An application of borane as a protecting group for pyridine

Article abstract of DOI:10.1021/jo801040m

(Chemical Equation Presented) Efforts to couple 4 with 12 employing base mediation are problematic due to the formation of 6. To circumvent this issue, 12 was converted to the pyridine borane complex (13). Alkylation of 4 with 13 provided 3 after removal of the borane under acidic conditions and saponification of the ester.

Full text of DOI:10.1021/jo801040m

SCHEME 1. Cyclization of Heterocyclic Intermediates
An Application of Borane As a Protecting Group
for Pyridine
Matthew A. Zajac
GlaxoSmithKline, Synthetic Chemistry 709 Swedeland Road,
P.O. Box 1539, King of Prussia, PennsylVania 19406
SCHEME 2
matthew.a.zajac@gsk.com
ReceiVed May 20, 2008
has required protecting the hetrocyclic nitrogen as an N-oxide
or completely excluding compounds such as 1 from the synthetic
route. Many times these types of intermediates (1) facilitate the
most direct pathway to the target compound. Because of the
problems associated with removing an N-oxide and handling
the intermediates on scale, we sought other potential solutions.
This problem was encountered during the development of a
scalable synthesis to the vitronectin inhibitor 3 (Scheme 2). A
straightforward disconnection of 3 reveals previously prepared
3a,b and pyridyl side chain 5.2b,4 However, as mentioned above,
Efforts to couple 4 with 12 employing base mediation are
problematic due to the formation of 6. To circumvent this
issue, 12 was converted to the pyridine borane complex (13).
Alkylation of 4 with 13 provided 3 after removal of the
borane under acidic conditions and saponification of the ester.
4
conversion of the hydroxyl group of 5 into a leaving group5
would lead to the formation of 6.2b,6 To make realistic use of
this route and 5, the lone pair of the pyridine moiety would
need to be occupied in some fashion to prevent cyclization.
A first generation synthesis3a,b of 3, partly shown in Scheme
3, relied on this disconnection and made use of a pyridine
N-oxide intermediate. Two alkylating agents were used with 4:
7 under Mitsunobu conditions and 8 using sodium hydroxide.3a,b
Unfortunately, the synthesis of both 7 and 8 required the use
of commercially available but expensive 2-chloropyridine
N-oxide. As an additional problem, this material was thermally
unstable, potentially leading to safety issues.7 Following alky-
lation, N-oxide 9 was treated with zinc dust to remove the oxide
protecting group and regain the free pyridine. This step proved
difficult to scale due to the density of zinc dust; variable results
were often obtained due to inconsistency in stirring, which
During the course of developing scalable syntheses to certain
pharmaceutical targets, we identified a problem installing
specific heterocyclic side chains containing a leaving group.
Compounds of form 1 (Scheme 1), where a nucleophilic nitrogen
is contained within an electron-rich heterocycle proximal to a
leaving group, would cyclize in many cases to give 2, which
does not function as an alkylating agent.1,2 Avoiding this issue
(1) Five membered systems: (a) Birman, V. B.; Li, X.; Jiang, H.; Uffman,
E. W. Tetrahedron 2006, 62, 285. (b) Copp, F. C.; Timmis, G. M. J. Chem.
Soc. 1955, 2021.
(2) Six membered system: (a) Klauschenz, E.; Hagan, V.; Wiesner, B.; Hagan,
A.; Reck, G.; Krause, E. G. Eur. J. Med. Chem. 1994, 29, 175. (b) Mandereau,
J.; Xuong, E. N.T.; Reynaud, P. Eur. J. Med. Chem. 1974, 9, 344.
