Efficient Mitsunobu reactions with N-phenylfluorenyl or N-trityl serine esters

Full text of DOI:10.1021/jo951958t

2544
J . Org. Chem. 1996, 61, 2544-2546
Sch em e 1
Efficien t Mitsu n obu Rea ction s w ith
N-P h en ylflu or en yl or N-Tr ityl Ser in e
Ester s
Robert J . Cherney* and Li Wang
Chemical and Physical Sciences,
The DuPont Merck Pharmaceutical Co.,
Experimental Station, Wilmington, Delaware 19880-0500
Sch em e 2
Received November 3, 1995
The Mitsunobu reaction1 is a mild way to convert an
alcohol into a wide range of functionality. During the
course of a SAR (structure-activity relationship) study,
we became interested in the Mitsunobu reaction of serine.
It was our intention to use the primary alcohol of L-serine
ester 1 as a handle, to which our desired functionality
could be attached to give a general derivative 2 (Scheme
1). However, we were concerned that a major problem
inherent with this approach would be the well docu-
mented1 â-elimination reaction to yield 4. In fact,
Wjciechowska and co-workers2 have reported this reac-
tion as an efficient method for producing didehydroala-
nine derivatives 4 from serine 1 through the action of
diethyl azodicarboxylate (DEAD) and PPh3 (Scheme 1
without an acidic HX present). A literature search on
serine-based Mitsunobu reactions yielded mainly refer-
ences for the intramolecular construction and chemistry
of â-lactones,3 â-lactams,4 and oxazolines.5 In fact inter-
molecular examples, as in Scheme 1, are limited and poor
yielding.6 An exception is the reaction7 using L-serine
with hydrazoic acid under the standard conditions (PPh3/
DEAD/solvent). This reaction converts the alcohol 1 into
an azide 5 (∼70% yield) with only a small amount of the
elimination product 4 (∼5%) being produced. Our inter-
ests were centered on a variety of different nucleophiles
(benzoic acids, phenols, and imides) other than hydrazoic
acid. We knew, in general, the Mitsunobu reaction with
4-nitrobenzoic acid was a very efficient8 reaction and thus
selected it as a starting point. As we suspected, N-Cbz-
L-serine 6 with 4-nitrobenzoic acid under standard Mit-
sunobu conditions (DEAD, PPh3, and PhH at rt) afforded
a large amount (53%) of the didehydroalanine 7 along
with a small amount (28%) of the nucleophilic substitu-
tion product 8 (Scheme 2).
Sch em e 3
to the same Mitsunobu reaction with 4-nitrobenzoic acid
at room temperature (Scheme 3), we were pleased to find
only the substitution product 10 in 86% yield. Hence, by
making a simple change in protecting groups, we were
able to prevent the elimination reaction completely.
To examine the scope and utility of this reaction, we
explored additional examples, the results of which are
summarized in Table 1. Entry A indicates that the
Mitsunobu reaction with N-trityl-L-serine methyl ester
and 4-nitrobenzoic acid behaves analogously to give the
substitution product 11A (90% yield) under identical
conditions. As mentioned above, we were interested in
other acidic moieties like imides and phenols. Phthal-
imide could also participate in the reaction to give only
the addition products 11C and 11D for N-phenylfluorenyl
(entry C) and N-trityl (entry D), respectively. One can
compare this example to entry B where N-Cbz-L-serine
methyl ester 6 gave primarily the elimination product 7
(73% yield) and some of the substitution product 11B
(13% yield).
This same trend was also noted for phenols. As
indicated in entry E, we did not observe any substitution
product for the Mitsunobu reaction between N-Boc-L-
serine methyl ester and methyl 4-hydroxybenzoate.
Again, an enhancement was observed with the bulky
N-protecting groups. When the same Mitsunobu reaction
was performed with N-trityl serine methyl ester (entry
F), the substitution product 11F was isolated in 63% yield
along with the aziridine 14 (Table 1). The aziridine 14
has previously been synthesized (via another route) by
Baldwin et al.,11 and our material compared exactly with
their reported characterization. The Mitsunobu reaction
with N-phenylfluorenyl serine 9 and methyl 4-hydroxy-
benzoate (entry G) was analogous to this, producing the
substitution product 11G (40% yield) and the correspond-
ing aziridine 15 (24% yield).
