Regiochemistry of the condensation of 2-aroyl-cyclohexanones and 2-cyanoacetamide: 13C-labeling studies and semiempirical MO calculations

Article abstract of DOI:10.1021/jo301138w

Hydroxy-aryl-5,6,7,8-tetrahydroisoquinoline-4-carbonitriles represent interesting chemical scaffolds, but synthetic access to these compounds is limited. The reaction of 2-aroyl-cyclohexanones with 2-cyanoacetamide and base in ethanol has been reported to

Full text of DOI:10.1021/jo301138w

Article
pubs.acs.org/joc
Regiochemistry of the Condensation of 2‑Aroyl-cyclohexanones and
13
2‑Cyanoacetamide: C‑Labeling Studies and Semiempirical MO
Calculations
Oscar P. J. van Linden, Maikel Wijtmans,* Luc Roumen, Lonneke Rotteveel, Rob Leurs,
and Iwan. J. P. de Esch*
Leiden/Amsterdam Center for Drug Research (LACDR), Division of Medicinal Chemistry, Faculty of Sciences, VU University
Amsterdam, De Boelelaan 1083, 1081 HV Amsterdam, The Netherlands
S
* Supporting Information
ABSTRACT: Hydroxy-aryl-5,6,7,8-tetrahydroisoquinoline-4-carbonitriles represent interesting chemical scaffolds, but synthetic
access to these compounds is limited. The reaction of 2-aroyl-cyclohexanones with 2-cyanoacetamide and base in ethanol has
been reported to lead to the formation of the tetrahydroisoquinoline isomer. We show that depending on the electronic nature of
the para-substituent on the aryl ring, formation of the regioisomeric tetrahydroquinoline isomer can significantly compete. The
electron-donating or -withdrawing properties of the para-substituent of the aryl ring determines the ratio of product isomers. A
series of 2-aroyl-cyclohexanones, with para-substituents ranging from electron-donating to electron-withdrawing, were reacted
with [2-13C]-cyanoacetamide. The product ratio and absolute regiochemistry were directly determined by quantitative 13C,
HMBC, and NOESY NMR spectroscopy on the reaction mixtures. A clear relationship between the regioisomeric product ratio
and the Hammett sigma values of the substituents is demonstrated. This is explained by the separate in situ yields, which reveal
that the pathway leading to the tetrahydroquinoline regioisomer is significantly more sensitive toward the electronic nature of the
para-substituent than the pathway leading to the tetrahydroisoquinoline. Semiempirical AM1 molecular orbital calculations on
the starting electrophile 2-aroyl-cyclohexanone support a correlation between the energy of the LUMOs and the regioisomeric
product ratio. Our results facilitate synthetic access to a range of these interesting synthetic intermediates.
different bases has been reported to yield the tetrahydroiso-
quinoline isomer 6 as the exclusive isolated product (Scheme
1).7,8 However, it has been shown that for nonaroyl substituted
5, the regioisomeric tetrahydroquinoline 7 is also formed, albeit
to a smaller extent.9 For example, the reaction of acetyl-
substituted cyclohexanones with 2-cyanoacetamide has been
reported to give rise to regioisomers depending on the
conditions.9 It can therefore be reasoned that the correspond-
ing 2-aroyl-cyclohexanones may give regioisomers as well,
giving either aryl-derivatized tetrahydroisoquinolines 3 or
regioisomeric tetrahydroquinolines 8 (Figure 1) as products.
In our previous work on kinase inhibitors,6 we prepared a series
INTRODUCTION
■
The pharmaceutically interesting scaffold 6,7,8,9-tetrahydro-
3H-pyrazolo[3,4-c]isoquinolin-1-amine 1 (Figure 1) has been
identified as an important moiety in, for example, the
development of kinase and phosphodiesterase inhibitors.1−5
We previously reported this attractive fragment in our search
for small-molecule inhibitors for the EPHA4 receptor tyrosine
kinase.6 For the synthesis of subsequently designed 5-aryl-
6,7,8,9-tetrahydro-3H-pyrazolo[3,4-c]isoquinolin-1-amines (2),
we explored 3-hydroxy-1-aryl-5,6,7,8-tetrahydroisoquinolines 3
as important intermediates.
Typically, such tetrahydroisoquinolines can be prepared via a
condensation reaction between 2-cyanoacetamide and sub-
stituted cyclohexanones. Indeed, the reaction of 2-cyanoaceta-
mide (4) with substituted 2-carbonyl-cyclohexanones (5) and
Received: June 2, 2012
Published: August 16, 2012
© 2012 American Chemical Society
7355
dx.doi.org/10.1021/jo301138w | J. Org. Chem. 2012, 77, 7355−7363

The Journal of Organic Chemistry
Article
Figure 1. Chemical structures of scaffold 1, EPHA4 kinase inhibitors 2, and intermediates 3 and 8.
Scheme 1. Proposed Reaction Pathways for the Formation of Regioisomers 6 and 7 from 2-Cyanoacetamide (4) and Substituted
2-Carbonyl-cyclohexanone (5)
a
Scheme 2. Reaction of [2-13C]-Cyanoacetamide (4) with Substituted 2-Aroyl-cyclohexanones (12)
a
Carbon and hydrogen atoms that are important for correct spectroscopic identification and quantification of products 3 and 8 are labeled.
of tetrahydroisoquinolines 3. The regiochemical identity of the
products was validated by advanced 2D-NMR techniques. At
the time we already noted that, depending on the nature of the
substituent on the phenyl ring, both 3 and 8 were present in the
crude reaction mixture. No molecular determinants underlying
the regiochemistry of such condensations of 2-aroyl-cyclo-
hexanones have been described so far in literature. We deemed
an in-depth description of such determinants of value to define
the synthetic scope and accessibility of these interesting
heteroaromatic motifs 3 and 8. Toward this end, we
investigated the effect of a strategically chosen set of para-
substituents on the in situ formations of tetrahydroisoquinoline
(3) and tetrahydroquinoline (8) products, both experimentally
(13C-labeling) as well as theoretically (AM1 calculations).
initial site of attack on electrophile 5 by enolate 9. It was
hypothesized that, for 2-aroyl-cyclohexanone 12 as the
electrophile (Scheme 2), the electronic nature of the para-
aryl-substituent R could play a major role in the outcome of this
reaction. That is, with R being an electron donating group
(EDG), the electron density on carbon atom Cb is likely to be
higher than the electron density on carbon atom Ca. This
would favor an attack of enolate 9 on the more electrophilic Ca
over an attack on Cb, ultimately resulting in the formation of
the regioisomer 3. If R is an electron-withdrawing group
(EWG), however, the electron density on Cb decreases,
resulting in greater electrophilicity and hence probably in
increased competition between attack by enolate 9 on Cb over
Ca. This would ultimately lead to higher amounts of
regioisomer 8. To investigate the effect of the electronic
properties of the R group on product formations, we intended
to react a number of substituted 2-aroyl-cyclohexanones 12
(with R-groups varying from electron-donating to electron-
withdrawing) with 4. The cyclohexanones 12 were synthesized
by reaction of aroyl-chlorides with the enolate of cyclo-
hexanone, albeit in low yields as a result of the purifications
needed to achieve the high quality required for our experi-
ments.
RESULTS AND DISCUSSION
■
Design of a 13C-Incorporation Approach. In the
proposed reaction mechanism, 4 is deprotonated to stabilized
enolate 9, which attacks either carbonyl carbon atom of 2-
carbonyl-cyclohexanone 5, giving the respective regioisomeric
Knoevenagel condensation intermediates 10 or 11 (Scheme 1).