(3) (a) Miller, W. H.; Alberts, D. P.; Bhatnager, P. K.; Bondinell, W. E.;
Callahan, J. F.; Calvo, R. R.; Cousins, R. D.; Erhard, K. F.; Heerding, D. A.;
Keenan, R. M.; Kwon, C.; Manley, P. J.; Newlander, K. A.; Ross, S. T.; Samanen,
J. M.; Uzinskas, I. N.; Venslavsky, J. W.; Yuan, C. C.-K.; Haltiwanger, R. C.;
Gowen, M.; Hwang, S.-M.; James, I. E.; Lark, M. W.; Rieman, D. J.; Glomic,
J. C.; Stroup, G. B.; Azzarano, L. M.; Salyers, K. L.; Smith, B. R.; Ward, K. W.;
Johanson, K. O.; Huffman, W. F. J. Med. Chem. 2000, 43, 22. (b) Wallace,
M. D.; McGuire, M. A.; Yu, M. S.; Goldfinger, L.; Liu, L.; Dai, W.; Shilcrat,
S. Org. Process Res. DeV. 2004, 8, 738.
(4) Tupin, T.; Raynaud, P.; Delaby, R. Bull. Soc. Chim. Fr. 1957, 721.
(5) In our hands, Mitsunobu conditions with 5, as in the following references,
provided 6 with only trace amounts of alkylation product:(a) Leonard, K.; Pan,
W.; Anaclerio, B.; Gushue, J. M.; Guo, Z.; DesJarlais, R. L.; Chaikin, M. A.;
Lattanze, J.; Crysler, C.; Manthey, C. L.; Tomczuk, B. E.; Marugan, J. J. Bioorg.
Med. Chem. Lett. 2005, 15, 2679. (b) Marugan, J. J.; Manthey, C.; Anaclerio,
B.; Lafrance, L.; Lu, T.; Markotan, T.; Leonard, K. A.; Crysler, C.; Eisennagel,
S.; Dasgupta, M.; Tomczuk, B. J. Med. Chem. 2005, 48, 926. Additionally, the
authors of the above paper include 1H NMR data of 5 that do not match our
data.
(6) It is interesting to note that treatment of 4 with 6 under acidic or basic
conditions gives no trace of desired product. For 6 see ref 2b.
10.1021/jo801040m CCC: $40.75
Published on Web 08/05/2008
2008 American Chemical Society
J. Org. Chem. 2008, 73, 6899–6901 6899

SCHEME 3. Previous N-Oxide Route3a,b
SCHEME 4. Borane Alkylating Agent
differed with reactor configuration. Hydrogenation conditions
were also screened (Pt or Pd/C heterogeneous catalysts) but gave
either no desired product or afforded byproducts.
SCHEME 5. Decomposition of Unprotected Pyridine
This opportunity was used to develop alternative blocking
groups that would avoid the three major problems with this
route: (1) 2-chloropyridine N-oxide is expensive and supplied
as an aqueous solution on scale, requiring a lengthy azeotropic
distillation before use, (2) zinc dust reduction of the N-oxide is
capricious on scale, and (3) 2-chloropyridine N-oxide is ther-
mally unstable.
It was hypothesized that borane might be used as a protecting
group instead of the N-oxide. Several precedents exist which
illustrate the strength of the borane-heterocycle bond. For
instance, oxazole-borane complexes are bench-stable and
require acidic conditions or Pd treatment for removal.8,9 These
complexes are completely stable to chromatography, air, and
moisture. Additionally, 4-dimethylaminopyridine (DMAP) quickly
and quantitatively forms a complex with borane that exhibits
stability similar to that of oxazole-borane complexes.10 We
felt that the system at hand should behave similarly to these
examples and form stable borane complexes.11 However, an
example where this type of complex thwarted the nucleophilicity
of a heteroaromatic system could not be found. On the basis of
the relevant literature examples of a DMAP borane complex10
and a 2-aminopyridine cyanoborane complex,11f the borane
would be expected to reside on the pyridine nitrogen.12
To test this strategy, a borane protected version of 7 or 8
was needed (Scheme 4). To this end, 10 was heated in
3-aminopropanol to give 5 in 64% yield after crystallization.
