Reduction of an enaminone: Synthesis of the diamino alcohol core of ritonavir

Article abstract of DOI:10.1021/op9802071

The reduction of (5S)-2-amino-5-dibenzylamino-4-oxo-l,6-diphenylhex-2-ene was optimized for diastereoselectivity and overall conversion to (2S,3S,5S)-5-amino-2-dibenzylamino-3-hydroxy-l,6-diphenylhexane (2a). A two-step reduction sequence is described wherein the enamine is reduced with a borane-sulfonate derivative followed by reduction of the resulting ketone with sodium borohydride. The desired 2a was obtained with 84% diastereoselectivity and an acyclic 1,4 stereoinduction ratio of 14:1. This methodology has been used to produce multikilogram quantities of the diamino alcohol core, of Ritonavir and should be general to the synthesis of related diamino hydroxyethylene isosteres.

Full text of DOI:10.1021/op9802071

Organic Process Research & Development 1999, 3, 94−100
Reduction of an Enaminone: Synthesis of the Diamino Alcohol Core of
Ritonavir
Anthony R. Haight,*,† Timothy L. Stuk,§ Michael S. Allen,† Lakshmi Bhagavatula,‡ Michael Fitzgerald,‡
Steven M. Hannick,‡ Francis A. J. Kerdesky,‡ Jerome A. Menzia,† Shyamal I. Parekh,† Timothy A. Robbins,†
David Scarpetti, and Jien-Heh J. Tien†
Chemical Process Research, D54P, and Process Research, D45L, Abbott Laboratories, 1401 Sheridan Road,
North Chicago, Illinois 60064
Abstract:
conditions to obtain improved diastereoselectivity for the
reduction of the readily available enaminone 3.6
The reduction of (5S)-2-amino-5-dibenzylamino-4-oxo-1,6-
diphenylhex-2-ene was optimized for diastereoselectivity and
overall conversion to (2S,3S,5S)-5-amino-2-dibenzylamino-3-
hydroxy-1,6-diphenylhexane (2a). A two-step reduction se-
quence is described wherein the enamine is reduced with a
borane-sulfonate derivative followed by reduction of the result-
ing ketone with sodium borohydride. The desired 2a was
obtained with 84% diastereoselectivity and an acyclic 1,4
stereoinduction ratio of 14:1. This methodology has been used
to produce multikilogram quantities of the diamino alcohol core
of Ritonavir and should be general to the synthesis of related
diamino hydroxyethylene isosteres.
Reduction of enaminones is an important method for
obtaining 1,3-amino alcohols.1 This can be accomplished by
using hydrogenation1,2 or hydride reduction1,3 methods.
Typically the products are obtained with varying levels of
cleavage products due to the forcing conditions required.2
To develop a scalable synthesis of the HIV protease inhibitor,
Ritonavir (Norvir, 1),4 a practical method of obtaining the
diamino alcohol 2a was required. Previously, we reported
on the preparation of 2a by a sequential reduction of the
enaminone 3 using a borohydride complex with methane-
sulfonic acid (MSA) for reduction of the enamine, followed
by treatment with trifluoroacetoxyborohydride to reduce the
carbonyl.5 In this paper, we report on the optimization of
Results and Discussion
Reduction of 3 using metal hydrides in inert solvents was
found to be ineffective. However, treatment of 3 with sodium
cyanoborohydride using acetic acid as the solvent was shown
to reduce 3 to yield a mixture of 2a-d (Scheme 1) in a 1:(3)
ratio [2a:(2b + 2c + 2d)] (Table 1, entry 1).7 Manipulation
of the acid or solvent in order to improve the selectivity
offered little enhancement (entries 2-5). By preforming a
trifluoroacetoxyborohydride complex with sodium borohy-
dride and trifluoroacetic acid (TFA),8 the selectivity could
be improved to favor the desired isomer with 44% diaste-
reoselectivity (entry 7). To obtain complete reduction,
however, it was necessary to add excess TFA (entries 6 vs
7).
Mineral acids such as sulfuric acid yielded amino alcohols
2a-d as a minor product, while the major products were
the amino ketones 4a and 4b (Scheme 2).9 Weaker acids
such as acetic acid gave almost no amino alcohol, yielding
instead the intermediate ketones 4a and 4b. The unreactive
nature of the carbonyl when using sodium borohydride
reducing agents is believed to be due to derivatization of
the carbonyl as a boron enolate. This complex most likely
is hydrolyzed by the addition of excess trifluoroacetic acid
(entry 7, Table 1).
* Corresponding author. E-mail: Anthony.Haight@ln.ssw.abbott.com. Fax:
847-937-3864.
† Chemical Process Research, D-54P.
‡ Process Research Department, D-45L.
§ Current address: Parke-Davis, 188 Howard Ave., Holland, MI 49424.
Current address: Guilford Pharmaceuticals, Inc., 6611 Tributary St.,
Baltimore, MD 21224.
(1) Greenhill, J. V. Chem. Soc. ReV. 1977, 6, 277.
(2) (a) Greenhill, J. V.; Ramli, M.; Tomassini, T. J. Chem. Soc., Perkin Trans.
1 1975, 588. (b) Martin, J. C.; Barton, K. R.; Gott, P. G.; Meen, R. H. J.
Org. Chem. 1966, 31, 943. (c) Occhiato, E. G.; Guarna, A.; Spinetti, L. M.
Tetrahedron 1993, 49, 10629.
(3) (a) Nilsson, L. Acta Chem. Scand. B 1979, 33, 547. (b) Melillo, D. G.;
Cvetovich, R. J.; Ryan, K. M.; Sletzinger, M. J. Org. Chem. 1986, 51, 1498.
(c) Cimarelli, C.; Palmieri, G. J. Org. Chem. 1996, 61, 5557.
(4) Kempf, D. J.; Marsh, K. C.; Denissen, J. F.; McDonald, E.; Vasavanonda,
S.; Flentge, C. A.; Green, B. E.; Fino, L.; Park, C. H.; Kong, X.-P.;
Wideburg, N. E.; Saldivar, A.; Ruiz, L.; Kati, W. M.; Sham, H. L.; Robins,
T.; Stewart, K. D.; Hsu, A.; Plattner, J. J.; Leonard, J. M.; Norbeck, D. W.
