N,N-1,2-benzenedisulfonylimide, a new cyclic leaving group for the stereoselective nucleophilic substitution of amines

Article abstract of DOI:10.1016/S0957-4166(98)00033-0

We hereby report the preparation and nucleophilic substitutions of the N,N-1,2-benzenedisulfonylimide derivatives la and 2a of the chiral amines 1 and 2. The nucleophilic attack of KNO₂ afforded the respective alcohols 3 and 4 with 84-90% inversion of configuration. Nucleophilic attack by the azide ion afforded the azide products 5 and 6 which were reduced to the corresponding inverted mines 1 and 2 (94-98.5% inversion). The improved leaving group ability of the N,N-1,2-benzenedisulfonylimides compared with previously reported N,N-disulfonylimides is discussed. Chiral GLC analysis of all products is summarized and the alternative chiral analysis of product 3 by ₁₃C NMR using heptakis(2,3,6-tri-O-methyl)-β-cyclodextrin as a chiral solvating agent (CSA) is discussed.

Full text of DOI:10.1016/S0957-4166(98)00033-0

Pergamon
Tetrahedron: Asymmetry 9 (1998) 681–689
N,N-1,2-Benzenedisulfonylimide, a new cyclic leaving group for
the stereoselective nucleophilic substitution of amines
Karsten Sørbye, Christoffer Tautermann, Per Carlsen and Anne Fiksdahl∗
Organic Chemistry Laboratories, Norwegian University of Science and Technology, N-7034 Trondheim, Norway
Received 18 December 1997; accepted 23 January 1998
Abstract
We hereby report the preparation and nucleophilic substitutions of the N,N-1,2-benzenedisulfonylimide deriva-
tives 1a and 2a of the chiral amines 1 and 2. The nucleophilic attack of KNO2 afforded the respective alcohols 3 and
4 with 84–90% inversion of configuration. Nucleophilic attack by the azide ion afforded the azide products 5 and
6 which were reduced to the corresponding inverted amines 1 and 2 (94–98.5% inversion). The improved leaving
group ability of the N,N-1,2-benzenedisulfonylimides compared with previously reported N,N-disulfonylimides is
discussed. Chiral GLC analysis of all products is summarized and the alternative chiral analysis of product 3 by
13C NMR using heptakis(2,3,6-tri-O-methyl)-β-cyclodextrinas a chiral solvating agent (CSA) is discussed. © 1998
Elsevier Science Ltd. All rights reserved.
1. Introduction
The development of methods for the inversion of configuration of chiral amines for the preparation
of new homochiral substances has not received the same attention as other substance groups such as
alcohols and their derivatives, epoxides and halides. In our ongoing effort to develop stereoselective trans-
formation reactions for chiral amines we have previously shown1–5 that some N,N-disulfonyl derivatives
of primary amines — the N,N-ditosylimides, the N,N-dimesylimides and the N,N-dinosylimides — may
be transformed by nucleophilic substitution reactions into the corresponding amines or alcohols with
inverted stereochemistry. The corresponding alcohols formed with retention of configuration have been
prepared via the diazonium intermediate.4
In the present study we report the preparation of the new cyclic N,N-1,2-benzenedisulfonylimide6
derivatives (1a and 2a) of the primary amines 1 and 2. The leaving group ability of the disulfonylimide
moiety in these compounds was studied by the nucleophilic attack by KNO2 or the azide ion. The
corresponding alcohols 3 and 4 and the azides 5 and 6 were formed. The azides were reduced to the
∗
Corresponding author. E-mail: Anne.Fiksdahl@chembio.ntnu.no
0957-4166/98/$19.00 © 1998 Elsevier Science Ltd. All rights reserved.
PII: S0957-4166(98)00033-0

682
K. Sørbye et al. / Tetrahedron: Asymmetry 9 (1998) 681–689
amines 1 and 2. The stereochemistry of the substitution reactions and the chiral analytical methods are
discussed.
