Determination of the enantiomeric purity and the configuration of β- aminoalcohols using (R)-2-fluorophenylacetic acid (AFPA) and fluorine-19 NMR: Application to β-blockers

Article abstract of DOI:10.1016/S0957-4166(00)00254-8

A method has been developed for determining the enantiomeric purity and the absolute configuration of β-aminoalcohols of type ArOCH₂CH(OH)CH₂NHR (R = iPr, tBu). To determine enantiomeric purity, the amine function was first protected by a benzyl group, then the compound formed was esterified using the acid chloride of (R)-2-fluorophenylacetic acid (AFPA). The ¹⁹F NMR analysis of the derivative obtained revealed the presence of two distinctly separate signals (~2.5 ppm), the one for the RS-SR pair being the most deshielded. The configuration was determined directly on the aminoalcohol by using the acid. In stoichiometric conditions, when R = iPr, the amide function was obtained very preponderantly. The ¹⁹F NMR spectrum of the amide presented four distinct signals when derivatization was carried out by means of a reaction between the (±)-β-aminoalcohol and the (R)-AFPA. The extreme signals, which were over 3.5 ppm apart, did not belong to the same diastereomer. With R = tBu essentially the ester function was obtained. The first studies revealed the presence of two signals, though not as clearly separated as in the previous cases. Each experiment was simple to perform, and purification was not necessary. Mosher's acid gave unsatisfactory results in each case. (C) 2000 Elsevier Science Ltd.

Full text of DOI:10.1016/S0957-4166(00)00254-8

Tetrahedron: Asymmetry 11 (2000) 2885±2898
Determination of the enantiomeric purity and the
con®guration of b-aminoalcohols using
(R)-2-¯uorophenylacetic acid (AFPA) and ¯uorine-19 NMR:
application to b-blockers
Marcel Apparu,* Younes Ben Tiba, Pierre-Marc Leo, Sylvain Hamman
and Christian Coulombeau
  Â
Laboratoire d'Etudes Dynamiques et Structurales de la Selectivite, Universite Joseph Fourier, BP 53,
38041 Grenoble Cedex 9, France
Received 10 April 2000; accepted 29 June 2000
Abstract
A method has been developed for determining the enantiomeric purity and the absolute con®guration of
b-aminoalcohols of type ArOCH2CH(OH)CH2NHR (R=iPr, tBu). To determine enantiomeric purity, the
amine function was ®rst protected by a benzyl group, then the compound formed was esteri®ed using the
acid chloride of (R)-2-¯uorophenylacetic acid (AFPA). The 19F NMR analysis of the derivative obtained
revealed the presence of two distinctly separate signals (ꢀ2.5 ppm), the one for the RS±SR pair being the
most deshielded. The con®guration was determined directly on the aminoalcohol by using the acid. In
stoichiometric conditions, when R=iPr, the amide function was obtained very preponderantly. The 19F
NMR spectrum of the amide presented four distinct signals when derivatization was carried out by means
of a reaction between the (þ)-b-aminoalcohol and the (R)-AFPA. The extreme signals, which were over 3.5
ppm apart, did not belong to the same diastereomer. With R=tBu essentially the ester function was
obtained. The ®rst studies revealed the presence of two signals, though not as clearly separated as in the
previous cases. Each experiment was simple to perform, and puri®cation was not necessary. Mosher's acid
gave unsatisfactory results in each case. # 2000 Elsevier Science Ltd. All rights reserved.
1. Introduction
Determination of the enantiomeric purity and the absolute con®guration of a compound is
now of particular importance when synthesizing chiral molecules where one enantiomer is of
biological interest. For example, in the case of b-blockers of type ArOCH2CH(OH)CH2NHR,
* Corresponding author.
0957-4166/00/$ - see front matter # 2000 Elsevier Science Ltd. All rights reserved.
PII: S0957-4166(00)00254-8

2886
M. Apparu et al. / Tetrahedron: Asymmetry 11 (2000) 2885±2898
which have to be synthesized, the carbon carrying the alcohol function is stereogenic and only the
S enantiomer is suciently active.
Numerous determination methods exist, using a variety of techniques such as polarimetry,
chromatography and NMR.1 In the latter case, the most interesting aspect is the condensation of
a chiral derivatization agent with the compound being studied. A certain number of chiral acids
or acid chlorides with a general structure ArCXYCOOH have proved to be very useful in studies
of chiral alcohols and amines.
o-Methylmandelic acid (X=CH3O, Y=H)2 6 and especially 2-methoxy-2-tri¯uorophenylacetic
acid (MTPA) (X=CH3O, Y=CF3) proposed by Mosher in 1969 have been widely used.2,4,7 12
The latter, which has the advantage of not causing racemization during derivatization, can also
be used in 19F NMR studies. This possibility is of interest for a number of reasons. In particular,
the di€erence in ÁꢀF chemical displacement is greater than ÁꢀH and the signals obtained are in a
clear zone.
However, MTPA also has certain disadvantages: its reactivity is not always sucient, for steric
reasons.10,16,17 In addition, the results obtained from NMR of the proton or ¯uorine are not
always reliable.10,18 Finally, on the one hand, the binding of ¯uorine atoms to a carbon adjacent
to the stereogenic carbon (rather than on it), and, on the other hand, long distance coupling
between the methoxy hydrogens and the ¯uorines (broadening of the signals)7,19 are not
favourable. 2-Fluorophenylacetic acid (AFPA) (X=F, Y=H) has the advantage of having a
¯uorine atom which is directly bound to the stereogenic carbon and consequently more sensitive
to diastereotopic magnetic di€erences. The rare studies performed, using the R con®guration of
this acid, have shown that the ÁꢀF values observed are in most cases very signi®cant and that
racemization does not occur during esteri®cation.20,21
In the context of our studies on b-blockers, we synthesized the S isomer of an iodinated
analogue 2 of practolol 1 via 3 (Scheme 1).22 Preliminary biological studies revealed this analogue
2 to be of considerable interest.23,24 In order to determine the enantiomeric purity obtained
during the synthesis of 2, we had to develop a determination method which would overcome a
problem of chemoselectivity inherent to ꢁ-aminoalcohols, namely the presence in the b position of
two often closely related reactivity functions. We used the chloride of 2-¯uorophenylacetic acid.
For comparative purposes, we also performed the same reactions with the chloride of Mosher
acid.
Scheme 1.
Moreover, from the studies already mentioned, it can be concluded that MTPA, AFPA and
o-methylmandelic acid can be used to determine the con®guration of secondary chiral alcohols
and aminoalcohols carrying a chiral carbon in the a position of the nitrogen. We therefore
decided to ®nd out whether this same method could also be used to determine the con®guration
in the case of b-aminoalcohols of which b-blockers are a part.

