Resolution of albuterol acetonide

Article abstract of DOI:10.1016/S0957-4166(99)00209-8

The (R)-enantiomer of albuterol has been isolated via resolution of albuterol acetonide with (2S,3S)-di-O-benzoyl- or (2S,3S)-di-O-toluoyltartaric acid. The absolute configuration of the resolved acetonide was assessed by ¹H NMR analysis of its (R)-Mosher's ester, and confirmed by an X-ray crystal structure determination of the (R)-phenylethylurea derivative of the (S)-enantiomer.

Full text of DOI:10.1016/S0957-4166(99)00209-8

Pergamon
Tetrahedron: Asymmetry 10 (1999) 2175–2189
Resolution of albuterol acetonide
Mino R. Caira, Roger Hunter,∗ Luigi R. Nassimbeni and Anne T. Stevens
Department of Chemistry, University of Cape Town, Rondebosch 7701, South Africa
Received 10 May 1999; accepted 21 May 1999
Abstract
The (R)-enantiomer of albuterol has been isolated via resolution of albuterol acetonide with (2S,3S)-di-O-
benzoyl- or (2S,3S)-di-O-toluoyltartaric acid. The absolute configuration of the resolved acetonide was assessed
1
by H NMR analysis of its (R)-Mosher’s ester, and confirmed by an X-ray crystal structure determination of the
(R)-phenylethylurea derivative of the (S)-enantiomer. © 1999 Elsevier Science Ltd. All rights reserved.
1. Introduction
Albuterol 1 is a β2-adrenergic receptor agonist and is currently one of the most prescribed broncho-
dilators for the treatment of bronchial asthma.1 The drug was developed in the late 1960s as a racemate
in a process that has been reviewed.2 The assignment of absolute configuration of the enantiomers of
the drug was reported in 1971, based on circular dichroism studies.3 More recently, the realisation that
the inactive (S)-enantiomer may be associated with toxic side effects,4–6 in addition to clinical trial data
confirming the improved efficacy of the drug when administered as the (R)-enantiomer,7 has spurred
efforts towards single enantiomer production. In August 1997 Sepracor Inc. submitted an NDA (New
Drug Application) for (R)-albuterol to the Food and Drug Administration (FDA) which received approval
in March 1999.8 This paper presents results on the successful resolution of albuterol using a resolution
of its acetonide derivative, as well as reporting the first X-ray absolute configurational determination via
single crystal analysis of the (R)-phenylethylurea derivative of the aforementioned ketal derivative.
2. Results and discussion
While (R)-albuterol has been successfully produced by asymmetric synthesis9 using enantioselective
reduction9,10 with an oxazaborolidine catalyst as a key step, the use of resolution has met with far less
∗
Corresponding author. E-mail: roger@psipsy.uct.ac.za
0957-4166/99/$ - see front matter © 1999 Elsevier Science Ltd. All rights reserved.
PII: S0957-4166(99)00209-8

2176
M. R. Caira et al. / Tetrahedron: Asymmetry 10 (1999) 2175–2189
success. The best results to date have been obtained with (S)-naproxen,1 which produced (R)-albuterol in
86% ee and 15% yield, and via salt formation with the metal complex (+)-Ba[CoEDTA]2.11 Various
benzylated derivatives have been shown to be superior substrates for resolution1,3,9,12 and a process
based on one of these has been adopted by Sepracor Inc. to produce (R)-albuterol hydrochloride in
multikilo quantities for clinical evaluation. In the light of these results we decided to study the resolution
of a derivative and guided by the criteria used by ten Hoeve and Wynberg in the design of synthetic
resolving agents,13 we opted to introduce rigidity into the substrate by ring formation. Mindful that
facile deprotection following resolution was an important criterion for the overall success of a scaled-up
process, condensation of the racemic drug with acetone under acidic conditions to afford an oxazolidine
or ketal was selected as our derivatisation method. In the event, reaction of albuterol with excess
acetone and BF3·OEt2 (2 equiv.) at 0°C for 1 h furnished a 95% yield of the acetonide 2 after column
chromatography. Upon scale-up, direct crystallisation of the crude material from acetone or acetonitrile
1
afforded high quality material in high yield (>80%) and pure as judged by H NMR. Optimisation
studies revealed that conc. H2SO4 (1 equiv.) under the same experimental conditions gave the same
result (Scheme 1).
Scheme 1.
1H and 13C NMR spectroscopy were used to assign the structure of 2 as an acetonide as opposed to
the oxazolidine alternative. Of note were the appearance of resonances at 1.505 and 1.510 ppm in the 1H
NMR spectrum (400 MHz) for the diastereotopic methyl resonances as well as a 13C signal for the ketal
quaternary carbon at 99.5 ppm which is in good agreement with the reported value for 2,2-dimethyl-1,3-
dioxanes (99 ppm) and significantly downfield to the analogous value for 2,2-dimethyl-1,3-oxazolidines
(81 ppm).14 Subsequently, single crystal X-ray analysis on resolved 2 proved this assignment to be
correct.
2.1. Resolution of 2
In order to determine whether (rac)-2 crystallised as a conglomerate which could be resolved by direct
crystallisation,15 a series of X-ray photographs were taken in an attempt to determine the space group.
These indicated two possible space groups, Cmma or Cm2a. Both space groups contain mirror planes
which necessitate the presence of both enantiomers of a chiral molecule in the crystal. Crystalline (rac)-2
was thus assigned as a true racemate and resolution employing diastereomer formation was the necessary
approach.
Owing to the well-documented lack of predictability in identifying a successful resolving agent for
classical resolution, a crystallisation array strategy as advocated by Wilen et al.16 was adopted. An
alternative combinatorial approach has since been proposed by Vries et al.17 and further discussed by
Collet.18 The following components were included in the study:
(i) resolution substrates (bases): albuterol 1 and albuterol acetonide 2;
(ii) resolving agents (acids): (1R)-(−)-camphor-10-sulfonic acid; (2R,3R)-(−)-di-O-benzoyltartaric acid
monohydrate (DOBT); (2R,3R)-(−)-di-O-toluoyltartaric acid (DOTT); (S)-(−)-malic acid; (2R,3R)-
(+)-tartaric acid; (S)-(+)-mandelic acid; (S)-(+)-1,10-binaphthylphosphoric acid;
(iii) solvents: methanol, ethanol, acetone and ethyl acetate.

