Peptide coupling in the presence of highly hindered tertiary amines

Article abstract of DOI:10.1021/jo950912x

Previously, 2,4,6-trimethylpyridine (collidine), due to steric shielding around the N-atom, was found to be an efficient base for effecting peptide segment coupling via azabenzotriazole-based onium-style coupling reagents. A number of even more highly hin

Full text of DOI:10.1021/jo950912x

2460
J . Org. Chem. 1996, 61, 2460-2465
P ep tid e Cou p lin g in th e P r esen ce of High ly Hin d er ed Ter tia r y
Am in es
Louis A. Carpino,* Dumitru Ionescu,† and Ayman El-Faham‡
Department of Chemistry, Box 34510, University of Massachusetts, Amherst, Massachusetts 01003-4510
Received May 16, 1995 (Revised Manuscript Received J anuary 4, 1996X)
Previously, 2,4,6-trimethylpyridine (collidine), due to steric shielding around the N-atom, was found
to be an efficient base for effecting peptide segment coupling via azabenzotriazole-based onium-
style coupling reagents. A number of even more highly hindered bases, including 2,3,5,6-
tetramethylpyridine, 2,6-di-tert-butyl-4-(dimethylamino)pyridine, triisopropylamine, and N-tert-
butylmorpholine, have been compared with collidine in such reactions. Some of the newer bases
showed advantages in terms of convenience in handling and maintenance of configuration during
segment coupling processes, although dramatic differences based on steric effects were not observed.
On the basis of results with a number of test peptides and many base-coupling reagent combinations,
it was noted that most efficient results are obtained if 1 equiv of HOAt is present as an additive
during the coupling process. For rapid activation of onium-style coupling reagents during stepwise
solid-phase coupling reactions, the stronger base 2,6-di-tert-butyl-4-(dimethylamino)pyridine was
more effective than collidine.
While we were examining the utility of a new class of
of the amino function relative to NMM (pKa 7.38), gave
the best results.4 Classic mechanistic studies5 of the
effect of base structure on C-H proton abstraction rates
provide a consistent rationale for the superiority of
collidine in these reactions. 2,6-Dimethylpyridine (2,6-
lutidine) gave results comparable to those of collidine,
but other dimethylpyridines bearing the methyl groups
at positions other than the R-position proved to be
unsuitable.
In the present work the effect of further increases in
the bulk of the pyridyl R-substituents has been examined.
Model test peptides 4-8,4,6,7 all of which have been
examined in earlier studies, were chosen for initial
testing with these newer bases. For tripeptides the
coupling reagents based on 7-aza-1-hydroxybenzotriazole
(HOAt)1,2 such as N-[(dimethylamino)-1H-1,2,3-triazolo-
[4,5-b]pyridin-1-ylmethylene]-N-methylmethanamin-
ium hexafluorophosphate N-oxide (HATU), 1, and O-(7-
a za ben zot r ia zol-1-yl)-1,1:3,3-bis(t et r a m et h ylen e)-
uronium hexafluorophosphate (HAPyU), 2, analogs of the
well-known reagent N-[(1H-benzotriazol-1-yl)(dimethyl-
amino)methylene]-N-methylmethanaminium hexafluo-
rophosphate N-oxide (HBTU) 3, it became apparent that
Model Peptides
the two bases most commonly recommended for peptide
coupling reactions, diisopropylethylamine (DIEA) and
N-methylmorpholine (NMM), often lead to unacceptable
levels of inversion of configuration at the activated
carboxyl function in the case of segment coupling. An
examination of other bases was begun, and among those
previously studied, 2,4,6-trimethylpyridine (collidine,
TMP, aqueous pKa 7.43),3 an amine of lowered basicity
relative to DIEA (pKa 10.1) and enhanced steric shielding
coupling position is shown by the vertical dotted line.
Passing from the 2,6-dimethylpyridine to the 2,6-di-
ethyl- ,8 diisopropyl-,8 and di-tert-butylpyridines9 was
† Fulbright Scholar. On leave of absence from the Department of
Organic Chemistry, Faculty of Chemistry, University of Bucharest,
70346 Bucharest, Romania.
(4) Carpino, L. A.; El-Faham, A. J . Org. Chem. 1994, 59, 695.
(5) (a) Gold, V. In Progress in Stereochemistry; de la Mare, P. B. D.,
Klyne, W., Eds.; Butterworths: Washington, 1962; Vol. 3, p 169. (b)
Feather, J . A.; Gold, V. Proc. Chem. Soc. 1963, 306; J . Chem. Soc. 1965,
1752. (c) Covitz, F.; Westheimer, F. H. J . Am. Chem. Soc. 1963, 85,
1773. (d) Hine, J .; Houston, J . G.; J ensen, J . H.; Mulders, J . J . Am.
Chem. Soc. 1965, 87, 5050. (e) Lewis, E. S.; Funderburk, L. H. J . Am.
Chem. Soc. 1967, 89, 2322. (f) Gutsche, C. D.; Redmore, D.; Buriks, R.
S.; Nowotny, K.; Grassner, H.; Armbruster, L. W. J . Am. Chem. Soc.
1967, 89, 1235. (g) Zollinger, H. Helv. Chim. Acta 1955, 38, 1623.
(6) Carpino, L. A.; El-Faham, A.; Albericio, F. J . Org. Chem. 1995,
60, 3561.
(7) Wenschuh, H.; Beyermann, M.; Haber, H.; Seydel, J . K.; Krause,
E.; Bienert, M.; Carpino, L. A.; El-Faham, A. J . Org. Chem. 1995, 60,
405.
(8) Balaban, A. T.; Nenitzescu, C. D. Liebigs Ann. Chem. 1959, 625,
74.
‡ On leave of absence from the Department of Chemistry, Faculty
of Science, Alexandria University, Alexandria, Egypt.