Upon subsequent ring closure via the amide nitrogen, followed
by the elimination of water, the tetrahydroisoquinoline 6 or
tetrahydroquinoline 7 product is formed. This description
implies that the ultimate regiochemistry is determined by the
Conventional methods that could be used to determine
product ratios as a result of regiochemistry (e.g., LC analysis)
7356
dx.doi.org/10.1021/jo301138w | J. Org. Chem. 2012, 77, 7355−7363

The Journal of Organic Chemistry
Article
Figure 2. Representative analysis on the reaction mixture of 12b (R = CN). (A) HMBC spectrum with observed couplings between Ci−Ha/Hb for
isomer 3b (blue circles) and Cq−Hh/Hc/Hd for isomer 8b (red circles). (B) NOESY spectrum with observed couplings between Hg−Hb for isomer
3b (blue circles) and Hh−Hd for isomer 8b (red circles). (C) Magnification of 1H NMR signals for Ha (in both 13C-3b and separately ran 12C-3b)
and for Hc. Signals of Hg appear underneath the signals of one of the NH2 hydrogen atoms of remaining 4. The large doublet around 3.6 ppm is the
CH2 group from remaining 4 (split by 13C).
have several drawbacks: (1) they require sample handling like
dilution and/or filtration; (2) they cause exposure of any
remaining starting materials or intermediates to eluents; and
(3) in order to ensure accurate quantification of both isomers,
by for example UV analysis, calibration curves for all
regioisomers need to be produced. These procedures are
quite elaborate and, more importantly, may cause deviation
between the measured and true product ratios. Therefore, we
strived to determine the product ratio directly, without any
sample handling, from the crude reaction mixtures. A significant
challenge associated with this approach is the requirement to
determine absolute product regiochemistries from the reaction
mixture. We envisioned that all requirements could be met in
an approach using 13C-labeling in combination with NMR
spectroscopy. That is, aroyl-cyclohexanones 12 were reacted
with excess 13C-enriched 2-cyanoacetamide (hereafter referred
to as 4) and Et3N base to form 13C-enriched tetrahydroisoqui-
noline and 13C-enriched tetrahydroquinoline products (here-
after referred to as 3 and 8, respectively, Scheme 2). Given the
only small difference in mass between 12C- and 13C-carbon
atoms, we anticipate that any kinetic isotope effects would not
lead to a significant deviation compared to use of 12C-4. Since
the reactions were run in DMSO-d6, we also ensured that none
of the CH2 protons of 4 exchanged for deuterium atoms
through base-assisted transfer from DMSO-d6 (not shown).
Reactions were run for 6 days at 80 °C and quantitative 13C,
HMBC, and NOESY NMR spectroscopy with appropriately set
parameters (such as sufficient relaxation time) were performed
on the reaction mixtures to get the necessary information on
assignment of regiochemistry, in situ yields and product ratio.
The area under the peak for the signal of carbon atom Cq (8)
and Ci (3) was standardized to mmol amounts by using an
external standard of 4 in DMSO-d6. Mass identities were
established using LC−MS. Regiochemistry was assigned as
illustrated in detail in the next paragraph.
Assignment of Regiochemistry. Figure 2 shows the
HMBC and NOESY NMR spectra of a representative reaction
mixture started with diketone 12b (R = CN). Quantitative 13C
NMR analysis on the chemically different 13C-enriched carbon
atoms of the regioisomers (i.e., Ci and Cq) enables direct
comparison of mmol amounts. The 13C NMR signals of Ci and
Cq are clearly visible as large peaks at ∼98 and ∼100 ppm and
are typically separated by 1.0−1.5 ppm (Figure 2A; peak widths
differ, see the Supporting Information). By means of HMBC
NMR analysis, coupling between Ci and hydrogen atoms Ha/
Hb is visible (Figure 2A). Coupling between Ci and aromatic Hg
is not observed because of the distance of six bond lengths
between these atoms. In contrast, for Cq a coupling is observed
with Hc/Hd as well as with aromatic Hh. Typically, couplings
over two to three and sometimes up to four bonds may be
visible in HMBC NMR. Because of the 13C-enrichment,
however, we also observed couplings over five bond lengths.
A NOESY NMR experiment pairs Hg with Hb and Hh with Hd
(Figure 2B), confirming the correct association of atoms Cq to
Hh and, hence, the correct assignment of Cq to tetrahydroqui-
noline 8. More broadly, confirmatory proof for this assignment
1
could be obtained from the H NMR spectrum. Atom Ha
couples with He but also with nearby Ci, resulting in a triplet of
doublet for Ha (3JCH ∼ 3.9 Hz) (Figure 2C). None of the other
benzylic methylene groups (Hb, Hc, Hd) show this extra
splitting, since they are further away from the 13C center
(exemplified for Hf-coupled Hc in Figure 2C). Moreover, this
extra splitting for Ha is absent when the same incubation is
performed with 12b and 12C-4 (Figure 2C, labeled Ha(12C-
7357
dx.doi.org/10.1021/jo301138w | J. Org. Chem. 2012, 77, 7355−7363

The Journal of Organic Chemistry
Article
Table 1. (Left) Regioisomeric Ratios and in Situ Yields for Products 3 and 8 and (Right) MO Energy Calculations on Starting
Materials 12
a
b
Regioisomeric ratios and in situ yield by 13C NMR spectroscopy were determined as outlined. Semiempirical MO calculations were performed
using AM1. σp = Hammett para-value.
3b)). This observation confirms the correct assignment of Ci to
3. The protocols described here could be used consistently and
with similar results for incubations on 12a−j. It should be
noted that with decreasing amounts (vide infra) of tetrahy-
droquinoline 8 being formed for compounds 12f−j, the HMBC
correlation of Cq to Hh became too weak to be clearly visible.
Assuringly, though, in these cases all discussed key HMBC
peaks to the benzylic groups (Ha, Hb, Hc and Hd) remained
consistently visible, as did the additional splitting of Ha by Ci.
Moreover, the chemical shifts of the benzylic groups (Ha, Hb,
Hc, Hd) and of 13C labels (Ci, Cq) were found not to change
significantly in the incubations on series 12a−j. All of this
assures correct assignment of Cq to 8 in the cases where
amounts of 8 were low. Last, analytical data for all major
isomers 3 in the incubations were in accord with data of
authentic pure 12C-3, prepared by us previously6 or synthesized
new (see the Experimental Section). A representative member
of the 8 class (i.e., 12C-8a) was isolated for the same purpose,
and its data also correspond to the analysis of 8a in the
corresponding incubation.
Effect of Varying R-Groups on Product Profiles. Given
that our initial hypothesis was based on electronic effects of R
most notably on Cb (Scheme 2), substituents were selected on
the basis of the Hammett σp values10 (Table 1, left). We used
the σp values as assembled by Hansch et al.11 Table 1 shows the
product profiles per R-group. It is apparent that a correlation
between the electronic properties of the para-substituent R and
the log of the observed regioisomeric ratio exists (Figure 3A).
Compared to R = H (12g), introduction of EWG (σp > 0) gives
a significant increase in amount of isomer 8 compared to the
amount of isomer 3. For the strongest EWG (R = NO2, 12a),
the regioisomeric ratio 8:3 is ∼1:2 (Table 1). Upon
introduction of EDG (σp < 0), the regioisomeric ratio decreases
slightly compared to R = H. Taking into account that two
7358
dx.doi.org/10.1021/jo301138w | J. Org. Chem. 2012, 77, 7355−7363

The Journal of Organic Chemistry
Article
extremely low amounts of 8 (and associated errors) with even
increased EDG character did not permit us to investigate
beyond this point.
