Enantioselective nucleophilic aromatic substitution with small-molecule chiral selectors

Article abstract of DOI:10.1002/1522-2675(200211)85:11<3605::AID-HLCA3605>3.0.CO;2-W

Small-molecule rationally designed chiral selectors have been shown to influence the stereochemical outcome of a variety of organic transformations. For instance, in a recent report, we demonstrated that a chiral selector (in conjunction with an achiral phase-transfer catalyst) could selectively inhibit one enantiomer of electron-deficient aromatic amides from forming Meisenheimer adducts (Scheme 2). We now extend this methodology to performing enantioselective nucleophilic aromatic substitutions. Initial studies involved biphasic kinetic resolutions with a chiral selector in conjunction with an achiral phase-transfer catalyst (Scheme 3). The results are consistent with previous data taken for biphasic reactions (e.g., Scheme 1) where the chiral selector effectively shields the more highly complexed enantiomer from reaction. With neutral nucleophiles such as amines, the enantioselective nucleophilic aromatic substitutions can also be conducted in single-phase systems. Several examples are given.

Full text of DOI:10.1002/1522-2675(200211)85:11<3605::AID-HLCA3605>3.0.CO;2-W

Helvetica Chimica Acta Vol. 85 (2002)
3605
Enantioselective Nucleophilic Aromatic Substitution with Small-Molecule
Chiral Selectors
by Seth E. Snyder*, Alex B. Shvets, and William H. Pirkle*
School of Chemical Sciences, University of Illinois at Urbana-Champaign, Urbana, Illinois, 61801
(tel.:217-333-3896; e-mail:sesnyder@uiuc.edu)
Dedicated to Professor Dieter Seebach on the occasion of his 65th birthday
Small-molecule rationally designed chiral selectors have been shown to influence the stereochemical
outcome of a variety of organic transformations. For instance, in a recent report, we demonstrated that a chiral
selector (in conjunction with an achiral phase-transfer catalyst) could selectively inhibit one enantiomer of
electron-deficient aromatic amides from forming Meisenheimer adducts (Scheme 2). We now extend this
methodology to performing enantioselective nucleophilic aromatic substitutions. Initial studies involved
biphasic kinetic resolutions with a chiral selector in conjunction with an achiral phase-transfer catalyst
(Scheme 3). The results are consistent with previous data taken for biphasic reactions (e.g., Scheme 1) where the
chiral selector effectively shields the more highly complexed enantiomer from reaction. With neutral
nucleophiles such as amines, the enantioselective nucleophilic aromatic substitutions can also be conducted in
single-phase systems. Several examples are given.
Introduction.
Principles of chiral recognition have been manipulated with
remarkable efficiency in the design of small-molecule chiral selectors used in the
preparation of chiral stationary phases (CSPs), leading to the ability to separate, with
significant selectivity factors, the enantiomers of many classes of racemic analytes [1].
Since early studies involving chiral separations were carried out in these laboratories, a
variety of CSPs have been produced that utilize readily available naturally occurring
enantiomers such as polysaccharides, cyclodextrins, proteins, alkaloids, and antibiotics
(for a review, see [2]). Nonetheless, the rationally designed CSPs developed from
small, relatively simple chiral selectors offer several advantages. For instance, the
selectors are inexpensive, quite stable, available in both enantiomeric forms, and can be
recovered for re-use. Furthermore, the selectors usually operate through well-defined
chiral-recognition mechanisms [3]. Chiral discrimination is rationalized through the
formation of energetically nonequivalent transient diastereoisomeric complexes and
generally relies on intermolecular interactions such as H-bonding, electrostatic forces,
steric effects, and p-p interactions. Since the total structure of each of these selectors
can be controlled, they can be tailored and optimized to the extent that one
understands the mechanisms by which they differentiate between enantiomers. These
principles will undoubtedly find application in the design of small-molecule enzyme-
like catalysts and inhibitors (for reviews, see [4]). In this context, we are investigating
the utilization of chiral selectors in a variety of enantioselective organic transformations.
In an earlier report, we described a method of performing highly enantioselective
kinetic resolutions in biphasic media [5]. When organic solutions of an N-(3,5-

3606
Helvetica Chimica Acta Vol. 85 (2002)
dinitrobenzoyl) ester of a racemic amino acid (e.g., (Æ)-1) and chiral selector (S)-2
were added to 2m NaOH, enantioselective hydrolysis ensued (Scheme 1). Enantioselec-
tivity increases with increasing amount of selector, a result that suggests selective inhi-
bition of the more strongly complexed enantiomer. Reaction rates and enantioselectivi-
ties of hydrolysis are greatest in nonpolar organic solvents such as hexane or CCl4.
Typically, the utility of a kinetic resolution (for reviews, see [6]) is measured by the stereo-
selectivity factor (s), which gives the relative rates of reaction of the two enantiomers.
Kinetic resolution is considered a practical and general route to optical resolution of
chiral compounds when s > 10. Stereoselectivity factors for hydrolysis of (Æ)-1 in hexane
(at 08) approach 100 (ee greater than 90% at 50% conversion), although one or more
molar equivalents of (S)-2 are required. Addition of a quaternary ammonium phase-
transfer catalyst (PTC) results in the formation of a burgundy colored, sludge-like salt
as a by-product of the reaction, initially suspected of being a Meisenheimer adduct.