M. R. Caira et al. / Tetrahedron: Asymmetry 10 (1999) 2175–2189
2177
Table 1
Specific rotations of material from salt formation experiments
Experiments with 1 as substrate failed to produce any solid material, thereby confirming literature
reports on the problematic nature of this compound in classical resolution. Albuterol acetonide 2 proved
to be slightly more accommodating and gave fine crystals with DOBT and DOTT using methanol or
ethanol as solvent. The 1H NMR spectrum of the filtered salt indicated a stoichiometry of acid:base, 1:2
which was confirmed by microanalysis and is advantageous industrially since only 0.5 equiv. of resolving
agent is required for the process. The infra-red spectra of both the DOBT and DOTT salts revealed a broad
band at 2700 cm−1 and a strong band at 1640 cm−1 assigned to ammonium N–H and carboxylate anion
stretching frequencies, respectively. These data present strong evidence in favour of the compounds being
salts rather than complexes. Moreover, a recent study19 on the diastereomers formed between pipecolic
acid anilides and DOBT showed that salts were formed with an acid:base ratio of 1:2 whereas the ratio
for complexes was 1:1.
From the optical rotations of the precipitates, mother liquors and resolving agents shown in Table 1, it
was concluded that some enrichment had taken place and that the n-salts, i.e. (+)-2/(−)-acid were being
formed.
The enriched acetonides were extracted from the salts using a standard acid/base extraction and a
specific rotation of +16 determined for each sample. Since the specific rotation of 2 was unknown,
various methods for evaluating the optical purity of 2 were used. Hydrolysis of the resolved acetonide
(+)-2 (see below) gave rise to (S)-albuterol acetate based on the sign of the optical rotation. Since
the (R)-enantiomer was desired ultimately, further work was undertaken using (2S,3S)-(+)-DOBT and
(2S,3S)-(+)-DOTT. Although the (2S,3S)-resolving agents are derivatives of unnatural tartaric acid
and hence more expensive, a recent resolution of DOBT by preferential crystallisation of its calcium
salt–methoxyethanol complex promises a cheaper source of this material.20
2.2. Determination of enantiomeric purity of 2
(R)- or (S)-O-Acetylmandelic acid (OAM) is well documented to form soluble diastereomeric salts
with a range of chiral amines and amino alcohols,21 thus permitting a direct measure of their enantiomeric
1
composition by H NMR spectroscopy. The spectra for (−)-2 isolated from both the DOBT and DOTT
resolutions with (R)-OAM were recorded. The integrals for the t-butyl singlets of the diastereomeric salts
(δ=1.14, 1.27 ppm/0°C; 1.15, 1.24 ppm/25°C), were used to calculate the enantiomeric excess of each
sample which were determined to be 90% in both cases.
More accurate ee values were obtained by HPLC on a Pirkle-type chiral stationary phase (Phenomenex
Chirex 3022). Separation of the enantiomers of albuterol on this stationary phase has been reported22 and
satisfactory separation of the enantiomers of 2 was obtained by reducing the polarity of the mobile phase.
Subsequently, separation of the enantiomers of albuterol and its acetonide under the same conditions
was demonstrated on a Chirobiotic Teicoplanin HPLC column. The (−)-2 obtained from resolution
experiments gave ee values of 89.5% (DOBT) and 92.1% (DOTT) which were in good agreement with
1
the H NMR results. Recrystallisation of (−)-2 from acetone produced a substantial increase in optical
purity as recorded in Table 2.

2178
M. R. Caira et al. / Tetrahedron: Asymmetry 10 (1999) 2175–2189
Table 2
Comparison of DOBT and DOTT resolutions of 2
2.3. Hydrolysis studies
In a recent publication describing an attempted synthesis of (R)-albuterol from optically active
cyanohydrins,23 hydrolysis (aq. HCl/∆) of acetals 3 and 4 resulted in racemisation and partial decompo-
sition of the product.
We were able to achieve complete hydrolysis of the isopropylidine ketal by heating it in aqueous
acetic acid (1:1) at 50°C for 6 h. Vacuum distillation at 40°C allowed removal of the reagents without
promoting any albuterol decomposition. The resultant oil was taken up in methanol, and albuterol acetate
monomethanolate 5 precipitated in 92% yield by the addition of ethyl acetate without any loss of
enantiomeric purity as judged by HPLC (Scheme 2). Further investigation of the hydrolysis conditions
revealed that refluxing in the same reagents for 2 h did cause a reduction in ee to 97.7%.
Scheme 2.
Characterisation of (R)-5 was not straightforward owing to its crystallisation as a methanol solvate.
Inconsistent microanalytical results were obtained due to partial desolvation on drying. Thermogravime-
tric analysis of a sample freshly removed from the mother liquor showed a mass loss of approximately
10% over the range 80–140°C which is as expected for the inclusion of one molecule of methanol per salt
formula unit (theoretical mass loss=9.7%). This is followed by mass loss due to decomposition of the salt
after melting of the desolvated crystal at 144°C. The trace produced by differential scanning calorimetry
over the same temperature range showed two endotherms confirming these two events (desolvation and
1
melting). H NMR of the same sample revealed a peak at δ=3.17 ppm integrating for three protons and
corresponding to that expected for the methyl protons on methanol, thus confirming the presence of a
1:1 methanol solvate. Microanalysis was performed on a sample dried at 120°C overnight and gave the
correct analysis for the desolvated salt. Specific rotation determinations were also carried out on the dried
salt.