M. Apparu et al. / Tetrahedron: Asymmetry 11 (2000) 2885±2898
2887
2. Results and discussion
2.1. Synthesis of compounds
The compounds synthesized for this study were obtained according to Scheme 2.
Scheme 2. Reagents: (i) NaH, DMF, 0ꢁC; (ii) (þ)- or (S)-glycidyl tosylate; (iii) amine, isopropanol, re¯ux
2.2. Determination of enantiomeric purity
The method was developed on the simple model compound (þ)-6 (Scheme 3). Obviously, it was
necessary during the derivatization to obtain a total reaction, and for one function only.
Determination of enantiomeric purity thus involved a preliminary step to protect one of the
functions. After various tests, protection of the amine function by the benzyl group was found to
give the best results.25,26
Esteri®cation was performed from the acid chlorides with an acid chloride:alcohol ratio of 3:2.
A ®rst experiment performed on (þ)-11 with the chloride of the racemic ¯uorophenylacetic acid
(Scheme 3a) showed, after workup, a very simple 19F NMR spectrum (recorded in chloroform):
two signals situated at ^15.78 and ^18.30 ppm, of di€erent intensity (54:46). A second experiment
performed using the same conditions with the R acid chloride revealed the presence of two

2888
M. Apparu et al. / Tetrahedron: Asymmetry 11 (2000) 2885±2898
Scheme 3. Reagents: (i) C6H5CH2Cl, K2CO3, CH3CN, re¯ux; (ii) (þ)-C6H5CFHCOCl, pyridine, rt; (iii) (R)-
C6H5CFHCOCl, pyridine, rt; (iiii) C6H5CH2Br, K2CO3, CH3CN, re¯ux
signals, of the same intensity in this case, at ^15.88 and ^18.36 ppm. Finally, a third experiment,
performed with the S acid chloride, gave the same results: the presence of two signals of the same
intensity at ^15.82 and ^18.32. Together, these three experiments con®rmed that the esteri®cation
was complete. It was thus quite possible, because of the signi®cant di€erence (2.5 ppm) between
the diastereomeric signals that were formed, to determine the enantiomeric purity (with the
precision of the NMR) of the starting b-aminoalcohols. The method leads to a very good
correlation when applied to 2 (^15.94 and ^18.36) or to 3 (^15.91 and ^18.40).26 The esteri®cation
performed from (S)-11 (Scheme 3b) made it possible to determine the position of the signal of
each pair of enantiomers RS±SR and RR±SS. The ®rst pair was found to have the most
deshielded signal.

M. Apparu et al. / Tetrahedron: Asymmetry 11 (2000) 2885±2898
2889
The chlorides of Mosher's acid did not give the desired esters, most likely for reasons of
bulkiness: the proton NMR spectrum of the crude reaction mixture recorded after 15 h revealed
no signal at 5.2 ppm (characteristic of >CHOCO).
A great many b-blockers have a tertiobutyl group bound to nitrogen in the place of the
isopropyl. By applying the reactions performed from (þ)-6 to (þ)-8 we were able to conclude that
it was also possible to determine the enantiomeric purity when the nitrogen carries a tertiobutyl
group: the 19F spectrum in CHCl3 of esters 14 obtained by the action of the R acid chloride on
(þ)-13 (Scheme 3c) presented two signals at ^16.39 and ^18.44 ppm, in other words very close to
those of (RS)-12 and (RR)-12. Their intensity ratio here was also 54:46. The position at ^16.40
ppm of the signal obtained after esteri®cation of (S)-8 (Scheme 3d) indicated that the enantiomer
pair RS±SR again had the most deshielded signal.
2.3. Determination of con®guration
In this case, the approach to the problem was simpli®ed: the intensity ratio of the two signals
did not have to be determined, but simply the position of the signal corresponding to each R or S
isomer. The identi®cation, discussed in the previous paragraph, of two distinctly separate signals
in the case of the N-benzylesters could also be used to determine the con®guration of the starting
b-aminoalcohol. But it was decided to try and develop a simpler method which would not require
protection of a function or the use of the acid chloride, which required preparation. To do this,
we started with the model compound (þ)-6.
Direct action of the ¯uorinated acid was ®rst envisaged, since it gives rise to the formation of
ammonium salts only. Unfortunately, in 19F NMR, these salts presented only one fairly broad
signal, both with Mosher's acid and with 2-¯uorophenylacetic acid, contrary to what may be
observed during the reaction between a-methylbenzylamine and AFPA.20
The action of the ¯uorinated acid on the b-aminoalcohol in the presence of dicyclohexyl-
carbodiimide (DCC) and a catalyst was then envisaged. In this case, we expected to observe the
simultaneous formation of the amide and ester functions. The results obtained would be positive
if, on the one hand, the zones where signals were present did not overlap, and, on the other hand,
at least one of the functions gave rise to the presence of distinctly separate signals. By reacting
(þ)-6 with a large excess of the chloride of racemic ¯uorophenylacetic acid, which led to 15, we
were able to identify the zones corresponding to the signals of the amides 16 and esters 17
(Scheme 4). In dimethylsulfoxide, used because of the low solubility of b-blockers often observed
in chloroform, these zones extended, respectively, from ^3.4 to ^7.9 ppm and from ^10.8 to ^14.5
Scheme 4.