Direct treatment of 5 with borane dimethylsulfide initially gave
a complex mixture;13 however, after aqueous workup only 5
and 11 were obtained (∼1:1 ratio). Alcohols 5 and 11 proved
somewhat difficult to separate and attempts at reaction optimiza-
tion did little to improve the ratio. Alternatively, 5 was treated
with thionyl bromide, successfully giving 12 in situ.
With 12 available, we studied its stability upon addition of
iPr2NEt (Scheme 5). When 12 was treated with 2 equiv of base
at -55 °C, we observed <1% decomposition after 2 h. At -45
°C, 5% loss was observed after 1 h whereas at -20 °C, 68%
had cyclized after 1 h. The decomposition of 12 at -20 °C is
shown graphically in Scheme 5. Additionally, the exponential
decay equation that fits the data is given. This gave us a clear
indication that cyclization of 12 would occur before alkylation
of 4 under basic conditions since the reaction of 4 with 8 is
relatively slow at room temperature. In the formation of 13 from
12, basification of 12 must occur either in the presence of the
borane source or at a temperature <-55 °C to avoid the
formation of 6.
Using this knowledge, we developed two procedures to
produce electrophile 13 (Scheme 4).12 Freebasing at tempera-
tures between -60 and -78 °C cleanly provided 13 after
treatment with borane dimethylsulfide. Alternately, the reaction
could be accomplished at -20 °C without forming appreciable
amounts of 6 by adding excess iPr2NEt to a solution of 12 and
borane dimethylsulfide. It appears that the iPr2NEt reacted faster
with 12 than the borane source and that the freebase of 12
reacted preferentially with the borane source rather than
cyclizing to 6. As we had hoped, isolated 13 was crystalline
and exhibited similar stability to water and oxygen as the
literature heterocycle-borane complexes.
(7) Shilcrat, S. Unpublished results, HCl salt: An exotherm was observed in
the range of 134.1-185.4 °C with an onset temperature of 136.1 °C and a heat
of reaction of-746.7 J/g. Freebase: Two exotherms were observed with the
first in the range of 42.7-108.1 °C with an onset of 46.6 °C and a heat of reaction
of-867.6 J/g. The second exotherm was in the range of 125.9-188 °C with an
onset of 146.6 °C and a heat of reaction of-6672.4 J/g.
(8) Monahan, S. D.; Vedejs, E. J. Org. Chem. 1996, 61, 5192.
(9) Zajac, M. A.; Vedejs, E. Org. Lett. 2001, 3, 2451.
(10) Shapland, P.; Vedejs, E. J. Org. Chem. 2006, 71, 6666.
(11) For examples of nonheteroaromatic amine borane complexes see:(a)
Schwartz, M. A.; Rose, B. F.; Vishnuvajjala, B. J. Am. Chem. Soc. 1973, 95,
612. (b) White, J. D.; Amedio, J. C.; Gut, S.; Jayasinghe, L. J. Org. Chem.
1989, 54, 4268. (c) Ferey, V.; Vedrenne, P.; Toupet, L.; Le Gall, T.; Mioskowski,
C. J. Org. Chem. 1996, 61, 7244. (d) Carboni, B.; Monnier, L. Tetrahedron
1999, 55, 1197. (e) Blakemorene, P. R.; Kim, S. K.; Schulze, V. K.; White,
J. D.; Yokochi, A. F. T. J. Chem. Soc., Perkin Trans. I 2001, 1831. For an
example of an R-amino pyridine borane complex see: (f) Tang, P. W.; Williams,
J. M. Carbohydr. Res. 1985, 136, 259.
Next, alkylation of 4 with 13 (Scheme 6) was carried out in
refluxing MeCN with K2CO3 and a catalytic amount of
tetrabutylammonium iodide.6 The resulting crude product (14)
after aqueous workup was first treated with 1.1 equiv of HCl
(aqueous, concentrated) to remove the borane protecting group
(15) and then with sodium hydroxide to saponify the ester. After
adjusting the pH to 5-5.5, 3 crystallized out of solution in 83%
(12) See the Supporting Information for spectral data of 13 and 14.