M. R. Caira et al. / Tetrahedron: Asymmetry 10 (1999) 2175–2189
2179
Figure 1. Molecular structure of (+)-2
Treatment of (R)-2 with 1 equiv. of sulfuric acid in methanol at ambient temperatures also led to
complete hydrolysis without any racemisation. Neutralisation of the reaction mixture with 1 equiv. of
chilled methanolic sodium hydroxide and subsequent filtration to remove the resulting salts allowed the
recovery of (R)-albuterol as a free base. However, yields were vastly inferior to those achieved with the
acetic acid hydrolysis. The use of sulfuric acid at elevated temperatures to effect racemisation, and thus
facilitate recycling of the unwanted enantiomer, was also demonstrated in our studies and is applicable
to an industrial scale-up.
2.4. Absolute configuration studies
The absolute configuration of albuterol was assigned by Hartley and Middlemiss in 19713 on the basis
of a negative Cotton effect in the circular dichroism spectrum of the (−)-enantiomer. This effect coincided
with that of (R)-octopamine, and (−)-albuterol was thus assigned as having the (R)-configuration. A
search of the Cambridge Structural Database24 revealed that no crystallographic studies on homochiral
albuterol or derivatives thereof had been reported and we thus decided to apply these techniques in order
to make an unambiguous assignment of the absolute configuration. Although the DOBT and DOTT salts
of albuterol acetonide were crystalline, no suitable crystals for structural determination could be obtained.
We thus decided to carry out the structural determination of resolved (+)-albuterol acetonide in order to
evaluate the Flack absolute structure parameter.25 Although the refined value of the Flack parameter was
not decisive in indicating the absolute configuration, even with a low temperature data-collection (223
K), the presence of the isopropylidine ketal, as was inferred from 13C NMR data, was confirmed, as
shown in Fig. 1.
We then decided to pursue the preparation of a covalent diastereomer of (−)-2, by reaction with
a homochiral reagent of known absolute stereochemistry, for structural determination. After several
failed derivatisations with some standard reagents including (S)-mandelic acid/DCC/HOBt26 and (1R)-
camphor-10-sulfonyl chloride/pyridine/DMF,27,28 (−)-2 was successfully chemoselectively O-acylated
by (S)-Mosher’s acid chloride29,30 to give the ester 6 (Scheme 3).
Assignment of 6 as a Mosher ester, as opposed to an amide, was based on the appearance of a C_O
stretching frequency at 1748 cm−1 in the infra-red spectrum, as well as a downfield shift in the benzylic
methine proton in the 1H NMR spectrum (δ=4.8 to 5.9 ppm), and the benzylic carbon (72 to 79 ppm) in
the 13C spectrum. Although 6 was an oil, thus preventing single crystal X-ray analysis, it did present
an opportunity to undertake a Mosher analysis31 for the configurational assignment. The analogous
1
derivatisation of (rac)-2 was carried out and the H NMR resonances of the two diastereomers 6 and
7 (from (+)-2) were identified, as shown in Table 3.

2180
M. R. Caira et al. / Tetrahedron: Asymmetry 10 (1999) 2175–2189
Scheme 3.
Table 3
1H chemical shift data for the (R)-Mosher esters of (+)- and (−)-2
Inspection of these data revealed the shielding of the phenyl protons of 6 and deshielding of its C-2 and
t-butyl protons relative to those on 7. Applying this outcome to the minimum energy conformation32 for
the Mosher ester resulted in the (R)-configuration being assigned to the (−)-acetonide, (−)-2, as illustrated
in Fig. 2.
Figure 2.
In the light of uncertainty cast on this analysis by the identification of three low energy
conformations,33 we decided to pursue the derivatisation to a crystalline diastereomer. We
were eventually gratified to discover a suitable urea derivative. Reaction of (rac)-2 with (R)-
phenylethylisocyanate prepared in situ from (R)-phenylethylamine34 and triphosgene35 gave a high

M. R. Caira et al. / Tetrahedron: Asymmetry 10 (1999) 2175–2189
2181
yield of the diastereomeric urea derivatives 8 and 9 which were separated. The same reaction with (−)-2
produced diastereomer 9. The assignment of these products as ureas was made on the basis of NMR
and infra-red spectra, and later corroborated by X-ray crystal structure determination. The observed
chemoselectivity, which is in contrast to that of the Mosher adduct, is probably due to the greater
‘softness’ of the isocyanate carbon compared to that of Mosher’s acid chloride (Scheme 4).
Scheme 4.
The urea arising from (+)-2, i.e. 8, produced material suitable for X-ray crystal structure determination.
The molecular structure, clearly showing the relative stereochemistry at the stereogenic centres is shown
in Fig. 3. Since the configuration of the phenylethylamine was known to be (R), that of the (+)-albuterol
acetonide moeity, which upon hydrolysis gave rise to (+)-albuterol acetate, was unambiguously assigned
as (S). This was in agreement with the original assignment made on the basis of circular dichroism
studies.3
Figure 3. Molecular structure of 8
3. Conclusion
A novel route to obtaining large quantities of (R)-albuterol has been devised by performing a low
cost, high yielding protection of (rac)-albuterol and resolving the resulting acetonide via diastereomeric