Proc. Natl. Acad. Sci. U.S.A. 1995, 92, 2484.
Addition of excess sodium borohydride to the in-situ
prepared amino ketones had no effect on reducing the
(5) Stuk, T. L.; Haight, A. R.; Scarpetti, D.; Allen, M. S.; Menzia, J. A.; Robbins,
T. A.; Parekh, S. I.; Langridge, D. C.; Tien, J.-H. J.; Pariza, R. J.; Kerdesky,
F. A. J. J. Org. Chem. 1994, 59, 4040.
(6) Haight, A. R.; Stuk, T. L.; Menzia, J. A.; Robbins, T. A. Tetrahedron Lett.
1997, 38, 4191.
(7) The product ratio was determined by comparison to independently
synthesized isomers of 2.
(8) Gribble, G. W.; Nutaitis, C. F. Org. Prep. Proc. Int. 1985, 17, 317.
(9) The amino ketones 4a/b are prone to â-amino elimination and are, therefore,
not routinely isolated.
94
•
Vol. 3, No. 2, 1999 / Organic Process Research & Development
10.1021/op9802071 CCC: $18.00 © 1999 American Chemical Society and Royal Society of Chemistry
Published on Web 01/08/1999

Scheme 1
consumption of 3 and formation of a mixture of 4a/b and
2a-d (entry 4). Treatment of the mixture with trifluoroacetic
acid (4 equiv) and sodium tris(trifluoroacetoxy)borohydride
(4 equiv) led to 98% conversion and a significant improve-
ment in the diastereoselectivity (83% de) (entry 5).
With the trifluoroacetoxyborohydride method available
to reduce in-situ prepared ketones 4a/b, we reexamined what
factors influenced the diastereoselectivity of the enamine
reduction. First it was necessary to quantitate the amounts
of the three undesired isomers being produced in the reaction
to determine which isomers were reducing from 1,4 stereo-
induction and which isomers were resulting from 1,2
stereoinduction. An HPLC method was developed which
separated all four diastereomers; however, only the retention
time of isomer 2a had been conclusively assigned.7
Scheme 2
Authentic samples of 2a, 2b, 2c, and 2d were obtained
by reducing 3 with sodium borohydride/acetic acid to give
4a and 4b (Scheme 1). Flash chromatography yielded the
separated unstable amino ketones 4a and 4b, which were
immediately protected as the Boc-carbamates 5a and 5b.
Reduction of 5a to 6a/b and deprotection led to formation
of a mixture of isomers 2a and 2b. By comparison of the
HPLC trace of independently prepared 2a with the reduction/
deprotection products of 5a, the relative retention time of
isomer 2b could be assigned. For assignment of the retention
times for isomers 2c and 2d, the characterization of at least
one isomer was necessary. Ketone 5b was reduced with
sodium borohydride, and the resulting Boc-carbamates 6c/d
were separated by chromatography (Scheme 3).11 Debenzyl-
ation of the 6c isomer under catalytic transfer conditions gave
the carbamate 7c. Treatment of the 1,2-amino alcohol 7c with
1,1′-carbonyldiimidazole resulted in formation of the ox-
azolidinone 8. With the aid of COSY at 500 MHz, the H4
and H5 protons were assigned at 4.16 and 3.66 ppm,
respectively (Scheme 3). The coupling constant between the
ring protons of 8 (J4-5 ) 4.7 Hz) was suggestive of a trans
arrangement.12 In the ROESY spectrum, an NOE was
observed between the H4 proton and the two methylene
protons at 2.78 and 2.70 ppm confirming that these groups
are cis to each other. Based on this, the stereochemistry of
6c, and also 2c, can be assigned as the 2S,3S,5R configu-
ration. This leaves the 2d isomer as the 2S,3R,5R isomer.
With the assignment of the isomers complete, a screening
of borane derivatives suggested that a higher diastereose-
lectivity for the enamine reduction was possible using a
borane generated from a sulfonic acid.10 Treatment of a
complex of sodium borohydride/sulfuric acid in THF fol-
lowed by sodium bis(trifluoroacetoxy)borohydride yielded
the desired 2a with 73% overall diastereoselectivity (Table
3, entry 1). Unfortunately, stirring problems developed when
preparing the borohydride/sulfuric acid mixture. The reaction
became extremely thick as the sulfuric acid was slowly added
to the sodium borohydride in THF. This is presumed to be
Table 1. Reduction of 3 with borohydrides
hydride
(equiv)
acid
(equiv)
condi- conv 2a:(2b +
tionsa (%)b 2c + 2d)c
entry
1
2
3
4
5
6
7
NaCNBH3 (3) AcOH (2)
NaCNBH3 (4) TFA (4)
NaCNBH3 (4) TFA (4)
NaCNBH3 (4) TFA (4)
NaCNBH3 (4) TFA (4)
A
>90
>90
>90
>90
>90
77
1.0:(3)
1.7:(1)
1.4:(1)
1.6:(1)
1.3:(1)
1.0:(1)
2.6:(1)
A
Ad
Ae
Af
B
NaBH4 (4)
NaBH4 (4)
TFA (8)
TFA (14)
B
>98
a Unless otherwise indicated, all reactions were conducted under a nitrogen
atmosphere using approximately 0.25-1.0 M THF solutions, between -10 and
10 °C. Method A: acid added to hydride/3 in THF. Method B: 3 added to
hydride complex. b Conversion calculated as HPLC peak area percent conversion.
c Determined by HPLC analysis. d Hexane used as solvent. e Toluene used as
solvent. f Methanol used as solvent.
Table 2. Stepwise Reduction of 3 with borohydride reagentsa
second
hydride
(equiv)
NaBH4
acid
additive
conv 2a:(2b +
(%)b 2c + 2d)c
entry (equiv) (equiv) (equiv)
1
2
3
4
5
4
4
4
4
4
TFA (8) none
TFA (8) TFA (4) none
TFA (8) TFA (4) NaBH4 (1)
H2SO4(4) none
NaBH4 (2)
77 1:(1)
95 3:(1)
>98 1.8:(1)
none
33 - -
H2SO4(5) TFA (4) NaBH(TFA)3 (4) 98 10.5:(1)
a Reactions were conducted under nitrogen by reacting 3 with borohydride
complex prepared in 0.15-0.25 M THF with NaBH4 and acid at -20 to 10 °C.