Scheme 1
Attempts to hydrolyze racemic amides such as (Æ)-3 and (Æ)-4 (Fig. 1) under
biphasic conditions in the presence of (S)-2 were not successful. However, when
reactions were carried out under PTC conditions in hexane, the burgundy colored,
sludge-like salt was formed. In halogenated solvents such as CH2Cl2 or CCl4, the
organic layer turns red, but no salt precipitates. In all cases, monitoring the unreacted
amide in the organic layer by chiral HPLC revealed that the process is quite
enantioselective, the more highly associated enantiomer once again being sequestered
from reaction. Upon acidic workup, the dark red color vanishes, and a significant
quantity of the starting amide is regenerated. Subsequent analysis shows that the
colored species is a Meisenheimer adduct 6 (Scheme 2). The unusual stability of the
complex is attributed to an ion-pairing interaction between the anionic Meisenheimer
adduct and the quaternary ammonium cation [7]. Indeed, the overall equilibrium is
dependent upon the quantity of PTC added. With acetonate as a nucleophile, the C-
adducts are formed even more readily, most likely a result of the added stability in
comparison with the corresponding O-adducts (e.g., 6) [8]. As with enantioselective

Helvetica Chimica Acta Vol. 85 (2002)
3607
Fig. 1. Racemic amides and chiral selector used in kinetic resolutions
Scheme 2
hydrolysis, s factors for these kinetic resolutions are substantial when at least 1 mol-
equiv. of the chiral selector is used, exceeding 100 for kinetic resolutions of (Æ)-3 and 10
for kinetic resolutions of (Æ)-4 (at 08). Furthermore, enantioselective Meisenheimer
adducts are formed enantioselectively with a variety of chiral selectors. With selector
(R)-5 (2 mol-equiv.), kinetic resolutions of (Æ)-4 give s factors exceeding 15. Here, we
report a novel method to perform enantioselective nucleophilic aromatic substitution
(SNAr), a reaction known to proceed through the formation of a Meisenheimer
complex (for reviews, see [9]).
Results and Discussion. Although NMR, EPR, and UV/VIS spectra of 6 are
consistent with previous data for Meisenheimer adducts [10], isolation of the ion pairs is
difficult owing to slow reversion to the starting amide or oxidation of the adduct
(particularly when reactions are not performed under O2-free conditions.) For
reactions in CCl4, where the organic layer remains homogeneous, the oxidative by-
products could be isolated. For instance, reaction of (Æ)-4 with acetonate as a
nucleophile results, after stirring for several hours, in oxidation of the C-adduct. When
reactions are performed in the presence of selector (R)-5, the isolated Janovsky
product (7; Fig. 2) is enriched in enantiomer from the less strongly complexed
enantiomer of (Æ)-4, while the residual unreacted amide (following acidic workup) is
enriched in the more strongly complexed enantiomer.
These results prompted us to investigate nucleophilic aromatic substitution (SNAr),
reactions that proceed through the formation of intermediate Meisenheimer adducts.
Nucleophilic aromatic substitution is an important step in the syntheses of a variety of
pharmaceuticals, including hypotensive drugs, antidepressants, and antitumor agents
[11]. An enantioselective approach to performing these reactions may prove to have

3608
Helvetica Chimica Acta Vol. 85 (2002)
Fig. 2. Isolated Janovsky product from kinetic resolution of (Æ)-4
some synthetic utility. As an initial demonstration, a CCl4 solution of the p-Cl-
substituted dinitrobenzamide (Æ)-8, when treated with a methanolic NaOH solution in
the presence of selector (R)-5 and tetrahexylammonium chloride (THAC), affords the
enantiomerically enriched MeO-substituted product 9 (Scheme 3). After 10 min, 50%
of (Æ)-8 had been converted to 9 as well as to a small quantity of the phenol resulting
from replacement of the Cl by OH. The ee of both the residual 8 and product 9 was ca.
60%, enriched in the most and least strongly complexed enantiomers respectively. At
72% conversion, the ee of residual 8 was 96%. The reaction can also be conducted with
catalytic amounts of THAC, giving nearly identical enantioselectivities. Reactions run
with NaOH solutions proceed with similar enantioselectively, although stoichiometric
amounts of THAC are required owing to the stable ion-pair formed between the
quaternary ammonium cation and the product phenolate anion. Thiols, which are
considerably more nucleophilic and acidic than the corresponding alcohols, react
readily with (Æ)-8 in a mildly basic saturated NaHCO3 solution (Scheme 4). Reactions
proceed cleanly under phase-transfer conditions (e.g., with 1.5-mol-% THAC) with no
catalyst poisoning owing to the lack of production of the phenolate by-product.
Stereoselectivity factors for kinetic resolutions appear to be independent of whether
the nucleophile is an alcohol or a thiol. To our knowledge, these are the first examples
of nonenzymatic enantioselective SnAr reactions yet reported.
Scheme 3
The results from enantioselective aromatic substitutions are consistent with those of
the enantioselective hydrolyses (Scheme 1). In both cases, the chiral selector
apparently inhibits one enantiomer from reaction. Although differential complexation
of enantiomers appeared to be the governing factor accounting for the enantioselec-
tivities observed in these biphasic reactions, one cannot rule out the occurrence of a

Helvetica Chimica Acta Vol. 85 (2002)
3609
Scheme 4
special interfacial orientation owing to the presence of the selector. Hence, reactions
performed in a monophasic environment were expected to be informative. Since SNAr
reactions can be achieved readily with neutral nucleophiles such as amines, this reaction
provides a convenient means of studying the monophasic counterpart to the biphasic
kinetic resolutions discussed above. An example is provided in Scheme 5. As with the
biphasic reactions, the more highly complexed enantiomer is sequestered from
reaction. Enantioselectivities are similar to those of the biphasic reaction, suggesting
that differential complexation is the dominant enantiomer-differentiating factor for
both the biphasic and the monophasic reactions. Interestingly, dilution or concentration
of reagents (with the same number of mol-equiv. of selector and racemic amide) does
not have a profound effect on enantioselectivity of these systems. Reactions run at 08
show increased enantioselectivity over those run at room temperature. However,
lowering the temperature to À 168 does not significantly further enhance enantiose-
lectivity. Kinetic resolutions of an electron-deficient amide derived from leucine ((Æ)-
12) with selector (S)-2 as a chiral complexing agent also show reasonably high s factors
(Scheme 6).