In order to circumvent this side reaction, we considered
the choice of N-protecting group. Both N-phenylfluorenyl
(PhF)9 and N-trityl (Tr)10 are known to protect the
R-center of amino acids from base-promoted racemization.
As a result, we postulated these groups might also protect
serine from the elimination reaction during a Mitsunobu
reaction. When N-phenylfluorenyl serine 9 was subjected
(1) (a) Mitsunobu, O. Synthesis 1981, 1. (b) Hughes, D. L. Org. React.
1992, 42, 335.
(2) Wojciechowska, H.; Pawlowicz, R.; Andruszkiewicz, R.; Grzy-
bowska, J . Tetrahedron Lett. 1978, 4063.
(3) Arnold, L. D.; Kalantar, T. H.; Vederas, J . C. J . Am. Chem. Soc.
1985, 107, 7105.
(4) Miller, M. J . Acc. Chem. Res. 1986, 19, 49.
(5) Galeotti, N.; Montagne, C.; Poncet, J .; J ouin, P. Tetrahedron Lett.
1992, 2807. Wipf, P.; Miller, C. P. Tetrahedron Lett. 1992, 6267.
(6) Nagai, U.; Kato, R.; Pavone, V. Peptide Chem. 1987, 187.
(7) (a) Golding, B. T.; Howes, C. J . Chem. Res. (S) 1984, 1. (b)
Otsuka, M.; Kittaka, A.; Iimori, T.; Yamashita, H.; Kobayashi, S.; Ohno,
M. Chem. Pharm. Bull. 1985, 33, 509. (c) Fabiano, E.; Golding, B. T.;
Sadeghi, M. M. Synthesis 1987, 190.
(8) Martin, S. F.; Dodge, J . A. Tetrahedron Lett. 1991, 32, 3017.
(9) Christie, B. D.; Rapoport, H. J . Org. Chem. 1985, 50, 1239.
(10) Guthrie, R. D.; Nicolas, E. C J . Am. Chem. Soc. 1981, 103, 4638.
(11) Baldwin, J . E.; Spivey, A. C.; Schofield, C. J .; Sweeney, J . B.
Tetrahedron 1993, 49, 6309.
0022-3263/96/1961-2544$12.00/0 © 1996 American Chemical Society

Notes
J . Org. Chem., Vol. 61, No. 7, 1996 2545
Ta ble 1. Mitsu n obu Rea ction s w ith N-P h F a n d N-Tr i-Ser in e Ester s
1.8, 8.8 Hz), 8.13 (dd, 2H, J ) 1.8, 8.8 Hz), 7.7 (t, 2H, J ) 10.2
Hz), 7.42-7.12 (m, 11H), 4.34 (dd, 1H, J ) 5.5, 10.6 Hz), 4.25
(dd, 1H, J ) 5.8, 10.9 Hz), 3.38 (s, 3H), 3.23 (br d, 1H, J ) 8.4
Hz), 3.04 (m, 1H); 13C NMR (CDCl3) δ 173.62, 164.03, 150.53,
148.60, 148.24, 143.84, 140.98, 139.89, 135.16, 130.69, 128.49,
128.27, 127.95, 127.51, 127.32, 125.97, 125.78, 125.09, 123.44,
120.07, 119.96, 72.71, 67.14, 54.76, 51.97; IR (neat) 3320, 3058,
2952, 1732, 1528, 1272 cm-1; ESMS m/z (rel intensity) 509 (M+
+ H, 52), 241 (100). Anal. Calcd for C30H24N2O6: C, 70.86; H,
4.77; N, 5.52. Found: C, 70.75; H, 4.75; N, 5.37.
Each of the N-trityl and N-phenylfluorenyl substitution
products 10, 11A, 11C, 11D, 11F , and 11G were checked
by Mosher amide12 analysis (1H NMR and 19F NMR) for
racemization. In each case, the D-serine enantiomer or
racemate was carried through the same Mitsunobu
reaction and then converted to the Mosher amide for
comparison. All the Mosher amides analyzed in this
fashion proved to be free of racemization and gave a
diastereomeric purity of g95%.