Stirred 10-18 h at ambient temperature, cooled to 0-10 °C, and additive added
followed by second borohydride. b Conversion calculated as HPLC peak area
percent conversion. c Determined by HPLC analysis.
carbonyls in the reaction mixture (Table 2, entry 1), whereas
addition of trifluoroacetic acid did allow for further reduction
to take place (entry 2). A combination of both trifluoroacetic
acid and sodium borohydride led to virtually complete
conversions to the 2a-d (entry 3).
(11) Due to the instability of 4a/b, a more practical method of obtaining larger
quantities of 6a-d was to reduce 3 to 4a/b and protect the crude reaction
mixture with Boc2O. Reduction of this mixture led to all four isomers 6a-
d, which could be separated by chromatography.
(12) Kempf, D. J.; Sowin, T. J.; Doherty, E. M.; Hannick, S. M.; Codavoci, L.;
Henr;y, R. F.; Green, B. E.; Spanton, S. G.; Norbeck, D. W. J. Org. Chem.
1992, 57, 5692 and references therein.
Formation of the borane complex from 4 equiv of sodium
borohydride with 4 equiv of sulfuric acid10 led to complete
(10) Abiko, A.; Masamune, S. Tetrahedron Lett. 1992, 33, 5517.
Vol. 3, No. 2, 1999 / Organic Process Research & Development
•
95

Scheme 3
ity is generally quite high (15-20:1), while the 3R,5R,3S,5R
isomers selectivity is typically low (1:1 to 4:1). the preference
for reduction of 4a presumably is due to a matched pairing,
while isomer 4b has a mismatched pairing. 1,2 Stereoinduc-
tion in both 4a and 4b favors the S configuration in the
resulting amino alcohols, in agreement with a Felkin-Ahn
type model.13 However, carbonyl reduction of the assumed
cyclic structures 9a and 9b (Figure 1) would come from the
pseudoaxial direction favoring the S isomer from 4a and the
R isomer from 4b (Figure 1).14
With the selectivity of the enamine reduction improved,
other methods of decomplexing the boronate ester were
studied. To compete with the bidentate chelation of 4a/b, a
bi- or tridentate chelating agent seemed necessary. Treatment
with ethanolamine (4 equiv) followed by sodium borohydride
in dimethylformamide gave >98% conversion to products
with 75% de (entry 5).15 Use of the tetradentate chelating
agent triethanolamine (TEOA)16 followed by sodium boro-
hydride in dimethylformamide yielded the desired product
with 80% diastereoselectivity (entry 6). While dimethylform-
amide worked well on a laboratory scale, the danger of using
this as a solvent for sodium borohydride on a large scale17
led us to replace this solvent with dimethylacetamide. Using
these conditions on a multikilogram scale yielded the desired
product in >98% conversion with 84% diastereoselectivity
(entry 7).
Debenzylation of the dibenzyldiamino alcohols 2a-d was
then carried out using hydrogen-transfer conditions. Precipi-
tation as the dihydrochloride salt yielded the desired hy-
droxyethylene isostere 10 in 60% overall yield from 3 in
>95% de (Scheme 4).
the result of ring opening of the solvent to give both ring-
opened products and polymers. Addition of water (7 equiv)
to the borohydride complex appeared to slow solvent
decomposition while also thinning the reaction mixture.
Unexpectedly, addition of water also increases the 1,4
stereoinduction and the overall diastereoselectivity of the
reaction to 92% and 91%, respectively, with 95% conversion
(entry 2). Replacement of sulfuric acid with MSA gave 78%
de for the 1,4 induction and 74% de overall with >98%
conversion (entry 3). On a laboratory scale, no problems were
observed with the MSA in THF. However, when these
reaction conditions (entry 3) were scaled up to 20 kg of 3,
problems with THF ring-opening byproducts were again
observed. In order to avoid this, several other ether solvents
were evaluated in the reaction. Dimethoxyethane (DME)
proved superior as a solvent relative to THF. Under the acidic
reaction conditions, DME did not appear to have a problem
with polymerization. Adding water to the DME borohydride
complex led to poor conversion to product; however,
2-propanol as a protic additive was found to improve the
diastereoselectivity (entry 4). The exact role of the protic
solvents in enhancing the diastereoselectivity is not known.
To obtain reproducible data regarding the diastereoselec-
tivity of the reaction, the reductions needed to be carried
out to completion. This was necessary because the stereo-
selectivity of the products decreases as the reaction progresses.
This correlation between conversion and selectivity suggests
that a kinetic preference is observed in the reduction of 4a/
b. Further evidence of this can be seen in the ratios of
diastereomers produced (Tables 2 and 3). The 2a:2b selectiv-
Conclusion
This methodology demonstrates the ability to synthesize
the diaminohydroxyethylene isostere 10 from l-phenylalanine
in high overall yield, involving only two isolated solids, and
using no chromatography. A method has been devised to
reduce 3 to the amino alcohols 2a-d with high diastereo-
selectivity. An acyclic 1,4 stereoinduction was optimized to
an impressive 14:1 selectivity by tuning the solvent and
acid.18 This method should prove general for the reduction
of enaminones in a syn-1,3 fashion to yield many types of
1,3-amino alcohols.
Experimental Section
General. (5S)-2-Amino-5-dibenzylamino-4-oxo-1,6-diphen-
ylhex-2-ene (3) was prepared as previously reported.6 Proton
(13) Reetz, M. T. Angew. Chem. 1991, 30, 1531.
(14) The attack of the hydride from the upper face leads to a chairlike transition
state. See: Deslongchamps, P. Stereoelectronic Effects in Organic Chem-
istry; Pergamon Press: Oxford, 1983; Chapter 6.
(15) NaBH4 was added as a solution in DMF to avoid adding as a solid. Solubility
data for NaBH4: 18.0 g/100 g of dimethylformamide or 14.0 g/100 g of
dimethylacetamide.
(16) Triethanolamine can sequester the boron completely as the triethanolamine
borate even in water. See: Sonoda, A.; Takagi, N.; Ooi, K.; Hirotsu, T.
Bull. Chem. Soc. Jpn. 1998, 71, 161.
(17) Liu, Y.; Schwartz, J. J. Org. Chem. 1993, 58, 5005. Ganem, B.; Osby, J.
