD spectrum seems to be negligible.
The reproduction of experimental data by DFT calculation is satisfactory
for most of the cases, with the exception of 4i, which,
indeed, represents a very complex problem. In fn class="Chemical">act, the best agreement
between the experimental ECD spectrum and that calculated is noticed
for one of the difficult cases, namely, 4c (see Figure c), although the
good agreement between experimental and theoretical data, obtained
for the most demanding imine 4f, is worth mentioning
(see Figure d).
For the particular case of imine 4c, the percentage
relation of conformers charn class="Chemical">acterized by the positive value of the
ω angle to that of the negative value of the ω angle and
optically inactive conformers (ω ≈ 180°) amounts
to 41% (+): 31% (−): 28% (0). In other cases, even if the abundance
of optically inactive conformers prevails, there is the conformer
(or conformers) in the population whose contribution to the ECD spectrum
is dominant. This seemingly high-conformational dynamics is not uncommon
for stereodynamic chirality probes and results directly from the principles
on which their mode of action is based. Finally, by neglecting the
contribution of optically inactive conformers, the mode of action
of probe 2 also determines the most typical equilibrium
between P- and M-helical diasteroisomers.
The naturally emerging question is how the chirality of the inducer
is transferred to the probe. The analysis of the available structural
data led to the conclusion that both steric and electrostatic interactions
take part in chirality transfer. The latter is easier to demonstrate.
The (N=C)H···π intern class="Chemical">actions between an
electron cloud of the naphthalene and positively charged methine proton
fix the anti conformation of β angles (see Chart c for definition). In all cases,
the β angles adapt anti conformation. The syn conformation of
H–C=N–C* moieties, typical of imines, associated
with the synperiplanar orientation of C*H and CH=N protons,
allowed for interaction of nitrogen lone pairs with aromatic protons
in ortho positions of the central unit B. The tilting
of naphthalene units is determined by very weak sterical interactions
with substituents flanking the stereogenic centers. In general, naphthalene
is tilted toward a less bulky substituent. For example, the opposite-sign
exciton CEs in 4d and 4h point to opposite
helicity of the chromophore. As both compounds are characterized by
the same R configuration of stereogenic centers,
the reversal of the signs is associated with steric demands of the
chiral substituent. In Figure , the low-energy but optically active conformers of 4d (conf. no. 14) and 4h (conf. no. 1) are shown,
which contribute the most to the overall, respective, ECD spectra.
The conformer no. 14 of 4d is characterized by a negative
value of the ω angle (ω = −31°), and the naphthalene
rings are tilted towards methyl groups. On the contrary, the ω
angle found for conformer no. 2 of 4h is positive (ω
= 31°), and the naphthalene rings are tilted toward the phenyl
substituents. Thus, the steric power of the methyl groups for induction
of dynamic chirality in 4h is larger than that of the
phenyl rings.
Figure 6
Top (upper panel) and side (lower panel) views of the
low-energy
conformers of (a) 4d (conf. no. 14) and (b) 4h (conf. no. 1). The naphthyl rings closer to the observer are in
green, whereas the naphthyl rings away from the observer are in red.
Top (upper panel) and side (lower panel) views of the
low-energy
conformers of (a) 4d (conf. no. 14) and (b) 4h (conf. no. 1). The n class="Chemical">naphthyl rings closer to the observer are in
green, whereas the naphthyl rings away from the observer are in red.
The calculated and ΔΔG-based and Boltzmann-averaged
ECD spectrum of 4i did not reproduce well the experimental
one. This is due to the complexity of the system and problems with
a correct mapping of energy relationships between conformers.At the final stage of our study, we would check the possibility
of the quantitative chirality sensing with probe 2.[26,33] Thus, we experimentally determined the chiroptical response of nonracemic
crude products 4a, obtained by condensation of 2 with n class="Chemical">3,3-dimethylbutan-2-amine varied in optical purity.
We observed linear relationships between the enantiomeric excess (ee)
of the free amine and the CEs appearing at 227 and 215 nm (Figure ). It is worth noting
that we found the same linear response for all optically active ECD
bands. This feature might be useful for samples where the intrinsic
rotatory strengths from the inductor obscure the induced circular
dichroism resulting from the point-to-axial chirality transmission.