X Abstract published in Advance ACS Abstracts, March 15, 1996.
(1) Carpino, L. A. J . Am. Chem. Soc. 1993, 115, 4397.
(2) For a list of some of these new reagents as well as current
structural assignments and nomenclature, see: (a) Carpino, L. A.; El-
Faham, A.; Minor, C. A.; Albericio, F. J . Chem. Soc., Chem. Commun.
1994, 201-203. (b) Abdelmoty, I.; Albericio, F.; Carpino, L. A.; Foxman,
B. M.; Kates, S. A. Lett. Peptide Sci. 1994, 1, 57-67. Whether HAPyU
crystallizes in the form shown (2) or in the isomeric guanidinium form
(comparable to structure 1 for HATU) is not yet known. In the absence
of appropriate X-ray data structure 2 is arbitrarily retained in the
present paper.
(3) Aqueous pKa data, unless otherwise indicated, are taken from:
Perrin, D. Dissociation Constants of Organic Bases in Aqueous Solution;
Butterworths: London, 1965; Supplement, 1972.
0022-3263/96/1961-2460$12.00/0 © 1996 American Chemical Society

Peptide Coupling in the Presence of Tertiary Amines
J . Org. Chem., Vol. 61, No. 7, 1996 2461
expected to lead to a progressive decrease in the extent
of epimerization or racemization at the activated car-
boxylic acid residue of the protected dipeptide or amino
acid involved in the reaction. Although such an effect
was observed for the very bulky tert-butyl substituent,
this was not the case for the ethyl or isopropyl deriva-
tives. Thus, for coupling leading to tripeptide 4 neither
2,6-diethyl- nor 2,6-diisopropylpyridine derivatives 9a or
9b, respectively, proved to be more effective than TMP.
of 10 and 11 steric effects may be significantly less for
the latter.17
For model peptides 4, 6, and 8, coupling efficiency for
pentamethylpyridine, 12,18 was slightly less than for
TEMP, and in addition, this liquid base was difficult to
handle under ordinary conditions due to its extremely
hygroscopic nature. Use of the hydrate itself caused
increased loss of configuration relative to the anhydrous
amine. For example, in the case of tripeptide 5, with
coupling effected by HAPyU, 2, the anhydrous amine and
its hydrate gave 1.8 and 4.8% of the DL form of Z-Gly-
Phe-Pro-NH2 (Z ) benzyloxycarbonyl), respectively. These
results led to a brief study of the influence of water on
the segment coupling process in general. It seems
possible that with a strong highly hindered base the
kinetically active base may not be the amine itself.19 In
the presence of water, hydroxide ion might play this role.
In addition, in the presence of large amounts of water,
the increase in solvent polarity may adversely affect the
coupling process.20
These two bases, however, appear to be relatively less
stable than TMP and become discolored on standing.
Freshly distilled material was used in all experiments,
but it is not certain that the results are unaffected by
unknown contaminants.
For most of the studies reported earlier,1,4,6 a com-
mercial grade of DMF (dimethylformamide, Fisher HPLC
grade) dried for at least 3 days over molecular sieves (3
Å) at room temperature was routinely used as solvent
for the test coupling reactions. This technique is known
not to provide totally anhydrous DMF.21 A more ef-
ficiently dried sample of DMF was obtained by a special
procedure which involved distillation from P2O5 followed
by treatment with molecular sieves (3 Å) which had been
freshly activated by heating over a free flame. The
resulting freshly prepared dry solvent was then used
directly in various coupling reactions and subsequently
diluted with increasing amounts of water for further
tests. The tertiary amines were freshly dried as de-
scribed previously.4 In reactions leading to model pep-
tides 4 and 6, for a number of bases both hindered and
nonhindered, the loss of configuration at valine first
increased with the addition of small amounts of water,
then decreased, and finally rose rapidly (Table 4, sup-
porting information). In a sample of DMF which had
simply been dried over ordinary molecular sieves, the
extent of epimerization at valine was in most cases
comparable to that at the point of reversal and thus lower
than that observed in the carefully dried medium. It
seems possible that these effects may be responsible for
the discrepancies often observed between various runs
before attempts were made to control more carefully the
reaction conditions. While the increased loss of config-
uration observed in the presence of large amounts of
water might be associated with the increased polarity of
the reaction medium, a rationale for the effect of adding
small amounts of water to the dried solvent is not
obvious. Whatever the reason, the effect allows for
simplification of the reaction protocol: DMF dried merely
For alkylated pyridines the addition of each new alkyl
group increases the basicity by about 0.5-0.85 pKa
units.10 2,3,5,6-Tetramethylpyridine, 10 (TEMP, pKa
7.90), and the analogous cycloalkyl derivative octahy-
droacridine, 11 (OHA, pKa 8.0911), were examined along
with pentamethylpyridine, 12 (pKa 8.75), as somewhat
more basic analogs of TMP. For these three amines best
results were obtained with TEMP and OHA which were
about as efficient as TMP. In addition, TEMP and OHA
have practical advantages over TMP in being stable
odorless solids. Not yet commercially available, in
contrast to OHA, TEMP was obtained for these studies
via a multistep procedure, each step of which, however,
proceeds readily.12 For the limited number of cases
examined, clear-cut differences among TMP, TEMP, and
OHA were not evident. For TEMP the steric effects of
the two R-methyl substituents might conceivably be
“buttressed”13 by the adjacent 3,5-dimethyl groups, and
judging from quaternization rates on the lower homologs
(9) Brown, H. C.; Kanner, B. J . Am. Chem. Soc. 1966, 88, 986.
(10) Ikekawa, N.; Sato, Y.; Maeda, T. Chem. Pharm. Bull. 1954, 2,
205.
(11) Thummel, R. P.; Kohli, D. K. J . Org. Chem. 1977, 42, 2742.