O. Chem. ReV. 1986, 86, 763.
(18) An example of 1,4 induction in a similar system was recently reported.
See: Captain, L. F.; Xia, X.; Liotta, D. C. Tetrahedron Lett. 1996, 37, 4293.
96
•
Vol. 3, No. 2, 1999 / Organic Process Research & Development

Table 3. Stepwise Reduction of 3 with Borane Sulfonate Complexesa
NaBH4
(equiv)
acid
(equiv)
additive
(equiv)
second reduction
(equiv)
conv
(%)b
2a-d
entry
1,4 drc
drc
1d
2e
3e
4
5
6
4
5
4
4
4
4
2.3
H2SO4 (4)
H2SO4 (5)
MSA (10)
MSA (10)
MSA (10)
MSA (10)
MSA (5)
TFA (6)
none
none
none
HO(CH2)2NH2 (4)
TEOA (4)
TEOA (2.2)
NaBH2(TFA)2 (4)
>98
95
nd
25:1
7:1
16:1
12:1
11:1
14:1
7:1
21:1
7:1f
14:1
7:1
8:1
12:1
NaBH3(TFA) (4), TFA (2)
NaBH3(TFA) (4), TFA(1)
NaBH3(TFA) (4), TFA (1)
NaBH4 (1), DMF
NaBH4 (4), DMF
NaBH4 (1.5), DMAC
>98
>98
>98
>98
>98
7
a Reactions were conducted under nitrogen by reacting 3 with NaBH4 precomplexed with RSO3H in DME between -10 and 0 °C. Solution of 3 added at <10 °C
in iPrOH/DME. Warmed to ambient temperature for 3-24 h. Additive added followed by borohydride at <10 °C. b Conversion calculated as HPLC peak area percent
ratio of 2a-d/2a-d + 3 + 4a/b. c Determined by HPLC analysis of crude reaction mixture. d Reaction performed in THF. e Reaction performed in THF with 7.7
equiv of H2O added to borohydride complex prior to addition of 3. f Measured after debenzylation.
were concentrated in Vacuo. The residue was diluted with
75 kg of EtOH and concentrated in Vacuo again to remove
other solvents. This residue was then crystallized from 150
kg of EtOH to give 33.85 kg (79%) of 3 (as a 1:1 EtOH
solvate): mp 101-102 °C; IR (CDCl3) 3630, 3500, 3110,
3060, 3030, 1620, 1595, 1520, 1495, 1450, 1300, 1250 cm-1;
1H NMR (CDCl3) δ 9.8 (br s, 1H), 7.2 (m, 20H), 5.1 (s,
1H), 4.9 (br s, 1H), 3.8 (d, 2H, J ) 14.7 Hz), 3.6 (d, 2H, J
) 14.7 Hz), 3.5 (m, 3H), 3.2 (dd, 1H, J ) 14.4, 7.5 Hz), 3.0
(dd, 1H, J ) 14.4, 6.6 Hz); 13C NMR (DEPT) (CDCl3) δ
198.0, 162.8, 140.2, 140.1, 136.0(+), 129.5(+), 129.3(+),
128.9(+), 128.7(+), 128.1(+), 128.0(+), 127.3(+), 126.7-
(+), 125.6(+), 96.9(+), 66.5(+), 54.3(-), 42.3(-),
32.4(-); MS (CI) m/z (relative intensity) 461 ([M + 1]+,
100).
Figure 1.
Scheme 4
Determination of the Enantiomeric Purity of 3. The
compound was analyzed by HPLC (Chiracel OD column
(Diacel Chemical), 20% 2-propanol/heptane, 210 nm). The
retention times at 1 mL/min are 10 and 13.5 min for the R
and S enantiomers, respectively. The ee of 3 is >99%.
General Procedure for Enaminone Reduction. A
suspension of NaBH4 (30 kg, 790 mol) in ethylene glycol
dimethyl ether (1176 kg) was cooled to less than -5 °C.
Methanesulfonic acid (192 kg, 2 kmol) in ethylene glycol
dimethyl ether (10 kg) was slowly added, keeping the
temperature below 5 °C. Caution: Borane is formed during
this addition. Proper Ventilation of the Vessel is pertinent!
Once the addition was complete, a solution of i-PrOH (140
L), 36 (122.5 kg, 266 mol), and ethylene glycol dimethyl
ether (270 kg) was slowly added. The mixture was stirred
for 12 h at 5 ( 5 °C. Triethanolamine (118 kg) was then
slowly added to the reaction while maintaining the temper-
ature below 5 °C. The solution was stirred at less than 5 °C
for 30 min. A solution of NaBH4 (25 kg, 660 mol) in
dimethylacetamide (184 kg) was slowly added. The resulting
suspension was stirred for 2 h at 15 ( 5 °C and then slowly
quenched with water (1355 L). The temperature was warmed
to ambient, and methyl tert-butyl ether 815 kg) was added.
The layers were separated and the organics washed succes-
sively with 1 N NaOH (1520 L), 18% (w/w) NH4Cl (1476
kg), and 7% (w/w) NaCl (1569 kg). The resulting organic
solution could then be stored or carried on in further
transformations. HPLC ratios of diastereomers 2a-d were
determined by analytical HPLC (475:525 [0.03]KH2PO4/CH3-
spectra were obtained at 300 and 500 MHz, and 13C NMR
spectra were obtained at 75 MHz.
(5S)-2-Amino-5-dibenzylamino-4-oxo-1,6-diphenylhex-
2-ene (3). Enaminone 3 was prepared as previously de-
scribed.6 To a vessel were added 14.0 kg (84.7 mol) of
l-phenylalanine, 27.0 kg (195 mol) of K2CO3, 4.8 kg (85.5
mol) of KOH, and 62.4 kg of tap water. This was stirred
until the solution became homogeneous before adding 35.8
kg (284 mol) of benzyl chloride. The solution was heated to
reflux for 5 h and cooled to 50 °C, and 55 kg of heptane
was added. The aqueous layer was drained, and the organics
were washed twice with 47 kg of H2O/MeOH (1:2 w/w).
The organics were concentrated in Vacuo to give an oil.