MO Calculations. Considering the competing reaction
mechanisms, we investigated whether FMO theory could
substantiate our experimental results. Toward this end, we
assume that the reaction occurs between enolate 9 and the
dicarbonyl tautomeric form of electrophiles 12 and not on any
enol form of 12.12 Given that Ca/Cb are the sites of attack
determining regiochemistry (Scheme 2), we calculated energy
levels of the 12a−j lowest unoccupied molecular orbitals
(LUMOs) in which Ca and Cb are significantly participating
(Table 1 and Figure 4). A representative visualization of the
LUMO of 12g (Figure 4B) shows that Cb is involved in the
LUMO, while Ca is not significantly. Therefore, we also
addressed the energies of the LUMO in which Ca is involved
(referred to as LUMO+, visualized for 12g in Figure 4C) in
compounds 12.
If the reactions would be dominated by HOMO−LUMO
effects, the HOMO of enolate 9 should preferably combine
with the LUMO of compounds 12. This would make Cb the
inherently preferred site of attack by 9 and should result in a
majority of tetrahydroquinoline 8. From our experimental
results, however, it is clear that the major product is always
tetrahydroisoquinoline 3, resulting from attack of 9 on carbon
Ca. LUMO+ (involving Ca) remains virtually constant in energy
in all starting materials 12 (Table 1 and Figure 4A). Thus,
although the energy level of the LUMO is consistently lower
than that of the LUMO+ in all diketones 12a−j, we
hypothesize that activation strain (i.e., reagent deforma-
tion)13−15 associated with the transition state may be
intrinsically higher for attack on Cb, arguably for steric and/
or conjugation effects. This may cause attack on Ca (i.e.,
combination with LUMO+) to dominate over attack on Cb.
Upon introduction of stronger EWG substituents, however, the
energy of the LUMO decreases to such an appreciable level
(Table 1 and Figure 4A) that carbon atom Cb gets increasingly
attacked by enolate 9. In all, the energy difference between
LUMO+ and LUMO (ΔLUMO) correlates reasonably well in
a positive fashion with the regioisomeric ratio (Figure 4A).
The results of the 13C-labeling experiments as well as AM1
calculations may be reconciled in one overall explanation. Key
is that two competing mechanisms can in principle operate:
initial attack on carbon atom Ca to give tetrahydroisoquinolines
or on Cb to give tetrahydroquinolines (shown in Scheme 2).
Our experimental data show that these two mechanisms have
different sensitivities toward the electronic properties of R: the
pathway through Ca is relatively insensitive, while the pathway
through Cb is significantly favored by EWG groups. This is
supported by our calculations, which show little variation of
LUMO+ energy levels but substantial lowering of LUMO
energy levels (in which Cb is involved) by introduction of EWG
groups. Practically, this means that with electron-donating
substituents, attack is almost exclusively at Ca giving
tetrahydroisoquinolines. As the EWG-capacity of R increases,
attack on Cb and thus tetrahydroquinoline formation
increasingly competes.
Figure 3. Correlation between the electronic properties of R and (A)
log of observed regioisomeric ratio (8:3); (B) in situ yield (%) of 8
and 3; and (C) log of in situ yield (%) of 8 and of 3. σp = Hammett
para-value.
mechanistic pathways are operative for the regioisomers
(Scheme 1), we dissected the Hammett plot into plots for
the (log of) in situ yield of each isomer (Figure 3B,C). Figure
3C reveals that formation of isoquinoline 3 is relatively
insensitive to electronic substituent effects. In contrast,
formation of quinoline 8 is favored by EWG substituents, as
indicated by the positive slope of the trend line. The relatively
lower total conversion with 12j is noteworthy, but the expected
CONCLUSION
■
In this work, we describe factors that govern the regioisomeric
outcome of the reaction of substituted 2-aroyl-cyclohexanones
with 2-cyanoacetamide. By incorporating a 13C-enriched carbon
atom into the two products (tetrahydroisoquinolines 3 and
7359
dx.doi.org/10.1021/jo301138w | J. Org. Chem. 2012, 77, 7355−7363

The Journal of Organic Chemistry
Article
Figure 4. (A) Correlation between the log of the observed regioisomeric ratio 8:3 and the calculated energy (eV) of the LUMO and LUMO+ in
starting materials 12. The difference in LUMO+ and LUMO energy (Δ LUMO) is indicated by red circles in the graph. (B) Representative
visualization of the LUMO in 12g (R = H). (C) Representative visualization of the LUMO in which Ca is involved (LUMO+) in 12g (R = H).
were dried over Na2SO4 and evaporated in vacuo to yield 744 mg (8.6
mmol, 90%) of the title compound as a colorless solid material: H
NMR (500 MHz, DMSO-d6) δ (ppm) 13.32 (br s, 1H, OH), 3.86 (d,
1JCH = 136.5 Hz, 2H); 13C NMR (126 MHz, DMSO-d6) δ (ppm)
165.8 (d, 1JCC = 57.6 Hz, CO), 115.6 (d, 1JCC= 60.5 Hz, CN), 24.7
(CH2, 13C-labeled, intense peak).
regioisomeric tetrahydroquinolines 8) we were able to
distinguish between both regioisomers in the reaction mixture
using HMBC and NOESY NMR spectroscopy. The in situ
product profile was determined using quantitative 13C NMR
spectroscopy. A correlation between the electronic properties
of the para-substituent and the regioisomeric product ratio is
demonstrated, which results from the distinctly differing
electronic sensitivities of the two involved competing pathways.
Semiempirical MO calculations on the starting materials 12
support the notion that a combination of activation strain and
of the energy gap between the LUMO and LUMO+ is
responsible for the observed regioisomeric ratios. From a
practical point of view, our work provides useful entries to this
interesting class of heterocycles and provides some guidelines
for synthetic planning en route to both regioisomeric forms.
For example, our results show that use of a nitro-substituent
followed by a reduction provides a route to the synthesis of an
otherwise difficult-to-synthesize amine-substituted tetrahydro-
quinoline isomer.
1
[2-13C]-Cyanoacetamide (4). [2-13C]-Cyanoacetic acid (700 mg,
8.13 mmol) was dissolved in dry DCM (25 mL). In small portions,
oxalyl chloride (2.0 M in DCM, 4.5 mL, 9.0 mmol) was added to the
mixture followed by two drops of DMF. The reaction mixture was
heated at reflux temperature for 3 h, during which it colored yellow.