2182
M. R. Caira et al. / Tetrahedron: Asymmetry 10 (1999) 2175–2189
salt formation with (2S,3S)-di-O-benzoyltartaric acid and (2S,3S)-di-O-toluoyltartaric acid. (R)-Albuterol
was obtained as its acetate salt in high yield and optical purity by acid hydrolysis of the resolved acetonide
derivative. The efficiency of the resolution step as well as the simplicity of the protection and deprotection
steps make this an attractive industrial alternative to the present resolution methods. The successful
structure elucidation of the (R)-phenylethylurea derivative of (S)-albuterol acetonide is the first structural
evidence available for confirmation of the assignment of the absolute configuration of the enantiomers of
albuterol.
4. Experimental
4.1. General
Melting points were measured using a Reichert–Jung Thermovar hot-stage microscope and are
uncorrected. Optical rotations were measured on a Perkin–Elmer 141 polarimeter at 24°C. Infra-red
spectra were recorded using a Perkin–Elmer Paragon 1000 FT-IR spectrometer over the range 4000–600
cm−1. Microanalyses were determined using a Fisons EA 1108 CHNS-O instrument. Mass spectra were
recorded on a VG micromass 16F spectrometer operating at 70 eV with an accelerating voltage of 4
1
kV. Accurate masses were determined on a VG-70E spectrometer. H NMR spectra were recorded on
a Varian VXR-200 (200 MHz) or a Varian Unity spectrometer (400 MHz). 13C NMR spectra were
recorded on the same instruments at 50 or 100 MHz. The relevant solvent peak (CHCl3 or DMSO)
was used as an internal standard in each case. Thermal gravimetric analysis and differential scanning
calorimetry were performed on a Perkin–Elmer PC7 series thermal analysis system. High performance
liquid chromatography was performed on a Hewlett–Packard 1090 system with a diode array detection
and a Hewlett–Packard 3393A integrator. Thin layer chromatography was performed on aluminium
backed silica gel 60 F254 plates. The plates were visualised under ultraviolet light and by spraying with
ceric ammonium sulfate in 8 mol dm−3 sulfuric acid and baking at 200°C. Column chromatography
was conducted with Merck Kieselgel 60, 70–230 mesh. Acetone was distilled from Drierite (anhydrous
CaSO4) and stored over 4 Å molecular sieves. Dichloromethane was freshly distilled from P2O5 prior
to use. Solvents used for HPLC were BDH HiperSolv grade. Albuterol was supplied by South African
Druggists, Port Elizabeth, South Africa. The resolving agents were purchased from the Aldrich Chemical
Company.
4.2. 2-(N-tert-Butylamino)-1-(2,2-dimethyl-1,3-benzodioxin-6-yl)ethanol 2
4.2.1. BF3·OEt2 procedure
Freshly distilled BF3·OEt2 (2.7 cm3, 3.1 g, 22.0 mmol) was added dropwise to a stirred mixture of
albuterol (2.39 g, 10.0 mmol) in dry acetone (50 cm3) under nitrogen at 0°C. The solution was stirred
at 0°C for 60 min after which it was slowly poured into 50 cm3 of cold aqueous Na2CO3. The excess
acetone was removed under reduced pressure and the mixture extracted into ethyl acetate. The combined
organic phases were dried (MgSO4) and the solvent removed in vacuo to give a crystalline crude product
(2.70 g). Column chromatography on silica gel with ethyl acetate:petroleum ether:triethylamine (50:45:5)
afforded 2 as a colourless crystalline product (2.67 g, 9.6 mmol, 96%), mp 91–92°C (acetone); found C,
68.9; H, 9.3; N, 5.0; C16H25NO3 requires C, 68.8, H, 9.0; N, 5.0%; νmax (CHCl3)/cm−1 3402 (O–H),
2969 (-CH2-), 1501 (aromatic C–C), 1118 (C–O); δH (400 MHz, CDCl3) 1.07 (9H, s, t-Bu), 1.505 (3H,
s, 20-CH3), 1.510 (3H, s, 20-CH3), 2.56 (1H, dd, J 11.7 and 9.0 Hz, 2-HA), 2.75 (2H, br s, OH/NH), 2.80

M. R. Caira et al. / Tetrahedron: Asymmetry 10 (1999) 2175–2189
2183
(1H, dd, J 11.7 and 3.5 Hz, 2-HB), 4.50 (1H, dd, J 9.0 and 3.5 Hz, 1-H), 4.81 (2H, s, 40-H2), 6.76 (1H,
d, J 8.3 Hz, 80-H), 6.99 (1H, d, J 1.6 Hz, 50-H), 7.11 (1H, dd, J 8.3 and 1.6 Hz, 70-H); δC (100 MHz,
CDCl3) 24.6 and 24.8 (20-CH3), 29.1 (t-Bu -CH3), 50.3 (C-2), 50.3 (t-Bu -C(CH3)3), 61.0 (C-40), 72.0
(C-1), 99.5 (C-20), 116.8 (C-80), 119.2 (C-40a), 122.0 (C-50), 125.7 (C-70), 134.9 (C-60), 150.5 (C-80a),
(found M+, 279, C16H25NO3 requires M, 279).
4.2.2. CuSO4/H2SO4 procedure
Acetone (50 cm3) was added to a flask containing albuterol (2.39 g, 10.0 mmol) and anhydrous copper
sulfate (1.65 g, 10.0 mmol) and the resulting mixture was cooled to 0°C under nitrogen. Concentrated
sulfuric acid (1.3 cm3, 22.0 mmol) was added dropwise and the solution was stirred for 60 min at 0°C.
The work-up procedure as used above resulted in 2.70 g crude material which was purified in an identical
manner to yield crystalline 2 (2.65 g, 9.5 mmol, 95%).
4.3. Diastereomeric salt formation
4.3.1. General procedure for pilot study
The relevant acid and base (0.2 mmol of each) were weighed into a glass vial. The required solvent
(1 cm3) was added and the solution heated sufficiently to allow dissolution of all solids (not achieved
in all cases). The vials were stoppered and left to stand at room temperature. Where no precipitate was
observed after 24 h, the lids were removed to allow slow evaporation. In each case either an oil, paste or
crystalline material was obtained. Where solid crystalline material was obtained, the melting point was
compared to that of the parent acid and base to indicate formation of a new species.
4.4. (2S,3S)-Di-O-benzoyl tartrate salt of (R)-2
To a refluxing solution of (rac)-2 (1.90 g, 6.8 mmol) in methanol (15 cm3) was added a hot solution
of (2S,3S)-di-O-benzoyltartaric acid monohydrate (1.32 g, 3.5 mmol) in methanol (10 cm3). Heating was
discontinued and the solution stirred until a solid cake of precipitate formed. A further 20 cm3 methanol
was added and the solution refluxed for a further 10 min whereafter it was stirred at room temperature for
2 h. The resulting precipitate was isolated by filtration, washed with ethyl acetate and dried under vacuum
to yield the diastereomeric salt (1.10 g, 1.20 mmol, 71% yield as a single enantiomer, 93.9% de†). A
portion of this salt (0.80 g, 0.87 mmol) was further purified by redissolving it in boiling methanol (70
cm3) followed by stirring at 0°C overnight. The salt was isolated and dried as above (0.50 g, 0.55 mmol,
63% for this step, 45% overall yield as a single enantiomer, >99.5% de†), mp 188–190°C (methanol);
[α]D +24.6 (c 0.19 methanol); found C, 65.5; H, 7.1; N, 3.0; C50H64N2O14 requires C, 65.5; H, 7.0; N,
3.1%; νmax (KBr)/cm−1 3411 (O–H), 2900–2400 (N+–H), 1723 (C_O ester), 1641 (C_O carboxylate
(s)), 1504 (C_O carboxylate (as)), 1118 (2° C–O); δH (400 MHz, DMSO-d6) 1.13 (9H, s, t-Bu), 1.44
(6H, s, 2×20-CH3), 2.68 (1H, dd, J 11.9 and 10.3 Hz, 2-HA), 2.90 (1H, dd, J 11.9 and 2.4 Hz, 2-HB),
3.36 (2H, br s, NH), 4.72 (1H, dd, J 10.3 and 2.4 Hz, 1-H), 4.76 (2H, s, 40-H2), 5.66 (1H, s, DOBT 2-H),
6.71 (1H, d, J 8.4 Hz, 80-H), 7.00 (1H, d, J 1.9 Hz, 50-H), 7.10 (1H, dd, J 8.4 and 1.9 Hz, 70-H), 7.50
(1H, m, DOBT 300-H and 500-H), 7.63 (1H, m, DOBT 400-H), 8.00 (1H, dd, J 7.8 and 1.4 Hz, DOBT
200-H and 600-H); δC (100 MHz, DMSO-d6) 24.4 and 24.5 (20-CH3), 25.1 (t-Bu -CH3), 48.1 (C-2), 54.7
(t-Bu -C(CH3)3), 60.0 (C-40), 68.2 (C-1), 74.9 (DOBT C-2), 99.1 (C-20), 116.0 (C-80), 119.0 (C-40a),
†
Note: diastereomeric excesses of the above salts were determined by liberation of 2 from the salt as outlined below and
subsequent chiral HPLC determination.