The crude l-N,N-dibenzylphenylalanine benzyl ester was
dissolved into 23 kg of methyl tert-butyl ether (MTBE) and
3.8 kg (92.6 mol) of CH3CN. This was then slowly added
to a slurry of 90% NaNH2 (8.2 kg, 189 mol) in 58 kg of
MTBE while keeping the temperature between -5 and 5
°C. The reaction was stirred for 2 h at 0-5 °C, and then
vacuum was applied for 30 min to remove the ammonia.
The solution was reconstituted with 17 kg of MTBE and
warmed to 25 °C, and a 2 M solution of benzylmagnesium
chloride in THF (72.4 kg, 140 mol) was slowly added. This
was stirred for 2 h. The reaction was cooled to 5 °C and
quenched by slow addition of 136 kg of 23% (w/w) aqueous
citric acid. The layers were separated and the organics
washed with 106 kg of 10% aqueous NaCl. The organics
Vol. 3, No. 2, 1999 / Organic Process Research & Development
•
97

CN, pH ) 3.0, C-8 5-µm Hypersil column, λ ) 205 nm, 1
mL/min and compared to standards; the mixture analyzed
as follows: 2S,3S,5R isomer (9.2 min, 4%), 2S,3R,5S (10.5
min, 6%), 2S,3S,5S (11.3 min, 83%), and 2S,3R,5R (14.3
min, 2%). An analytical sample of the 2S,3S,5S isomer 2a
was prepared by flash chromatography [silica; 9:1:0.1
hexanes/i-PrOH/NH4OH(aqueous)] to afford 2a as a clear
oil: IR (CDCl3) 3400-2800 (br), 3090, 3060, 3030, 1600,
was then diluted with CH2Cl2 (50 mL), water (150 mL), and
50% NaOH (100 mL). The separated aqueous layer was
extracted twice with CH2Cl2 (50 mL). The organics were
combined, dried over Na2SO4, and concentrated in Vacuo to
yield 10.63 g of a yellow syrup. The residue was subjected
to column chromatography (silica; 1% MeOH/CH2Cl2) to
give 4a (420 mg, 4% yield) and 4b (220 mg, 2% yield) as
unstable oils (ca. 95% diastereomeric purity). These were
carried on immediately. Extensive decomposition due to
elimination prevented complete characterization of these
intermediates.
4a: 1H NMR (CDCl3) δ 7.40-7.00 (m, 20H), 3.82 (d,
2H, J ) 13.6 Hz), 3.59 (d, 2H, J ) 13.6 Hz), 3.55-3.35
(m, 1H), 3.25-3.15 (m, 1H), 3.15 (dd, 1H, J ) 13.5, 9.6
Hz), 2.92 (dd, 1H, J ) 13.5, 4.5 Hz), 2.65-2.30 (m, 4H),
1.55 (br s, 2H); MS (DCI/NH3) m/z (relative intensity) 463
([M + H]+, 20), 446 ([M - NH3]+, 15), 198 ([Bn2NH2]+,
100).
1
1580, 1490, 1450, 1370, 1300, 1100, 1070, 1030 cm-1; H
NMR (CDCl3) δ 7.2 (m, 20H), 4.15 (d, 2H, J ) 14 Hz),
3.65 (ddd, 1H, J ) 10, 5, 2 Hz), 3.5 (d, 2H, J ) 14 Hz),
3.50-2.49 (m, 8H) 2.48 (dd, 1H, J ) 14, 7 Hz), 1.60 (dt,
1H, J ) 14, 10 Hz), 1.25 (dt, 1H, J ) 14, 2 Hz); 13C NMR
(DEPT) (CDCl3) δ 140.9, 140.2, 138.5, 129.4(+), 129.4-
(+), 128.9(+), 128.4(+), 128.3(+), 128.2(+), 126.8(+),
126.3(+), 125.7(+), 72.1(+), 63.7(+), 55.0(-), 53.3(+),
46.4(-), 40.3(-), 30.3(-); MS (CI) m/z (relative intensity)
465 ([M + 1]+, 100).
Flash chromatography also yielded samples of the other
isomers, 2b-d.
4b: MS (DCI/NH3) m/z (relative intensity) 446 ([M -
NH3]+, 85), 198 ([Bn2NH2]+, 100).
(2S,3R,5S)-5-Amino-2-dibenzylamino-3-hydroxy-1,6-
diphenylhexane (2b): H NMR (CDCl3) δ 7.4-7.0 (m,
(2S,5S)-5-(tert-Butyloxycarbamoyl)amino-2-dibenzyl-
amino-3-oxo-1,6-diphenylhexane (5a). A solution of amine
4a (420 mg, 0.90 mmol) in THF (20 mL) was treated with
di-tert-butyl dicarbonate (239 mg, 1.1 mmol) at rt for 1 h.
The reaction mixture was diluted with ethyl acetate (ca. 50
mL), washed with 1 N NaOH (ca. 50 mL), dried over
MgSO4, and concentrated in Vacuo. The residue was
subjected to column chromatography (silica; 15:1 hexanes/
1
20H), 4.25-4.18 (m, 1H), 3.8-3.5 (m, 6H), 3.28-3.15 (m,
1H), 3.11-2.95 (m, 1H), 2.95-2.82 (m, 1H), 2.75 (dd, 1H,
J ) 13.0, 6.5 Hz), 2.70-2.40 (m, 3H), 1.75-1.55 (m, 2H);
13C NMR (DEPT) (CDCl3) δ 142.0, 140.1, 138.7, 129.6-
(+), 129.3(+), 128.9(+), 128.6(+), 128.1(+), 126.8(+),
126.5(+), 125.6(+), 69.5(+), 67.1, 63.9(+), 54.7(-), 50.5-
(+), 43.6(-), 38.7(-), 32.4(-); MS (CI) m/z (relative
intensity) 465 ([M + 1]+, 100).
1
EtOAc) to give 270 mg (0.48 mmol, 53%) of 5a: H NMR
(CDCl3) δ 7.35-6.95 (m, 20H), 4.72-4.65 (m, 1H), 4.08-
3.95 (m, 1H), 3.77 (d, 2H, J ) 13.6 Hz), 3.58 (d, 2H, J )
13.6 Hz), 3.51 (m, 1H), 3.12 (dd, 1H, J ) 13.4, 9.0 Hz),
2.91 (dd, 1H, J ) 13.6, 4.6 Hz), 2.82 (dd, 1H, J ) 18.0, 6.0
Hz), 2.75-2.70 (m, 2H), 2.33 (dd, 1H, J ) 18.0, 5.4 Hz),
1.34 (s, 9H); MS (DCI/NH3) m/z (relative intensity) 563 ([M
+ 1]+, 100).