In summary, we have shown that with N-phenylfluo-
renyl or N-trityl protection, Mitsunobu reactions can be
performed quite efficiently with serine. In the cases
shown, benzoic acids and imides showed good reactivity,
whereas phenols were somewhat less reactive.
The following compounds (see Table 1) were synthesized
according to the general procedure:
N-(Tr ip h en ylm eth yl)-O-(4-n itr oben zoyl)-L-ser in e m eth yl
ester (11A): [R]25 +58.9° (c 0.21, CH2Cl2); 1H NMR (DMSO-
D
d6) δ 8.33 (d, 2H, J ) 9.2 Hz), 8.11 (d, 2H, J ) 9.2 Hz), 7.42 (d,
6H, J ) 7.7 Hz), 7.29-7.15 (m, 9H), 4.51 (dd, 1H, J ) 5.9, 11.0
Hz), 4.38 (dd, 1H, J ) 7.5, 11.0 Hz), 3.61 (m, 1H), 3.19 (d, 1H,
J ) 10.2 Hz), 3.12 (s, 3H); 13C NMR (CDCl3) δ 172.69, 164.20,
150.25, 145.34, 135.10, 120.69, 128.57, 127.88, 126.57, 123.50,
70.00, 67.28, 55.36, 51.94; IR (KBr) 3324, 3084, 1734, 1528, 1272
cm-1; CIMS m/z (rel intensity) 511 (M+ + H, 65), 243 (100). Anal.
Calcd for C30H26N2O6: C, 70.58; H, 5.13; N, 5.50. Found: C,
70.50; H, 5.13; N, 5.37.
Exp er im en ta l Section
Gen er a l. All reactions were performed under a nitrogen
atmosphere at room temperature unless otherwise noted. The
amino acid derivative N-phenylfluorenyl-L-serine methyl ester
9 was synthesized according to the procedure of Rapoport.13 In
addition, N-Trityl-L-serine methyl ester was synthesized as
reported by Baldwin.11 All other reagents and solvents were
purchased from commercial sources and used as received unless
otherwise noted. The 1H NMR and 13C NMR were recorded at
300 and 75 MHz, respectively. The chemical shifts are reported
in ppm (δ) relative to TMS in the solvent listed. ESMS is the
abbreviation for electrospray ionization mass spectroscopy.
Gen er a l P r oced u r e for N-P h F or N-Tr -ser in e Mitsu n obu
Rea ction s: N-(9-P h en ylflu or en -9-yl)-O-(4-n itr oben zoyl)-L-
ser in e Meth yl Ester (10). The N-PhF-L-serine methyl ester
9 (474.4 mg, 1.3 mmol) was dissolved in benzene (15 mL) prior
to the addition of PPh3 (380.8 mg, 1.4 mmol) and 4-nitrobenzoic
acid (330.8 mg, 1.9 mmol). After 5 min, DEAD (0.23 mL, 1.4
mmol) was added dropwise. The reaction was stirred for 15 h,
after which time the solid urea was removed via filtration. The
filtrate was concentrated to an oil, which was purified by flash
chromatography (1:5 EtOAc/hexane) to provide the desired
N-(9-P h en ylflu or en -9-yl)-â-(ph th alim ido)-L-alan in e m eth -
yl ester (11C): [R]25D -280.0° (c 0.16, CH2Cl2); 1H NMR (DMSO-
d6) δ 7.90-7.85 (m, 4H), 7.79 (dd, 2H, J ) 7.7, 13.9 Hz), 7.36 (t,
1H, J ) 7.3 Hz), 7.28-7.12 (m, 8H), 6.67 (d, 2H, J ) 3.4 Hz),
3.64 (dd, 1H, J ) 9.1, 13.5 Hz), 3.50 (dd, 1H, J ) 5.5, 13.5 Hz),
3.38 (d, 1H, J ) 9.89 Hz), 3.34 (s, 3H), 2.74 (ddd, 1H, J ) 5.48,
9.1, 9.89 Hz); 13C NMR (CDCl3) δ 174.31, 167.84, 133.91, 128.40,
128.31, 127.19, 126.02, 123.29, 120.02, 72.73, 54.42, 52.01, 41.16;
IR (KBr) 3472, 3060, 1718, 1394 cm-1; CIMS m/z (rel intensity)
489 (M+ + H, 100). Anal. Calcd for C31H24N2O4: C, 76.21; H,
4.95; N, 5.73. Found: C, 75.88; H, 4.80; N, 5.59.