2184
M. R. Caira et al. / Tetrahedron: Asymmetry 10 (1999) 2175–2189
122.4 (C-50), 125.5 (C-70), 128.3 (DOBT C-300 and C-500), 129.1 (DOBT C-200 and C-600), 130.6 (DOBT
C-100), 132.8 (DOBT C-400), 134.1 (C-60), 149.9 (C-80a), 165.1 (DOBT C-10), 170.2 (DOBT C-1).
4.5. (2S,3S)-Di-O-toluoyl tartrate salt of (R)-2
(rac)-2 (7.50 g, 26.9 mmol) was dissolved in methanol (60 cm3) and heated to reflux. A hot solution
of (2S,3S)-di-O-toluoyltartaric acid (5.30 g, 13.7 mmol) in methanol (40 cm3) was slowly added to the
refluxing solution. Precipitation started during the addition and a solid cake formed on complete addition
of the resolving agent. A further 100 cm3 methanol was added and the resulting slurry refluxed for 10
min followed by stirring at room temperature for 2 h. The precipitate was filtered off, washed with ethyl
acetate and dried under vacuum to yield the salt (5.75 g, 6.1 mmol, 91% as a single enantiomer, 90.0%
de†). This salt (5.25 g, 5.6 mmol) was added to boiling methanol (450 cm3) and stirred at 0°C overnight.
The solid material was again isolated by filtration, washed with ethyl acetate and vacuum dried (4.19
g, 4.4 mmol; 80% for this step, 73% overall yield as a single enantiomer, 95.2% de†), mp 191–192°C
(methanol); [α]D +38.5 (c 0.15 methanol); found C, 65.8; H, 7.3; N, 2.9; C52H68N2O14 requires C, 66.1;
H, 7.2; N, 3.0%; νmax (KBr)/cm−1 3409 (O–H), 2900–2400 (N+–H), 1721 (C_O ester), 1641 (C_O
carboxylate (s)), 1503 (C_O carboxylate (as)), 1123 (2° C–O); δH (400 MHz, DMSO-d6) 1.14 (9H, s,
t-Bu), 1.45 (6H, s, 2×20-CH3), 2.37 (3H, s, DOTT 400-CH3), 2.67 (1H, dd, J 12.0 and 10.4 Hz, 2-HA),
2.89 (1H, dd, J 12.0 and 2.2 Hz, 2-HB), 3.35 (2H, br s, NH), 4.71 (1H, dd, J 10.4 and 2.2 Hz, 1-H), 4.76
(2H, s, 40-H2), 5.61 (1H, s, DOTT 2-H), 6.71 (1H, d, J 8.4 Hz, 80-H), 6.99 (1H, d, J 2.0 Hz, 50-H), 7.09
(1H, dd, J 8.4 and 2.0 Hz, 70-H); 7.29 (1H, d, J 8.2 Hz, DOTT 300-H and 500-H), 7.87 (1H, d, J 8.2 Hz,
DOTT 200-H and 600-H); δC (100 MHz, DMSO-d6) 21.1 (DOTT 400-CH3), 24.4 and 24.5 (20-CH3), 25.0
(t-Bu -CH3), 48.1 (C-2), 54.7 (t-Bu -C(CH3)3), 60.0 (C-40), 68.2 (C-1), 74.8 (DOTT C-2), 99.1 (C-20),
116.0 (C-80), 118.9 (C-40a), 122.4 (C-50), 125.5 (C-70), 128.0 (DOTT C-100), 128.8 (DOTT C-300 and
C-500), 129.2 (DOTT C-200 and C-600), 134.1 (C-60), 142.9 (DOTT C-400), 149.9 (C-80a), 165.2 (DOTT
C-10), 170.6 (DOTT C-1).
4.6. (R)-2-(N-tert-Butylamino)-1-(2,2-dimethyl-1,3-benzodioxin-6-yl)ethanol (R)-2
To a vigorously stirring mixture of ethyl acetate (200 cm3) and aqueous Na2CO3 (200 cm3) was added
the DOTT salt (4.19 g, 4.4 mmol). Stirring was continued until no solid material remained. The organic
layer was removed and the aqueous phase extracted twice with equal volumes of ethyl acetate. The
combined organic phases were dried (MgSO4) and the solvent removed under reduced pressure to yield a
colourless crystalline solid (2.43 g, 8.7 mmol, 98%). Recrystallisation from acetone was used to improve
the optical purity as follows (by chiral HPLC): crude extract 95.2% ee; recrystallisation 1 gave 99.5% ee;
recrystallisation 2 gave >99.5% ee and recrystallisation 3 gave 100% ee, mp 102–103°C (acetone); [α]D
−18.9 (c 1.5 methanol); found C, 68.5; H, 9.6; N, 5.1; C16H25NO3 requires C, 68.8; H, 9.0; N, 5.0%; δH
(400 MHz, CDCl3) 1.11 (9H, s, t-Bu), 1.53 (3H, s, 20-CH3), 1.54 (3H, s, 20-CH3), 2.56 (1H, dd, J 11.7
and 9.1 Hz, 2-HA), 2.12 (2H, br s, OH/NH), 2.87 (1H, dd, J 11.7 and 3.5 Hz, 2-HB), 4.51 (1H, dd, J 9.0
and 3.5 Hz, 1-H), 4.84 (2H, s, 40-H2), 6.79 (1H, d, J 8.4 Hz, 80-H), 7.01 (1H, d, J 1.5 Hz, 50-H), 7.13
(1H, dd, J 8.4 and 1.5 Hz, 70-H).
Liberation of (R)-2 from the analogous DOBT salt was performed in an identical fashion and the same
% yield was achieved.