6900 J. Org. Chem. Vol. 73, No. 17, 2008

SCHEME 6. Preparation of 3
filtration. Wash the solid twice with cold (0-5 °C) methanol (2 ×
1.5 L). The wet product is dried at 20-25 °C under reduced
pressure. This afforded 676 g of white solid (13, 75% yield) of
sufficient purity for the next step. 1H NMR (400 MHz, THF-d8) δ
(ppm) 8.12 (1H, d, J ) 5.9 Hz), 7.67 (1H, dd, J ) 7.3, 8.8 Hz),
6.82 (1H, d, J ) 8.8 Hz), 6.65-6.53 (2H, m), 3.61-3.50 (4H, m),
2.60-1.75 (3H, br m), 2.19 (2H, m). 13C NMR (100 MHz, THF-
d8) δ (ppm) 156.0, 147.2, 140.7, 112.0, 108.1, 41.5, 32.8, 31.3.
Found (LR-ESI) 227 and 229 [M - H]+, C8H13BBrN2 requires
227 and 229.
Pyridine 14. Charge 4 (600 g, 1.81 mol, 1.0 equiv), potassium
carbonate (1.0 kg, 7.24 mol, 4.0 equiv), tetrabutylammonium iodide
(33.4 g, 90.6 mmol, 0.05 equiv), 13 (498 g, 2.17 mol, 1.2 equiv),
and acetonitrile (3.0 L) at 18-23 °C to the reactor. Stir the mixture
vigorously and heat to 49-52 °C under a nitrogen atmosphere. After
the reaction is complete (typical hold is 16-24 h), cool the mixture
to 18-23 °C and concentrate to 2.0-2.5 L. Add TBME (6.0 L)
and mix thoroughly. Add water (6.0 L) and mix thoroughly. Collect
the TBME layer and back extract the water layer with TBME (6.0
L). Wash the pooled TBME with water (3.0 L) and concentrate
the TBME layer to 2.0-2.5 L. Add methanol (6.0 L) and
concentrate to 2.0-2.5 L. This solution is then used directly in the
next step. For reference purposes, an analytically pure sample was
obtained by concentrating a portion of this solution to near dryness
and subjecting the residue to silca gel column chromatography
eluting with 0 to 100% EtOAc/hexanes. 1H NMR (400 MHz,
toluene-d8) δ (ppm) 8.04 (1H, d, J ) 5.7 Hz), 7.07 (1H, dd, J )
7.7, 8.7 Hz), 6.78-6.70 (2H, m), 6.67-6.60 (2H, m), 6.15 (1H, d,
J ) 8.7 Hz), 6.09 (1H, dd, J ) 5.7, 7.7 Hz), 4.90 (1H, d, J ) 16.8
Hz), 4.05-3.90 (1H, m), 3.80-3.60 (4H, m), 3.52 (1H, m), 3.44
(3H, s), 3.06-2.98 (2H, m), 2.89 (1H, dd, J ) 8.8, 16.8 Hz),
3.00-2.50 (5H, m), 2.17 (1H, dd, J ) 4.9, 16.8 Hz), 1.73 (2H, m).
13C NMR (100 MHz, toluene-d8) δ (ppm) 174.7, 172.3, 157.0,
155.2, 146.5, 139.8, 135.1, 131.9, 128.7, 125.1 (q, J ) 280.7 Hz),
115.4, 114.2, 111.4, 107.2, 65.8, 53.0, 51.3, 47.4 (q, J ) 33.2 Hz),
40.2, 37.0, 36.7, 34.7, 28.8. Found (LR-ESI) 478 [M - H]+,
C23H28BF3N3O4 requires 478.