2. Results and discussion
Less vigorous conditions were required for the preparation of the cyclic N,N-1,2-
benzenedisulfonylimide derivatives 1a and 2a from the primary amines 1 and 2 and benzene-
1,2-disulfonyl chloride compared with the methods previously reported for the synthesis of N,N-
disulfonylimides.1–5 A distinctly lower reaction temperature was needed. Dichloromethane was used as
the solvent, while NEt3 was sufficient as base, compared with NaH in previous methods. Comparable
yields of the crystalline imides (% yields, see Scheme 1) were obtained. Purification via crystallization
procedures proved as good as or better than chromatographic methods. The N,N0-bis-byproduct,
the disulfonamide, R–NH–SO2–Ph–SO2–NH–R, could also be isolated from all the reactions and
identified after characterization. The reagent, benzene-1,2-disulfonyl chloride, was prepared in a
three-step procedure (in 21% overall yield) from 2-aminobenzenesulfonic acid by PCl5 treatment of the
diazotization/SO2 product.7
Scheme 1.
Nucleophilic attack on the N,N-1,2-benzenedisulfonylimides 1a and 2a by KNO2 and NaN3 afforded
the alcohols 3, 4 and azides 5, 6 with 84–90% and 94–98.5% inversion of configuration respectively. (See
Table 1 for the % ee and R or S configuration of the substrates 1a, 2a and the products 3–6 as well as the
% yields.)
As can be seen from Table 1, the benzylic disulphonylimide substrate 1a is much more easily
substituted with both nucleophiles (nitrite and azide) than the corresponding cyclohexyl intermediate 2a.
This is demonstrated by both a lower reaction temperature and a shorter reaction time required for the
conversion of 1a than for 2a and also compared with our previously reported N,N-disulfonylimides.1–5
This expected difference in reactivity between the benzylic and the cyclohexyl substrate has not been
observed with acyclic disulfonylimide leaving groups.1–5
Our results show that the highest stereoselectivity was obtained when the reactions were conducted
in DMSO as demonstrated by comparing entries 4 and 7. As expected for SN2 reactions, signifi-
cantly increased stereoselectivity was obtained by decreasing the reaction temperature (see entries 4–6,
90-→20°C), especially in combination with a DMSO based solvent (30% DMF was added in order to

K. Sørbye et al. / Tetrahedron: Asymmetry 9 (1998) 681–689
683
Table 1
Results for the reactions shown in Scheme 1

684
K. Sørbye et al. / Tetrahedron: Asymmetry 9 (1998) 681–689
avoid freezing, 0°C, see entries 2 and 8). No increase in the degree of inversion was observed when
changing the concentrations of the reactants, and no product could be isolated by changing to other
solvents.
Reacting the sulfonimides 1a and 2a with KNO2 in DMF, yielded the alcohols 3 and 4 together
with the corresponding 1-phenylethyl and 1-cyclohexylethyl formate ester byproducts. The formate ester
formation can be rationalized by the competing reaction where DMF functions as an O-nucleophile
followed by hydrolysis of the iminium ester intermediate. Especially for the cyclohexyl substrate 2a
the stereoselectivity in the formate ester formation was low (20–25% ee). The formate was isolated and
separately hydrolysed to the alcohol product. Alkaline work-up of the crude product mixture hydrolyses
the formate ester byproduct to the alcohol and includes the contribution from the formate of low
enantiomeric excess, giving lower stereoselectivity and higher yields compared with a neutral work-
up for the alcohol 4 (see entries 10a and 10b). For the benzylic substrate 1a, the stereoselectivity in the
formate formation was identical to the alcohol 3. The formate ester formation was avoided by changing
to DMSO as solvent. However, dry solvents were important, otherwise acetophenone was identified as
a byproduct for the 1a-→3 alcohol reaction and both acetophenone and 1-phenylethanol were formed in
addition to the azide main product 5 in the 1a-→5 reaction.
2.1. Chiral analysis
The chiral analysis of the products and byproducts was based on GLC chromatography. The cyclohexyl
azide product 6 was indirectly analysed after derivatization to the diastereomeric amides of the reduction
amine product 2. The products 3, 4 and 5 and the formate byproducts were analysed directly using
a chiral stationary phase. In Table 2 is shown the chromatographic separation parameters and the
chromatographic conditions applied.
A preliminary study of an alternative chiral analytical method was also tested for the alcohol 3.