2890
M. Apparu et al. / Tetrahedron: Asymmetry 11 (2000) 2885±2898
ppm using C6F6 as a reference. When we worked with stoichiometric amounts of (R)-2-¯uoro-
phenylacetic acid in the presence of DCC and hydroxybenzotriazole (HOBT) the very
preponderant formation of the (RR)-16 and (RS)-16 amides was observed. The spectra were also
recorded in other solvents (Table 1). In each case, the presence of four distinct signals was
observed after the suppression of the H±F coupling. The subsequent studies were carried out in
methanol. Each RR or RS con®guration of the amide 16 was thus demonstrated by the presence
of two ¯uorine-NMR signals. They correspond to the Z and E conformers (Scheme 5).5,19 The
latter result from a rotation, which is slow in relation to the NMR time scale, around the amide
bond. In order to determine the position of the signals corresponding to each enantiomer of (þ)-6,
an experiment was carried out with (S)-6. The two signals of the (RS)-16 isomer obtained were
situated at ^3.04 and ^3.68 ppm, in other words more deshielded than those of the (RR)-16 iso-
mer. A study of molecular modelling, which is in the course of publication, allowed us to partly
justify the position of the di€erent signals: the signal at ^6.59 ppm corresponds to the RR(Z) form
and the one at ^4.24 ppm to the RR(E) form. This last one is clearly predominant and its
percentage varies with the solvent: methanol (75), acetone (87.5), acetonitrile (84.5).
The reaction carried out under the same conditions with Mosher's acid gave much less decisive
results. No matter what solvent was used, the four good signals were never completely separate.
The biggest di€erence between the extremes was 0.71 ppm in methanol.
Table 1
ꢀF values of signals (and ÁꢀF di€erences between these signals) observed for amides 16 (Ar=C6H5) in di€erent
solvents. Internal reference: C6F6
Scheme 5.

M. Apparu et al. / Tetrahedron: Asymmetry 11 (2000) 2885±2898
2891
These results, which can be used for establishing a correlation between the dF values observed
and the R and S con®gurations of the b-aminoalcohols, needed to be con®rmed. The aryl group
of b-blockers is in fact generally substituted, which was not the case for (þ)-6; this may in¯uence
the results. The reaction was thus extended to six b-blockers substituted in various ways: practolol 1
and oxprenolol 10 (in racemic and S forms) prepared according to Scheme 1, propranolol
(racemic and S), atenolol, metoprolol and pindolol, which are available commercially. The results
are shown in Table 2. It can be seen from Table 2 that for the di€erent derivatives each signal was
situated in a narrow ꢀF range and there was a correlation between the ꢀF values and the absolute
con®guration: the RS-isomer presented the most deshielded signals (except for 10). Both of the
extreme signals thus corresponded to a di€erent con®guration. The di€erence between these
signalsÐat least 3.5 ppmÐwas such that they could be used to determine the R or S con®guration
of the b-aminoalcohols and in particular of the b-blockers with the structure ArOCH2-
CHOHCH2NHiPr.
Table 2
Chemical displacement of ¯uorine signals in amides 16. Solvent: methanol. Internal reference: C6F6
Here again, as with 6, the E conformer clearly predominates in the case of the RR con®guration.
One cannot, however, establish a valid correlation between the importance of the E:Z ratio and
the substituent (nature or position) in the cycle, as the results below show. With the para
substituent the following percentages of E conformer can be observed: 81 (practolol), 74 (ateno-
lol) and 72 (metoprolol). With the ortho substituent the results are very similar: 80 (propranolol),
73 (oxprenolol) and 70 (pindolol). In the case of the RS isomer, for which the determination of
the position of each conformer is in progress, one arrives at the same conclusion. In e€ect, there is
practically equality of percentage between the two forms for oxprenolol (50), propranolol (51),
atenolol and 6 (52). The other three compounds show a slightly greater di€erence (57% of
predominant conformer). In each case the most deshielded signal is the less important. As with
the RR isomer, it is therefore not possible to establish a correlation between the percentage of the
predominant form and the substitution in the aromatic cycle.