After cooling to room temperature, NH3 gas was bubbled through the
mixture for 10 min, and the formed suspension was stirred at room
temperature for 30 min. The solvent was evaporated in vacuo, and the
yellow colored solid material was suspended in distilled THF, followed
by filtration and subsequent washing of the residue with THF. The
wash fractions were combined, and the solvent was evaporated in
vacuo to yield 667 mg of the crude product as a yellow colored solid
material. The product was recrystallized from EtOH (95%) in several
batches to yield 400 mg (4.7 mmol, 58%) of the title compound as a
light-yellow colored solid material: 1H NMR (500 MHz, DMSO-d6) δ
(ppm) 7.64 (s, 1H), 7.34 (s, 1H), 3.58 (d, 1JCH = 135.5 Hz, 2H); 13C
NMR (126 MHz, DMSO-d6) δ (ppm) 164.1 (d, 1JCC = 47.9 Hz, CN),
EXPERIMENTAL SECTION
■
1
116.3 (d, JCC = 60.1 Hz, CO), 25.3 (CH2, 13C-labeled, intense
General Experimental Procedures. Chemicals and reagents
were obtained from commercial suppliers and were used without
further purification. 1H and 13C NMR chemical shifts (δ) were
internally referenced to the residual solvent peak. LC−MS analyses
were performed using the following conditions: Xbridge (C18) 5 μm
column (100 × 4.6 mm) with solvent A (water with 0.1% formic acid)
and solvent B (acetonitrile with 0.1% formic acid), flow rate of 1.0
mL/min, start 5% B, linear gradient to 90% B in 5 min, then 5 min at
90% B, then 5 min at 5% A, total run time of 15 min. HRMS detection
was performed with an ESI hybrid quadrupole time of flight mass
spectrometer, except for compound 12b, which was analyzed with an
ion trap time of flight mass spectrometer.
peak).
General Method A. The following method is representative for
the synthesis of substituted 2-aroyl-cyclohexanones 12a−j. 1H and 13
C
NMR experiments on the products were conducted in CDCl3. Under
those conditions, all products were >95% in the keto form (labeled
1
KETO) or enol form (labeled ENOL, 12b only) as judged from H
NMR spectra.
2-(4-Nitrobenzoyl)cyclohexanone (12a). Under nitrogen at-
mosphere, 7.5 mL of LDA (2.0 M in THF, 15 mmol) was added to dry
THF (25 mL) at −78 °C. Cyclohexanone (1.47 g, 15 mmol) dissolved
in 5 mL of dry THF was added dropwise to the mixture. After stirring
the solution for 45 min at −78 °C, 4-nitrobenzoyl chloride (2.78 g, 15
mmol) dissolved in 5 mL of dry THF was added dropwise. The
mixture was allowed to warm up and stirred for 2 h at room
temperature. The mixture was quenched with 15 mL of HCl (2M).
Water (20 mL) was added, and the product was extracted with diethyl
ether (2 × 20 mL). The combined organic layers were washed with 1.0
M aq NaHCO3, dried over Na2SO4 and concentrated in vacuo to yield
a brown colored oil. Purification by flash chromatography (75%
petroleum ether/23% CHCl3/2% TEA) and subsequent recrystalliza-
tion from petroleum ether (60−80) yielded 150 mg (0.61 mmol, 4%)
[2-13C]-Cyanoacetic acid.16 [2-13C]-Bromoacetic acid (1.34 g, 9.6
mmol) was dissolved in water (30 mL). In small portions, potassium
carbonate (1.33 g, 9.6 mmol) was added. Potassium cyanide (625 mg,
9.6 mmol) was added, and the mixture was heated at reflux
temperature for three hours and stirred overnight at room temper-
ature. The flask was cooled in an ice bath, and concentrated aq HCl (2
mL) was added to the mixture dropwise (Caution! Potential release of
HCN gas!). At room temperature, solid sodium chloride was added to
the mixture until the solution was saturated. The resulting solution was
extracted with EtOAc (6 × 50 mL). The combined organic phases
7360
dx.doi.org/10.1021/jo301138w | J. Org. Chem. 2012, 77, 7355−7363

The Journal of Organic Chemistry
Article
1
of the title compound as white crystals: H NMR KETO (500 MHz,
2-(4-Methylbenzoyl)cyclohexanone (12h). Synthesized from 4-
methylbenzoyl chloride (2.81 g, 18 mmol) according to General
Method A. Yield 1.04 g (4.8 mmol, 26%) of the title compound as
white crystals: mp 98.9−103.7 °C; 1H NMR KETO (500 MHz,
CDCl3) δ (ppm) 7.79 (d, J = 8.0 Hz, 2H), 7.24 (d, J = 7.9 Hz, 2H),
4.35 (t, J = 7.0 Hz, 1H), 2.63−2.52 (m, 1H), 2.52−2.43 (m, 1H), 2.39
(s, 3H), 2.35−2.24 (m, 1H), 2.15−2.04 (m, 1H), 2.04−1.86 (m, 3H),
1.81−1.68 (m, 1H); 13C NMR KETO (126 MHz, CDCl3) δ 208.9,
197.2, 144.3, 134.2, 129.5, 128.8, 58.9, 42.4, 30.2, 27.4, 23.1, 21.8;
HRMS (EIMS) m/z calcd for C14H16NaO2 239.1048 [M + Na]+,
found 239.1043.
CDCl3) δ (ppm) 8.29 (d, J = 8.8 Hz, 2H), 8.02 (d, J = 8.8 Hz, 2H),
4.36 (dd, J = 9.5, 5.7 Hz, 1H), 2.53 (t, J = 6.1 Hz, 2H), 2.35−2.24 (m,
1H), 2.23−2.12 (m, 1H), 2.12−1.99 (m, 2H), 1.96−1.83 (m, 1H),
1.83−1.71 (m, 1H); 13C NMR KETO (126 MHz, CDCl3) δ 208.2,
196.5, 150.4, 141.3, 129.6, 124.0, 59.5, 42.6, 29.7, 27.4, 23.5; LC−MS
(ESI) m/z calcd for C13H14NO4 248.09 [M + H]+, found 248.1.
4-(2-Oxocyclohexanecarbonyl)benzonitrile (12b). Synthe-
sized from 4-cyanobenzoyl chloride (2.5 g, 15 mmol) according to
General Method A. Yield 340 mg (1.5 mmol, 10%) of the title
1
compound as white crystals: mp 101.3−103.5 °C; H NMR ENOL
2-(4-Methoxybenzoyl)cyclohexanone (12i). Synthesized from
4-methoxybenzoyl chloride (2.6 g, 15 mmol) according to General
Method A. Yield 761 mg (3.3 mmol, 21%) of the title compound as
white crystals: 1H NMR KETO (500 MHz, CDCl3) δ (ppm) 7.87 (d, J
= 8.9 Hz, 2H), 6.90 (d, J = 8.9 Hz, 2H), 4.32 (t, J = 6.5 Hz, 1H), 3.84
(s, 3H), 2.61−2.50 (m, 1H), 2.50−2.40 (m, 1H), 2.34−2.24 (m, 1H),
2.11−1.86 (m, 4H), 1.78−1.66 (m, 1H); 13C NMR KETO (126 MHz,
CDCl3) δ 208.9, 195.9, 163.7, 131.0, 129.7, 113.9, 58.6, 55.6, 42.3,
30.2, 27.4, 23.1; HRMS (EIMS) m/z calcd for C14H16NaO3 255.0997
[M + Na]+, found 255.0992.
(500 MHz, CDCl3) δ (ppm) 16.47 (s, 1H), 7.72 (d, J = 8.1 Hz, 2H),
7.62 (d, J = 8.2 Hz, 2H), 2.50 (t, J = 6.7 Hz, 2H), 2.33 (t, J = 6.1 Hz,
2H), 1.82−1.73 (m, 2H), 1.67−1.59 (m, 2H); 13C NMR ENOL (126
MHz, CDCl3) δ 191.1, 188.7, 141.6, 132.2, 128.3, 118.3, 114.0, 107.3,
32.9, 26.3, 23.4, 21.8; HRMS (EIMS) m/z calcd for C14H12NO2
226.0874 [M − H]−, found 226.0886.