M. R. Caira et al. / Tetrahedron: Asymmetry 10 (1999) 2175–2189
2185
4.7. (R)-Albuterol acetate monomethanolate (R)-5
Glacial acetic acid (0.4 cm3, 7.2 mmol) was added to a mixture of (R)-2 (90 mg, 0.32 mmol, 95.5%
ee‡) in H2O (0.4 cm3, 0.4 g, 22 mmol). The resulting solution was stirred at 60°C for 5 h. The solvents
were removed by vacuum distillation to leave a colourless tacky oil which was taken up in methanol and
chilled. Ethyl acetate was added to this solution resulting in the precipitation of (R)-5 (98 mg, 0.30 mmol,
92%, 97.0% ee‡), mp 143–144°C (methanol/ethyl acetate) (lit. 144.3°C); δH (400 MHz, DMSO) 1.15
(9H, s, t-Bu), 1.82 (3H, s, CH3COO−), 2.66 (1H, dd, J 11.7 and 9.6 Hz, 2-HA), 2.75 (1H, dd, J 11.7 and
3.4 Hz, 2-HB), 3.17 (3H, s, CH3OH), 4.48 (2H, s, 30-CH2OH), 4.60 (1H, dd, J 9.6 and 3.4 Hz, 1-H), 5.22
(3H, br s, OH/NH), 6.73 (1H, d, J 8.1 Hz, 50-H), 7.03 (1H, dd, J 8.1 and 1.9 Hz, 60-H), 7.27 (1H, d, J 1.9
Hz, 20-H), δC (200 MHz, DMSO) 23.5 (CH3COO−), 27.6 (t-Bu -CH3), 49.2 (CH3OH), 50.2 (C-2), 53.1
((t-Bu -C(CH3)3), 59.0 (30-CH2OH), 71.1 (C-1), 114.8 (C-50), 125.5 (C-60), 125.7 (C-20), 128.7 (C-30),
134.1 (C-10), 154.1 (C-40), 174.5 (CH3COO−). Since desolvation of the salt occurred on drying at room
temperature, the following analyses were performed on the fully desolvated material: [α]D −37.9 (c 0.5
methanol); found C, 60.4; H, 8.6; N, 4.7; C15H25NO5 requires C, 60.2; H, 8.4; N, 4.7%.
4.8. Racemisation of (R)-2
A solution of (R)-2 (4.18 g, 15.00 mmol, 68% ee‡) in 0.50 M aqueous sulfuric acid (150 cm3) was
stirred for 7 h while the temperature was maintained in the range 57–67°C. The solution was cooled and
a cold aqueous solution of sodium hydroxide added. The resulting solution was then re-acidified to pH 3
with 0.5 M sulfuric acid and the solvent removed by azeotropic distillation with toluene under reduced
pressure at 50°C. The resulting slurry was stirred in methanol (70 cm3) whilst the pH was adjusted to 9.7
with methanolic sodium hydroxide. The solid material was removed by filtration and the solvent removed
from the filtrate to yield an orange oil. Redissolution of the oil in acetone followed by removal of the
solvent under reduced pressure afforded crude, racemised albuterol (3.07 g, 12.85 mmol, 86%, 10% ee‡).
4.9. (1R,200R)-2-(N-tert-Butylamino)-1-(2,2-dimethyl-1,3-benzodioxin-6-yl)ethyl-(2-methoxy-2-
trifluoromethyl)phenylacetate 6
To a solution of (R)-2 (280 mg, 1.0 mmol) in dichloromethane (5 cm3) at room temperature and under
nitrogen was added N,N-dimethylaminopyridine (127 mg, 1.0 mmol), triethylamine (0.7 cm3, 511 mg,
5.0 mmol), and (R)-(2-methoxy-2-trifluoromethyl) phenylacetyl chloride (0.23 cm3, 311 mg, 1.2 mmol).
The resulting yellow solution was stirred for 90 min followed by quenching with aqueous Na2CO3. The
mixture was extracted into dichloromethane and the combined organic phases dried (MgSO4) and the
solvent removed under reduced pressure giving an oil (500 mg) which was adsorbed onto silica and
chromatographed using ethyl acetate:petroleum ether (20:80) as eluent to give the product 33 as an oil
(416 mg, 0.8 mmol, 84%), [α]D 1.7 (c 1.5 methanol); νmax (CHCl3)/cm−1 1748 (C_O ester), δH (400
MHz, CDCl3) 1.06 (9H, s, t-Bu), 1.53 (6H, s, 2×20-CH3), 2.83 (1H, dd, J 12.3 and 4.1 Hz, 2-HA), 3.01
(1H, dd, J 12.3 and 9.3 Hz, 2-HB), 3.55 (3H, s, 200-OCH3), 4.72 (1H, d, J 15.0 Hz, 40-HA), 4.79 (1H, d, J
15.0 Hz, 40-HB), 5.87 (1H, dd, J 9.3 and 4.1 Hz, 1-H), 6.75 (1H, d, J 8.5 Hz, 80-H), 6.83 (1H, d, J 2.2 Hz,
50-H), 7.07 (1H, dd, J 8.5 and 2.2 Hz, 70-H), 7.2–7.4 (3H, m, 3000-H, 4000-H, 5000-H), 7.46 (2H, d, J 7.4
Hz, 2000-H, 6000-H); δC (100 MHz, CDCl3) 24.6 and 24.7 (20-CH3), 28.8 (t-Bu -CH3), 47.8 (C-2), 50.3
‡
Note: enantiomeric excesses determined by HPLC.