2892
M. Apparu et al. / Tetrahedron: Asymmetry 11 (2000) 2885±2898
When the nitrogen carried a tertiobutyl, as in the case of (þ)-8, it was observed that the
derivatization carried out in the same conditions as for (þ)-6 led to the extremely preponderant
formation of the RS and RR esters 18 (Scheme 4), for obvious steric reasons. The spectra plotted
in the same solvents as previously gave the values shown in Table 3. It can be seen that the results
are not as clear as in the case of amides 16. Only acetone, chloroform and DMSO gave sucient
di€erences between the two expected signals for there to be some hope of being able to determine
the con®guration of the initial aminoalcohol. Unlike the case of the amides, it is not possible to
generalize on the basis of these results. Further studies are needed on compounds where the
aromatic cycle is substituted in di€erent ways. Such studies are now under way. Even so, these
®rst results are very encouraging.
Table 3
Fluorine chemical shift in esters 18
3. Conclusion
We developed a method for determining the enantiomeric purity and con®guration of
b-aminoalcohols with the structure ArOCH2CH(OH)CH2NHR (R=iPr or tBu) using 19F NMR.
The derivatization agent used, 2-¯uorophenylacetic acid or its chloride, made it possible to obtain
very simple spectra with distinctly separate signals. For determination of enantiomeric purity,
after protection of the amine function by a benzyl group, the ester spectrum obtained presented
essentially two signals separated by at least 2.5 ppm. For determination of the con®guration, the
direct action of the acid on the aminoalcohol led to the very preponderant formation of amide
(R=iPr) or ester (R=tBu). In the ®rst case, the spectrum presented four distinctly separate signals
in various solvents. The two extreme signals, which did not belong to the same diastereomer
compound, were at least 3.5 ppm apart. The presence or absence of one of these two signals can
thus be used to determine the con®guration of the b-aminoalcohol unambiguously. The method
was successfully applied to six b-blockers. With R=tBu, the ®rst results obtained were positive,
but they must be completed with further studies. From a practical standpoint, a complete
derivatization reaction was not required to determine the con®guration; the running of this
reaction and workup were very simple, without ®nal puri®cation.

M. Apparu et al. / Tetrahedron: Asymmetry 11 (2000) 2885±2898
2893
4. Experimental
Tetrahydrofuran was dried by distillation from sodium benzophenone ketyl under argon
immediately prior to use. Pyridine and acetonitrile were dried by re¯ux over and distillation from
CaH2. DMF was freshly distilled from BaO. All epoxidations, amide and ester formations were
conducted under an atmosphere of argon. TLC was performed on Merck aluminium precoated
plates of silica gel 60F-254 (detection by UV and by spraying with phosphomolybdic acidÐ5% in
ethanolÐfollowed by heating). Column chromatography was carried out on a Merck kieselgel 60
(0.063±0.20 mm) using the same eluting system as for TLC, unless otherwise noted.
Uncorrected melting points were determined on a Buchi melting point apparatus. Optical
rotations were determined at the sodium D line with a Perkin±Elmer 341 polarimeter. Infrared
spectra were recorded on a Nicolet impact 400 spectrophotometer. NMR spectra were recorded on
a Bruker AC 200 spectrometer using Me4Si as internal reference for 1H NMR spectra, CDCl3 for
13C NMR and C6F6 for 19F NMR spectra. Coupling constants (J) are given in hertz. Elemental
analyses were determined by the Service Central d'Analyses du CNRS (Vernaison, France).
General procedure for the epoxidation reaction. To a slurry of NaH in dry DMF at 0ꢁC was
added dropwise a solution of the phenol in DMF. When the H2 evolution stopped, glycidyl
tosylate in DMF was added dropwise. After complete consumption of the phenol at rt, the
solvent was very carefully removed and the resulting crude product dissolved in EtOAc. The
organic layer was washed three times with water, then dried, concentrated in vacuo and the
residue puri®ed by column chromatography.
4.1. (þ)-3-(4-Acetamidophenoxy)-1,2-epoxypropane (þ)-427,28
4-Acetamidophenol (7.55 g, 50 mmol), NaH (2.20 g, 55 mmol), (þ)-glycidyl tosylate (11.40 g,
50 mmol) and DMF (100 mL) gave, after column chromatography (CHCl3:MeOH, 97:3), 7.7 g
ꢁ
ꢁ
1
(75%) of epoxide (þ)-4 as a white solid: mp 112±113 C (lit.:27,28 110 C); H NMR (200 MHz,
CDCl3) ꢀ 2.13 (s, 3H), 2.73±2.77 (dd, H, J=4.8 and 2.6 Hz), 2.88±2.92 (2 overlapping d, H), 3.30±
3.38 (m, H), 3.86±3.95 (dd, H, J=10.9 and 5.8 Hz), 4.16±4.24 (dd, H, J=10.9 and 3.1 Hz), 6.82±
6.87 (d, 2H, J=8.9 Hz), 7.36±7.41 (d, 2H, J=8.9 Hz), 7.48 (s, H); 13C NMR (50 MHz, CDCl3) ꢀ
24.2, 44.6, 50.1, 69.0, 114.9, 121.8, 131.5, 155.1, 168.6.
4.2. (S)-3-(4-Acetamidophenoxy)-1,2-epoxypropane (S)-429
4-Acetamidophenol (0.755 g, 5 mmol), NaH (0.22 g, 5.5 mmol) and (S)-glycidyl tosylate (1.14
g, 5 mmol) in DMF (15 mL) furnished 0.700 g (66%) of (S)-4 as a white solid: mp 108±110ꢁC
(lit.:29 104±106ꢁC for the R isomer); ‰ꢂŠD20=9.6 (c 1.02, MeOH) [lit.:29 [ꢂ]23655=^18.5 (c 2.0, EtOH)
for the R isomer].
4.3. (S)-1,2-Epoxy-3-phenoxypropane (S)-5
Phenol (0.5 g, 5.32 mmol), NaH (0.23 g, 6 mmol) and (S)-glycidyl tosylate (1.21 g, 5.30 mmol)
in DMF (15 mL) gave (S)-5 (0.573 g, 73%) as an oil after chromatography (cyclohexane:EtOAc,
6:2); [ꢂ]D=5.2 (c 1.15, CHCl3).