The method is based on the 13C NMR study of the compound after the addition of heptakis(2,3,6-
tri-O-methyl)-β-cyclodextrin (tri-OMe-β-CD) as a chiral solvating agent (CSA). Cyclodextrins have
shown enantiodiscriminating ability toward chiral substrates, and their use as chiral solvating agents
for NMR spectroscopy has been proposed.8 The effect of both the native α-, β-, γ-cyclodextrins
(CD) and a variety of their derivatives have been explored. Both inclusion complex theories and non-
inclusion complexes based on superficial interactions have been discussed. The substrate concentrations
are typically 1–100 mM, and the best shift effects are obtained with CD:substrate ratios higher than 1:1
(often 3:1). Increased ∆δ values are reported in the NMR spectra at low temperatures. The ∆δ values of
the signals from either the CD or the substrates have been studied. The sign and the extent of the shift
effects are often unpredictable. The highest ∆δ values observed are about 0.4–0.5 ppm. However, for an
enantiomeric purity (% ee) determination high ∆(∆δ) values=∆δR−∆δS are required. Better resolutions
(∆δR−∆δS) than have so far been reported are necessary for enantiodeterminations. Alternatively, more
simplified NMR signals; singlets instead of multiplets, would allow a better resolution and, consequently,
enantiomeric excess determination. Our study is therefore based on 13C NMR spectroscopy instead of
1H NMR. For the quantitative analysis it is necessary that similar nuclei have similar relaxation.
1
In our preliminary H NMR study (using tri-OMe-β-CD:substate ratios 1:1, 2:1 and 3:1 as well as
temperatures in the range of +15°C to −50°C), the ∆(∆δ) values (δR−δS) of the 1-methine signal of
the alcohol 3 in CD3OD increased from 0.005 ppm to 0.016 ppm with decreasing temperature. However,
because of the quartet nature of the signal, the resolution was not sufficient for an enantiodetermination.
Because the R- and S-signals overlap, the R:S ratios were not available by integration. No shifts were
observed using CDCl3 as the solvent. Improved results were obtained in a 13C NMR study: fully

K. Sørbye et al. / Tetrahedron: Asymmetry 9 (1998) 681–689
685
Table 2
Chiral GLC analysis. Chromatographic separation parameters for R:S enantioseparations (column:
CP-Chirasil-DEX-CB)
decoupled spectra of (i) the racemic and (ii) a 72:28 (R:S) mixture of the alcohol 3 in CD3OD with
two or three equivalents of tri-OMe-β-CD added showed a splitting of the methyl signal into two (i)
50:50% and (ii) 72:28% signals respectively. These data, confirming the chiral GLC results, prove that
there is no difference in relaxation between the two enantiomeric carbons, that is: the two diastereomeric
carbons in the two tri-OMe-β-CD complexes. The measured 13C NMR ∆(∆δ) values were 0.104–0.112
ppm and the resolutions were sufficient for enantiodetermination based on integrals. These preliminary
results show that tri-OMe-β-CD used as CSAs in 13C NMR spectroscopy represent an improvement
relative to the 1H NMR measurements for the determination of enantiomeric excess. The method will be
further investigated.
In conclusion, less vigorous conditions were needed both for the preparation and the nucleophilic
substitution of the new hereby reported N,N-1,2-benzenedisulfonylimides, 1a and 2a. The highest degree
of inversion (84% for the alcohol 3 and 94% for the azide 5), for the benzylic substrate 1a were
obtained using a lower reaction temperature (0°C) and a shorter reaction time (24 h) than for the
analogous cyclohexyl intermediate 2a (90% degree of inversion for the alcohol 4 and 98% for the azide
6) (2a-→alcohol 4 and azide 6; 80–90°C, 3–4 days). For the former, DMSO as solvent showed a definitive

686
K. Sørbye et al. / Tetrahedron: Asymmetry 9 (1998) 681–689
positive effect on the stereoselectivity. In the development of heptakis(2,3,6-tri-O-methyl)-β-cyclodextrin
(tri-OMe-β-CD) as a chiral solvating agent for 13C NMR as an alternative method to the chiral GLC
analysis for the determination of enantiomeric excess, our preliminary study seems promising.