Methyl 4-(2-oxocyclohexanecarbonyl)benzoate (12c). Syn-
thesized from methyl 4-(chlorocarbonyl)benzoate (2.0 g, 10 mmol)
according to General Method A. Yield 190 mg (0.73 mmol, 7%) of the
title compound as white crystals: 1H NMR KETO (500 MHz, CDCl3)
δ (ppm) 8.10 (d, J = 8.4 Hz, 2H), 7.92 (d, J = 8.4 Hz, 2H), 4.37 (dd, J
= 8.8, 5.8 Hz, 1H), 3.94 (s, 3H), 2.60−2.45 (m, 2H), 2.38−2.24 (m,
1H), 2.19−2.09 (m, 1H), 2.07−1.96 (m, 2H), 1.96−1.84 (m, 1H),
1.82−1.69 (m, 1H); 13C NMR KETO (126 MHz, CDCl3) δ 208.4,
197.4, 166.3, 139.9, 134.1, 130.0, 128.5, 59.2, 52.6, 42.5, 29.9, 27.4,
23.3; HRMS (EIMS) m/z calcd for C15H16NaO4 283.0946 [M + Na]+,
found 283.0941.
2-(4-Bromobenzoyl)cyclohexanone (12d). Synthesized from 4-
bromobenzoyl chloride (3.28 g, 15 mmol) according to General
Method A. Yield 610 mg (2.17 mmol, 15%) of the title compound as
white crystals: 1H NMR KETO (500 MHz, CDCl3) δ (ppm) 7.75 (d, J
= 8.5 Hz, 2H), 7.59 (d, J = 8.6 Hz, 2H), 4.30 (dd, J = 8.4, 5.9 Hz, 1H),
2.59−2.42 (m, 2H), 2.34−2.23 (m, 1H), 2.16−2.06 (m, 1H), 2.06−
1.84 (m, 3H), 1.80−1.68 (m, 1H); 13C NMR KETO (126 MHz,
CDCl3) δ 208.5, 196.6, 135.4, 132.1, 130.1, 128.6, 58.9, 42.4, 29.9,
27.4, 23.3; HRMS (EIMS) m/z calcd for C13H13BrNaO2 302.9997 [M
+ Na]+, found 302.9991.
2-(4-Chlorobenzoyl)cyclohexanone (12e). Synthesized from 4-
chlorobenzoyl chloride (1.81 g, 10 mmol) according to General
Method A. Yield 984 mg (3.01 mmol, 29%) of the title compound as
white crystals: 1H NMR KETO (500 MHz, CDCl3) δ (ppm) 7.82 (d, J
= 8.6 Hz, 2H), 7.42 (d, J = 8.6 Hz, 2H), 4.31 (dd, J = 8.4, 5.9 Hz, 1H),
2.59−2.44 (m, 2H), 2.35−2.22 (m, 1H), 2.16−2.07 (m, 1H), 2.07−
1.84 (m, 3H), 1.82−1.67 (m, 1H); 13C NMR KETO (126 MHz,
CDCl3) δ 208.5, 196.4, 139.9, 135.0, 130.1, 129.1, 59.0, 42.4, 30.0,
27.4, 23.3; HRMS (EIMS) m/z calcd for C13H13ClNaO2 259.0502 [M
+ Na]+, found 259.0496.
2-(4-Isopropoxybenzoyl)cyclohexanone (12j). Synthesized
from 4-isopropoxybenzoyl chloride (2.08 g, 26.8 mmol) according
to General Method A. Yield 683 mg (2.62 mmol, 25%) of the title
1
compound as white crystals: mp 105.1−106.7 °C; H NMR KETO
(500 MHz, CDCl3) δ 7.86 (d, J = 8.8 Hz, 2H), 6.88 (d, J = 8.8 Hz,
2H), 4.68−4.58 (m, 1H), 4.32 (t, J = 6.5 Hz, 1H), 2.61−2.53 (m, 1H),
2.51−2.42 (m, 1H), 2.37−2.26 (m, 1H), 2.11−1.97 (m, 2H), 1.97−
1.89 (m, 2H), 1.78−1.67 (m, 1H), 1.35 (d, J = 6.1 Hz, 6H); 13C NMR
KETO (126 MHz, CDCl3) δ 209.0, 195.8, 162.3, 131.1, 129.2, 115.3,
70.3, 58.7, 42.3, 30.3, 27.4, 23.1, 22.0; HRMS (EIMS) m/z calcd for
C16H20NaO3 283.1310 [M + Na]+, found 283.1305.
General Method B. The following method is representative for
the reactions of 2-aroyl-cyclohexanones 12 with [2-13C]-cyanoaceta-
mide 4.
Reaction of [2-13C]-Cyanoacetamide (4) with 2-(4-
Nitrobenzoyl)cyclohexanone (12a) to Yield 3a and 8a. In a
sealable NMR tube, 2-(4-nitrobenzoyl)cyclohexanone (12a) (8.68 mg,
0.035 mmol) was dissolved in 442 μL of DMSO-d6. Then, 50 μL of
[2-13C]-cyanoacetamide (4) stock solution (0.30 mg/μL of DMSO-d6,
15 mg, 0.176 mmol) and triethylamine (8.0 μL, 0.057 mmol) were
added (total volume 500 μL). The headspace above the mixture was
shortly flushed with nitrogen gas, and the NMR tube was sealed with a
screw cap. Samples were heated at 80 °C for 6 days. All samples
remained clear during incubations. After that, samples were cooled to
room temperature (all samples remained clear). NMR experiments
were recorded on the reaction mixtures to assign absolute
regiochemistries (1H, HMBC, and NOESY NMR spectroscopy) and
to determine the peak areas for Ci and Cq (quantitative 13C NMR
spectroscopy with a long D1 parameter of 80 s; spectra are available in
the Supporting Information). Peak area standardization was done by
DMSO-normalized comparison of peak areas to the peak area
observed in a 13C NMR spectrum (D1 = 80 s) of an external standard
of [2-13C]-cyanoacetamide 4 (15 mg) in DMSO-d6 (500 μL). LC−MS
experiments were conducted on the mixtures to confirm the masses of
the 13C-enriched products (two baseline separated peaks are observed
for 3a and 8a, [M + H]+ and retention time (min) are reported): 13C
NMR (only signals of 13C-enriched Ci (3a) and Cq (8a) are reported)
(126 MHz, CDCl3) δ 99.86, 97.74; LC−MS (ESI) m/z calcd for
C1513CH14N3O3 297.11 [M + H]+, found 297.0 (4.89 min) and 297.0
(5.06 min); HRMS (EIMS) m/z calcd for C1513CH14N3O3 297.1063
[M + H]+, found 297.1056.
2-(4-Fluorobenzoyl)cyclohexanone (12f). Synthesized from 4-
fluorobenzoyl chloride (2.38 g, 15 mmol) according to General
Method A. Yield 770 mg (3.5 mmol, 23%) of the title compound as
white crystals: 1H NMR KETO (500 MHz, CDCl3) δ (ppm) 7.92 (dd,
3JHH = 8.8 Hz, 4JHF = 5.4 Hz, 2H), 7.11 (t, J = 8.6 Hz, 2H), 4.33 (dd, J
= 7.9, 6.1 Hz, 1H), 2.61−2.43 (m, 2H), 2.37−2.25 (m, 1H), 2.16−2.06
(m, 1H), 2.06−1.85 (m, 3H), 1.81−1.68 (m, 1H); 13C NMR KETO
1
(126 MHz, CDCl3) δ 208.6, 196.0, 165.9 (d, JCF = 255.3 Hz), 133.2
(d, 4JCF = 3.0 Hz), 131.3 (d, 3JCF = 9.4 Hz), 115.9 (d, 2JCF = 21.9 Hz),
58.9, 42.4, 30.0, 27.4, 23.2; HRMS (EIMS) m/z calcd for
C13H13FNaO2 243.0797 [M + Na]+, found 243.0792.