2894
M. Apparu et al. / Tetrahedron: Asymmetry 11 (2000) 2885±2898
4.4. (þ)-3-(4-Iodophenoxy)-1,2-epoxypropane (þ)-730
4-Iodophenol (3.3 g, 15 mmol), NaH (0.72 g, 18 mmol) and (þ)-glycidyl tosylate (4 g, 18 mmol)
in DMF (25 mL) furnished after column chromatography (CHCl3:MeOH, 9:1) (þ)-7 (3.32 g,
80%) as a white solid: mp 67±68ꢁC; 1H NMR (200 MHz, CDCl3) ꢀ 2.71±2.76 (dd, H, J=4.8 and
2.7 Hz), 2.87±2.92 (overlapping dd, H), 3.28±3.37 (m, H), 3.85±3.93 (dd, H, J=11.0 and 5.8 Hz),
4.16±4.23 (dd, H, J=11.0 Hz and 3.1 Hz), 6.67±6.71 (d, 2H, J=8.6 Hz), 7.51±7.57 (d, 2H, J=8.2
Hz); 13C NMR (50 MHz, CDCl3) ꢀ 44.58, 49.9, 68.8, 83.4, 117.0, 138.2, 158.3.
4.5. (S)-3-(4-Iodophenoxy)-1,2-epoxypropane (S)-7
4-Iodophenol (0.44 g, 2 mmol), NaH (0.088 g, 2.2 mmol) and (S)-glycidyl tosylate (0.456 g, 2
mmol) in DMF (25 mL) furnished after puri®cation 0.41 g (74%) of (S)-7 as a solid: mp 58±
60ꢁC; ‰ꢂŠ2D0=9.1 (c 1.01, MeOH).
4.6. (þ)-3-(2-Allyloxyphenoxy)-1,2-epoxypropane (þ)-931,32
Catechol monoallylether33 (1.5 g, 10 mmol), HNa (0.44 g, 11 mmol) and (þ)-glycidyl tosylate
(2.28 g, 10 mmol) in DMF (50 mL in all) furnished after chromatography (cyclohexane:EtOAc,
1
8:2) 1.65 g (80%) of (þ)-9 as an oil. H NMR (200 MHz, CDCl3) ꢀ 2.73±2.77 (dd, H, J=5.0 and
2.6 Hz), 2.85±2.90 (overlapping dd, H), 3.31±3.42 (m, H), 3.99±4.07 (dd, H, J=11.3 and 5.5 Hz),
4.21±4.29 (dd, H, J=11.3 and 3.4 Hz), 5.29±5.47 (m, 2H), 5.98±6.18 (m, H), 6.88±6.99 (m, 4H);
13C NMR (50MHz, CDCl3) ꢀ 44.5, 50.1, 69.7, 70.1, 114.3, 114.9, 117.3, 121.2, 121.8, 133.3, 148.4,
148.6.
4.7. (S)-3-(2-Allyloxyphenoxy)-1,2-epoxypropane (S)-9
Starting from the same quantities as for (þ)-9, 1.41 g (68.5%) of (S)-9 was obtained as an oil;
‰ꢂŠ2D0=13.55 (c 1.07, MeOH).
Preparation of the aminoalcohols: general procedure. A solution of the epoxide and the chosen
amine (10 equiv.) in isopropanol was heated to re¯ux. After complete consumption of the epoxide
(TLC), the solvent was evaporated to dryness, the residue powdered, and stirred in ether or pen-
tane at rt. After ®ltration the compound was recovered as a white powder. The yield was always
higher than 90%. The aminoalcohol could also be puri®ed by column chromatography.
4.8. (þ)-1-Isopropylamino-3-phenoxypropan-2-ol (þ)-634,35
(þ)-1,2-Epoxy-3-phenoxypropane (22.5 g, 0.15 mol) and isopropylamine (60 g)ꢁin isopropanol
(150 mL) gave after treatment with pentane 30.1 g (96%) of (þ)-6: mp 92.5±94 C.36 1H NMR
(200 MHz, CDCl3) ꢀ 1.03 and 1.06, (d, 6H, J=6.2 Hz), 2.64±2.95 (m, 5H), 3.91±4.09 (m, 3H),
6.85±7.00 (m, 3H), 7.21±7.34 (m, 2H); 13C NMR (50 MHz, CDCl3) ꢀ 22.9, 23.1, 48.9, 49.3, 68.3,
70.4, 114.6, 120.9, 129.4, 158.6.