(2S,3S,5R)-5-Amino-2-dibenzylamino-3-hydroxy-1,6-
1
diphenylhexane (2c): H NMR (CDCl3) δ 7.4-6.9 (m,
20H), 3.97 (d, 2H, J ) 13.5 Hz), 3.85 (ddd, 1H, J ) 8.3,
8.3, 3 Hz), 3.40 (d, 2H, J ) 13.5 Hz), 3.25-3.15 (m, 1H),
3.08 (dd, 1H, J ) 13.5, 6.0 Hz), 2.88-2.75 (m, 1H), 2.70-
2.45 (m, 6H), 1.45-1.25 (m, 2H); 13C NMR (DEPT) (CDCl3)
δ 140.5, 139.1, 129.3(+), 129.0(+), 128.5(+), 128.4(+),
127.2(+), 126.2(+), 68.1(+), 64.2(+), 54.2(-), 49.6(+),
44.3(+), 40.5(-), 32.1(-); MS (CI) m/z (relative intensity)
465 ([M + 1]+, 100).
(2S,5R)-5-(tert-Butyloxycarbamoyl)amino-2-dibenzyl-
amino-3-oxo-1,6-diphenylhexane (5b). A solution of amine
4b (220 mg, 0.47 mmol) in THF (20 mL) was treated with
di-tert-butyl dicarbonate (125 mg, 0.57 mmol) at rt for 3 h.
The reaction mixture was diluted with EtOAc (ca. 50 mL),
washed with 1 N NaOH (ca. 50 mL), dried over MgSO4,
and concentrated in Vacuo to give 190 mg (0.34 mmol, 72%)
(2S,3R,5R)-5-Amino-2-dibenzylamino-3-hydroxy-1,6-
1
diphenylhexane (2d): H NMR (CDCl3) δ 7.4-6.9 (m,
20H), 4.05 (ddd, 1H, J ) 12.0, 6.0, 1.2 Hz), 3.8-3.5 (m,
5H), 3.15-3.05 (m, 1H), 3.00 (d, 2H, J ) 7.5 Hz), 3.0-2.6
(m, 4H), 2.58 (dd, 1H, J ) 14.4, 8.1 Hz), 1.90 (d, 1H, J )
14.7 Hz), 1.1-0.9 (m, 1H); 13C NMR (DEPT) (CDCl3) δ
141.9, 140.2, 138.2, 129.7(+), 129.3 (+), 128.8(+), 128.6-
(+), 128.1(+),128.0(+), 126.6(+), 125.5(+), 72.7(+), 64.3-
(+), 54.7(-), 54.5(+), 47.3(-), 39.9(-), 32.2(-); MS (CI)
m/z (relative intensity) 465 ([M + 1]+, 100).
(2S,5S/R)-5-Amino-2-dibenzylamino-3-oxo-1,6-diphen-
ylhexane (4a/b). To a THF (62.5 mL) solution of NaBH4
(3.55 g, 94 mmol) at -10 °C was added a precooled solution
of acetic acid (62.5 mL, 1.09 mol) over 10 min. This was
stirred at rt for 30 min. After addition of 3 (10 g, 21.7 mmol)
in two portions over 1 h, the reaction mixture was stirred
for 14 h. The reaction was quenched with concentrated HCl
(11.5 mL) and allowed to stir at rt for 30 min. The reaction
1
of 5b: H NMR (CDCl3) δ 7.35-7.05 (m, 18H), 6.82-6.72
(m, 2H), 5.20-5.10 (m, 1H), 4.05-3.90 (m, 1H), 3.77 (d,
2H, J ) 13.6 Hz), 3.54 (d, 2H, J ) 13.6 Hz), 3.41 (dd, 1H,
J ) 9.6, 3.7 Hz), 3.14 (dd, 1H, J ) 13.2, 9.9 Hz), 2.89 (dd,
1H, J ) 12.9, 3.3 Hz), 2.79 (dd, 1H, J ) 17.6, 4.8 Hz),
2.75-2.70 (m, 1H), 2.50 (dd, 1H, J ) 13.2, 8.5 Hz), 2.25
(dd, 1H, J ) 17.6, 5.1 Hz), 1.38 (s, 9H); MS (DCI/NH3)
m/z (relative intensity) 563 ([M + 1]+, 100).
(2S,3S/R,5R)-5-(tert-Butyloxycarbamoyl)amino-2-diben-
zylamino-3-hydroxy-1,6-diphenylhexane (6c/d). To a THF
(5 mL) and MeOH (1 mL) solution of 5b (190 mg, 0.34
mmol) was added NaBH4 (20 mg, 0.53 mmol) at 20-25
°C. The reaction mixture was stirred at rt for 3 h. Additional
NaBH4 (8 mg, 0.21 mmol) was added and the reaction
mixture stirred at rt for 30 min. The reaction was diluted
98
•
Vol. 3, No. 2, 1999 / Organic Process Research & Development

with CH2Cl2 (ca. 20 mL), washed with water and then brine,
and dried over MgSO4. The solvent was concentrated in
Vacuo to yield an oil (160 mg).
140.8, 139.9, 138.2, 129.4(+), 129.3(+), 129.0(+), 129.0-
(+), 128.4(+), 128.3(+), 126.9(+), 126.4(+), 125.8(+),
79.3, 67.4(+), 63.9(+), 54.9(-), 49.0(+), 41.2(-),
38.8(-), 30.6(-), 28.4(+); MS (CI m/z (relative intensity)
565 ([M + 1]+, 100).