2-Benzoylcyclohexanone (12g). Synthesized from benzoyl
chloride (2.62 g, 15 mmol) according to General Method A. Yield
1
Reaction of [2-13C]-Cyanoacetamide (4) with 4-(2-
Oxocyclohexanecarbonyl)benzonitrile (12b) to Yield 3b and
8b. Performed according to General Method B: 13C NMR (only
signals of 13C enriched Ci (3) and Cq (8) are reported) (126 MHz,
CDCl3) δ 99.93, 98.28; LC−MS (ESI) m/z calcd for C1613CH14N3O
277.12 [M + H]+, found 277.0 (4.64 min) and 277.0 (4.89 min);
HRMS (EIMS) m/z calcd for C1613CH13N3NaO 299.0984 [M + Na]+,
found 299.0980.
830 mg (4.1 mmol, 27%) of the title compound as white crystals: H
NMR KETO (500 MHz, CDCl3) δ (ppm) 7.89 (d, J = 8.2 Hz, 2H),
7.55 (t, J = 7.4 Hz, 1H), 7.45 (t, J = 7.7 Hz, 2H), 4.37 (dd, J = 7.6, 6.4
Hz, 1H), 2.63−2.44 (m, 2H), 2.37−2.25 (m, 1H), 2.15−2.05 (m, 1H),
2.05−1.85 (m, 3H), 1.80−1.69 (m, 1H); 13C NMR KETO (126 MHz,
CDCl3) δ 208.7, 197.7, 136.7, 133.4, 128.8, 128.6, 58.9, 42.4, 30.1,
27.4, 23.2; HRMS (EIMS) m/z calcd for C13H14NaO2 225.0891 [M +
Na]+, found 225.0886.
7361
dx.doi.org/10.1021/jo301138w | J. Org. Chem. 2012, 77, 7355−7363

The Journal of Organic Chemistry
Article
Reaction of [2-13C]-Cyanoacetamide (4) with Methyl 4-(2-
Oxocyclohexanecarbonyl)benzoate (12c) to Yield 3c and 8c.
Performed according to General Method B: 13C NMR (only signals of
13C enriched Ci (3) and Cq (8) are reported) (126 MHz, CDCl3) δ
HMBC-spectra are annotated accordingly. Compounds 12C-3e−i have
been previously described by us.6
3-Hydroxy-1-(4-nitrophenyl)-5,6,7,8-tetrahydroisoquino-
line-4-carbonitrile (12C-3a) and 2-Hydroxy-4-(4-nitrophenyl)-
5,6,7,8-tetrahydroquinoline-3-carbonitrile (12C-8a). In a 100 mL
flask, 2-(4-nitrobenzoyl)cyclohexanone 12a (1.0 g, 4.0 mmol) and
cyanoacetamide (680 mg, 8.1 mmol) were dissolved in EtOH (30
mL). Triethylamine (1.13 mL, 8.1 mmol) was added, and the mixture
was heated at reflux temperature overnight. The solvent was
evaporated to yield the crude product as a solid material. From the
crude product, 100 mg was washed with ice-cold EtOH (3 × 5 mL)
and Et2O (3 × 5 mL). Flash chromatography (EtOAc) on the
resulting solid yielded 12 mg of 12C-3a (0.041 mmol) and 10 mg 12C-
8a (0.034 mmol) as yellow solids. 12C-3a: 1H NMR (500 MHz,
DMSO) δ (ppm) 12.59 (s, 1H), 8.33 (d, J = 8.7 Hz, 2H), 7.77 (d, J =
8.7 Hz, 2H), 2.86 (t, J = 6.5 Hz, 2H), 2.27 (t, J = 6.0 Hz, 2H), 1.78−
1.68 (m, 2H), 1.62−1.52 (m, 2H); 13C NMR (126 MHz, DMSO) δ
160.3 (HMBC), 159.8, 148.0, 147.8 (HMBC), 139.3 (HMBC), 130.7,
123.5, 115.6, 114.6 (HMBC), 99.6, 29.0, 24.9, 21.8, 20.9; LC−MS 5.06
min, LC-purity > 99%; HRMS (EIMS) m/z calcd for C16H12N3O3
99.97, 98.46; LC−MS (ESI) m/z calcd for C1713CH17N2O3 310.13 [M
+ H]+, found 310.0 (4.85 min) and 310.0 (5.08 min); HRMS (EIMS)
m/z calcd for C1713CH17N2O3 310.1267 [M + H]+, found 310.1263.
Reaction of [2-13C]-Cyanoacetamide (4) with 2-(4-
Bromobenzoyl)cyclohexanone (12d) to Yield 3d and 8d.
Performed according to General Method B: 13C NMR (only signals
of 13C enriched Ci (3) and Cq (8) are reported) (126 MHz, CDCl3) δ
100.18, 98.72; LC−MS (ESI) m/z calcd for C1513CH14BrN2O 330.03/
332.03 [M + H]+, found 329.9/331.9 (5.44 min) and 329.9/331.9
(5.56 min); HRMS (EIMS) m/z calcd for C1513CH13BrN2NaO
352.0137 [M + Na]+, found 352.0139.
Reaction of [2-13C]-Cyanoacetamide (4) with 2-(4-
Chlorobenzoyl)cyclohexanone (12e) to Yield 3e and 8e.
Performed according to General Method B: 13C NMR (only signals
of 13C enriched Ci (3) and Cq (8) are reported) (126 MHz, CDCl3) δ
100.30, 98.98; LC−MS (ESI) m/z calcd for C1513CH14ClN2O 286.08
[M + H]+, found 285.9 (5.35 min) and 286.0 (5.47 min); HRMS
(EIMS) m/z calcd for C1513CH13ClN2NaO 308.0642 [M + Na]+,
found 308.0628.
1
294.0884 [M − H]−, found 294.0874. 12C-8a: H NMR (500 MHz,
DMSO) δ (ppm) 12.56 (s, 1H), 8.37 (d, J = 8.7 Hz, 2H), 7.67 (d, J =
8.7 Hz, 2H), 2.65 (t, J = 6.3 Hz, 2H), 2.00 (t, J = 6.1 Hz, 2H), 1.73−
1.64 (m, 2H), 1.61−1.51 (m, 2H); 13C NMR (126 MHz, DMSO) δ
159.8, 159.6, 151.0, 147.9, 142.1, 129.4, 123.9, 115.8, 111.9, 100.1,
27.3, 24.7, 21.7, 20.5; LC−MS 4.89 min, LC-purity 98%; HRMS
(EIMS) m/z calcd for C16H12N3O3 294.0884 [M − H]−, found
294.0881.
Reaction of [2-13C]-Cyanoacetamide (4) with 2-(4-
Fluorobenzoyl)cyclohexanone (12f) to Yield 3f and 8f.
Performed according to General Method B: 13C NMR (only signals
of 13C enriched Ci (3) and Cq (8) are reported) (126 MHz, CDCl3) δ
100.52, 99.09; LC−MS (ESI) m/z calcd for C1513CH14FN2O 270.11
[M + H]+, found 270.0 (5.01 min) and 270.0 (5.13 min); HRMS
(EIMS) m/z calcd for C1513CH14FN2O 270.1118 [M + H]+, found
270.1109.