M. Apparu et al. / Tetrahedron: Asymmetry 11 (2000) 2885±2898
2895
4.9. (S)-1-Isopropylamino-3-phenoxypropan-2-ol (S)-6
(S)-5 (570 mg, 3.8 mmol) and isopropylamine (3.36 g, 57 mmol) in isopropanol (30 mL) gave,
after chromatography (EtOAc:MeOH, 6:4), (S)-6 (575 mg, 72%) as a white powder: mp 42±
43.5ꢁC; ‰ꢂŠ2D5=^7.8 (c 1.05, CHCl3).
4.10. (þ)-3-(4-Acetamidophenoxy)-1-(isopropylamino)propan-2-ol (practolol) (þ)-128
ꢁ
ꢁ
1
Mp 140±142 C (lit.:28 142±143 C); H NMR (200 MHz, CDCl3) ꢀ 1.06±1.09 (d, 6H, J=6.2
Hz), 2.14 (s, 3H), 2.40 (br s, 2H), 2.65±3.00 (m, 3H), 3.80±4.10 (m, 3H), 6,85 (d, 2H, J=9.0 Hz),
7.32 (s, H), 7.36 (d, 2H, J=9.0 Hz); 13C NMR (50 Hz, CDCl3) ꢀ 22.4, 22.6, 23.6, 50.3, 50.9, 69.6,
72.2, 115.6, 122.9, 133.2, 156.9, 171.3.
4.11. (S)-3-(4-Acetamidophenoxy)-1-(isopropylamino)propan-2-ol (S)-129
Mp 132±134ꢁC; ‰ꢂŠ2D5=^2.42 (c 0.99, EtOH) [lit.:29 ‰ꢂŠ2D5=þ4.3 (c 1.0, EtOH) for R isomer].
4.12. (þ)-1-tert-Butylamino-3-(4-iodophenoxy)propan-2-ol (þ)-836
Mp 96±97ꢁC; 1H NMR (200 MHz, CDCl3) ꢀ 1.13 (s, H), 2.61±2.74 (m, 2H), 2.78±2.91 (m, 4H),
3.86±4.02 (m, 2H), 6.64±6.73 (dt, 2H, J=8,9 Hz and 2.0 Hz), 7.50±7.59 (dt, 2H, J=8.9 and 2.0
Hz); 13C NMR (50 MHz, CDCl3) ꢀ 28.8, 44.7, 50.6, 68.3, 70.6, 82.9, 116.8, 138.0, 158.5.
4.13. (S)-1-tert-Butylamino-3-(4-iodophenoxy)propan-2-ol (S)-8
Mp 72±74ꢁC; ‰ꢂŠ2D0=2.61 (c 1.07, CHCl3).
4.14. (þ)-3-(2-Allyloxyphenoxy)-1-(isopropylamino)propan-2-ol (oxprenolol) (þ)-10 32
Mp 77±78ꢁC (lit.:32 78±80ꢁC); H NMR (200 MHz, CDCl3) ꢀ 1.05±1.09 (d, 6H, J=6.2 Hz),
1
2.69±2.88 (m, 5H), 3.94±4.20 (m, 3H), 4.54±4.58 (m, 2H), 5.22±5.47 (m, 2H), 5.97±6.16 (m, 4H),
6.83±6.98 (m, 4H); 13C NMR (50 MHZ, CDCl3) ꢀ 22.9, 48.8, 49.3, 68.4, 69.8, 73.0, 114.2, 115.3,
117.7, 121.5, 121.9, 133.3, 148.7, 148.9.
4.15. (S)-3-(2-Allyloxyphenoxy)-1-(isopropylamino)propan-2-ol (S)-1031
Mp 56±58ꢁC; ‰ꢂŠ2D0=^2.57 (c 1.05, CHCl3).
4.16. (þ)-1-(N-Benzyl)isopropylamino-3-phenoxypropan-2-ol (þ)-11
A solution of (þ)-6 (1 g, 4.78 mmol) and benzyl chloride (0.847 g, 6.69 mmol) in anhydrous
acetonitrile (30 mL) was re¯uxed under argon in the presence of dry potassium carbonate (0.926
g, 6.69 mmol). After completion of the reaction, and monitoring by TLC (cyclohexane:acetone,
4.5:0.5), the solid was ®ltered and rinsed with CH2Cl2. The ®ltrate was concentrated in vacuo and
the residue puri®ed by chromatography to give 1.30 g (91%) of (þ)-11 as a viscous yellowish oil.
1
IR (neat) 3444 cm^1; H NMR (200 MHz, CDCl3) ꢀ 1.01±1.09 (2d, 6H, J=6.6 Hz), 2.58±2.64