3. Experimental
3.1. Chemicals
(S)-1-Phenylethylamine 1, Hexel Chemical Products; (R)-1-cyclohexylethylamine 2, phosphorus pen-
tachloride, 18-crown-6, (S)-α-methoxyphenylacetyl chloride from (S)-α-methoxyphenylacetic acid,
Fluka (purum); 2-aminobenzenesulfonic acid (orthanilic acid), Fluka (pract.); sodium nitrite, Merck
(>99%); potassium nitrite, Acros (>97%). Solvents: p.a. quality. DMF and DMSO were dried over
activated molecular sieve (4 Å). TLC: DC-Fertigplatten Kieselgel 60 F254 (0.25 mm). Detection: UV
light at 254 nm or preferentially by 5% alcoholic molybdatophosphoric acid and heating. Flash chro-
matography: Kieselgel 60 (230–400 mesh) Merck. GLC: Carlo Erba Model 8130; injector: split (100
ml/min, T=300°C), hydrogen, detector: FID (T=270°C), column: Chrompack CP-Sil 5CB fused silica
WCOT (25 m). Chiral GLC analysis: Chrompack CP-CHIRADEX-CB fused silica WCOT (25 m, 0.32
mm; 0.32 µm), carrier gas pressure 5–5.5 p.s.i. M.p.s are uncorrected, and were measured on a Büchi
1
apparatus. H NMR: Bruker Avance DPX 300 MHz and 400 MHz NMR spectrometer, chemical shifts
are reported in ppm downfield from TMS. MS: AEI MS-902. IR: Nicolet 20SXC FT-IR spectrometer.
[α]D: Perkin–Elmer 241 polarimeter (10 cm cell with a total volume of 1 ml).
3.2. Preparation of benzene-1,2-disulfonyl chloride from 2-aminobenzenesulfonic acid7
A mixture of orthanilic acid (2-aminobenzenesulfonic acid, 50 g, 0.24 mol, 83% pure), anhydrous
sodium carbonate (12.7 g, 0.14 mol) and water (250 ml) was heated and stirred until all the orthanilic
acid had dissolved before being cooled to around 15°C. A solution of sodium nitrite (19.8 g, 0.29 mol)
in water (50 ml) was added dropwise. The resulting solution was poured onto a mixture of concentrated
hydrochloric acid (55 ml) and crushed ice (300 g). Cooling for 1 h and filtering afforded pink crystals
of 2-benzendiazonium sulfonate which were immediately dissolved in concentrated hydrochloric acid
(300 ml). A suspension of Cu(I)Cl (7 g, 0.07 mol) in sulfur dioxide–acetic acid (475 ml, 30% sulfur
dioxide) was added with stirring, and the temperature was slowly raised to 42°C. After 4 h evolution of
nitrogen had ceased, and the solvents were stripped off. The resulting solid was stirred and heated with
a small amount of saturated sodium chloride (10–20 ml) to yield a green porridge which was cooled,
filtered, washed with methanol and dried to yield the disodium salt of 1,2-benzenedisulfonic acid (19
g, 0.67 mol). This material was heated overnight with phosphorus pentachloride (35.1 g, 0.168 mol,
2.5 eq.). The resulting phosphorous oxychloride and excess phosphorus pentachloride were distilled off
under vacuum to leave a dry yellowish cake. Chloroform (230 ml) and water (80 ml) were added and
the mixture was heated and stirred until all the solid had dissolved. The phases were separated whilst
still hot, and the organic phase was concentrated and cooled to afford clear, off-white crystals (13.6 g,
0.05 mol, 21% overall yield from orthanilic acid). M.p. 141–146°C. 1H NMR (300 MHz, CDCl3): δ 8.03
(m, 2H), 8.47 (m, 2H). 13C NMR (75.47 MHz, CDCl3): δ 132.3, 136.0, 141.3. MS [m/z (% rel. int.)]:
278/276/274 (M, 0.5/2.0/3.5%), 241 (44%), 239 (100%), 156 (16%), 113 (14%), 111 (38%), 64 (54%).
IR (KBr, cm−1): 3106 (w), 1375 (s), 1183 (s), 787 (w), 747 (w), 888 (m), 566 (m), 540 (m).