Figure 7
(a) ECD
spectra of the crude imine samples, obtained from 2 and
3,3-dimethylbutan-2-amine of varying ee. (b) Linear
relationships between CE amplitude at 227 nm (blue line) and 215 nm
(red line) and the sample ee.
(a) ECD
spectra of the crude imine samples, obtained from 2 and
n class="Chemical">3,3-dimethylbutan-2-amine of varying ee. (b) Linear
relationships between CE amplitude at 227 nm (blue line) and 215 nm
(red line) and the sample ee.
Conclusions
In conclusion, we have designed and proven the usefulness of stereodynamic
2,5-di(1-naphthalene)-terephthalaldehyde (2) for qualitative
chirality sensing of n class="Chemical">amines. The sensor is readily available from
nonexpensive and commercial substrates and is smoothly converted into
corresponding diimines through simple condensation reactions. In fact,
there are only three simple synthetic steps from terephtalaldehyde
to the imine, approachable to even nonspecialists in organic synthesis.
Hence, the cost of synthesis, compared to, e.g., the synthesis of
porphyrin-based probes is small.
The ECD spectra measured for
crude and purified by crystallization
samples did not show any visible differences. Thus, purification of
the condensation product is not necessary, which further simplifies
the whole procedure. The additional values of the studied compounds
are their solubility in nonpolar solvents of the hydrocarbon type
and easy transformation into more chemically resistant n class="Chemical">amines by reduction.
The latter provide the ECD-active products as well; however, the chiroptical
response of amines is smaller than that established for the parent
imines.
The most important feature of probe 2 is
its high
sensitivity toward inductors characterized by the very small structural
diversity at the stereogenic center.In principle, taken into
account the G values,
it is possible to estimate the relative size of groups flanking stereogenic
center. It should be noted that this approach does not provide values
regarding the absolute size of the substituent. Instead, it allows
us to estimate the steric power for dynamic induction of chirality
of one substituent relative to the other flanking the stereogenic
center(s). The weak point of this approach is its limitation to aliphatic
amines and those in which the aryl substituent does not interfere
with the induced circular dichroism. The results obtained from the
analysis suggest that the relative size of the substituents in structurally
similar imines 4a–d increases as
follows: t-Bu > Cy > i-Pr
> Et >
Me. For compounds of this type, the positive exciton couplet observed
at around 220 nm corresponds to R absolute configuration
and vice versa. For imines characterized by the presence of stereogenic
center in the cycle, these relationships are −C*(Me)–
> −CH2–; −CH2O–
> −CH2CH2–; −C(Ar)– > −CH2–, respectively, for 4e–g. The larger steric power of the methyl
group with respect to phenyl and naphthyl groups in 4h and 4i, respectively, is apparently the reason for
the reversal of correlation between the sign of exciton couplet and
the absolute configuration of aliphatic vs aromatic substituents.To make this analysis more comprehensive, we have compared sensitivity
factors calculated from the data available in the literature with
the results obtained in this study.[39,50,51,62,82] We have chosen only the data, which are given in ε and Δε
units, since their expressions include both the concentration and
path length. For the reasons given above, only inducers with high
structural variability (preferably n class="Chemical">2-amino-3,3-dimethylbutane) were
selected. The structures of inductor–reporter systems used
in this analysis are shown in Chart .
Chart 3
Exemplary Sterodynamic Inductor–Reporter Systems
and Their
Calculated Sensitivity Factors G
As one can see, all of the probes shown in Chart are characterized by comparable
values of
sensitivity fn class="Chemical">actors. Among the analyzed systems, none of the sensors
definitely outperforms the others; however, there is a slight advantage
of probes containing extended chromophoric systems.
Hence, we
can rank these compounds toward decreasing sensitivity
in the following order: 7 > 11 > 4a > 10 > 9 > 8.Although on the basis of X-ray results the direct correlation
between
the chromophore structure and experimentally observed induced optical
activity is not visible, one can bear in mind that the solid-state
structure of flexible compounds is determined by the way of pn class="Chemical">acking
of individual molecules in the crystal lattice. Hence, the direct
inference on the structure of a given compound in solution, based
on crystallographic data solely, may lead to erroneous results.
Even in the case of inductor characterized by low structural diversity,
the agreement between experimental and calculated ECD spectra is good
to excellent. Thus, this alternative theoretical appron class="Chemical">ach can be used
for determining the absolute configuration of the inducers.