Alternative Preparation of 6a-d. To a THF (100 mL)
solution of NaBH4 (7.10 g, 188 mmol) at -10 °C was added
a precooled solution of acetic acid (125 mL, 2.2 mol) over
10 min. This was stirred at rt for 15 min. After addition of
3 (20 g, 43.5 mmol), the reaction mixture was stirred for 3
h at rt. The reaction was quenched with concentrated HCl
(23 mL) and allowed to stir at rt for 15 min. The reaction
was then diluted with water (100 mL) and 50% NaOH (110
mL). The layers were separated, and the aqueous layer was
extracted twice with EtOAc (100 mL). The organics were
combined, dried over Na2SO4, and concentrated in Vacuo to
yield a yellow syrup. The residue was subjected to column
chromatography (silica; 1% CH3OH/CH2Cl2) to give a
mixture of 4a and 4b as well as traces of the alcohols 2a-d
(15.71 g). The residue was taken up in THF (250 mL) and
treated with di-tert-butyl dicarbonate (8.91 g, 41 mmol) at
rt for 4 h. The reaction mixture was diluted with EtOAc (100
mL) and concentrated in Vacuo. The residue was subjected
to column chromatography (silica; 16:1 hexanes/EtOAc) to
give 14.0 g (24.9 mmol, 57%) of 5a/b. The mixture was
diluted with THF (250 mL) and MeOH (50 mL) and treated
with NaBH4 (1.14 g, 30 mmol). The reaction was stirred for
30 min at rt, concentrated in Vacuo, and diluted with EtOAc
(ca. 100 mL). This was washed successively with water and
brine and dried over MgSO4. Concentration in Vacuo gave
13.9 g (25 mmol) of 6a-d. The isomers were separated by
eluting through a silica column with nitrogen pressure using
a gradient of MTBE/hexane (5:95 to 50:50). The isomers
eluted in the order 6c (4.59 g), 6b (0.96 g), 6a (6.80 g), 6d
(0.77 g). Mixed fractions were discarded.
(6d): 1H NMR (CDCl3) δ 7.35-7.10 (m, 18H), 7.05-
6.90 (m, 2H), 4.95-4.85 (m, 1H), 3.75-3.65 (m, 4H), 3.55
(d, 2H, J ) 15.0 Hz), 3.10-2.85 (m, 3H), 2.75-2.55 (m,
3H), 1.95 (ddd, 1H, J ) 13.5, 2.2, 2.2 Hz), 1.40 (s, 9H),
1.40-1.25 (m, 1H); 13C NMR (DEPT) (CDCl3) δ 155.8,
140.3, 139.5, 138.3, 129.5(+), 129.2(+), 128.9(+), 128.9-
(+), 128.4(+), 128.3(+), 127.0(+), 126.4(+), 126.0(+),
79.4, 70.7(+), 63.7(+), 55.0(-), 52.2(+), 41.7(-),
38.1(-), 31.8(-), 28.4(+); MS (CI) m/z (relative intensity)
565 ([M + 1]+, 100).
(2S,3S,5R)-2-Amino-5-(tert-butyloxycarbamoyl)amino-
3-hydroxy-1,6-diphenylhexane (7c). An ethanol (2.3 mL)
solution of 6c (140.2 mg, 0.248 mmol) was slurried with
10% palladium on carbon (24 mg). To this mixture was
added NH4CO2H (80.1 mg) in water (280 µL). This was then
heated to reflux for 3 h, cooled to rt for 12 h, and filtered
through a pad of Celite. The filtrate was concentrated in
Vacuo, diluted with EtOAc (20 mL), and washed successively
with water and brine. The organics were dried over MgSO4
and concentrated to dryness to give 7c as a white solid (87.2
1
mg, 0.227 mmol, 92%): H NMR (DMSO-d6) δ 7.30-7.10
(m, 10H), 6.62 (d, 1H, J ) 9 Hz), 4.39 (m, 1H), 3.75 (m,
1H), 3.40-3.30 (m, 2H), 2.75-2.25 (m, 5H), 1.50 (m, 3H),
1.30 (s, 9H); 13C NMR (DMSO-d6) δ 155.3, 140.3, 139.2,
129.1, 129.1, 128.0, 127.9, 125.7, 125.5, 77.3, 69.2, 57.2,
49.1, 41.3, 40.6, 38.2, 28.2; MS (DCI/NH3) m/z (relative
intensity) 385 ([M + 1]+, 100).
(4S,5S,2′R)-4-Benzyl-5-[(2R)-2-(amino-tert-butyloxy-
carbamyl)-3-phenylpropyl]oxazolidinone (8). A THF (10
mL) solution of 7c (100.3 mg, 0.26 mmol) was treated at rt
with CDI (55.0 mg, 0.34 mmol) and DMAP (6.6 mg, 0.05
mmol). The reaction mixture was stirred at rt for 3 h. The
reaction was diluted with EtOAc (30 mL) and washed
successively with 1 M HCl and brine. The organics were
dried over MgSO4 and concentrated in Vacuo to a white
powder (100.1 mg). Purification by preparative TLC (10%
1
(6a): H NMR (CDCl3) δ 7.35-7.00 (m, 20H), 4.82 (br
s, 1H), 4.32 (s, 1H), 3.90 (d, 2H, J ) 13.5 Hz), 3.79 (apparent
dq, 1H, J ) 6.8, 6.8 Hz), 3.65-3.55 (m, 1H), 3.38 (d, 2H,
J ) 13.5 Hz), 3.05 (dd, 1H, J ) 15.0, 6.0 Hz), 2.85-2.72
(m, 2H), 2.71-2.55 (m, 2H), 1.55-1.45 (m, 1H), 1.39 (s,
9H), 1.30-1.15 (m, 1H); 13C NMR (DEPT) (CDCl3) δ 155.6,
140.2, 138.9, 138.3, 129.6(+), 129.1(+), 129.0(+), 128.6-
(+), 128.5(+), 128.2(+), 127.2(+), 126.2(+), 126.1(+),
78.9, 69.4(+), 64.3(+), 54.0(-), 51.3(+), 41.2(-),
38.3(-), 32.0(-), 28.4(+); MS (CI) m/z (relative intensity)
565 ([M + 1]+, 100).