K. Sørbye et al. / Tetrahedron: Asymmetry 9 (1998) 681–689
687
3.3. (S)-N,N-1,2-Benzenedisulfonyl-1-phenylethylamine 1a from (S)-phenylethylamine 1
Benzene-1,2-disulfonyl chloride (2.0 g, 7.3 mmol) was dissolved in methylene chloride (200 ml) and
brought to reflux. A solution of S-(+)-1-phenylethylamine (1, 0.93 ml, 7.3 mmol) and triethylamine (2.1
ml, 15 mmol, 2.1 eq.) in methylene chloride (50 ml) was added continuously over 30 h. The reaction was
allowed to reflux for at least another 2 h. The solvent was stripped off to yield 4.39 g of solid material
which was extracted with hot acetone (30 ml), cooled, filtered and stripped of solvent. The resulting
solid was dissolved in hot cyclohexane (150 ml), cooled to ambient temperature and immediately
filtered to remove the byproduct N,N0-bis((S)-1-phenylethyl)-1,2-benzenedisulfonamide. Concentration
1
of the mother liquor, cooling and filtration afforded 1a (80–90% pure by H NMR analysis). By
repeating the cyclohexane recrystallization procedure, pure crystalline (S)-N,N-1,2-benzenedisulfonyl-
1-phenylethylamine (1a, 1.1 g, 47%) could be obtained. M.p. 105–106°C. 1H NMR (300 MHz, CDCl3):
δ 2.08 (d, 3H), 5.43 (q, 1H), 7.38 (m, 3H), 7.60 (m, 2H), 7.88 (m, 2H), 7.95 (m, 2H). 13C NMR (75.47
MHz, CDCl3): δ 19.4, 56.7, 121.8, 128.1, 128.6, 128.8, 134.6, 135.4, 137.2. MS [m/z (% rel. int.)]: 323
(M, 3%), 308 (12%), 219 (22%), 194 (4%), 180 (3%), 156 (16%), 105 (81%), 104 (100%). IR (KBr,
cm−1): 3085 (w), 1498 (w), 1447 (m), 1349 (s), 1195 (m), 1172 (s), 1041 (m), 984 (m), 884 (m), 756
(m), 695 (m). [α]D −6.3 (c=1, CHCl3).
3.3.1. N,N0-Bis((S)-1-phenylethyl)-1,2-benzenedisulfonamide
6%; m.p. 214–217°C. 1H NMR (300 MHz, CDCl3): δ 1.44 (d, 6H), 4.55 (m, 2H), 6.67 (NH, d, 2H),
6.91 (m, 10H), 7.04 (m, 2H), 7.42 (m, 2H). 13C NMR (75.47 MHz, CDCl3): δ 23.9, 55.0, 126.1, 127.2,
128.1, 130.4, 131.7, 137.5, 140.7. MS [m/z (% rel. int.)]: 444 (0.02%), 429 (6%), 325 (36%), 308 (8%),
207 (17%), 120 (100%), 105 (86%), 77 (11%). IR (KBr, cm−1): 3305 (s), 3290 (s), 2990 (w), 1458 (m),
1400 (s), 1330 (s), 1322 (s), 1170 (s), 1155 (s), 1080 (m), 1040 (m), 1010 (m), 950 (m), 870 (w). [α]D
+1.5 (c=1.25, CHCl3).
3.4. (R)-N,N-1,2-Benzenedisulfonyl-1-cyclohexylethylamine 2a from (R)-1-cyclohexylethylamine 2
Benzene-1,2-disulfonyl chloride (2.6 g, 9.5 mmol) was dissolved in methylene chloride (400 ml) and
brought to reflux. A solution of S-(+)-1-phenylethylamine (2, 1.4 ml, 9.5 mmol) and triethylamine (2.6
ml, 19 mmol, 2.0 eq.) in methylene chloride (7.5 ml) was added continuously over 48 h. The reaction
was allowed to reflux for at least another 2 h. The solvent was stripped off and the remaining solid
material was extracted with hot cyclohexane (2×100 ml) and then continuous Soxhlet extraction with
cyclohexane for 20 h. The combined extracts were stripped of solvent to yield 3.3 g of solid material.
Repeated recrystallisation from hot cyclohexane (25–30 ml) afforded 1.2 g (38%) pure crystalline 2a.
M.p. 133.5–136°C. 1H NMR (300 MHz, CDCl3): δ 1.65 (d, 3H), 0.9–2.2 (m, 11H), 4.0 (m, 1H), 7.89
(m, 2H), 7.98 (m, 2H). 13C NMR (75.47 MHz, CDCl3): δ 18.1, 25.5, 25.8, 26.0, 30.4, 30.8, 41.4, 61.3,
121.8, 134.4, 135.7. MS [m/z (% rel. int.)]: 329 (M, 0.2%), 314 (0.3%), 246 (100%), 168 (11%), 110
(17%). IR (KBr, cm−1): 3737 (w), 3101 (w), 2921 (s), 2850 (s), 1446 (m), 1384 (s), 1203 (m), 1161 (m),
1133 (m), 1045 (m), 899 (m), 870 (m), 760 (s), 728 (s), 589 (m), 560 (s). [α]D −41 (c=1, CHCl3).