Finally, we have demonstrated a linear relationship between ee
of the amine and ECD signal; thus, the probe might be used for screening
purposes, e.g., for quick checking of optical n class="Chemical">activity of product
of stereoselective processes.
Experimental Section
General
Information
Unless otherwise noted, all reactions
were carried out in air. n class="Chemical">Deuterated chloroform (CDCl3),
solvents, and other chemicals were purchased from commercial suppliers
and used as received without further purification. Intermediates:
2,5-dibromoterephthalaldehyde and (1R,2R)-N-phthaloyl-1,2-diaminohexane were synthesized
according to the literature procedures.[72,83−85]
1H and n class="Chemical">13C{H} NMR spectra were recorded
on a Varian 400 MHz spectrometer at room temperature. Chemical shifts
are reported in parts per million (ppm). Spectra are referenced using
an internal reference (trimethylsilane or CDCl3 residual
solvent peak). Data is described as follows: chemical shift, multiplicity
(s = singlet, d = doublet, t = triplet, q = quartet, quint = quintet,
h = sextet, m = multiplet, dd = doublet of doublets, dt = doublet
of triplets, dq = doublet of quartets, td = triplet of doublets, tt
= triplet of triplets, qd = quartet of doublets, quintd = quintet
of doublets, ddd = doublet of doublets of doublets and br = broad),
coupling constants (Hz), and integration. Column chromatography was
performed on silica gel of pore size 60 Å, 70–230 mesh,
63–200 μm. (Fluka). Thin-layer chromatography (TLC) was
carried out using Sigma-Aldrich precoated TLC plates (60 Å medium
pore diameter with a 254 nm fluorescence indicator).
Melting
points were measured on a BUCHI B-545 apparatus. High-resolution
mass spectra (HRMS) were measured using a Bruker Impact n class="Disease">HD spectrometer.
Optical rotations were recorded on a Jasco P-2000 polarimeter at 20
°C.
The ECD and UV spectra were measured using a Jasco
J-810 spectropolarimeter
at room temperature in cyclohexane and n class="Chemical">acetonitrile solutions and
with the use of a quartz cell of 0.1 cm optical lengths. The concentration
of analytes ranged from 1.0 to 2.0 × 10–4 mol
L–1. Background spectra of the pure solvents were
recorded from 400 to 185 nm with a scan speed of 100 nm min–1. The ECD spectra of analytes were measured with eight accumulations.
IR spectra were recorded on a Jasco FT-IR 4600 spectrophotometer
with ATR PRO ONE using a diamond crystal.
2-(1-Naphthyl)-benzaldehyde
(1a)
In a 50 mL round-bottom flask, K2CO3 (1.310
g, 9.46 mmol, 2.5 equiv) was dissolved in 20 mL of n class="Chemical">H2O
and the resulting solution was sparged with argon for 30 min. A 100
mL round-bottom flask containing a mixture of toluene (30 mL) and
EtOH (20 mL) was sparged with argon for 30 min. Then, 2-bromobenzlaldehyde
(0.44 mL, 3.78 mmol, 1 equiv), 1-naphthyl boronic acid (0.780 g, 4.54
mmol, 1.2 equiv), Pd(PPh3)4 (0.306 g, 0.27 mmol,
7 mol %), and water solution of K2CO3 were added
to the flask. The resulting mixture was refluxed overnight under an
argon atmosphere and using a heating mantle. Then, the dark mixture
was cooled to room temperature, diluted with CHCl3, and
filtered through Celite. The filtrate was washed twice with water
and brine, dried (Na2SO4), and concentrated
under reduced pressure, resulting in dark thick oil. The product was
separated using column chromatography (CHCl3/n-hexane 1:1). Yellow oil, 88% yield (772 mg).
IR (thin film,
cm–1): 3368, 3054, 2840, 2749, 1943, 1823, 1740,
1690, 1596, 1504, 1479, 1447, 1387, 1337, 1298, 1261, 1246, 1194,
1159, 1119, 1039, 1014, 957, 926, 883, 870, 836, 817, 802, 780, 765,
738, 704, 636, 618, 578, 565, 477, 444, 419.1H NMR
(400 MHz, n class="Chemical">CDCl3): δ 9.63 (d, J =
0.8 Hz, 1H), 8.11 (dd, J = 7.8, 1.1
Hz, 1H), 7.97–7.88 (m, 2H), 7.69 (td, J =
7.5, 1.5 Hz, 1H), 7.63–7.37 (m, 7H).