N-(Tr ip h en ylm eth yl)-â-(p h th a lim id o)-L-a la n in e m eth yl
ester (11D): [R]25 +16.9° (c 0.11, CH2Cl2); 1H NMR (DMSO-
D
d6) δ 7.93-7.85 (m, 4H), 7.38-7.14 (m, 15H), 3.83 (d, 2H, J )
7.0 Hz), 3.48 (m, 1H), 3.04-3.01 (m, 4H); 13C NMR (CDCl3) δ
172.89, 167.83, 145.42, 133.93, 128.61, 127.75, 126.36, 123.49,
123.26, 80.90, 55.38, 51.88, 41.62; IR (KBr) 3202, 2950, 1734,
1718, 1394 cm-1; CIMS m/z (rel intensity) 491 (M+ + H, 100).
product 10 (568.1 mg, 1.1 mmol, 86%) as a white foam: [R]25
D
Anal. Calcd for
C31H26N2O4: C, 75.90; H, 5.34; N, 5.71.
1
-162.5° (c 0.25, CH2Cl2); H NMR (CDCl3) δ 8.27 (dd, 2H, J )
Found: C, 75.51; H, 5.25; N, 5.59.
N -(T r ip h e n y lm e t h y l)-O -(4-c a r b o m e t h o x y p h e n y l)-L-
(12) The substitution product was heated in TFA and CH2Cl2 to
remove the N-phenylfluorenyl or N-trityl protecting group. The solution
was concentrated and reacted with the (R)-Mosher acid chloride (Fluka)
to give the corresponding amide: Dale, J . A.; Dull, D. L.; Mosher, H.
S. J . Org. Chem. 1969, 34, 2543.
ser in e m eth yl ester (11F ): [R]25 +96.1° (c 0.15, CH2Cl2); 1H
D
NMR (CDCl3) δ 7.89 (d, 2H, J ) 8.9 Hz), 7.43 (d, 6H, J ) 7.3
Hz), 7.29 (t, 6H, J ) 7.5 Hz), 7.20 (t, 3H, J ) 7.1 Hz), 6.97 (d,
2H, J ) 8.9 Hz), 4.28 (dd, 1H, J ) 4.9, 9.5 Hz), 4.05 (dd, 1H, J
) 6.6, 9.5 Hz), 3.81 (s, 3H), 3.56 (ddd, 1H, J ) 4.9, 6.6, 10.2 Hz),
(13) Lubell, W.; Rapoport, H. J . Org. Chem. 1989, 54, 3824.

2546 J . Org. Chem., Vol. 61, No. 7, 1996
Notes
3.23 (s, 3H), 2.90 (d, 1H, J ) 10.2 Hz); 13C NMR (CDCl3) δ
173.24, 162.14, 145.63, 131.55, 128.72, 127.94, 126.58, 123.03,
114.23, 70.99, 70.35, 56.03, 51.96, 51.87; IR (KBr) 3412, 3030,
2950, 1718, 1606 cm-1; CIMS m/z (rel intensity) 496 (M+ + H,
7), 243 (100). Anal. Calcd for C31H29NO5: C, 75.13; H, 5.91; N,
2.84. Found: C, 75.11; H, 5.73; N, 2.75.
128.27, 128.13, 105.98, 66.95, 52.79; IR (KBr) 3412, 1742, 1500
cm-1; CIMS (NH3) m/z (rel intensity) 253 (M+ + NH4, 100), 236
(M+ + H, 19); HRMS calcd for C12H14NO4 (M+ + H), 236.092283;
found 236.091810.