2186
M. R. Caira et al. / Tetrahedron: Asymmetry 10 (1999) 2175–2189
(t-Bu -C(CH3)3), 55.5 (200-OCH3), 60.7 (C-40), 79.0 (C-1), 84.6 (q, 200-CF3), 99.6 (C-20), 117.0 (C-80),
119.2 (C-40a), 123.3 (C-50), 126.6 (C-70), 127.5 (C-2000, C-6000), 128.2 (C-3000, C-5000), 129.4 (C-4000),
129.6 (C-1000), 132.3 (C-60), 151.3 (C-80a), 165.7 (C-100); (HRMS: found M+ 495.2226, C26H32NO5F3
requires M, 495.2231).
4.10. (1R,200R)-2-(N-tert-Butylamino-1-(2,2-dimethyl-1,3-benzodioxin-6-yl)ethyl-(2-methoxy-2-
trifluoromethyl)phenylacetate 6 and its (1S,200R)-diastereomer 7
Reaction of (rac)-2 under the conditions described above resulted in an inseparable mixture of
diastereomers. These were purified as above and analysed by 1H NMR (Table 4).
Table 4
1H-NMR shift data for the diastereomeric (R)-Mosher esters 6 and 7
4.11. (1S,100R)-2-[N-tert-Butyl-N-(1-phenethylamido)amino]-1-(2,2-dimethyl-1,3-benzodioxin-6-yl)-
ethanol 8
A solution of triphosgene (220 mg, 0.7 mmol) in dichloromethane (4 cm3) was stirred at room
temperature under nitrogen while a solution of (R)-phenylethylamine (0.26 cm3, 247 mg, 2.0 mmol)
and N,N-diisopropylethylamine (0.40 cm3, 296 mg, 2.3 mmol) in dichloromethane (3.5 cm3) was added
over 45 min. The resulting solution was stirred for a further 10 min followed by the addition of a solution
of (rac)-16 (558 mg, 2.0 mmol) in dichloromethane (4 cm3). After 10 min of stirring the solution was
evaporated to dryness under reduced pressure and the residue taken up in ethyl acetate (100 cm3), washed
with equal volumes of 10% NaHCO3 and brine and dried (MgSO4). The solvent was removed under
reduced pressure to produce 768 mg crude material (largely comprising two major products and starting
material). The material was adsorbed onto silica and chromatographed using a mixture of ethyl acetate
and petroleum ether (10:90 followed by 20:80) as eluent. Only the more polar of the products could
be isolated cleanly to give 8 (156 mg, 0.4 mmol, 37% as a single diastereomer), mp 149–150°C (ethyl
acetate/hexane); [α]D −3.5 (c 1.1 methanol); found C, 70.1; H, 8.6; N, 6.7; C25H34N2O4 requires C, 70.4;

M. R. Caira et al. / Tetrahedron: Asymmetry 10 (1999) 2175–2189
2187
H, 8.0; N, 6.6%; νmax (CHCl3)/cm−1 1652 (C_O urea); δH (400 MHz, CDCl3) 1.40 (9H, s, t-Bu), 1.47
(3H, d, J 6.8 Hz, 200-H), 1.54 (6H, s, 2×20-CH3), 3.23 (1H, dd, J 16.3 and 2.3 Hz, 2-HA), 3.44 (1H, dd,
J 16.3 and 9.6 Hz, 2-HB), 4.76 (1H, dd, J 9.6 and 2.3 Hz, 1-H), 4.82 (2H, s, 40-H2), 4.94 (1H, m, 100-H),
6.60 (1H, d, J 7.4 Hz, 100-NH), 6.80 (1H, d, J 8.4 Hz, 80-H), 6.93 (1H, d, J 1.9 Hz, 50-H), 7.09 (1H, dd,
J 8.4 and 1.9 Hz, 70-H), 7.2–7.4 (5H, m, 2000-6000-H); δC (100 MHz, CDCl3) 23.3 (C-200), 24.7 and 24.8
(20-CH3), 29.1 (t-Bu -CH3), 49.9 (C-100), 54.4 (C-2), 55.9 (t-Bu -C(CH3)3), 60.9 (C-40), 76.7 (C-1), 99.7
(C-20), 117.3 (C-80), 119.7 (C-40a), 121.9 (C-50), 125.5 (C-70), 126.0, 126.7 and 128.5 (C-2000-C-6000),
134.1 (C-60), 145.3 (C-1000), 151.1 (C-80a), 160.7 (C_O); (HRMS: found M+ 426.2507, C25H34N2O4
requires M 426.2518).
4.12. Analyses
4.12.1. NMR determination of optical purity of (R)-albuterol acetonide, (R)-2
Typically, 2 (5 mg) and (R)- or (S)-O-acetylmandelic acid (6 mg) were dissolved in CDCl3 (1 cm3)
and the 1H NMR spectra recorded (400 MHz). Several of the protons in 2 gave well distinguished peaks
for each diastereomer formed in solution and integration of these pairs (usually the t-Bu peaks) allowed
determination of the enantiomeric excess.
4.12.2. Chiral high performance liquid chromatography on (R)-2 and (R)-5
The conditions used for the chromatographic determinations are shown in Table 5.
Table 5
Chiral high performance liquid chromatography on (R)-2 and (R)-5
4.13. Crystal structure determinations
For both (+)-2 and 8, preliminary unit cell and space group data were determined from X-ray
precession photographs using Ni-filtered CuKα-radiation (λ=1.5418 Å). Intensity data were collected
using graphite-monochromated MoKα-radiation (λ=0.7107 Å) and corrected for Lorentz-polarisation
effects. The structures were solved using program SHELXS-8636 and refined on F2 with SHELXL-
9337 with H atoms in idealised positions and treated isotropically while non-H atoms were refined
anisotropically. Details for the individual crystals follow.