(12) Tsuda, K.; Ikekawa, N.; Mishima, H.; Iino, A. Chem. Pharm.
Bull. 1953, 1, 122.
(13) Kinetic examination of the Menschutkin reaction has often been
used to compare the steric encumbrance around the N-atom of pyridine
derivatives.14 According to this criterion the effect of buttressing has
been demonstrated by comparison of the 2,5- and 2,3-dimethylpyr-
idines, the former being 2.5 as reactive as the latter.15 On the other
hand, the analogy between the Menschutkin reaction and steric effects
toward the deprotonation of weak carbon acids may be inexact in view
of the radical differences between the expected transition states for
the two reactions. Furthermore, while a highly hindered base such as
2,6-di-tert-butylpyridine reacts with potent methylating agents only
very slowly, it gives a stable N-methylpyridinium salt which has been
shown not to be deformed according to X-ray crystal analysis.16 Thus,
it may not be unexpected that differences among R-substituted
pyridines in the reactions examined here will be muted except for the
most bulky substituents.
(16) Weber, H.; Pant, J .; Liedigk, M.; Wunderlich, H. Chem. Ber.
1981, 114, 1455.
(17) Seeman, J . I.; Galzerno, R.; Curtis, K.; Schug, J . C.; Viers, J .
W. J . Am. Chem. Soc. 1981, 103, 5982.
(18) Balaban, A. T.; Bota, A.; Chiraleur, F.; Sliam, E.; Hanes, A.;
Draghici, C. Rev. Roum. Chim. 1977, 22, 1003.
(19) An example of such an effect has been reported for proton
sponge. See: Alder, R. W.; Goode, N. C.; Miller, N.; Hibbert, F.; Hunte,
K. P. P.; Robbins, H. J . J . Chem. Soc., Chem. Commun. 1978, 89.
(20) Although increases in solvent polarity are often associated with
increased loss of configuration in peptide coupling the true nature of
the effect is more complex. For a review see: Benoiton, N. L. In The
Peptides; Gross, E., Meienhofer, J ., Eds.; Academic: New York, 1983;
Vol. 5, Part B, p 276.
(14) Shug, J . C.; Viers, J . W.; Seeman, J . I. J . Org. Chem. 1983, 48,
4892.
(15) Seeman, J . I.; Schug, J . C.; Viers, J . W. J . Org. Chem. 1983,
48, 2399.
(21) Burfield, D. R.; Smithers, R. H. J . Org. Chem. 1978, 43, 3966.

2462 J . Org. Chem., Vol. 61, No. 7, 1996
Carpino et al.
by standing over molecular sieves at room temperature
has been adopted for all routine coupling tests.
gated in the case of this highly sensitive system (Table
1). In the case of HAPyU, 2, yields were rather low
unless reaction times were extended or 3 or more equiv
of base was used, although normal yields were observed
with HATU, 1, and HBTU, 3.
Previously4 it was reported that 2,6-di-tert-butylpyri-
dine (pKa 4.9522,23) was too weakly basic to induce
coupling between Z-Phe-Val-OH and H-Ala-OMe via
HAPyU, 2. The unusually low basicity in water of this
base has been the subject of many studies since the first
observations of Brown and co-workers.9 The effect seems
best ascribed to steric effects on solvation and loss of
entropy of the cation formed upon protonation in solu-
tion.23 In contrast to effects observed in water, the gas
phase basicity of 2,6-di-tert-butylpyridine is enhanced
over that of pyridine as expected on the basis of the
inductive effects of the two alkyl residues.24
Since the amino acid derivative to which coupling is
made, proline amide (pKa 8.82) in the case of 4, is itself
basic enough to activate the coupling reagent, the reac-
tions were also carried out in the absence of any added
tertiary amine. As shown in Table 1, in the presence of
a total of 2 or 3 equiv of proline amide, the extent of
epimerization was 1.8 and 6.9%, respectively. With 1 or
2 equiv of DB(DMAP) in place of the extra proline amide,
epimerization was reduced to 1.3 or 1.7%, respectively.
The results were drastically different in the case of 2,6-
dimethyl-4-(dimethylamino)pyridine (2,6-diMe-DMAP),28
the collidine analog of DMAP, where 16.1% of the epimer
was obtained in the presence of 2 equiv of this base.
Other examples confirm that the high basicity of 2,6-
diMe-DMAP outweighs any steric protection provided by
the two R-methyl substituents (Tables 5 and 6, support-
ing information). Due to its low melting point and
extreme hygroscopic nature, 2,6-diMe-DMAP was a dif-
ficult base with which to work. The even more basic
(estimated pKa about 14),29 relatively unhindered guani-
dine derivative, N-tert-butyltetramethylguanidine (t-Bu-
TMG), generally led to extensive loss of configuration.
Even in the case of tripeptide 7, shown in a previous
study6 to be quite insensitive, the use of HAPyU/t-Bu-
TMG gave 16.1% of the LDL form. Generally, for this
system less than 0.1-0.8% LDL isomer is obtained when
coupling was effected by HAPyU, 2, in the presence of
DIEA, NMM, or the various other bases described here.
Introduction of a 4-(dialkylamino) residue into such
systems causes a significant increase in basicity: the pKa
(7.45) measured25 for the 4-diethylamino derivative 13b
in methoxyethanol/water (80/20) is only slightly lower
than that observed for DMAP (7.70) under the same
conditions. The latter corresponds to a pKa of about 9.22
in water, suggesting a pKa of about 9 in water for the
analogous amine 13a ,26 for which no pKa measurements
were performed during the course of the present work.