Data for 8: [R]25D +38.0° (c 0.67, CH2Cl2); 1H NMR (CDCl3) δ
8.27 (dd, 2H, J ) 1.8, 7.0 Hz), 8.13 (dd, 2H, J ) 2.2, 7.0 Hz),
7.35 (s, 5H), 5.64 (br d, 1H, J ) 7.7 Hz), 5.13 (s, 2H), 4.81 (m,
1H), 4.69 (m, 2H), 3.81 (s, 3H); 13C NMR (CDCl3) δ 169.66,
164.11, 155.64, 150.77, 135.88, 134.64, 130.81, 128.57, 128.35,
128.16, 123.66, 67.37, 65.37, 53.34, 53.05; IR (KBr) 3424, 1754,
1730, 1718, 1536 cm-1; ESMS m/z (rel intensity) 403 (M+ + H,
100). Anal. Calcd for C19H18N2O8: C, 56.72; H, 4.52; N, 6.96.
Found: C, 56.53; H, 4.58; N, 6.83.
Mitsu n obu Rea ction betw een N-Cbz-L-Ser in e Meth yl
Ester (6) a n d P h th a lim id e. The reaction was performed as
described in the general procedure. Starting with N-Cbz-L-
Serine methyl ester (6) (443.2 mg, 1.7 mmol), phthalimide (382.5
mg, 2.6 mmol), PPh3 (505.8 mg, 1.9 mmol), and DEAD (0.31 mL,
1.9 mmol) gave the â-elimination product 7 (301.6 mg, 1.3 mmol,
73%) and the substitution product 11B (87 mg, 0.22 mmol, 13%).
Data for 7: See above.
Data for 11B: [R]25D +12.0° (c 0.22, CH2Cl2); 1H NMR (CDCl3)
δ 7.84 (m, 2H), 7.73 (m, 2H), 7.31 (s, 5H), 5.69 (br d, 1H, J ) 7.3
Hz), 5.07 (s, 2H), 4.68 (br q, 1H, J ) 5.5 Hz), 4.22-4.04 (m, 2H),
3.79 (s, 3H); 13C NMR (CDCl3) δ 170.13, 168.02, 155.70, 136.06,
134.11, 131.65, 128.34, 127.98, 127.91, 123.46, 66.95, 53.27,
52.80, 39.03; IR (KBr) 3312, 1772, 1752, 1726, 1684 cm-1; CIMS
(NH3) m/z (rel intensity) 400 (M+ + NH4, 100), 383 (M+ + H,
27); HRMS calcd for C20H19N2O6 (M+ + H), 383.124312; found
383.125009.
N-(Tr iph en ylm eth yl)azir idin e-(2S)-car boxylic acid m eth -
1
yl ester (14): [R]25 -85.9° (c 0.09, CH2Cl2); H NMR (CDCl3)
D
δ 7.51-7.19 (m, 15H), 3.76 (s, 3H), 2.26 (dd, 1H, J ) 1.4, 2.6
Hz), 1.89 (dd, 1H, J ) 2.6, 6.2 Hz), 1.41 (dd, 1H, J ) 1.4, 6.2
Hz); 13C NMR (CDCl3) δ 172.20, 143.53, 129.24, 128.98, 128.55,
128.36, 127.90, 127.83, 127.75, 127.56, 126.84, 126.55, 74.30,
51.98, 31.61, 28.57; IR (KBr) 3434, 3060, 1744, 1490, 1448, 1284,
1242 cm-1; CIMS m/z (rel intensity) 344 (M+ + H, 21), 243 (100);
HRMS calcd for C23H22NO2 (M+ + H), 344.1650; found 344.1664.
N-(9-P h en ylflu or en -9-yl)-O-(4-ca r bom eth oxyp h en yl)-L-
ser in e m eth yl ester (11G): [R]25 -62.5° (c 0.11, CH2Cl2); 1H
D
NMR (CDCl3) δ 7.90 (dd, 2H, J ) 1.8, 8.8 Hz), 7.69 (t, 2H, J )
6.6 Hz), 7.44-7.16 (m, 11H), 6.71 (dd, 2H, J ) 2.2, 9.1 Hz), 4.00
(dd, 1H, J ) 5.1, 9.5 Hz), 3.92 (dd, 1H, J ) 2.9, 6.6 Hz), 3.87 (s,
3H), 3.36 (s, 3H), 3.20 (br d, 1H, J ) 8.8 Hz), 3.05 (m, 1H); 13C
NMR (CDCl3) δ 174.06, 166.66, 161.97, 148.90, 148.46, 143.99,
140.88, 140.07, 131.39, 128.46, 128.24, 127.94, 127.54, 127.29,
126.00, 125.71, 125.05, 122.78, 120.07, 119.96, 114.06, 72.74,
70.00, 54.94, 51.83, 51.72; IR (film) 3320, 2950, 1716 cm-1; ESMS
m/z (rel intensity) 494 (M+ + H, 41), 241 (100). Anal. Calcd for
C
31H27NO5: C, 75.44; H, 5.51; N, 2.85. Found: C, 75.57; H, 5.41;
N, 2.80.