3.4.1. N,N0-Bis((S)-1-cyclohexylethyl)-1,2-benzenedisulfonamide
M.p. 112–113°C. 1H NMR (300 MHz, CDCl3): δ 0.77 (d, 6H), 0.8–1.7 (m, 22H), 3.22 (m, 2H), 6.07
(d, NH, 2H), 7.70 (m, 2H), 8.22 (m, 2H). 13C NMR (75.47 MHz, CDCl3): δ 17.5, 26.0, 26.1, 26.3, 28.4,
28.6, 43.6, 55.3, 131.0, 132.7, 139.0. MS [m/z (% rel. int.)]: 456 (M, 0.06%), 455 (0.1%), 441 (0.8%),
373 (100%), 330 (13%), 263 (55%), 246 (15%), 220 (18%), 168 (9%), 111 (25%). IR (film, cm−1): 3305

688
K. Sørbye et al. / Tetrahedron: Asymmetry 9 (1998) 681–689
(s), 3285 (s), 2920 (s), 2850 (s), 1430 (s), 1345 (s), 1200 (s), 1180 (s), 1125 (m), 1110 (m), 1001 (m),
785 (m), 760 (m), 700 (s). [α]D +49 (c=1, CHCl3).
3.5. Nucleophilic substitution reactions for the preparation of (R)-1-phenylethanol 3 and (S)-1-cyclo-
hexylethanol 4
The nucleophilic substitution reactions for the preparation of (R)-1-phenylethanol 3 from (S)-
N,N-1,2-benzenedisulfonyl-1-phenylethylamine 1a and (S)-1-cyclohexylethanol 4 from (R)-N,N-1,2-
benzenedisulfonyl-1-cyclohexylethylamine 2a were carried out using KNO2/18-crown-6 mainly as de-
scribed elsewhere.4 Specific reaction conditions are listed in Table 1, including yields and the degree of
1
inversion. The alcohol products 3 and 4 were characterized by H NMR, MS and [α]D, giving data
in accordance with data for these substances published previously.3–5 The alcohol products 3 and 4
coeluted on GLC (both on unpolar methylsilicone and chiral cyclodextrin stationary phases) with the
respective alcohols prepared previously and characterized elsewhere.3–5 The % enantiomeric excesses
for the alcohol products 3 and 4 were based on chiral GLC analysis; for chromatographic separation
parameters and chromatographic conditions, see Table 2. For alcohol 3 a chiral 13C NMR method using
heptakis(2,3,6-tri-O-methyl)-β-cyclodextrin (tri-OMe-β-CD) as CSA confirmed the chiral GLC results;
1
to a 70 mM NMR sample in CD3OD was added 1–3 equivalents of tri-OMe-β-CD. H NMR and 13C
NMR spectra were recorded at temperatures from +15°C to −50°C and the ∆(∆δ) values (δR−δR) were
measured. The δR−δR value for the methine proton quartet in the 1H NMR spectra was 0.005–0.016 ppm,
while the respective data for 13C NMR was 0.104–0.112 ppm, the values increasing with a decrease in
temperature. For both reactions in DMF the formate ester byproduct could also be isolated by flash
chromatography and characterized when the reaction work-up procedure did not include the alcalic
treatment as described above. The % ee of the respective formates was 40–68 (identical to the alcohol
product 3) and 20–25 (compared with 80% ee for the alcohol 4). The isolated formate esters could be
hydrolysed (KOH/MeOH) to give the alcohols 3 and 4. No formate byproduct was observed in the DMSO
based reactions (added 30% DMF) at 0°C.
3.5.1. (R)-1-Phenylethyl formate
1H NMR (300 MHz, CDCl3): δ 1.6 (d, 3H), 6.1 (m, 1H), 7.2–7.4 (m, 5H), 8.1 (s, 1H). MS [m/z (%
rel.int.)]: 150 (M, 25%), 120 (6%), 107 (11%), 105 (53%), 104 (100%), 103 (25%), 91 (5%), 77 (37%).
IR (film, cm−1): 2983 (w), 2932 (w), 1723 (s), 1173 (s), 761 (m), 699 (m).