In spite of the relatively high basicity of these substi-
tuted DMAP derivatives the rate of proton transfer from
the R-position of an activated amino acid may be expected
to be less than for other pyridine bases including TMP
and TEMP. Proton transfer rates between 2,6-di-tert-
butylpyridine and hydronium ion are 50-70 times lower
than for simple pyridines.27
2,6-Di-tert-butyl-4-(dimethylamino)pyridine [DB(D-
MAP), 13a ] was first synthesized by modification of the
method described by Potts and Winslow25 for the analog
13b. A second, much more convenient synthesis was
based on direct reaction of tert-butyllithium with DMAP.26
This highly hindered amine effected rapid and complete
coupling of Z-Phe-Val-OH with H-Pro-NH2, and while
retention of configuration was not complete, it was
somewhat better than for other tertiary amines investi-
In addition to the pyridine bases studied here which
were designed to push the collidine example to its limit,
two more highly hindered aliphatic analogs of DIEA and
NMM were also examined. Triisopropylamine is an
amine of unique structure, being planar according to gas
phase electron diffraction and 14N/15N NMR data.30
While triisopropylamine proved generally to be more
effective than DIEA for the systems examined (Table 1;
Tables 5-8, supporting information; Table 2 of ref 4; and
Tables 2 and 3 of ref 6), it was less efficient than TEMP.
In the morpholine series the N-tert-butyl derivative31
showed similar advantages over the commonly used
NMM (Table 1; Tables 5-8, supporting information; and
Table 2 of ref 4).
Following these preliminary studies with simple di-
and tripeptide models 4-8, a test hexapeptide 14 was
assembled. The [3 + 3] coupling to give 14 had previ-
ously been shown4 to be a sensitive test for the nature of
both coupling reagent and base. Results for TEMP and
DB(DMAP) are shown in Table 2. Both bases give
(22) Although 2,6-di-tert-butylpyridine is a weak base in water it is
unusually weak in DMSO. In water it differs from pyridine by only
0.2 pKa units whereas in DMSO it is 2.5 units less basic [Benoit, R.
L.; Fre´chette, M.; Lefebvre, D. Can. J . Chem. 1988, 66, 1159]. Although
not examined in DMF, relative pKa’s are expected to be similar in
solvents such as DMF and DMSO (Bordwell, F. G. Acc. Chem. Res.
1988, 21, 456. Bordwell, F. G.; Branca, J . C.; Hughes, D. L.; Olmstead,
W. N. J . Org. Chem. 1980, 45, 3305). A unique example of the difference
which may be observed in water and organic solvents is that of the
equilibrium between benzoic acid and triethylamine which is said to
show complete ionization in water (K ) 106.5) but little ionization in
DMSO (K ) 10-2.1) (Ritchie, C. D. In Physical Organic Chemistry, 2nd
ed.; Dekker: New York, 1990; p 209). Relationships of this type may
be responsible for the poor activation of carboxylic acids sometimes
observed in DMF which may show effects similar to those seen in
DMSO.
(23) Hopkins, H. P., J r.; J ahagirdar, D. V.; Moulik, P. S.; Aue, D.
H.; Webb, H. M.; Davidson, W. R.; Pedley, M. D. J . Am. Chem. Soc.
1984, 106, 4341.
(24) Arnett, E. M.; Chawla, B. J . Am. Chem. Soc. 1979, 101, 7141.
(25) Potts, K. T.; Winslow, P. A. Synthesis 1987, 839.
(26) Deutsch, E.; Cheung, N. K. V. J . Org. Chem. 1973, 38, 1123.
(27) Bernasconi, C. F.; Carre´, D. J . J . Am. Chem. Soc. 1979, 101,
2707.
(28) Although
a number of references to the use (but not the
preparation) of 2,6-dimethyl-DMAP N-oxide could be located (e.g. Dega-
Szafran, Z.; Hrynio, A.; Szafran, M. J . Chem. Soc., Perkin Trans. 2
1991, 1161) no reference to 2,6-dimethyl-DMAP itself could be found.
The method used here was adapted from that of related compounds
as noted in the experimental section. The closest analog for which a
pKa has been reported is that of 13b. For pKa values of related amines
see: Essery, J . M.; Schofield, K. J . Chem. Soc. 1961, 3939.
(29) Barton, D. H. R.; Elliott, J . D.; Gero, S. D. J . Chem. Soc., Perkin
Trans. 1 1982, 2085.
(30) (a) Kuffner, F.; Koechlin, W. Monatsh. Chem. 1962, 93, 476.
(b) Bock, H.; Goebel, I.; Havlas, Z.; Liedle, S.; Oberhammer, H. Angew.
Chem., Int. Ed. Engl. 1991, 30, 187. (c) Wong, T. C.; Collazo, L. R.;
Guziec, F. S., J r. Tetrahedron 1995, 51, 649.
(31) Cook, M. J .; Katritzky, A. R. Man˜os, M. M. J . Chem. Soc. 1971,
1330.