Scheme 5
The synthetic utility and generality of the enantioselective aromatic substitution
reaction would be extended significantly if reactions could be performed on reagents
containing only one NO2 group on the aromatic moiety. Fluorinated aromatic
compounds containing but one NO2 substituent tend to react readily with a variety
of nucleophiles. Addition of BuNH2 to (Æ)-14 in the presence of selector (S)-2

3610
Helvetica Chimica Acta Vol. 85 (2002)
Scheme 6
(Scheme 7) gives the aromatic amine 15 with modest enantioselectivity (s factor ca.
4.5). The reduced enantioselectivity of the reaction, relative to the kinetic resolution of
(Æ)-12, is likely the result of reduced complexation of (Æ)-14 with the selector. These
types of reactions are currently being investigated with other racemic substrates and
chiral selectors.
Scheme 7
Conclusions. Results presented in this manuscript display our initial attempts to
utilize small-molecule chiral selectors to effect enantioselective processes. Although
substrates were chosen to demonstrate −proof of principle×, it is clear that the impressive
stereoselectivity factors obtained can potentially bring these reactions into the realm of
practical kinetic resolutions. Furthermore, the methodology provides a simplistic
means of screening numerous chiral selectors that could find application as both CSPs
and chiral catalysts. As understanding of the requirements for high levels of chiral
recognition advances, the sophistication and generality in the design and utilization of
chiral selectors is expected to follow suit.
The enantioselective SNAr reactions discussed above are a significant advance.
Besides the potential practical value, the reactions provide a means of comparing the
biphasic enantioselective reactions with single-phase reactions. Obviously, kinetic
studies of monophasic reactions are far less complicated than the corresponding
biphasic reactions, a fact that could be important in elucidating the mechanistic basis of
the observed enantioselectivities. Ongoing studies in this area are focused on expanding
the scope of substrates (e.g., heterocyclic aromatic amines) and organic reactions that
can be influenced by chiral selectors.

Helvetica Chimica Acta Vol. 85 (2002)
3611
Experimental Part
General. The following compounds and solvents used in the studies were purchased from the
manufacturers:tetrahexylammonium chloride (THAC), triethylamine (NEt 3), Et2NH, leucine, di(tert-butyl)
dicarbonate, dicyclohexylcarbodiimide (DCC), 4-fluoro-3-nitrobenzoic acid from Aldrich; NaOH, NaHCO3,
THF, DMF, CH2Cl2, CCl4 (old supply), acetone, MeOH from Fisher. 4-Chloro-3,5-dinitrobenzoic acid was
available from previous studies and used in the synthesis without further purification:(m.p. 161 163 8; [12]:
159 1628). All solvents were distilled prior to use. Anh. CH2Cl2 and Et3N were distilled from CaH2. THF was
distilled from Na, benzophenone was added as an indicator. DMF was distilled from CaO and stored over
activated molecular sieves (4 ä). Distilled H2O was obtained from a Millipore water purification system. H2O
was collected when resistivity reached 14 MW. TLC:0.2-mm silica-gel glass-coated plates ( Merck) with F254
indicator. Flash chromatography (FC): Merck 40 63-mm silica gel. HPLC was conducted with the following
equipment:injectors: Beckman, Rheodyne; pumps: Rainin Rabbit-HPX, Beckman 110B, Alcott 760; fixed-
wavelength UV detectors: LDC/Milton Roy, Altech 450 at 254 nm, variable-wavelength UV detector: Linear
UVIS 200; recorder/integrator: HP 3390A. Chromatographic runs were conducted at ambient temp. with the
flow rate of 2 ml/min, dimensions of all anal. HPLC columns were 24 cm  4.6 mm (unless otherwise indicated).
The void volume was determined by injecting 1,3,5-tris(tert-butyl)benzene. The retention factor (kn) was
calculated on the basis of the following equation:
kn ˆ (tn À t0)/t0
where tn is the retention time of an analyte (enantiomer) and t0 is the retention time of 1,3,5-tris(tert-
butyl)benzene. The separation factor (a) is a ratio of the retention factors of two enantiomers and calculated
according to the equation:
a ˆ k2/k1 ˆ (t2 À t0)/(t1 À t0)
Calculation of the stereoselectivity factor (s) was accomplished according to the equation:
s ˆ ln [1 À C (1 ‡ ee)]/ln [1 À C (1 À ee)]
where C is the extent of conversion of starting material and ee is enantiomeric excess of the remaining starting
material. For all chromatographic assays, conversion was determined with the selector as an internal standard.
1H- and 13C-NMR spectra: Varian Unity 400 or 500 spectrometers at −The Varian Oxford Instruments Center for
Excellence in NMR Laboratory× at the University of Illinois School of Chemical Science; all chemical shifts are
reported in ppm (d) with the solvent reference CDCl3 (d 7.26 ppm) for 1H and CDCl3 (d 77.0) for 13C spectra;
coupling constants (J) in Hz; 13C-NMR proton-decoupled. Low- (ZAB-SE instrument) and high-resolution (70-
SE-4F instrument) fast atom bombardment (FAB) mass spectra (MS) as well as matrix-assisted laser desorption
ionization (MALDI, −Voyager× instrument) MS were obtained in the Mass Spectrometry Laboratory, School of
Chemical Sciences, University of Illinois. Elemental analyses were performed at the Analytical Laboratories of
the University of Illinois, School of Chemical Science.
Reaction of 3,5-Dinitro-N-(1,2,3,4-tetrahydrophenanthren-4-yl)benzamide (4) with Acetone. Racemic 4
(0.52 mmol) and THAC (0.4 mmol) were dissolved in 10 ml of benzene in a scintillation flask equipped with a
Teflon lining and a magnetic stir bar. Acetone (30 ml) was added via syringe followed with 10 ml of 2m NaOH.
The mixture turned dark blue immediately on mixing. The vial was capped and vigorously stirred for 4 h. Phases
were separated. The aq. phase was acidified with 2m HCl. The brown gel precipitate was removed by filtration
through −Whatman× #4 paper. The solid was dissolved in CH2Cl2, dried (Na2SO4), and analyzed by MALDI-MS.