1
CH3OH/CH2Cl2) gave 72.2 mg (67%) 8: H NMR (DMSO-
d6) δ 7.66 (s, 1H), 7.30-7.11 (m, 10H), 6.71 (d, 1H, J )
7.5 Hz), 4.16 (ddd, 1H, J ) 9.5, 4.7, 4.7 Hz), 3.75 (m, 1H),
3.66 (m, 1H), 2.78 (dd, 1H, J ) 12.5, 5.0 Hz), 2.70 (dd, 1H,
J ) 12.5, 5.0 Hz), 2.70-2.60 (m, 2H), 1.70-1.55 (m, 2H),
1.33 (s, 9H); 13C NMR (DEPT) (DMSO-d6) δ 157.6, 155.0,
138.5, 136.4, 129.5(+), 129.1(+), 128.2(+), 128.0 (+),
126.4(+), 125.9(+), 77.4, 77.0(+), 57.7(+), 48.6(+), 40.9-
(-), 39.5(-), 38.7(-), 28.2(+); HRMS m/z [M + K]+ calcd
for C24H30N2O4K1, 449.1843, found 449.1840.
(2S,3S,5S)-2,5-Diamino-3-hydroxy-1,6-diphenylhex-
ane Dihydrochloride (10). A solution of crude 5-amino-2-
(dibenzylamino)-3-hydroxy-1,6-diphenylhexanes, 2a-d (20
kg, 43 mol), MeOH (320 L), aqueous ammonium formate
(13.6 kg in 23 L of water), and 5% palladium on carbon
(4.0 kg, 50-60% water by weight) was heated to reflux for
6 h. The cooled suspension was filtered through a bed of
1
(6b): H NMR (CDCl3) δ 7.35-7.05 (m, 20H), 4.39 (d,
1H, J ) 9.6 Hz), 4.15-4.00 (m, 1H), 3.85-3.75 (m, 2H),
3.68-3.58 (m, 4H), 3.08-2.95 (m, 1H), 2.95-2.85 (m, 2H),
2.75 (d, 2H, J ) 6.3 Hz), 1.55-1.30 (m, 2H), 1.42 (s, 9H);
13C NMR (DEPT) (CDCl3) δ 157.0, 141.5, 140.1, 137.7,
129.6(+), 129.3(+), 128.8(+), 128.5(+), 128.1(+), 128.1-
(+), 126.7(+), 126.5(+), 125.7(+), 79.8, 67.8(+), 63.7(+),
54.9(-), 48.1(+), 41.5(-), 41.2(-), 32.5(-), 28.4(+); MS
(CI) m/z (relative intensity) 565 ([M + 1]+, 100).
1
(6c): H NMR (CDCl3) δ 7.35-7.05 (m, 20H), 4.70 (d,
1H, J ) 9.0 Hz), 4.25 (s, 1H), 4.10-3.95 (m, 3H), 3.57 (br
s, 1H), 3.40 (d, 2H, J ) 15 Hz), 3.05-2.75 (m, 3H), 2.55
(dd, 2H, J ) 15, 9 Hz), 1.85-1.70 (m, 1H), 1.35 (s, 9H),
0.95-0.85 (m, 1H); 13C NMR (DEPT) (CDCl3) δ 156.2,
Vol. 3, No. 2, 1999 / Organic Process Research & Development
•
99

diatomaceous earth, and the cake was washed with MeOH
(2 × 30 kg). The filtrtae was concentrated in Vacuo to an
oil. The residue was dissolved in EtOAc (175 L) and washed
successively with 1 N NaOH (200 L), 20% brine (195 L),
and water (97 L). The organics were concentrated in Vacuo.
To the oil was added i-PrOH (205 L) and concentrated HCl
(aqueous, 17 L). The suspension was heated to reflux for 1
h, cooled over 6 h to 25 °C, and held at that temperature for
12 h. The slurry was filtered, and the cake was washed with
EtOAc (30 L). The product could be recrystallized if greater
purity was desired. The solids were resuspended in i-PrOH
(120 L) and water (6.3 L). This slurry was heated at reflux
for 1 h, cooled to 25 °C, and stirred for 12 h. The slurry
was filtered, and the cake was washed with i-PrOH (15 kg).
This afforded 8.9 kg (60% from 3) (>99% HPLC purity) of
a white powder: Rf 0.25 (34:30:25:8:4 CHCl3/EtOAc/CH3-
(ddd, 1H, J ) 14.5, 3.0, 3.0 Hz); 13C NMR (DEPT) (CD3-
OD) δ 136.8, 136.6, 130.5(+), 134.4(+), 130.2(+), 128.6-
(+), 128.5(+), 68.9(+), 58.5(+), 53.2(+), 39.9(-),
36.9(-), 36.7(-).
The data for the free base are as follows: 1H NMR
(CDCl3) δ 7.3 (m, 10H), 3.7 (ddd, 1H, J ) 10.5, 3.0, 3.0
Hz), 3.1 (m, 1H), 2.8 (m, 3H), 2.55 (dd, 1H, J ) 9.0, 13.5
Hz), 2.5 (d, 1H, J ) 8.4, 13.5 Hz), 1.7 (ddd, 1H, J ) 15,
4.5, 4.5 Hz), 1.55 (ddd, 1H, J ) 15.0, 12.0, 12.0 Hz); 13C
NMR (DEPT) (CDCl3) δ 139.7, 138.3, 129.3(+), 129.2(+),
128.5(+), 128.4(+), 126.4(+), 74.5(+), 57.5(+), 54.0(+),
47.3(-), 41.3(-), 39.1(-); MS (CI) m/z (relative intensity)
285 ([M + 1]+, 100). Anal. Calcd for C18H24N2O: C, 76.02;
H, 8.51; N, 9.85. Found: C, 76.07; H, 8.12; N, 9.82.
Supporting Information Available
Copies of NMR spectra (27 pages). See any current
masthead page for Web access instructions.
OH/H2O/AcOH); mp > 300 °C; [R]20 +80° (c 0.2, CH3-
D
OH); IR (Nujol) 3090, 3060, 3000-2800 (br), 2825, 2050,
1
1735, 1650, 1600, 1590, 1495, 1445, 1045 cm-1; H NMR
(CD3OD) δ 7.3 (m, 10H), 3.83 (ddd, 1H, J ) 11.0, 3.3, 3.3
Hz), 3.63 (m, 1H), 3.38 (ddd, 1H, J ) 7.2, 7.2, 3.5 Hz),
2.95 (m, 4H), 1.90 (ddd, 1H, J ) 14.5, 11.0, 7.3 Hz), 1.76
Received for review June 25, 1998.
OP9802071
100
•
Vol. 3, No. 2, 1999 / Organic Process Research & Development