2188
M. R. Caira et al. / Tetrahedron: Asymmetry 10 (1999) 2175–2189
(+)-2: MW=297.38, C16H25NO3, crystal size 0.2×0.4×0.3 mm, orthorhombic, P212121 (No. 19),
a=5.958(1), b=8.716(2), c=30.481(3) Å, V=1582.9(4) Å3, Z=4, Dcalc=1.172 g cm−3. Intensity data in
the range 1°<θ<30° were measured on an Enraf–Nonius CAD4 diffractometer with the crystal cooled to
−50ꢀ1°C. A total of 4612 reflection intensities were measured of which 3950 were unique (Rint=0.029)
and 2507 had I>2σ(I). Final R indices were R1=0.0488 and wR2=0.3352 for 185 parameters.
8: MW=426.56, C25H34N2O4, crystal size 0.3×0.4×0.4 mm, orthorhombic, P212121 (No. 19),
a=6.749(1), b=13.860(2), c=25.628(3) Å, V=2397.3(4) Å3, Z=4, Dcalc=1.182 g cm−3. Intensity data
in the range 3°<θ<28° were measured at 20ꢀ1°C on a Nonius Kappa CCD diffractometer using the
following strategy: 200 frames, 40 s per frame, χ=0°, 1° rotation in φ per frame; 37 frames, 40 s per
frame, χ=−90°, 1° rotation in ω per frame. Data were processed with the DENZO-SMN package.38 Of
the 5919 reflections measured, 5750 were unique (Rint=0.024) and 4765 had I>2σ(I). Final refinement
involved two restraints and 290 parameters, yielding R1=0.0396 and wR2=0.1173.
Acknowledgements
We thank South African Druggists, Port Elizabeth and Fine Chemicals Corporation, Cape Town for
financial support.
References
1. Bakale, R. P. Speciality Chem. Prod., Market Applic. 1995, 15, 249.
2. Lunts, L. H. C. Salbutamol: A Selective β2-Stimulant Bronchodilator; Roberts, S. M.; Price, B. J., Eds.; Medicinal
Chemistry: The Role of Organic Chemistry in Drug Research; Academic Press: London, 1985; pp. 49–91.
3. Hartley, D.; Middlemiss, D. J. Med. Chem. 1971, 14, 895.
4. Boulton, D. W.; Fawcett, J. P. Clin. Rev. Allerg. Immunol. 1996, 14, 115.
5. Perrin-Fayolle, M.; Blum, P. S.; Morley, J.; Grosclaude, M.; Chambe, M. T. Clin. Rev. Allerg. Immunol. 1996, 14, 139.
6. Templeton, A. G.; Chapman, I. D.; Chilvers, E. R.; Moreley, J.; Handley, D. A. Pulm. Pharmacol. Ther. 1998, 11, 1.
7. Nelson, H. S.; Bensch, G.; Pleskow, W. W.; Disantostefano, R.; DeGraw, S.; Reasner, D. S.; Rollins, T. E.; Rubin, P. D. J.
Allergy Clin. Immunol. 1998, 102, 943.
8. FDA approval no: NDA 20-837; http://www.fda.gov./cder/approval/main2.htm
9. Bakale, R. P.; Wald, S. A.; Butler, H. T.; Gao, Y.; Hong, Y.; Nie, X.; Zepp, C. M. Clin. Rev. Allerg. Immunol. 1996, 14, 7.
10. Gao, Y.; Hong, Y.; Zepp, C. M. U.S. Patent # 5,442,118, 1995, 15 August.
11. Hawkins, C. J.; Klease, G. T. J. Med. Chem. 1973, 16, 856.
12. Gao, Y.; Zepp, C. M. U.S. Patent # 5,545,745, 1996, 13 August.
13. ten Hoeve, W.; Wynberg, H. J. Org. Chem. 1985, 50, 4508.
14. Pihlaja, K.; Kleinpeter, E. Carbon-13 NMR Chemical Shifts in Structural and Stereochemical Analysis; VCH: New York,
1994; p. 175.
15. Jacques, J.; Collet, A.; Wilen, S. H. Enantiomers, Racemates and Resolutions; John Wiley: New York, 1981; pp. 217–250.
16. Wilen, S. H.; Collet, A.; Jacques, J. Tetrahedron 1977, 33, 2725.
17. Vries, T.; Wynberg, H.; van Echten, E.; Koek, J.; ten Hoeve, W.; Kellogg, R. M.; Broxterman, Q. B.; Minnaard, A.;
Kaptein, B.; van der Sluis, S.; Hulshof, L.; Kooistra, J. Angew. Chem., Int. Ed. Engl. 1998, 37, 2349.
18. Collet, A. Angew. Chem., Int. Ed. Engl. 1998, 37, 3239.
19. Nemák, K.; Ács, M.; Jászay, Z. M.; Kozma, D.; Fogassy, E. Tetrahedron 1996, 52, 1637.
20. Mravik, A.; Lepp, Z.; Fogassy, E. Tetrahedron: Asymmetry 1996, 7, 2387.
21. Parker, D.; Taylor, R. J. Tetrahedron 1987, 43, 5451.
22. Adams, A. G.; Stewart, J. T. J. Liquid Chromatog. 1993, 16, 3863.
23. Effenberger, F.; Jäger, J. J. Org. Chem. 1997, 62, 3867.
24. Cambridge Structural Database and Cambridge Structural Database System Version 5.11 (April 1998), Cambridge
Crystallographic Data Centre, University Chemical Laboratory, Cambridge, England.

M. R. Caira et al. / Tetrahedron: Asymmetry 10 (1999) 2175–2189
2189
25. Flack, H. D. Acta Cryst. 1983, A39, 876.
26. Tung, R. D.; Rich, D. H. J. Am. Chem. Soc. 1985, 107, 4342.
27. Halterman, R. L.; Vollhardt, K. P. C.; Welker, M. E. J. Am. Chem. Soc. 1987, 109, 8105.
28. Dehmlow, E. V.; Westerheide, R. Synthesis 1992, 947.
29. Dale, J. A.; Dull, D. L.; Mosher, H. S. J. Org. Chem. 1969, 34, 2543.
30. Ward, D. E.; Rhee, C. K. Tetrahedron Lett. 1991, 32, 7165.
31. Dale, J. A.; Mosher, H. S. J. Am. Chem. Soc. 1973, 95, 512.
32. Latypov, Sh. K.; Seco, J. M.; Quiñoá, E.; Riguera, R. J. Org. Chem. 1995, 60, 504 and references cited therein.
33. Latypov, Sh. K.; Seco, J. M.; Quiñoá, E.; Riguera, R. J. Org. Chem. 1996, 61, 8569.
34. Nowick, J. S.; Holmes, D. L.; Noronha, G.; Smith, E. M.; Nguyen, T. M.; Huang, S.-L. J. Org. Chem. 1996, 61, 3929.
35. Eckert, H.; Forster, B. Angew. Chem., Int. Ed. Engl. 1987, 26, 894.
36. Sheldrick, G. M. Acta Crystallogr. 1990, A46, 467.
37. Sheldrick, G. M. SHELXL-93. Program for the Refinement of Crystal Structures; University of Göttingen: Germany, 1993.
38. Otwinowski, Z.; Minor, W. Processing of X-Ray Diffraction Data in Oscillation Mode; Carter, C. W.; Sweet, R. M., Eds.;
Methods in Enzymology; Academic Press: New York, 1996; Vol. 276, pp. 307–326.