Clusters of peaks were observed, these being separated by molecular-weight increments close to the molecular
weight of the starting material. This indicates that some oligomerization is occurring. The org. phase was
quenched with 2m HCl for ca. 50 min. The color very slowly changed from dark blue to red and finally to yellow.
The org. phase was dried (Na2SO4) and concentrated under vacuum. The majority of the starting material was
recovered by crystallization from THF/hexane. The mother liquor was collected, concentrated under vacuum,
and chromatographed on silica gel with pure CH2Cl2, followed with 10% acetone in CH2Cl2. All recovered
starting material was combined, yielding a total of 172 mg (86%). Janovsky product 7 (5 mg) was isolated as a
white solid.

3612
Helvetica Chimica Acta Vol. 85 (2002)
3,5-Dinitro-2-(2-oxopropyl)-N-(1,2,3,4-tetrahydrophenanthren-4-yl)benzamide ((Æ)-7). M.p. > 2398 (dec.).
1H-NMR (400 MHz, CDCl3):1.58 ( s, 3 H); 1.82 (m, 1 H); 2.05 (m, 2 H); 2.32 (m, 1 H); 2.98 (m, 2 H); 3.84
(d, JAB ˆ 17.1, 1 H); 3.99 (d, JAB ˆ 17.1, 1 H); 6.00 (m, 1 H); 6.98 (d, J ˆ 9.0, 1 H); 7.24 (d, J ˆ 8.5, 1 H); 7.50
(ddd, J ˆ 7.1, 7.1, 0.7, 1 H); 7.59 (ddd, J ˆ 7.7, 7.7, 1.2, 1 H); 7.73 (d, J ˆ 8.5, 1 H); 7.84 (d, J ˆ 7.8, 1 H); 8.06 (d, J ˆ
8.5, 1 H); 8.58 (d, J ˆ 2.4, 1 H); 8.81 (d, J ˆ 2.4, 1 H). 13C-NMR (100 MHz, CDCl3); 17.8; 29.1; 29.5; 29.7; 43.8;
44.37; 120.9; 122.5; 125.5; 126.97; 127.03; 128.36; 128.41; 128.8; 129.1; 131.6; 132.6; 133.7; 136.2; 142.9; 146.8;
‡
149.4; 163.6; 205.0. HR-FAB-MS 448.1509; ([M ‡ 1] , C24H21N3O6; calc. 448.1509).
4-Chloro-3,5-dinitro-N-(1,2,3,4-tetrahydrophenanthren-4-yl)benzamide ((Æ)-8). 2.0 g of 4-chloro-3,5-dini-
trobenzoic acid was suspended in 10 ml of SOCl2. 0.1 ml of dry DMF was added, and the mixture was brought to
reflux for 5 h. SOCl2 was distilled and residual liquid was poured into dry hexane. However, the acid chloride
oiled out and did not crystallize. Hexane and residual SOCl2 were removed under vacuum to yield 2.3 g of
viscous oil as a product. The crude acid chloride 355 mg (1.34 mmol) was combined with the HCl salt of the
amine (1,2,3,4-tetrahydrophenanthren-4-amine; 0.67 mmol) [1] and suspended in 3 ml of dry CH2Cl2. The
mixture was cooled in an ice/NaCl bath to À 108 and 0.19 ml of dry Et3N was added dropwise via syringe. All
solid material dissolved, and the solution turned a clear light yellow color. A precipitate began to develop after
ca. 10 min. The soln. was stirred under N2 in an ice bath for 1 h. A thick slurry formed. The resulting suspension
was diluted with ca. 3 ml of hexane and filtered. The resulting solid (product along with HCl salt of Et3N) was
dissolved in ca. 100 ml of CHCl3 and extracted in sequence with 5% NaHCO3, 1m HCl, sat. NaCl soln. and
distilled H2O. The org. phase was dried (MgSO4), filtered, and concentrated under vacuum. The resulting solid
was recrystallized from CH2Cl2/hexane to yield 201 mg (70%) of (Æ)-8. White solid. M.p. > 2488 (dec.).
1H-NMR (400 MHZ, CDCl3):1.89 ( m, 1 H); 2.03 (m, 2 H); 2.41 (m, 1 H); 3.03 (m, 2 H); 5.95 (m, 1 H); 6.50
(d, J ˆ 7.3, 1 H); 7.27 (d, J ˆ 8.5, 1 H); 7.45 (ddd, J ˆ 7.6, 6.2, 1.4, 1 H); 7.50 (ddd, J ˆ 8.1, 6.8, 1.6, 1 H); 7.77
(d, J ˆ 8.5, 1 H); 7.82 (dd, J ˆ 7.8, 1.5, 1 H); 7.84 (d, J ˆ 8.3, 1 H); 8.30 (s, 2 H). 13C-NMR (100 MHZ, 1,4-
(D8)dioxane):18.3; 30.1; 30.8; 45.1; 123.1; 123.6; 126.1; 127.3; 127.7; 128.7; 129.1; 129.4; 130.6; 133.2; 133.5;
‡
136.0; 136.9; 150.4; 161.7. HR-FAB-MS:426.0856 ([ M ‡ 1] , C21H16ClN3O5; calc. 426.0857). Anal. calc. for
C21H15ClN3O5:C 59.23, H 3.79, N 9.87, Cl 8.33; found:C 59.41, H 3.86, N 9.54, Cl 8.41.
When conditions are not controlled as specified above, some formation of an adduct of the starting amine at
the p-position of 8 is observed (by 1H-NMR). This by-product can be separated from the product by
chromatography on a semi-prep. −Pirkle Covalent d,l-Phenylglycine× HPLC column (Regis Technologies) with
hexane/CH2Cl2 1 :1 at a flow rate of 8.5 ml/min and ambient temp.