1-(4-Cyanophenyl)-3-hydroxy-5,6,7,8-tetrahydroisoquino-
line-4-carbonitrile (12C-3b). In a 100 mL flask, crude 4-(2-
oxocyclohexanecarbonyl)benzonitrile 12b (5.07 g, 22 mmol), which
was synthesized as previously described,6 and cyanoacetamide (2.79 g,
33 mmol) were dissolved in EtOH (50 mL) by applying brief heating.
After cooling to room temperature, triethylamine (3.1 mL, 22 mmol)
was added, and the mixture was stirred at room temperature for 24 h.
The precipitated solid was filtered, washed with cold EtOH and
recrystallized from EtOH to yield 860 mg 12C-3b (3.1 mmol, 10%) as
Reaction of [2-13C]-Cyanoacetamide (4) with 2-Benzoylcy-
clohexanone (12g) to Yield 3g and 8g. Performed according to
General Method B: 13C NMR (only signals of 13C enriched Ci (3) and
Cq (8) are reported) (126 MHz, CDCl3) δ 100.30, 99.21; LC−MS
(ESI) m/z calcd for C1513CH15N2O 252.12 [M + H]+, found 252.0
(4.92 min) and 252.0 (5.03 min); HRMS (EIMS) m/z calcd for
C1513CH15N2O 252.1212 [M + H]+, found 252.1206.
1
a yellow solid: mp 281.1−283.9 °C; H NMR (500 MHz, DMSO) δ
(ppm) 12.54 (s, 1H), 7.99 (d, J = 8.4 Hz, 2H), 7.68 (d, J = 8.4 Hz,
2H), 2.84 (t, J = 6.5 Hz, 2H), 2.25 (t, J = 6.1 Hz, 2H), 1.76−1.68 (m,
2H), 1.61−1.53 (m, 2H); 13C NMR (126 MHz, DMSO) δ 160.3
(HMBC), 159.8, 148.2 (HMBC), 136.9 (HMBC), 132.4, 130.0, 118.4,
115.6, 113.6 (HMBC), 112.4, 100.8 (HMBC), 29.0, 24.9, 21.8, 20.9;
LC−MS 4.85 min, LC-purity 98%; HRMS (EIMS) m/z calcd for
C17H12N3O 274.0986 [M − H]−, found 274.0973.
Reaction of [2-13C]-Cyanoacetamide (4) with 2-(4-
Methylbenzoyl)cyclohexanone (12h) to Yield 3h and 8h.
Performed according to General Method B: 13C NMR (only signals
of 13C enriched Ci (3) and Cq (8) are reported) (126 MHz, CDCl3) δ
100.33, 99.20; LC−MS (ESI) m/z calcd for C1613CH17N2O 266.14 [M
+ H]+, found 266.0 (5.26 min) and 266.0 (5.41 min); HRMS (EIMS)
m/z calcd for C1613CH17N2O 266.1369 [M + H]+, found 266.1358.
Reaction of [2-13C]-Cyanoacetamide (4) with 2-(4-
Methoxybenzoyl)cyclohexanone (12i) to Yield 3i and 8i.
Performed according to General Method B: 13C NMR (only signals
of 13C enriched Ci (3) and Cq (8) are reported) (126 MHz, CDCl3) δ
100.44, 99.04; LC−MS (ESI) m/z calcd for C1613CH17N2O2 282.13
[M + H]+, found 282.0 (4.93 min) and 282.0 (5.13 min); HRMS
(EIMS) m/z calcd for C1613CH17N2O2 282.1318 [M + H]+, found
282.1322.
Methyl 4-(4-cyano-3-hydroxy-5,6,7,8-tetrahydroisoquinolin-
1-yl)benzoate (12C-3c). In a 100 mL flask, methyl 4-(2-
oxocyclohexanecarbonyl)benzoate 12c (500 mg, 1.9 mmol) and
cyanoacetamide (323 mg, 3.8 mmol) were dissolved in MeOH (50
mL). Triethylamine (535 μL, 3.8 mmol) was added, and the mixture
was heated at reflux temperature for four days. After evaporation of
most of the solvent, 100 mg of the precipitated solid material was
collected and purified using flash chromatography (1:1 EtOAc/
1
Heptane) to yield 16 mg 12C-3c (0.05 mmol) as a white solid: H
Reaction of [2-13C]-Cyanoacetamide (4) with 2-(4-
Isopropoxybenzoyl)cyclohexanone (12j) to Yield 3j and 8j.
Performed according to General Method B: 13C NMR (only signals of
13C enriched Ci (3) and Cq (8) are reported) (126 MHz, CDCl3) δ
NMR (500 MHz, DMSO) δ (ppm) 12.53 (s, 1H), 8.05 (d, J = 8.3 Hz,
2H), 7.61 (d, J = 8.3 Hz, 2H), 3.89 (s, 3H), 2.84 (t, J = 6.5 Hz, 2H),
2.26 (t, J = 6.3 Hz, 2H), 1.76−1.68 (m, 2H), 1.60−1.52 (m, 2H); 13C
NMR (126 MHz, DMSO) δ 165.8, 160.9 (HMBC), 159.8, 148.2
(HMBC), 136.9 (HMBC), 130.6, 129.5, 129.2, 115.8, 113.6 (HMBC),
100.8 (HMBC), 52.5, 29.0, 25.0, 21.8, 20.9; LC−MS 5.09 min, LC-
purity 95%; HRMS (EIMS) m/z calcd for C18H15N2O3 307.1088 [M
− H]−, found 307.1081.
1-(4-Bromophenyl)-3-hydroxy-5,6,7,8-tetrahydroisoquino-
line-4-carbonitrile (12C-3d). In a 50 mL flask, 2-(4-bromobenzoyl)-
cyclohexanone 12d (300 mg, 1.1 mmol) and cyanoacetamide (179 mg,
2.1 mmol) were dissolved in EtOH (25 mL). Triethylamine (297 μL,
2.1 mmol) was added, and the mixture was heated at reflux for two
days. The reaction mixture was cooled to room temperature. The solid
material was filtered and washed with ice-cold EtOH (3 × 2 mL) and
Et2O (3 × 2 mL) to yield 180 mg 12C-3d (0.54 mmol, 51%): 1H NMR
100.38, 99.04; LC−MS (ESI) m/z calcd for C1813CH21N2O2 310.16
[M + H]+, found 310.0 (5.56 min) and 310.0 (5.77 min); HRMS
(EIMS) m/z calcd for C1813CH21N2O2 310.1631 [M + H]+, found
310.1630.
Synthesis and Characterization of Purified Isomers 12C-3a−j
and 12C-8a. Compounds 3a−d, 3j, and 8a were synthesized with 12C-
cyanoacetamide, as described below. For these compounds the
quaternary aromatic 13C-signals proved hard to visualize with regular
13C NMR analysis, even after extensive scanning of concentrated
solutions. However, in the HMBC-spectra, these 13C-signals are clearly
visible. The reported 13C-signals are therefore a result of combined 13
C
and HMBC NMR analyses. The 13C-signals obtained from the
7362
dx.doi.org/10.1021/jo301138w | J. Org. Chem. 2012, 77, 7355−7363

The Journal of Organic Chemistry
Article
(500 MHz, DMSO) δ (ppm) 12.45 (s, 1H), 7.71 (d, J = 8.5 Hz, 2H),
7.41 (d, J = 8.5 Hz, 2H), 2.83 (t, J = 6.5 Hz, 2H), 2.27 (t, J = 6.2 Hz,
2H), 1.77−1.66 (m, 2H), 1.62−1.51 (m, 2H); 13C NMR (126 MHz,
DMSO) δ 160.6 (HMBC), 159.9, 148.6 (HMBC), 132.4 (HMBC),
131.4, 131.1, 123.4, 115.8, 113.6 (HMBC), 100.2 (HMBC), 29.0, 25.0,
21.9, 20.9; LC−MS 5.55 min, LC-purity >99%; HRMS (EIMS) m/z
calcd for C16H12BrN2O 327.0138 [M − H]−, found 327.0136.