Peptide Coupling in the Presence of Tertiary Amines
J . Org. Chem., Vol. 61, No. 7, 1996 2463
Ta ble 1. Effect of Cou p in g Rea gen t a n d Ba se on th e Exten t of In ver sion a t Va lin e Du r in g F or m a tion of
Z-P h e-Va l-P r o-NH2 in DMF a
LDL, % (yield, %)c,d
baseb
additiveb
HOAt (1)
HAPyU
38.1 (81.8)
6.5 (93.5)
2.7 (89.1)
3.5 (74.1)
3.3 (100)
HATU
HBTU
19.2 (98)
t-BuTMG (2)
i-Pr3N (2)
i-Pr3N (2)
7.1 (89.5)
5.4 (86.8)
9.9 (94.7)
TMP (1)
TMP (2)
TMP (1)
TMP (2)
NMM (2)
NMM (2)
NBM (2)
NBM (2)
TEMP (1)
TEMP (2)
HOAt (1)
HOAt (1)
3.4 (66)
2.8, 1.4, 1.3 (82.8)
12.2 (90.2)
3.3 (71.2)
5.5 (86.5)
1.9 (70.9)
2.9 (90.2)
2.5, 3.5
HOAt (1)
HOAt (1)
8.9 (100)
5.4, 5.3 (86.8)
17.6 (86.8)
6.7 (84.2)
3.5, 2.8 (76)
TEMP (1)
TEMP (2)
HOAt (1)
HOAt (1)
3.7 (76.0)
1.3 (66.7)
4.7, 5.4 (84)
1.9, 3.8 (79.6)
4.9 (75.2)
2.7 (63.7)
4.6 (88.5)
2.8 (54.8)
3.9 (84.7)
1.4 (67)
3.7, 4.1 (70.4)
2,6-diEt-4-Me-Py (2)
2,6-diEt-4-Me-Py (2)
2,6-di-i-Pr-4-Me-Py (2)
2,6-di-i-Pr-4-Me-Py (2)
OHA (2)
OHA (2)
Me5Py (2)
Me5Py (2)
3.2, 3.6, 3.0 (100)
4.2, 4.8, 5.9 (84.2)
5.6, 4.2 (89.5)
8.8 (39.5)
9.7, 8.3, 10.6 (94.7)
5.1 (73.7)
8.5 (81.6)
HOAt (1)
HOAt (1)
HOAt (1)
HOAt (1)
6.1 (71)
Me5Py‚1/2H2O (2)
4.4 (90.9)
DMAN (2)
3.3, 3.8 (100)
4.1, 5.1 (89.5)
3.6, 3.0 (100)
3.0, 3.7 (84.2)
3.9 (81.6)
5.4, 5.6 (89.5)
2.1, 2.9 (92.9)
1.8 (61.9)
3.2, 3.5 (84.2)
2.2 (85)
2.7, 2.2 (100)
2.2 (85)
13.2 (100)
4.2 (80.7)
12.7 (94.8)
6.5 (93.5)
17.6 (94.7)
8.4 (84.2)
11.4 (93)
3.1 (79.6)
15.9 (74.7)
3.6 (81.6)
9.5, 9.8, 8.9 (97.4)
3.7, 3.5 (73.7)
DMAN (2)
HOAt (1)
HOAt (1)
HOAt (1)
HOAt (1)
HOAt (1)
HOAt (1)
HOAt (1)
HOAt (1)
DMAN (1)/TEMP (1)
DMAN (1)/TEMP (1)
iPr3N (1)/TEMP (1)
iPr3N (1)/TEMP (1)
DB(DMAP) (3)
DB(DMAP) (3)
DB(DMAP) (2)
DB(DMAP) (2)
DB(DMAP) (1)
DB(DMAP (1)
DB(DMAP) (0.5)
DB(DMAP) (0.5)
DB(DMAP) (4)
DB(DMAP) (4)
2,6-diMe-DMAP (2)
H-Pro-NH2 (1)e
H-Pro-NH2 (1)e
H-Pro-NH2 (2)e
1.7 (76.1)
0.9 (61.9)
1.7, 2.5 (70.8)
1.1, 1.0, 0.9 (35.4)
1.3 (66.4)
0.9 (44.3)
1.1 (36.9)
1.6 (52.6)
2.5, 1.9 (88.5)
1.0 (70.8)
2.3, 2.3 (99.5)
1.5 (84)
16.1 (79.7)
1.7 (88.5)
0.7 (60)
3.2 (88.4)
3.2 (73)
4.9 (73.5)
HOAt (1)
6.9 (59.8)
a
Standard protocol: test couplings were carried out by adding the coupling reagent (0.125 mmol) to a stirred and ice-cooled solution
of the protected dipeptide acid (0.125 mmol), H-Pro-NH2 (0.125 mmol), and the chosen base (0.25 mmol) in 1 mL of DMF. After 1 h the
ice bath was removed and stirring was continued for 2-3 h. In the case of reactions involving an amino acid ester hydrochloride instead
of H-Pro-NH2, 0.375 mmol of base was used and the volume of DMF was increased to 1.5 mL. The reaction mixture was diluted with 25
mL of EtOAc, and the organic layer was washed twice with 10-mL portions of 1 N HCl, twice with 10-mL portions of 10% NaHCO3, and
twice with 10-mL portions of saturated NaCl solution. The organic solution was dried (MgSO4) and filtered, the solvent removed in vacuo
at 30-35 °C, and the resulting residue dissolved in 50 mL of CH3CN for direct injection onto a C-18 column for HPLC analysis of yield
b
and extent of LDL isomer. The LLL and LDL forms of the test tripeptide have been described elsewhere.6 The number in parentheses
refers to the number of equivalents of base or additive used. In some cases the amount of base was less than that used in the standard
protocol. c The number in parentheses refers to the yield of tripeptide, generally based on HPLC examination. In many cases evaporation
of solvent gave the crude peptide as a solid for which the yield data based on weight and that obtained by HPLC analysis were roughly
d
comparable. Where more than one number is given for % LDL, the results are of separate runs. The first number given is from the run
for which the yield is cited. e In this case the figure in parentheses refers to the number of equivalents of H-Pro-NH2 used in excess of the
1 equiv needed for the acylation reaction.
relatively little epimerization in contrast to the case
involving DIEA (HAPyU/DIEA: 5.3% DL).4
Two final examples of the use of these new bases
involved stepwise solid phase coupling reactions. One
involved the solid phase coupling of Fmoc-Leu-OH onto
proline which had been attached to a PAL-PEG-polysty-
rene resin 15.
Following removal of the protected dipeptide amide
from the resin by means of trifluoroacetic acid, analysis
was carried out according to a method described previ-
ously for the coupling of Fmoc-Leu-F with proline amide
in solution.7 The results are presented in Table 3. In
this case, the new coupling reagent, tetramethylfluoro-
formamidinium hexafluorophosphate (TFFH),32 was ex-

2464 J . Org. Chem., Vol. 61, No. 7, 1996
Carpino et al.