4-Methoxy-3,5-dinitro-N-(1,2,3,4-tetrahydrophenanthren-4-yl)benzamide ((Æ)-9). Racemic 8 (0.05 mmol),
chiral selector (R)-5 (0.1 mmol), and THAC (0.025 mmol) were dissolved in 0.5 ml of CH2Cl2 and 4.5 ml of
CCl4. The reaction was carried out in a 25-ml scintillation vial with a Teflon cover under the cap. The soln. was
cooled in an ice bath. A separate soln. of 1m NaOH (4 ml) and pure MeOH (1 ml) was prepared and also cooled.
The cold aq. soln. was added to the org. soln. all at once, and the mixture was stirred vigorously in a capped vial.
Color slowly developed, at first light orange, then changing to raspberry-pink. The soln. remained transparent
throughout the reaction. The reaction was monitored by chiral HPLC with a column available from Regis
Technologies. CSP:( S,S)-ULMO, mobile phase:15% of i-PrOH in hexane, detection at l ˆ 254 nm, starting
material: k1 ˆ 3.24, k2 ˆ 5.05; product: k1 ˆ 4.01, k2 ˆ 6.84. Selector (R)-5 [7] was used as an internal standard.
HPLC Assay after 15 min reaction time indicated 68% conversion. After 30 min, the phases were separated, and
the org. phase was quenched with 1m HCl. The org. phase turned light yellow. The org. phase was dried (Na2SO4)
and concentrated under vacuum. The crude mixture was chromatographed on silica gel in 1% Et2O/CH2Cl2.
Starting amide 8 was recovered:5.2 mg (24%) with ee of 96%. Product 9 (15.7 mg, 72%) was obtained as a
white solid with ee 36%. M.p. > 2368 (dec.) (ee 36%). 1H-NMR (400 MHz, CDCl3); 1.90 (m, 1 H); 2.02
(m, 2 H); 2.39 (m, 1 H); 3.02 (m, 2 H); 4.08 (s, 3 H); 6.46 (d, J ˆ 7.6, 1 H); 7.27 (d, J ˆ 9.3, 1 H); 7.45 (ddd, J ˆ
7.3, 7.3, 1.0, 1 H); 7.49 (ddd, J ˆ 6.7, 6.7, 1.3, 1 H); 7.82 (dd, J ˆ 8.3, 1.5, 1 H); 7.86 (d, J ˆ 8.3, 1 H); 8.37 (s, 2 H).
13C-NMR (100 MHz, CDCl3); 18.1; 29.1; 29.9; 44.7; 65.0; 114.7; 122.4; 125.5; 127.2; 127.5; 127.9; 128.7; 128.79;
‡
128.81; 130.2; 131.8; 132.5; 136.3; 145.0; 161.1. HR-FAB-MS:422.1354 ([ M ‡ H] , C22H20N3O6; calc. 422.1353).
The percent of remaining starting material calculated from HPLC data obtained just prior to isolation was 26%.
A strongly adsorbed yellow band was eluted from silica gel after the initial chromatography with pure MeOH,
followed by pure CH2Cl2. The yellow org. soln. was dried (Na2SO4) and concentrated under vacuum to yield
5.5 mg of a mixture of THAC and p-OH derivative of 8. The by-product was analyzed by 1H-NMR and negative
FAB-MS. Although the alkyl proton region is obscured with signals from THAC, the aromatic region is similar
to that of the starting material. MS Data indicates that the by-product is a phenol derived from the starting
‡
material. FAB-MS:406.0 ([ M À 1]À; calc. for C21H17N3O6 (M ) 407.1).

Helvetica Chimica Acta Vol. 85 (2002)
3613
General Procedure for the Synthesis of Racemic 12 and 14. Synthesis began from racemic leucine by first
protecting the amine functionality with a Boc group ((Boc)2O) [13], followed by DCC-mediated coupling to
diethylamide [14], deprotection with TFA [14] and finally acylation with one of the activated acids (analogously
to preparation of (Æ)-8). 4-Fluoro-3-nitrobenzoyl chloride was prepared similarly to 4-chloro-3,5-dinitroben-
zoyl chloride as in the preparation of 8 and used in the reaction without further purification.
4-Chloro-N-[1-(diethylcarbamoyl)-3-methylbutyl]-3,5-dinitrobenzamide ((Æ)-12) was obtained as a white
solid. M.p. 183 1868 . 1H-NMR (400 MHz, CDCl3):0.95 ( d, J ˆ 6.6, 3 H); 1.01 (d, J ˆ 6.6, 3 H); 1.14 (t, J ˆ 7.1,
3 H); 1.38 (t, J ˆ 7.1, 3 H); 1.41 (m, 1 H); 1.80 (m, 2 H); 3.18 (dq, J ˆ 13.8, 7.0, 1 H); 3.38 (dq, J ˆ 14.8, 7.4, 1 H);
3.48 (dq, J ˆ 15.0, 7.5, 1 H); 3.67 (dq, J ˆ 13.9, 7.0, 1 H); 5.09 (ddd, J ˆ 11.1, 8.1, 2.9, 1 H); 8.43 (s, 2 H). 13C-NMR
(100 MHz):12.7; 14.0; 21.2; 23.4; 24.9; 40.9; 41.0; 42.3; 49.3; 123.2; 126.3; 133.7; 149.2; 161.2; 172.5. FAB-MS:
‡
415.1 ([M ‡ 1]1; calc. for M , 414.13). Anal. calc. for C17H23ClN4O6:C 49.22, H 5.59, N 13.51; found:C 49.05, H
5.52, 13.25.