1-(4-Chlorophenyl)-3-hydroxy-5,6,7,8-tetrahydroisoquino-
line-4-carbonitrile (12C-3e). This compound has been described by
us previously.6
1-(4-Fluorophenyl)-3-hydroxy-5,6,7,8-tetrahydroisoquino-
line-4-carbonitrile (12C-3f). This compound has been described by
us previously.6
3-Hydroxy-1-phenyl-5,6,7,8-tetrahydroisoquinoline-4-car-
bonitrile (12C-3g). This compound has been described by us
previously.6
ACKNOWLEDGMENTS
We thank Dr. Frans de Kanter and Prof. Matthias Bickelhaupt
for valuable advice and Maarten Sijm for technical assistance.
■
REFERENCES
■
(1) Cavasotto, C. N.; Ortiz, M. A.; Abagyan, R. A.; Piedrafita, F. J.
Bioorg. Med. Chem. Lett. 2006, 16, 1969−1974.
(2) Paronikyan, E. G.; Sirakanyan, S. N.; Noravyan, A. S.; Paronikyan,
R. G.; Dzhagatspanyan, I. A. Pharm. Chem. J. 2001, 35, 8−10.
(3) Rajapakse, H. A.; Moore, K. P.; Nantermet, P. G.; Sanders, J. M.;
Parmentier-Batteur, S.; Mark, R. J. US20100113415. 2010.
(4) Taltavull Moll, J.; Gracia Ferrer, J. WO2010/097172. 2010.
(5) Woller, J.; Spindler, K.; Sarodnick, G.; Kempter, G. Pharmazie
1996, 51, 937−940.
(6) van Linden, O. P. J.; Farenc, C.; Zoutman, W. H.; Hameetman,
L.; Wijtmans, M.; Leurs, R.; Tensen, C. P.; Siegal, G.; de Esch, I. J. P.
Eur. J. Med. Chem. 2012, 47, 493−500.
3-Hydroxy-1-p-tolyl-5,6,7,8-tetrahydroisoquinoline-4-car-
bonitrile (12C-3h). This compound has been described by us
previously.6
3-Hydroxy-1-(4-methoxyphenyl)-5,6,7,8-tetrahydroisoqui-
noline-4-carbonitrile (12C-3i). This compound has been described
by us previously.6
(7) Freeman, F.; Fraquhar, D. K.; Walker, R. L. J. Org. Chem. 1968,
33, 3648−3650.
(8) Rosowsky, A.; Papathanasopoulos, N. J. Med. Chem. 1974, 17,
1272−1276.
3-Hydroxy-1-(4-isopropoxyphenyl)-5,6,7,8-tetrahydroiso-
quinoline-4-carbonitrile (12C-3j). In a 50 mL flask, 2-(4-
isopropoxybenzoyl)cyclohexanone 12j (200 mg, 0.8 mmol) and
cyanoacetamide (129 mg, 1.5 mmol) were dissolved in EtOH (20
mL). Triethylamine (214 μL, 1.5 mmol) was added, and the mixture
was heated at reflux for four days. The reaction mixture was cooled to
room temperature, and the solid material was filtered. Thirty
milligrams of the crude material was washed with ice-cold EtOH (3
× 2 mL) and Et2O (3 × 2 mL). The remaining solid was recrystallized
from EtOH to yield 14 mg 12C-3j (0.045 mmol) as a yellow solid: mp
(9) Kasturi, T. R.; Sharma, V. K. Tetrahedron. 1975, 31, 527−531.
(10) Hammett, L. P. J. Am. Chem. Soc. 1937, 59, 96−103.
(11) Hansch, C.; Leo, A.; Taft, R. W. Chem. Rev. 1991, 91, 165−195.
(12) Campbell, R. D.; Gilow, H. M. J. Am. Chem. Soc. 1960, 82,
2389−2393.
(13) Bickelhaupt, F. M. J. Comput. Chem. 1999, 20, 114−128.
(14) de Jong, G. T.; Bickelhaupt, F. M. ChemPhysChem. 2007, 8,
1170−1181.
(15) van Zeist, W.-J.; Bickelhaupt, F. M. Org. Biomol. Chem. 2010, 8,
3118−3127.
1
293.1−295.4 °C; H NMR (500 MHz, DMSO) δ (ppm) 12.28 (s,
(16) Olszewski, J. D.; May, M. K.; Berger, D. M. US20070208164.
2007.
1H), 7.37 (d, J = 8.7 Hz, 2H), 7.01 (d, J = 8.7 Hz, 2H), 4.74−4.65 (m,
1H), 2.82 (t, J = 6.5 Hz, 2H), 2.32 (t, J = 6.2 Hz, 2H), 1.77−1.67 (m,
2H), 1.60−1.52 (m, 2H), 1.29 (d, J = 6.0 Hz, 6H); 13C NMR (126
MHz, DMSO) δ 160.4 (HMBC), 160.0, 158.6, 149.7 (HMBC), 130.6,
124.3 (HMBC), 116.1, 115.1, 113.1 (HMBC), 99.5 (HMBC), 69.4,
29.0, 25.3, 22.0, 21.8, 21.0; LC−MS 5.56 min, LC-purity >99%;
HRMS (EIMS) m/z calcd for C19H19N2O2 307.1452 [M − H]−, found
307.1444.
Semiempirical MO Calculations. Starting materials 12 were
subjected to a stochastic conformational search using MOE2011.10 to
find the global minimum structures (forcefield: MMFF94x, rejection
limit = 500, energy window = 7, rms gradient = 0.005). The lowest
energy conformations were used as starting point for an SCF
calculation, using AM1 semiempirical method (engine = MOPAC,
basis Set: AM1, solvent: none) to calculate the energies of the
molecular orbitals.
ASSOCIATED CONTENT
* Supporting Information
■
S
1H and 13C NMR spectra for compounds 4 and 12a−j.
Quantitative 13C and HMBC NMR spectra of the reaction
mixtures of 2-aroyl-cyclohexanones 12a−j with [2-13C]-
cyanoacetamide 4. 1H, 13C, HMBC, and NOESY NMR spectra
for reference compounds 12C-3a−d, 12C-3j, and 12C-8a. LC−
MS tracings for reference compounds 12C-3a−d, 12C-3j, and
12C-8a. Cartesian coordinates and total energies of the lowest
energy conformations of 12a−j. This material is available free
of charge via the Internet at http://pubs.acs.org.
AUTHOR INFORMATION
Corresponding Author
*Phone: +31(0)205987600. Fax: +31(0)205987610. E-mail: i.
de.esch@vu.nl, m.wijtmans@vu.nl.
■
Notes
The authors declare no competing financial interest.
7363
dx.doi.org/10.1021/jo301138w | J. Org. Chem. 2012, 77, 7355−7363