N-9-P h en ylflu or en -9-yla zir id in e-(2S)-ca r b oxylic a cid
m eth yl ester (15): [R]25 -124.1° (c 0.94, CH2Cl2); 1H NMR
D
(CDCl3) δ 7.69 (td, 2H, J ) 2.5, 4.8 Hz), 7.41-7.33 (m, 4H), 7.28-
7.18 (m, 7H), 3.69 (s, 3H), 2.19 (dd, 1H, J ) 2.9, 6.2 Hz), 2.09
(dd, 1H, J ) 1.1, 2.9 Hz), 1.55 (dd, 1H, J ) 1.1, 6.2 Hz); 13C
NMR (CDCl3) δ 171.40, 147.24, 146.05, 142.95, 141.20, 140.29,
128.90, 128.71, 128.25, 127.77, 127.64, 127.14, 126.90, 126.21,
126.23, 119.95, 119.90, 75.72, 52.11, 33.18, 29.76; IR (KBr) 3058,
1748, 1448 cm-1; HRMS calcd for C23H20NO2 (M+ + H),
342.149404; found 342.147554.
Mitsu n obu Rea ction betw een N-Boc-L-Ser in e Ben zyl
Ester (en tr y E) a n d Meth yl 4-Hyd r oxyben zoa te. The
reaction was performed as described in the general procedure.
Starting with N-Boc-L-serine benzyl ester (517.7 mg, 1.7 mmol),
methyl 4-hydroxybenzoate (399.4 mg, 2.6 mmol), PPh3 (505.8
mg, 1.9 mmol), and DEAD (0.31 mL, 1.9 mmol) gave the
â-elimination product 12 (442.5 mg, 1.6 mmol, 91%): 1H NMR
(CDCl3) δ 7.37 (m, 5H), 7.03 (br s, 1H), 6.17 (br s, 1H), 5.79 (d,
1H, J ) 1.5 Hz), 5.26 (s, 2H), 1.48 (s, 9H); 13C NMR (CDCl3) δ
163.86, 152.53, 135.21, 131.38, 128.63, 128.50, 128.18, 105.43,
80.67, 67.60, 28.23; IR (KBr) 3456, 3420, 2978, 1718 cm-1; HRMS
calcd for C15H20NO4 (M+ + H), 278.139233; found 278.138973.
Mitsu n obu Rea ction betw een N-Cbz-L-Ser in e Meth yl
Ester (6) a n d 4-Nitr oben zoic Acid . The reaction was per-
formed as described in the general procedure. Starting with
N-Cbz-L-serine methyl ester (6) (283.0 mg, 1.1 mmol), 4-ni-
trobenzoic acid (280.2 mg, 1.7 mmol), PPh3 (322.5 mg, 1.2 mmol),
and DEAD (0.19 mL, 1.2 mmol) gave the â-elimination product
7 (141 mg, 0.6 mmol, 53%) and the substitution product 8 (126
mg, 0.3 mmol, 28%).
Ack n ow led gm en t. The authors would like to thank
Professor Henry Rapoport and Dr. Arthur D. Patten for
informative discussions.
Data for 7: 1H NMR (CDCl3) δ 7.37 (m, 5H), 7.24 (br s, 1H),
6.25 (br s, 1H), 5.79 (d, 1H, J ) 1.5 Hz), 5.16 (s, 2H), 3.83 (s,
3H); 13C NMR (CDCl3) δ 164.07, 153.02, 135.76, 130.91, 128.51,
J O951958T