N-[1-(Diethylcarbamoyl)-3-methylbutyl]-4-fluoro-3-nitrobenzamide ((Æ)-14) was obtained as a white
solid:57% yield. M.p. 109 110 8. 1H-NMR (400 MHz, CDCl3,):0.95 ( d, J ˆ 6.6, 3 H); 1.03 (d, J ˆ 6.6, 3 H); 1.15
(t, J ˆ 7.2, 3 H); 1.34 (t, J ˆ 7.2, 3 H); 1.42 (m, 1 H); 1.79 (m, 2 H); 3.23 (dq, J ˆ 13.8, 7.0); 3.41 (dq, J ˆ 14.7, 7.4);
3.47 (dq, J ˆ 14.7, 7.4); 3.63 (dq, J ˆ 13.8, 7.0); 5.11 (ddd, J ˆ 11.1, 8.1, 2.9); 7.25 (dd, J ˆ 10.1, 8.7, 1 H); 7.97
(d, J ˆ 8.1, 1 H); 8.05 (ddd, J ˆ 8.7, 4.2, 2.3, 1 H); 8.52 (dd, J ˆ 7.0, 2.3, 1 H). 13C-NMR (100 MHz, ‡ 408):13.1;
14.6; 21.9; 23.7; 25.2; 40.9; 42.3; 42.7; 48.8; 118.7 (d, J(C,F) ˆ 21.4); 125.6 (d, J(C,F) ˆ 1.5); 131.2 (d, J(C,F) ˆ
3.8); 134.4 (d, J(C,F) ˆ 9.2); 137.5 (d, J(C,F) ˆ 5.8); 157.3 (d, J(C,F) ˆ 270.1); 163.7; 172.2. FAB-MS:354 ([ M ‡
‡
‡
1] ; calc. for M , 353.18). Anal. calc. for C17H24FN3O4:C 57.78, H 6.85, N 11.89; found:C 57.45, H 6.70, N 11.82.
General Procedure for the Syntheses of Racemic 10, 11, 13, and 15. The starting material (compounds 8, 12,
and 14) was dissolved in dry CH2Cl2 (concentration range:0.025 0.15 m) and placed in a dry sealed tube
equipped with a stir bar. Nucleophile (BuNH2 or PrSH) was added in 10-fold excess. Dry K2CO3 was used as a
base (in equimolar quantity to thiol) for preparation of 10. The color of the reaction solns. became quickly
yellow as product was being formed. The reactions were stirred at r.t. for 10 12 h in a sealed tube. The reaction
temp. for preparation of 15 was 408. At the end of the reaction time, the contents of the reaction soln. were
diluted with 20 ml of CH2Cl2 and extracted sequentially with 1m HCl, brine, and distilled H2O. Org. phase was
dried (MgSO4), filtered, and concentrated under vacuum. Compounds 10 and 11 were directly recrystallized
from hexane/CH2Cl2. Compounds 13 and 15 were chromatographed on silica gel in 1% Et2O/5% Acetone/
CH2Cl2, followed by crystallization from hexane/CH2Cl2. Reported yields are after crystallization.
3,5-Dinitro-4-(propylsulfanyl)-N-(1,2,3,4-tetrahydrophenantren-4-yl)benzamide ((Æ)-10) was obtained
1
from (Æ)-8 as a pale yellow solid:68% yield. M.p. > 1808 (dec.). H-NMR (500 MHz, CDCl3):0.93 ( t, J ˆ 7.4,
3 H); 1.54 (tq, J ˆ 7.7, 7.3, 2 H); 1.57 (m, 1 H); 1.89 (m, 1 H); 2.02 (m, 2 H); 2.38 (d, J ˆ 12.9, 1 H); 2.81 (t, J ˆ 7.4,
2 H); 3.02 (m, 2 H); 5.93 (m, 1 H); 6.53 (d, J ˆ 7.1, 1 H); 7.25 (d, J ˆ 8.4, 1 H); 7.44 (dd, J ˆ 7.3, 7.3, 1 H); 7.49
(ddd, J ˆ 7.7, 7.7, 1.4, 1 H); 7.74 (d, J ˆ 8.4, 1 H); 7.80 (d, J ˆ 7.7, 1 H); 7.85 (d, J ˆ 8.4, 1 H); 8.15 (s, 2 H).
13C-NMR (125 MHz):13.0; 18.1; 22.8; 29.0; 29.8; 38.5; 44.7; 122.4; 124.8; 125.5; 127.2; 127.9; 128.6; 128.79;
‡
‡
128.80; 131.8; 132.5; 135.2; 136.4; 154.3; 161.1; 169.0. FAB-MS:466.1 ([ M ‡ 1] ; calc. for M , 465.14).
4-(Butylamino)-3,5-dinitro-N-(1,2,3,4-tetrahydrophenantren-4-yl)benzamide ((Æ)-11) was obtained from
1
(Æ)-8 as a yellow solid:78% yield. M.p. > 2258 (dec.). H-NMR (400 MHz, CDCl3):0.92 ( t, J ˆ 7.3, 3 H); 1.38
(tq, J ˆ 7.5, 7.5, 2 H); 1.64 (tt, J ˆ 7.3, 7.3, 2 H); 1.95 (m, 2 H); 2.36 (d, J ˆ 10.5, 1 H); 5.93 (m, 1 H); 6.45 (d, J ˆ
7.6, 1 H); 7.25 (d, J ˆ 7.6, 1 H); 7.42 (ddd, J ˆ 7.4, 7.4, 1.1, 1 H); 7.47 (ddd, J ˆ 7.6, 7.6, 1.5, 1 H); 7.73 (d, J ˆ 8.5,
1 H); 7.79 (dd, J ˆ 7.8, 1.2, 1 H); 7.87 (d, J ˆ 8.3); 8.52 (s, 2 H); 8.58 (br. s, 1 H). 13C-NMR (100 MHz):13.6; 18.1;
19.8; 29.2; 29.9; 31.9; 44.3; 46.3; 119.6; 122.6; 125.3; 127.1; 127.9; 128.5; 128.6; 129.1; 130.6; 131.9; 132.5; 136.2;
‡
‡
136.8; 141.3; 161.8. FAB-MS:463.5 ([ M ‡ 1] ; calc. for M ; 462.50).