2896
M. Apparu et al. / Tetrahedron: Asymmetry 11 (2000) 2885±2898
(m, 2H), 2.98 (sept, H, J=6.6 Hz), 3.52±3.75 (2d, 2H, J=13.8 Hz), 3.87±3.95 (m, 3H), 6.85±6.96
(m, 3H), 7.19±7.31 (m, 7H); 13C NMR (50 MHz, CDCl3) ꢀ 16.1, 19.6, 50.1, 51.9, 54.8, 66.1, 70.3,
114.5, 120.8, 127.1, 128.4, 128.6, 129.3, 158.8. Anal. calcd for C19H25NO2: C, 76.22; H, 8.42; N,
4.68. Found: C, 76.01; H, 8.60; N, 4.54.
4.17. (S)-1-(N-Benzyl)isopropylamino-3-phenoxypropan-2-ol (S)-11
Aminoalcohol (S)-6 (100 mg, 0.48 mmol), benzyl chloride (84.7 mg, 0.67 mmol), acetonitrile
(5 mL) and dry K2CO3 (93 mg, 0.67 mmol) gave (S)-11 (120 mg, 84%) after chromatography
(cyclohexane:EtOAc:MeOH, 8.5:1:0.5) as a viscous oil. ‰ꢂŠ2D0=^17.2 (c 1.15, CHCl3).
4.18. (þ)-1-(N-Benzyl)tert-butylamino-3-(4-iodophenoxy)propan-2-ol (þ)-13
The procedure was the same as for (þ)-11. Aminoalcohol (þ)-8 (80 mg, 0.21 mmol), benzyl
bromide (70 mg, 0.42 mmol), acetonitrile (8 mL) and dry K2CO3 (57 mg, 0.42 mmol) furnished
after chromatography (cyclohexane:EtOAc, 4.5:0.5) 82 mg (91%) of (þ)-13 as an oil. IR (neat)
3432; 1H NMR (200 MHz, CDCl3) ꢀ 1.19 (s, 9H), 2.64±2.85 (m, 2H), 3.2 (s, H), 3.40±3.49 (m, H),
3.52±3.59 (d, H, J=14.4 Hz), 3.65±3.68 (m, 2H), 3.68±3.82 (d, H, J=14.4 Hz), 6.55±6.61 (d, 2H,
J=9 Hz), 7.20±7.30 (m, 5H), 7.44±7.52 (d, 2H, J=9 Hz); 13C NMR (50 MHz, CDCl3) ꢀ 27.4,
53.6, 55.8, 66.9, 70.3, 82.8, 116.9, 127.0, 128.5, 138.0, 141.6, 158.6. Anal. calcd for C20H26INO2:
54.67; H, 5.92; I, 28.93; N, 3.19. Found: C, 54.83; H, 6.14; I, 27.32; N, 3.11.
4.19. (S)-1-(N-Benzyl)tert-butylamino-3-(4-iodophenoxy)propan-2-ol (S)-13
Aminoalcohol (S)-8 (100 mg, 0.286 mmol), benzyl bromide (73 mg, 0.42 mmol), acetonitrile (5
mL) and dry K2CO3 (60 mg, 0.43 mmol) gave (S)-13 (100 mg, 88%) after chromatography
(cyclohexane:EtOAc:MeOH, 8.5:1:0.5) as a viscous oil. ‰ꢂŠ2D0=5.0 (c 1.00, CHCl3).
4.20. 1-(N-Benzyl)isopropylamino-3-phenoxyprop-2-yl 2-¯uorophenylacetate 12
To a solution of (þ)-11 (35 mg, 0.12 mmol) in anhydrous CH2Cl2 at rt, under argon, was added
(R)-2-¯uorophenylacetyl chloride (40 mg, 0.23 mmol) and pyridine (9 mL). After complete
consumption of the starting material (TLC; cyclohexane:acetone, 9:1) the reaction mixture was
washed ®rst with an NaHCO3 saturated aqueous solution, then with water until neutrality. The
organic layer was dried and the solvent evaporated. The residue was dissolved in the NMR
solvent. The esteri®cation of (þ)-13 or (S)-13 followed the same way to give isomers 14.
4.21. 1-(N-Isopropyl-ꢂ-¯uorotoluenecarboxamido)-3-phenoxyprop-2-yl 2-¯uorophenyl acetate 15
To a solution of (þ)-6 (42 mg, 0.2 mmol) and (R)-a-¯uorophenylacetyl chloride (173 mg, 1
mmol) in THF:acetonitrile (1:1, 10 mL) was added triethylamine (101 mg, 1 mmol). A
precipitate appeared and the reactionÐmonitored by TLC (methanol:CHCl3, 3:7)Ðwas ®nished
after 1 hour. The precipitate was ®ltered and the solvent removed in vacuo to dryness. The
residue was dissolved in CH2Cl2 (20 mL) and the solution washed with 5% NaHCO3. After
drying and evaporation of the solvent, the residue was dissolved in deuterated DMSO for NMR
measures.