Ta ble 2. Effect of Cou p lin g Rea gen t a n d Ba se on Loss of
Con figu r a tion a t Va lin e Du r in g F or m a tion of
of the collected data was found to be 0.03%. For the coupling
reactions, data for which are collected in Tables 1-4, some
scatter was noted in measuring the loss of configuration.
Where separate runs were carried out the data are listed in
Tables 1-4. Measured differences may be due to variable
amounts of water present in solvents or reagents, inability to
control the presence of basic impurities which might serve to
increase racemization or epimerization, and the presence of
extraneous materials which might overlap the regions of
interest in the HPLC traces. On the basis of experience in
carrying out several hundred such determinations, it appears
that closer correspondence is not to be expected when the
reactions are completed over a period of several months using
different samples of solvents, reagents, tertiary amines, etc.
a
Hexa p ep tid e 14 in DMF
coupling reagent
base
TEMP
DB(DMAP)
TEMP
yield, %
DL, %
HATU
HATU
HAPyU
HAPyU
81.2
91.2
84.6
86.2
0.5
0.4
0.4
0.5
DB(DMAP)
a
Tripeptide segments needed for the assembly of 14 were
obtained and coupling reactions performed as described in Table
5 of ref 4, footnote a.
Ta ble 3. Effect of Cou p lin g Rea gen t a n d Ba se on
Ra cem iza tion of F m oc-Leu -OH u p on Cou p lin g on to
H-P r o-P AL-P EG-P S in DMF
Tr iisop r op yla m in e. The preparation followed that de-
scribed.30 In the synthesis of the intermediate, R-(diisopropyl-
amino)propionitrile, it is important to avoid overheating in the
presence of water which is difficult to remove completely from
the crude product. In that case most of the propionitrile is
converted to lactonitrile. Indeed, lactonitrile, and i-Pr2NH
could be converted to R-(diisopropylamino)propionitrile in
38.5% yield by heating in benzene at reflux temperature with
removal of water as the azeotrope, followed by distillation at
70-79 °C (13 mm). Conversion of the nitrile to triisopropyl-
amine by the method described gave a yield of 55% (overall
yield 21.2%): bp 46-48 °C (14 mm) (lit.30b bp 47 °C (14 mm));
1H-NMR (CDCl3) δ 0.99 (d, 18); 3.12 (sept, 3).
coupling reagent
base
DIEA
TMP
TEMP
DL, %
TFFH
TFFH
TFFH
TFFH
0.8
<0.1
0.1
DB(DMAP)
0.2
a
Fmoc-Leu-OH was coupled to H-Pro-PAL-PEG-PS (0.2 mmol/
g) using 3 equiv of protected amino acid, 3 equiv of coupling
reagent, and 6 equiv of base. Solvent: DMF; preactivation time:
7 min; coupling time: 30 min. The protected dipeptide was
removed from the resin by cleavage with TFA for 1 h and the
analysis carried out according to the method described previously
in Table 2, footnotes a and b, of ref 7.
N-ter t-Bu tylm or p h olin e. The synthetic method followed
the described31 procedure which involved refluxing of bis(2-
chloroethyl) ether and tert-butylamine in ethanol for 7 days:
amined. Results are comparable to those observed with
HATU as coupling reagent (DIEA: 0.8% DL; TMP: 0.2%
DL). Thus, in spite of its significantly greater basicity,
DB(DMAP) is comparable to collidine in maintaining
configuration during activation of the carboxylic acid
residue.
A second example involved the stepwise assembly of
tripeptide 4. For practical solid phase syntheses quick
activation of the amino acid is essential. Collidine is
insufficiently basic for such purposes, leading in the case
of 4 (TFFH as coupling reagent) to contamination by
39.5% of the des-Val product under conditions where the
more basic amine, DB(DMAP), provides for a tripeptide
of good quality.
1
yield 31.8%; bp 175-176 °C (lit.31 bp 174-176 °C); H-NMR
(CDCl3) δ 1.06 (s, 9), 2.54 (t, 4), 3.72 (t, 4).
4-(N,N-Dim eth yla m in o)-2,6-d i-ter t-bu tylp yr id in e (13a ).
(A) F r om P in a colon e. The method followed that of Potts
and Winslow25 except that dimethylamine was used instead
of diethylamine for replacement of the methylthio group.
Some discrepancies with the literature description were noted.
Pinacolone (5 g, 0.05 mol) dissolved in 200 mL of anhydrous
THF was stirred vigorously with 11 g (0.1 mol) of KO-t-Bu at
room temperature for 1 h, and to the resulting orange-colored
reaction mixture was added 10 g (0.05 mol) of 5,5-bis-
(methylthio)-2,2-dimethyl-4-penten-3-one. Stirring was con-
tinued for 16 h, and the resulting crude orange crystalline
potassium salt was filtered and used in the next step without
purification. The crude salt (21 g, 0.071 mole), 470 mL of 50%
aqueous HBF4, and 80 mL of CH2Cl2 were stirred together at
reflux temperature for 3 h. The aqueous layer was extracted
with CHCl3 (3 × 50 mL). The combined organic layers were
dried (MgSO4), solvent evaporated, and the residue dissolved
in a small amount of acetone. Addition of ether gave 5.7 g
(24.6%) of 2,6-di-tert-butyl-4-(methylthio)pyrrylium tetrafluo-
roborate as a colorless crystalline mass, mp 113-114 °C. To
a stirred solution prepared from 1.5-2 g (0.033-0.044 mole)
of anhydrous dimethylamine in 50 mL of dry MeOH was added
Exp er im en ta l Section
Gen er a l. Coupling reactions were carried out as described
in footnote a of Table 1. Bases were purified as noted in
footnote a, Table 2, ref 4. Fmoc-PAL-PEG-PS resin, HATU,
and HAPyU were obtained from Perseptive Biosystems, Inc.