4-(Butylamino)-N-[1-(diethylcarbamoyl)-3-methylbutyl]-3,5-dinitrobenzamide ((Æ)-13) was obtained from
(Æ)-12 as a yellow solid:53% yield. M.p. 135 137 8. 1H-NMR (400 MHz, CDCl3):0.94 ( d, J ˆ 6.1, 6 H); 0.97
(t, J ˆ 7.3, 3 H); 1.14 (t, J ˆ 7.1, 3 H); 1.30 (td, J ˆ 10.9, 2.8, 1 H); 1.44 (t, J ˆ 7.1, 3 H); 1.46 (m, 2 H); 1.72 (tt, J ˆ
7.4, 7.4, 2 H); 1.86 (dtd, J ˆ 10.0, 6.7, 3.5, 1 H); 1.95 (ddd, J ˆ 13.8, 12.1, 3.4, 1 H); 2.96 (m, 1 H); 3.05 (m, 2 H);
3.34 (dq, J ˆ 14.6, 7.3, 1 H); 3.52 (dq, J ˆ 14.6, 7.3, 1 H); 3.97 (dq, J ˆ 13.9, 7.0, 1 H); 5.04 (ddd, J ˆ 11.1, 8.2, 2.9,
1 H); 8.47 (s, 2 H); 8.58 (t, J ˆ 4.5, 1 H); 9.25 (d, J ˆ 7.6, 1 H). 13C-NMR (100 MHz):12.7; 13.7; 13.9; 19.9; 21.1;
‡
23.5; 24.9; 31.9; 40.7; 40.9; 42.1; 46.2; 49.0; 118.0; 130.6; 136.4; 140.9; 162.6; 173.1. FAB-MS:452.2 ([ M ‡ 1] ;
‡
calc. for M , 451.24). Anal. calc. for C21H33N5O6:C 55.86, H 7.37, N 15.51; found:C 55.53, H 7.41, N 15.13.
4-(Butylamino)-N-[1-(diethylcarbamoyl)-3-methylbutyl]-3-nitrobenzamide ((Æ)-15) was obtained from
1
(Æ)-14 as a pale orange solid:45% yield. M.p. 112 113 8. H-NMR (400 MHz, CDCl3):0.94 ( d, J ˆ 6.6, 3 H);
0.98 (t, J ˆ 7.3, 3 H); 1.02 (d, J ˆ 6.6, 3 H); 1.13 (t, J ˆ 7.1, 3 H); 1.31 (t, J ˆ 7.2, 3 H); 1.42 (m, 1 H); 1.47
(tq, J ˆ 7.5, 7.5, 2 H); 1.73 (m, 3 H); 3.24 (dq, J ˆ 13.7, 6.9, 1 H); 3.32 (q, J ˆ 6.2, 2 H); 3.44 (q, J ˆ 7.2, 2 H); 3.56

3614
Helvetica Chimica Acta Vol. 85 (2002)
(dq, J ˆ 13.6, 6.9, 1 H); 5.14 (ddd, J ˆ 11.1, 7.9, 3.0, 1 H); 6.79 (d, J ˆ 9.0, 1 H); 7.33 (d, J ˆ 7.6, 1 H); 7.88 (dd, J ˆ
8.9, 1.6, 1 H); 8.23 (t, J ˆ 4.4, 1 H); 8.64 (d, J ˆ 2.0, 1 H). 13C-NMR (100 MHz):12.9; 13.7; 14.4; 20.1; 21.7; 23.5;
24.8; 30.8; 40.6; 42.0; 42.7; 42.9; 47.9; 113.5; 120.6; 126.5; 130.9; 134.6; 147.0; 164.9; 172.2. FAB-MS:407.2 ([ M ‡
‡
‡
1] ; calc. for M , 406.26). Anal. calc. for C21H34N4O4:C 62.04, H 8.43, N 13.78; found:C 61.84, H 8.43, N 13.57.
General Procedure for Biphasic Enantioselective Nucleophilic Aromatic Substitution of (Æ)-8 with PrSH
(Scheme 4). To a stirred soln. of (Æ)-8 (0.009 mmol), (R)-5 (0.018 mmol), and THAC (0.0002 mmol) in 2 ml of
CCl4 were added 1.0 ml of sat. NaHCO3 soln. and 0.05 ml of PrSH at 08. The reaction progress was monitored by
the FUSE-4 chiral stationary phase1) developed in these laboratories under the following conditions:mobile
phase 12% i-PrOH in hexane, detection at l ˆ 254 nm, (R)-5: k ˆ 0.8, (R)-8: k ˆ 3.7, (S)-8: k ˆ 9.8, (S)-10: k ˆ
2.8, (R)-10: k ˆ 9.8. The extent of conversion was monitored by HPLC with (R)-5 as an internal standard. After
2 h (at ca. 50% conversion), the soln. was diluted with CCl4 and H2O, and the layers were separated. The org.
layer was washed with 2m HCl and H2O, and then dried (MgSO4).
General Procedure for Monophasic Enantioselective Nucleophilic Aromatic Substitution of (Æ)-8 with
BuNH2 (Scheme 5). To a stirred soln. of (Æ)-8 (0.009 mmol) and (R)-5 (0.018 mmol) in 1 ml of CCl4 and 0.02 ml
of CH2Cl2 was added 0.013 mmol of BuNH2. The soln. was stirred for 30 min, diluted with CCl4, and washed
sequentially with 2m HCl and H2O. The extent of conversion was determined with a Pirkle Covalent d,l-
Phenylglycine column (20% i-PrOH in hexane) available from Regis Technologies. Selector (R)-5 was used an
internal standard. Enantiomeric excess was determined with an (R,R)-Whelk O1 column (Regis Technologies)
under the following conditions:mobile phase 25% i-PrOH in hexane, detection at d ˆ 254 nm, (R)-5: k ˆ 58,
(R)-8: k ˆ 2.1, (S)-8: k ˆ 2.5, (S)-11: k ˆ 4.5, (R)-11: k ˆ 5.7.