M. Apparu et al. / Tetrahedron: Asymmetry 11 (2000) 2885±2898
2897
Reaction of AFPA (or MTPA) with the aminoalcohols in the presence of DCC and HOBT:
general procedure. A solution of the aminoalcohol (0.1 mmol), AFPA (0.1 mmol), DCC (0.11
mmol) and HOBT (0.11 mmol) in THF:acetonitrile (1:1) was stirred at rt. After consumption of
the starting aminoalcohol, monitored by TLC (methanol:CHCl3, 3:7), the reaction mixture was
®ltered and the solvents evaporated to dryness. The residue was dissolved in the proper solvent
for 19F NMR.
References
1. Parker, D. Chem. Rev. 1991, 91, 1441±1457.
2. Dale, J. A.; Mosher, H. S. J. Am. Chem. Soc. 1973, 95, 512±519.
3. Trost, B. M.; Belletire, J. L.; Godleski, S.; McDougal, P. G.; Balkovec, J. M.; Baldwin, J. J.; Christy, M. E.;
Ponticello, G. S.; Varga, S. L.; Springer, J. P. J. Org. Chem. 1986, 51, 2370±2374.
4. Latypov, S. K.; Seco, J. M.; Quinoa, E.; Riguera, R. J. Org. Chem. 1996, 61, 8569±8577, and references cited
therein.
5. Latypov, S. K.; Seco, J. M.; Quinoa, E.; Riguera, R. J. Org. Chem. 1995, 60, 1538±1545, and references cited
therein.
6. Trost, B. M.; Bunt, R. C.; Pulley, S. R. J. Org. Chem. 1994, 59, 4202±4205.
7. Dale, J. A.; Dull, D. L.; Mosher, H. S. J. Org. Chem. 1969, 34, 2543±2549.
8. Sullivan, G. R.; Dale, J. A.; Mosher, H. S. J. Org. Chem. 1973, 38, 2143±2147.
9. Hoye, T. R.; Renner, M. K. J. Org. Chem. 1996, 61, 8489±8495.
10. Ohtani, I.; Kusumi T.; Kashman, Y.; Kakisawa, H. J. Am. Chem. Soc. 1991, 113, 4092±4096, and references cited
therein.
11. Kusumi, T.; Fukushima, T.; Ohtani, I.; Kakisawa, H. Tetrahedron Lett. 1991, 32, 2939±2942.
15
12. Other acids or chlorides have also been proposed.13
13. Parker, D. J. Chem. Soc., Perkin Trans. 2 1983, 83±88.
14. Heumann, A.; Faure, F. J. Org. Chem. 1993, 58, 1276±1279, and references cited therein.
15. Fukushi, Y.; Yajima, S.; Mizutani, J. Tetrahedron Lett. 1994, 599±602.
16. Maddaluno, J.; Corruble, A.; Leroux, V.; Ple, G.; Duhamel, P. Tetrahedron: Asymmetry 1992, 3, 1239±1242.
17. Brown, H. C.; Salunkhe, A. M.; Singaram, B. J. Org. Chem. 1991, 56, 1170±1175.
18. Joshi, B. S.; Newton, M. G.; Lee, D. W.; Barber, A. D.; Pelletier, S. W. Tetrahedron: Asymmetry 1996, 7, 25±28.
19. Khan, M. A.; Tavares, D. F.; Rauk, A. Can. J. Chem. 1982, 60, 2451±2455.
20. Hamman, S.; Barrelle, M.; Tetaz, F.; Beguin, C. G. J. Fluorine Chem. 1987, 37, 85±93.
21. Barrelle, M.; Hamman, S. J. Chem. Res. 1990 (S), 100±101, 1990 (M), 701±709.
22. Apparu, M.; Ben Tiba, Y.; Leo, P. M.; Mathieu, J. P.; Mauclaire, L. J. Labelled Cpd. Radiopharm. 1999, 42, 1195±
1202.
23. Apparu, M.; Demenge, P.; Fagret, D.; Ghezzi, C.; Majid, S.; Mathieu, J. P.; Mauclaire, L.; Pasqualini, R.; Vidal,
M. International Pat. WO96/00717; Chem. Abstr. 1996, 124, 288981v.
24. Apparu, M.; Ben Tiba, Y.; Leo, P. M.; Dao Boulanger, T. H.; Mauclaire, L.; Mathieu, J. P.; Carbonnelle, E.;
Fagret, D. 4th International Conference on Nuclear and Radiochemistry, St Malo, Sept. 8±13, 1996.
25. The signal corresponding to the H proton in the structure RR0CHOR00 (R00=C6H5CH2) is completely absent from
the NMR spectrum of 11. On the other hand, the two doublets of the only AB system present are perfectly
separated between 3.51 and 3.75 ppm.
26. Ben Tiba, Y. Thesis, Universite Joseph Fournier, Grenoble, 1997.
27. Imperial Chemical Industries Ltd., British Patent 1,078,852, 1968; Chem. Abstr. 1966, 65, 7099f.
28. Crowther, A. F.; Howe, R.; Smith, L. H. J. Med. Chem. 1971, 14, 511±513.
29. Danilewicz, J. C.; Kemp, E. G. J. Med. Chem. 1973, 16, 168±169.
30. Korn, N.; Gibson, J. K.; Kni€en, F. J.; Lucchesi, B. R.; Ranade, V. V.; Mimnaugh, M.; Yu, T.; Counsell, R. E.
J. Pharm. Sci. 1980, 69, 1010±1013.

2898
M. Apparu et al. / Tetrahedron: Asymmetry 11 (2000) 2885±2898
31. Nelson, W. L.; Burke Jr., T. R. J. Org. Chem. 1978, 43, 3641±3645.
32. Belgian Patent 669402; Chem. Abstr. 1966, 65, 5402d.
33. Laskina, E. D.; Devitskaya, T. A.; Belov, V. N. Tr. Vses. Nauchn. Issled. Inst. Sintetich i Natural'n. Dushistykh
Veshchestv. 1963, 6, 31±37; Chem. Abstr. 1964, 61, 11919b.
34. Backvall, J. E.; Bystrom, S. E. J. Org. Chem. 1982, 47, 1126±1128.
35. Shtacher, G.; Rubinstein, R.; Somani, P. J. Med. Chem. 1978, 21, 678±683.
36. Kunz, W.; Jacobi, H.; Koch, K. German Patent 1967, 1236523; Chem. Abstr. 1967, 66, 64046k.