13C{H}
NMR (101 MHz, n class="Chemical">CDCl3): δ 192.0,
192.0, 144.2, 135.4, 134.7, 133.6, 133.3, 132.6, 131.7, 128.6, 128.3,
128.1, 127.0, 126.7, 126.1, 125.7, 125.0.
HRMS (electrospray
ionization quadrupole time-of-flight (ESI-Q-TOF)), m/z: [M + Na]+ calcd for n class="Chemical">C17H12NaO, 255.0786; found, 255.0796.
2-(9-Anthracene)-benzaldehyde
(1b)
In a 50 mL round-bottom flask, K2CO3 (1.310
g, 9.46 mmol, 2.5 equiv) was dissolved in 20 mL of n class="Chemical">H2O
and the resulting solution was sparged with argon for 30 min. A 100
mL round-bottom flask containing a mixture of toluene (30 mL) and
EtOH (20 mL) was sparged with argon for 30 min. Then, 2-bromobenzaldehyde
(0.44 mL, 3.78 mmol, 1 equiv), 9-anthraceneboronic acid (1.010 g,
4.54 mmol, 1.2 equiv), Pd(PPh3)4 (0.306 g, 0.27
mmol, 7 mol %), and water solution of K2CO3 were
added to the flask. The resulting mixture was refluxed overnight in
an argon atmosphere in the dark and with the use of a heating mantle
as the heat source. Then, the dark mixture was cooled to room temperature,
diluted with toluene, and filtered through Celite. The filtrate was
washed twice with water and brine, dried (Na2SO4), and concentrated under reduced pressure, resulting in dark thick
oil. The product was separated using column chromatography (CHCl3/n-hexane 9:1). Isolation of 2-(9-anthracene)-benzaldehyde
should not be unnecessarily prolonged due to its highly light-sensitiveness.
The pure product should be stored in covered glassware in a dark place.
Solidifying yellow-green oil, 64% yield (682 mg).
IR (thin film,
cm–1): 3380, 3056, 2858, 2763, 1950, 1928, 1843,
1821, 1787, 1725, 1695, 1621, 1595, 1517, 1474, 1440, 1396, 1354,
1287, 1265, 1221, 1199, 1168, 1141, 1092, 1037, 1013, 957, 936, 896,
859, 849, 838, 820, 793, 760, 736, 694, 655, 630, 610, 576, 553, 467,
442, 407.1H NMR (400 MHz, n class="Chemical">CDCl3): δ
9.32 (d, J = 0.9 Hz, 1H), 8.56 (s, 1H), 8.26–8.19
(m, 1H),
8.10–8.04 (m, 2H), 7.78 (td, J = 7.5, 1.5
Hz, 1H), 7.68 (tt, J = 7.8, 1.1 Hz, 1H), 7.51–7.41
(m, 5H), 7.37 (dd, J = 6.3, 1.2 Hz, 1H), 7.35 (dd, J = 6.4, 1.3 Hz, 1H).
13C{H} NMR (101 MHz,
n class="Chemical">CDCl3): δ 191.9,
142.8, 135.6, 134.0, 132.6, 131.5, 131.1, 131.0, 128.5, 128.5, 127.7,
127.2, 126.3, 126.1, 125.3.
HRMS (ESI-Q-TOF), m/z: [M + Na]+ calcd for n class="Chemical">C21H14NaO, 305.0942; found,
305.0947.
2,5-Di-(1-naphthyl)-terephthalaldehyde (2)
The general synthesis method was based
on the procedure described
by Prusinowska et al. and Frederickson et al.[72,84] In a 50 mL round-bottom flask, K2CO3 (2.49
g, 17.99 mmol, 5 equiv) was dissolved in 30 mL of n class="Chemical">H2O and
the resulting mixture was sparged with argon for 30 min. A 250 mL
round-bottom flask containing a mixture of toluene (40 mL) and EtOH
(30 mL) was sparged with argon for 30 min. Then, 2,5-dibromoterephthalaldehyde
(1.05 g, 3.6 mmol, 1 equiv), 1-naphthyl boronic acid (1.56 g, 8.99
mmol, 2.5 equiv), Pd(PPh3)4 (416 mg, 0.36 mmol,
10 mol %), and water solution of K2CO3 were
added to the flask. The resulting mixture was refluxed overnight in
an argon atmosphere and with the use of a heating mantle as the heat
source. The dark mixture was cooled to room temperature, diluted with
CHCl3, and filtered through Celite. The filtrate was washed
twice with water and brine, dried (Na2SO4),
and concentrated under reduced pressure, resulting in dark thick oil.