General Procedure for Monophasic Enantioselective Nucleophilic Aromatic Substitution of (Æ)-12 with
BuNH2 (Scheme 6). To a stirred soln. of (Æ)-12 (0.009 mmol) and (S)-2 (0.018 mmol) in 1 ml of CCl4 and 0.02 ml
of CH2Cl2 was added 0.013 mmol of BuNH2. The soln. was stirred for 30 min, diluted with CCl4, and washed
sequentially with 2m HCl and H2O. The extent of conversion and the enantiomeric excess were determined with
an (R,R)-Whelk O1 column (Regis Technologies) under the following conditions:mobile phase 3% i-PrOH in
97% hexane, detection at l ˆ 280 nm, (S)-2: k ˆ 7.9, (S)-12: k ˆ 2.96, (R)-12: k ˆ 5.1, (R)-13: k ˆ 6.4, (S)-13:
k ˆ 12.3. Selector (S)-2 was used as an internal standard.
General Procedure for Monophasic Enantioselective Nucleophilic Aromatic Substitution of (Æ)-14 with
BuNH2 (Scheme 7). To a stirred soln. of (Æ)-14 (0.009 mmol) and (S)-2 (0.018 mmol) in 1 ml of CCl4 and 0.02 ml
of CH2Cl2 was added 0.09 mmol BuNH2. The soln. was stirred for 40 h, diluted with CCl4, and washed
sequentially with 2m HCl and H2O. Assays were analyzed periodically by HPLC with an (S)-a-Burke column
(Regis Technologies) under the following conditions:mobile phase 3% i-PrOH in 97% hexane, detection at l ˆ
248 nm, (S)-2: k ˆ 9.5, (S)-14: k ˆ 3.2, (R)-14: k ˆ 4.0, (R)-15: k ˆ 5.3, (S)-15: k ˆ 6.9. Selector (S)-2 was used as
an internal standard.
Determination of Absolute Configuration. Abs. configurations were assigned by comparison with authentic
enantiomerically pure samples (1, 3, and 4). When authentic enantiomerically pure samples were not available,
it was assumed that chromatographic behavior of structural analogues of 3 (e.g., 12, 13, 14, and 15) and of 4 (e.g.,
7, 8, 9, 10, and 11) resembled the parent compounds 3 and 4. Structural analogues of 3 were resolved on the CSP
derived from (S)-2 [15] with large enantiomer-separation factors. Similarly, analogues of 4 were resolved on the
CSP derived from (R)-5 (FUSE-3)2)with large enantioseparation factors.
REFERENCES
[1] J. Welch, Chromatogr. A 1994, 3, 666.
[2] A. Pryde, in −Chiral Liquid Chromatography×, Ed. W. J. Lough, Chapman and Hall, New York, 1989, p. 23.
[3] W. H. Pirkle, T. C. Pochapsky, Chem. Rev. 1989, 89, 347.
[4] a) R. Breslow, Acc. Chem. Res. 1995, 28, 146; b) Y. Murakami, J. Kikuchi, Y. Hiseada, O. Hayashida, Chem.
Rev. 1996, 96, 721.
1
)
Chiral stationary phase FUSE-4 was made from an analogue of Betti amine ((2-(Methoxynaphthalen-1-
yl)benzylamine) according to an earlier described procedure [15], resolved with d-malic acid [15], and
immobilized as carboxamide of 2,2-dimethylpent-4-enoic acid by a previously described method [16].
Chiral stationary phase FUSE-3 was prepared from commercially available (‡)-(R)-1-(2-Naphthyl)ethy-
lamine (Lancaster) and immobilized as carboxamide of 2,2-dimethylpent-4-enoic acid by a previously
described method [15].
2
)

Helvetica Chimica Acta Vol. 85 (2002)
[5] S. E. Snyder, W. H. Pirkle, Org. Lett. 2002, 4, 3283.
3615
[6] a) J. M. Keith, J. F. Larrow, E. N. Jacobsen, Adv. Synth. Catal. 2001, 343, 5; b) H. B. Kagan, J. C. Fiaud, Top.
Stereochem. 1988, 18, 249.
[7] A. B. Shvets, S. E. Snyder, W. H. Pirkle, manuscript submitted.
[8] E. Buncel, J. M. Dust, Can. J. Chem. 1994, 72, 1709.
[9] a) F. Terrier, −Nucleophilic Aromatic Displacement. The Influence of the Nitro Group×, VCH, New York,
1991; b) J. F. Bunnett, R. E. Zahler, Chem. Rev. 1951, 49, 272; c) M. Makosza, Russ. Chem. Bull. 1996, 45,
491.
[10] a) M. R. Crampton, H. A. Khan, J. Chem. Soc., Perkin Trans. 2 1972, 710; b) B. Gibson, M. R. Crampton, J.
Chem. Soc., Perkin Trans. 2 1978, 648; c) C. A. Fyfe, M. Cocivera, S. W. H. Damji, J. Am. Chem. Soc. 1975,
97, 5707; d) R. Bacaloglu, A. Blasko, C. Bunton, E. Dorwin, F. Ortega, C. Zucco, J. Am. Chem. Soc. 1991,
113, 238; e) S. M. Shein, V. V. Brovko, A. D. Khmelinskaya, J. Org. Chem. USSR 1970, 6, 784.
[11] E. Buncel, J. M. Dust, F. Terrier, Chem. Rev. 1995, 95, 2261.
[12] M. C. Etter, G. M. Frankenbach, J. Bernstein, Tetrahedron Lett. 1989, 30, 3617.
[13] M. Bodanszky, A. Bodanszky, −The Practice of Peptide Synthesis×, Springer-Verlag, Heidelberg, 1984, p. 20.
[14] J. Christoffers, A. Mann, Chem. Eur. J. 2001, 7, 1014.
[15] C. Cardellicchio, G. Ciccarella, F. Naso, E. Schingaro, F. Scordari, Tetrahedron: Asymmetry 1998, 9, 3667.
[16] W. H. Pirkle, M. E. Koscho, J. Chromatogr., A. 1999, 840, 151.
Received June 1, 2002