Pyridine 3. Dilute the methanol solution of 14 (100% assumed
yield on the previous stage) up to 6.0 L with methanol. Cool to
0-5 °C and add concentrated aqueous hydrochloric acid (181 mL,
2.17 mol, 1.2 equiv, 12 M solution) while maintaining the
temperature at <15 °C. Warm the mixture to 20-25 °C and hold
for up to 3 h. After the reaction is complete, cool to 0-5 °C and
add 3 M NaOH (4.23 L, 7.0 equiv) while maintaining the
temperature at <15 °C. Warm the mixture to 20-25 °C and hold
for up to 3 h. After the reaction is complete, add enough
concentrated hydrochloric acid (∼4.8 equiv) to bring the pH to
6.1-6.5 while maintaining the temperature at <25 °C. Hold for
16-24 h at 18-25 °C, during which time 3 precipitates out of
solution. Isolate 3 by filtration and wash with water (2 × 4.3 L).
Dry the solid product in a vacuum oven at 50-60 °C for up to
12 h. This afforded 683 g of 3 as an off-white solid (83% yield
from 4). All spectral data matched the previously reported data.3a,b
yield from 4. This entire sequence from 10 was completed on
a kilogram scale, providing 1.32 kg of 3 in two runs.14 In
conclusion, we have developed a procedure that prevents the
cyclization of compound 12 using borane as a blocking group.
Borane is both easy and economical to install and can be readily
cleaved by using an acid workup.
Experimental Section
(3-Bromopropyl)(1-boropyridin-2-yl)amine (13). Charge 5
(600 g, 3.94 mol, 1.0 equiv) followed by methylene chloride (3.0
L). Mix the contents of the reactor thoroughly for 10 min. Cool
the mixture to 0-5 °C. Add thionyl bromide (820 g, 3.94 mol, 1.0
equiv) slowly over 5-25 min to keep the internal temperature <15
°C. Warm the yellow mixture to 20-25 °C and hold for 1.5-2 h
under a nitrogen atmosphere. After the reaction is complete,
concentrate the solution in vacuo to 2.0-2.5 L with the jacket
temperature at ca. 40 °C and add methylene chloride (3.0 L). Cool
the solution to -60 to -70 °C and add diisopropylethylamine (561
g, 756 mL, 4.34 mol, 1.1 equiv) directly to the reaction mixture
while maintaining the internal temperature below -55 °C. After
addition is complete, borane dimethylsulfide complex (330 g, 412
mL, 4.34 mol, 1.1 equiv) is added directly to the reaction mixture
over <25 min. The temperature of the mixture is kept below -50
°C during addition. Warm the solution to 10-15 °C and add a
saturated aqueous solution of sodium bicarbonate (3.0 L) and water
(3.0 L), mix thoroughly. Collect the organic layer and extract the
aqueous layer with methylene chloride (3.0 L). Concentrate the
combined organic layers in vacuo to 2.0-2.5 L. Add methanol (4.2
L) to dissolve/suspend the mixture and stir for 15 min. Concentrate
the mixture to 3.0-3.5 L to remove any residual methylene chloride.
Cool the mixture to 5-10 °C and isolate the resulting solid by
(13) It seems that the rate of reaction of borane at the pyridine nitrogen
to give 11 is about as fast as the reaction with the hydroxyl function of 5. If
the reaction occurs at the hydroxy site, either or both of the following two
compounds could be present and would give 5 after aqueous workup:
Acknowledgment. The author thanks Alan Freyer for taking
NMR spectra of 5, 13, and 14. Assistance from the Analytical
Sciences group at GlaxoSmithKline has been appreciated.
Supporting Information Available: Experimental proce-
dures and full spectroscopic data for all new compounds. This
material is available free of charge via the Internet at
http://pubs.acs.org.
(14) The best cost comparison that can be made between the previous route
(Scheme 3) and the new borane route is in the price difference of the starting
materials on a kilogram scale. Both routes use 4, thionyl bromide, and
3-aminopropanol, which can be left out of consideration. The 2007 pricing for
2-chloropyridine N-oxide is about $400/mol whereas the price of 2-chloropyridine
is about $5/mol and that of borane dimethylsulfide is about $21/mol.
JO801040M
J. Org. Chem. Vol. 73, No. 17, 2008 6901