All model peptides have been reported previously. For Z-Phe-
Val-Pro-NH2,6 Z-Gly-Phe-Pro-NH2,6 and Z-Phe-Val-Ala-OMe4
HPLC analysis was carried out on a C-18, 4 µm Waters
Novapak column, 3.9 × 150 mm, flow rate 1 mL/min, detection
at 220 nm with a Waters 996 PDA detector using a linear
gradient 25-50% CH3CN/H2O 0.1% TFA. For Z-Gly-Phe-Val-
3.5 g (0.01 mole) of the crystalline pyrrylium salt at
a
temperature of -30 °C. The reaction vessel was then closed,
the dry ice/acetone bath removed, and the mixture stirred at
room temperature for 24 h. Removal of MeOH in vacuo gave
an orange crystalline residue which was triturated with ether.
Filtration and washing with ether gave 3.4 g (97.4%) of the
aminopyrrylium salt as colorless crystals, mp 147-148 °C. An
analytical sample was obtained by crystallization from acetone-
ether: 1H NMR (CDCl3) δ 6.67 (s, 2), 3.45 (s, 6), 1.39 (s, 18).
7
OMe6 and Z-Phg-Pro-NH2 all parameters were the same
except that for separation isocratic systems of 40% CH3CN/
60% H2O (0.1% TFA) and 25% CH3CN/75% H2O (0.1% TFA),
respectively, were used. For retention times and data on
characterization of the model peptides see the references cited.
As a check on the ability of the system to quantify the minor
diastereomer a method similar to that described in footnote
a, Table 2, ref 4 was used. The system Z-Phe-Val-Pro-NH2
was selected, and authentic samples of the diastereomeric
mixture were weighed out and analyzed using the PDA
detector: % LDL form via weight ratio/HPLC-determined figure
50.54/50.95; 1.19/1.20; 0.633/0.63; 0.477/0.49; 0.401/0.40; 0.358/
0.36; 0.318/0.33; 0.256/0.26. For the system containing 0.401%
by weight of the LDL isomer, for six runs the standard deviation
Anal. Calcd for C15H26BF4NO: C, 55.73; H, 8.10; N, 4.33.
Found: C, 55.90; H, 8.18; N, 4.39.
A solution of 7.5 g (0.023 mole) of the pyrrylium salt in 200
mL of MeOH was added dropwise to a stirred solution of 225
mL of 30% ammonium hydroxide with cooling in an ice bath.
The resulting mixture was stirred at room temperature for
12 h and the crystalline product collected by filtration. Re-
crystallization from MeOH-H2O gave 5 g (92.5%) of the amine
as colorless needles: mp 65-66 °C (lit.26 mp 65-66 °C); 1H
NMR (CDCl3) δ 1.33 (s, 18), 2.97 (s, 6), 6.38 (s, 2). The overall
yield from pinacolone was 17.8%.
(32) Carpino, L. A.; El-Faham, A. J . Am. Chem. Soc. 1995, 117,
5401-5402.

Peptide Coupling in the Presence of Tertiary Amines
J . Org. Chem., Vol. 61, No. 7, 1996 2465
(B) F r om DMAP . The described method26 was followed
using tert-butyllithium and led to a yield of 37.5%. The
physical properties of the sample agreed with those described
above.
and 10 mmol of HBTU in 33 mL of CH2Cl2 for 2 h. The
resulting resin was washed with CH2Cl2 (5 × 20 mL), 95%
ethanol (1 × 20 mL), and anhydrous ether (2 × 20 mL) and
dried in vacuo, and 200-mg portions were used in the sequen-
tial coupling via Fmoc-Val-OH and Z-Phe-OH. The resin (200
mg) was washed with DMF (5 × 5 mL), deprotected with 20%
piperidine in DMF (5 mL) for 7 min, and washed with DMF
(5 × 5 mL). Preactivation was carried out for 7 min using
40.7 mg of Fmoc-Val-OH and 56.3 mg of DB(DMAP) or 31.8
µL of TMP and 31.7 mg of TFFH in 0.4 mL of DMF. Following
the requisite preactivation period (7 min), the solution of the
activated amino acid was added to the resin and reaction
allowed to proceed for 30 min. Washing and deblocking of the
Fmoc group was followed by an analogous coupling step with
Z-Phe-OH (39.5 mg). After the second coupling the resin was
washed with DMF (5 × 5 mL), CH2Cl2 (5 × 5 mL), 95% EtOH
(1 × 5 mL), and anhydrous ether (2 × 5 mL) and dried for 30
min in vacuo. The dried resin was treated with TFA (4 mL)
for 1 h, filtered, and washed on the filter with 2 mL of TFA
and CH2Cl2 (2 × 5 mL). The combined filtrates were distilled
in vacuo at room temperature, and the resulting residue was
dissolved in 4 mL of CH3CN and injected directly onto an
HPLC column for analysis. The HPLC traces are shown in
Figure 1 (supporting information). Yields as estimated by the
intensities of the HPLC curves were DB(DMAP), 80%, and
TMP, 65%. The HPLC conditions were as described6 and
showed that for DB(DMAP) and TMP the amount of the des-
Val product (Z-Phe-Pro-NH2, tR ) 12.4 min, confirmed by
authentic synthesis) amounted to 3.1% and 35.9%, respec-
tively. Scanning the same HPLC curve for the presence of
the LDL or DLL diastereomers (tR 16.9 and 16.2 min, respec-
tively) showed that no significant stereomutation had occurred
at either valine or phenylalanine (<0.2%) during the assembly
of 4. The coupling of Fmoc-Leu-OH to proline amide was
examined in the same manner. See Table 3.