3.5.2. (S)-1-Cyclohexylethyl formate
1H NMR (300 MHz, CDCl3): δ 1.2 (d, 3H), 0.8–1.8 (m, 11H), 4.9 (m, 1H), 8.1 (s, 1H). 13C NMR
(75.47 MHz, CDCl3): δ 17.1, 25.9, 26.0, 26.3, 28.3, 28.4, 42.4, 74.7, 161.0. MS [m/z (% rel.int.)]: 156
(M, 1%), 155 (4%), 127 (8%), 126 (36%), 111 (16%), 109 (43%), 83 (35%), 81 (100%). IR (film, cm−1):
2929 (s), 2855 (m), 1725 (s), 1189 (s).
3.6. Nucleophilic substitution reactions for the preparation of (R)-1-phenylethylazide 5 and (S)-1-
cyclohexylethylazide 6
The nucleophilic substitution reactions for the preparation of (R)-1-phenylethylazide 5 from (S)-
N,N-1,2-benzenedisulfonyl-1-phenylethylamine 1a and (S)-1-cyclohexylethylazide 6 from (R)-N,N-
1,2-benzenedisulfonyl-1-cyclohexylethylamine 2a were carried out using NaN3 mainly as described
elsewhere.1,2 Specific reaction conditions are listed in Table 1, including yields and the degree of

K. Sørbye et al. / Tetrahedron: Asymmetry 9 (1998) 681–689
689
inversion. The azide products 5 and 6 were characterized by 1H and 13C NMR, MS and IR giving data in
accordance with data for these substances published previously.3–5 Chiral analysis of the azide product
5 could be carried out directly on the chiral cyclodextrin GLC column, see Table 2. The azide product 6
was reduced by hydrogenolysis to the respective amine 2 and chiral analysis was carried out indirectly on
an unpolar methylsilicone stationary phase after derivatization with (S)-α-methoxyphenylacetyl chloride
as described elsewhere.1–3 The yields and % enantiomeric excess for the azide products 5 and 6 are given
in Table 1.
References
1. Seljestokken, B.; Fiksdahl, A. Acta Chem. Scand. 1993, 47, 1050.
2. Johansen, C.; Fiksdahl, A. Chirality 1994, 6, 161.
3. Oppedal, H.; Tveit, I. C.; Fiksdahl, A. Tetrahedron: Asymmetry 1994, 5, 895.
4. Ileby, N.; Kuzma, M.; Heggvik, L. R.; Sørbye, K.; Fiksdahl, A. Tetrahedron: Asymmetry 1997, 8, 2193.
5. Heggvik, L. R.; Fiksdahl, A. Tetrahedron: Asymmetry 1997, 8, 2189.
6. Carlsen, P. H. J. Tetrahedron Lett. 1998, in press.
7. Meerwein, H.; Dittmar, G.; Göllner, R.; Hafner, K.; Mencsh, F.; Steinfort, O. Chem. Ber. 1957, 90, 841. Adams, R.; Marvel,
C. S. In Organic Syntheses, 2nd edn, vol. I, Wiley & Sons: New York, 1967, p. 84. Conant, J. B.; Corson, B. B. In Organic
Syntheses, 2nd edn, vol. II, Wiley & Sons: New York, 1967, p. 33.
8. Casy, A. F.; Mercer, A. D. Magnetic Resonance in Chemistry 1988, 26, 765. Richards, J. J.; Webb, M. L. Analytical
Proceedings 1992, 29, 251. Uccello-Barretta, G.; Balzano, F.; Caporuzzo, A. M.; Iodice, A.; Salvadori, P. J. Org. Chem.
1995, 60, 2227. Uccello-Barretta, G.; Balzano, F.; Caporuzzo, A. M.; Salvadori, P. J. Org. Chem. 1994, 59, 836. Uccello-
Barretta, G.; Balzano, F.; Menicagli, R.; Salvadori, P. J. Org. Chem. 1996, 61, 363. Yoo, S.; Kim, S.-I. Bull. Korean Chem.
Soc. 1996, 17, 673. Park, K. K.; Park, J. M. Bull. Korean Chem. Soc. 1996, 17, 1052. Botsi, A.; Perly, B.; Hadjoudis, E. J.
Chem. Soc., Perkin Trans. 2 1997, 89. Uccello-Barretta, G.; Cuzzola, A.; Balzano, F.; Menicagli, R.; Iuliano, A.; Salvadori,
P. J. Org. Chem. 1997, 62, 827.