The product was separated using column chromatography and crystallization.
Green-to-yellow crystals, mp 245–247 °C, 65% yield
(601 mg).IR (thin film, cm–1): 3352, 3060,
3041, 3012,
2881, 2749, 1746, 1679, 1592, 1509, 1483, 1438, 1406, 1396, 1365,
1267, 1239, 1146, 1114, 1021, 1008, 977, 911, 870, 855, 799, 772,
658, 632, 522, 456, 426.1H NMR (400 MHz, n class="Chemical">CDCl3): δ 9.77 (s,
2H), 8.23 (s, 2H), 8.06–7.94 (m, 4H), 7.71–7.47 (m,
10H).
13C{H} NMR (101 MHz, n class="Chemical">CDCl3): δ
191.3,
143.4, 143.4, 137.7, 137.7, 134.0, 133.5, 133.4, 132.5, 132.4, 130.6,
130.6, 129.2, 128.6, 128.6, 128.5, 128.4, 127.2, 127.1, 126.4, 126.4,
125.4, 125.4, 125.1.
HRMS (ESI-Q-TOF), m/z: [M + Na]+ calcd for n class="Chemical">C28H18NaO2, 409.1204;
found, 409.1196.
The larger-scale synthesis was repeated starting
from 3.15 g of
2,5-dibromoterephthalaldehyde (10.8 mmol), n class="Chemical">1-naphthyl boronic acid
(4.68 g, 27 mmol) and reduced amount of Pd(PPh3)4 (624 mg, 0.54 mmol, 5 mol %). The reaction time was prolonged to
36 h. After workup and purification, the pure product was obtained
with 62% yield (2.58 g).
General Procedure for the
Synthesis of Imines 3a and 3b
To
a 25 mL round-bottom flask containing
aldehyde 1a or 1b (1 equiv, 0.44 mmol) and
chiral n class="Chemical">amine (1.2 equiv, 0.53 mmol), toluene (8 mL) was added. The
resulting mixture was stirred overnight under reflux using a Dean–Stark
apparatus and a heating mantle as the heat source. Then, the solvent
was removed in vacuo and the product was used as received.
General Procedure for the Synthesis of Imines 4a–4j
To a 25 mL round-bottom flask containing
dialdehyde 2 (1 equiv, 0.26 mmol) and chiral n class="Chemical">amine (2.3
equiv, 0.60 mmol),
toluene (8 mL) was added. The resulting mixture was stirred overnight
under reflux using a Dean–Stark apparatus and a heating mantle
as the heat source. Then, the solvent was removed in vacuo and the product was crystallized if necessary.
General Procedure
for the Reduction of Imines 4a–j to
Amines 6a–j
To a 25 mL round-bottomed
flask containing a solution of
respective imine (0.08 mmol, 1 equiv) inn class="Chemical">CHCl3 (6 mL),
MeOH was added (6 mL) and then NaBH4 (0.4 mmol, 5 equiv)
was added. The mixture was stirred overnight at room temperature and
quenched with 6 N solution of HCl till pH ≈ 1. Then, the mixture
was neutralized with a saturated water solution of K2CO3 to basic pH. The whole mixture was diluted with CHCl3 (20 mL) and washed twice with water and brine, dried (Na2SO4), and concentrated under reduced pressure,
giving a white solid. The crude product was further crystallized or
used as such.
Authors: Roly J Armstrong; Meganathan Nandakumar; Rafael M P Dias; Adam Noble; Eddie L Myers; Varinder K Aggarwal Journal: Angew Chem Int Ed Engl Date: 2018-05-30 Impact factor: 15.336
Chart 1
Figure 1
Chart 2
Figure 2
Figure 3
Figure 4
Scheme 1
Figure 5
Figure 6
Figure 7
Chart 3