N,N,2,6-Tetr a m eth yl-4-a m in op yr id in e (2,6-Dim eth yl-
DMAP , 13c). The method followed the latter steps of the
procedure described under method A for DB(DMAP) except
that the 2,6-dimethyl-4-methoxypyrrylium iodide (45%, mp 114
°C (lit.33 mp 110 °C) was obtained from 2,6-dimethyl-4-
pyrone.34 Reaction of the pyrrylium iodide with dimethyl-
amine gave 2,6-dimethyl-4-(dimethylamino)pyrrylium iodide
(89.6%, mp 220 °C dec): 1H NMR (CDCl3) δ 2.6 (s, 6), 3.55 (s,
6), 7.25 (s, 2). Previously, the corresponding perchlorate salt
(mp 184-185 °C dec) had been characterized.33 Finally,
treatment of the 4-(dimethylamino)pyrrylium salt with NH3/
35
(NH4)2CO3 for 15 min gave 13c in 78.4% yield, mp 38-39
°C, after recrystallization from hexane: 1H NMR (CDCl3) δ
2.41 (s, 6), 2.94 (s, 6), 6.21 (s, 2). Operations with 2,6-dimethyl-
DMAP had to be carried out in a dry glove bag under N2 due
to its extreme hygroscopic nature.
Anal. Calcd for C9H14N2: C, 71.95; H, 9.39; N, 18.65.
Found: C, 71.91; H, 9.41; N, 18.61.
2,3,4,5,6-P en ta m eth ylp yr id in e (9). The literature meth-
od18 was followed exactly: yield 12.1%; bp 98-100 °C (10 mm);
1H-NMR (CDCl3) δ 2.17 (s, 9), 2.48 (s, 6). The corresponding
1
half hydrate had: mp 49-50 °C; H-NMR (CDCl3) δ 2.18 (s,
9), 2.46 (s, 6).
2,6-Dieth yl-4-m eth ylp yr id in e. The literature method8
1
was followed exactly: yield 17%; bp 94-95 °C (20 mm); H-
NMR (CDCl3) δ 1.27 (t, 6), 2.28 (s, 3), 2.76 (q, 4), 6.79 (s, 2).
2,6-Diisopr opyl-4-m eth ylpyr idin e. The literature method8
was followed exactly: yield 22%; bp 105 °C (20 mm); 1H-NMR
(CDCl3) δ 1.27 (d, 12), 2.29 (s, 3), 2.99 (sept, 2), 6.79 (s, 2).
2,3,5,6-Tetr a m eth ylp yr id in e (7). The method followed
exactly that described in the literature.12,36 Although a mul-
tistep procedure, each step works well. The various interme-
diates were characterized by 1H NMR analysis: (1) 2,6-
Dimethyl-3,5-dicarbethoxypyridine (CDCl3) δ 1.41 (t, 6), 2.84
(s, 6), 4.40 (q, 4), 8.68 (s, 1); (2) 2,6-dimethyl-3,5-bis(hydroxy-
methyl)pyridine (CDCl3) δ 2.54 (s, 6), 4.73 (s, 4), 7.37 (s, 1);
(3) 2,6-dimethyl-3,5-bis(chloromethyl)pyridine (CDCl3) δ 2.61
(s, 6), 4.58 (s, 4), 7.56 (s, 1); (4) TEMP (CDCl3) δ 2.19 (s, 6),
2.42 (s, 6); 7.1 (s, 1), 28.9% (overall from ethyl acetoacetate);
mp 75-76 °C (lit.12 mp 76 °C).
Ack n ow led gm en t. We are indebted to the National
Science Foundation (NSF CHE-9003192), the National
Institutes of Health (GM-09706), and Perseptive Bio-
systems, Inc., for support of this work. Dr. Eberhard
Krause is thanked for discussion and evaluation of
methods of ensuring accuracy in the HPLC technique
of determining small amounts of one diastereomer in
the presence of large amounts of another. The National
Science Foundation is also thanked for grants used to
purchase the high-field NMR spectrometers used in this
research.
Assem bly of Z-P h e-Va l-P r o-NH2 a n d F m oc-Leu -P r o-
NH2 by Step w ise Solid P h a se Cou p lin g. A 10-g sample
of amide resin (Fmoc-PAL-PEG-PS; 0.2 mmol/g) was washed
with DMF (5 × 20 mL), deprotected with 20% piperidine in
DMF (20 mL) for 7 min, washed with DMF (5 × 20 mL), and
then treated with 10 mmol of Fmoc-Pro-OH, 20 mmol of DIEA,
Su p p or tin g In for m a tion Ava ila ble: Additional data
comparable to that of Table 1 detailing the effect of water, base,
additive, and coupling reagent on model peptide coupling and
HPLC data demonstrating rapid activation of a protected
amino acid by TFFH in the presence of DB(DMAP) (6 pages).
This material is contained in libraries on microfiche, im-
mediately follows this article in the microfilm version of the
journal, and can be ordered from the ACS; see any current
masthead page for ordering information.
(33) King, L. C.; Ozog, F. J . J . Org. Chem. 1955, 20, 448.
(34) King, L. C.; Ozog, F. J .; Moffat, J . J . Am. Chem. Soc. 1951, 73,
300.
(35) The conditions described for the synthesis of the 4-piperidi-
nopyrrylium salt and its reaction with ammonia/ammonium carbonate
were adopted. See: Anker, R. M.; Cook. A. H. J . Chem. Soc. 1946, 117.
(36) For the precursor (1,4-dihydro-3,5-dicarbethoxy-2,6-dimeth-
ylpyridine), see: Singer, A.; McElvain, S. M. In Organic Syntheses;
Blatt, A. H., Ed.; Wiley: New York; 1943; Collect. Vol. 2, p 214.
J O950912X