Phosphonate-free phosphorylation of alcohols using bis-(tert-butyl) phosphoramidite with imidazole·hydrochloride and imidazole as the activator

Article abstract of DOI:10.1016/j.tetlet.2011.05.020

A variety of commercially available alcohols were converted to their bis-(tert-butyl) protected mono-phosphate pro-drugs. The improved procedure uses the unique combination of imidazole·hydrochloride and imidazole as the activator to suppress the formatio

Full text of DOI:10.1016/j.tetlet.2011.05.020

Tetrahedron Letters 52 (2011) 3625–3629
Contents lists available at ScienceDirect
Tetrahedron Letters
journal homepage: www.elsevier.com/locate/tetlet
Phosphonate-free phosphorylation of alcohols using bis-(tert-butyl)
phosphoramidite with imidazoleÁhydrochloride and imidazole as the activator
⇑
Bruno Haché , Lisa Brett, Sagar Shakya
Array BioPharma Inc., 2620 Trade Centre Ave, Longmont, CO 80503, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
A variety of commercially available alcohols were converted to their bis-(tert-butyl) protected mono-
phosphate pro-drugs. The improved procedure uses the unique combination of imidazoleÁhydrochloride
and imidazole as the activator to suppress the formation of undesired phosphonate by-product fre-
quently encountered with the bis-(tert-butyl) phosphoramidite reagents. The new activation procedure
eliminates the need to use excess reagents.
Received 8 April 2011
Revised 28 April 2011
Accepted 5 May 2011
Available online 14 May 2011
Ó 2011 Elsevier Ltd. All rights reserved.
1. Discussion
The benzyl groups can also be cleaved under TMSBr and TMSI con-
ditions5 to name a few. Under these conditions, concomitant for-
Mono-alkyl phosphates are recognized to play important roles
in many biological processes.1a Derivatization of an alcohol moiety
as the mono-phosphate esters have been employed as pro-drugs
since they are readily ionizable, significantly increase aqueous sol-
ubility,1b–d are sufficiently stable in both the solid state and in
aqueous solution for drug product manipulation and are readily
cleaved in vivo by nonselective alkaline phosphatases, thus releas-
ing the parent drug in vivo.1d Early synthetic methods involved the
electrophilic dialkyl phosphoryl halides in their P(V) oxidation
state (Scheme 1).
Although these methods gave direct access to the desired pro-
tected phosphates, many of the phosphoryl halide methods gave
complex mixtures2a and/or low yields,2b or required toxic1c or exo-
tic reagents.2c Protecting groups other than benzyl, tert-butyl or
trimethylsilyl-ethoxy gave incomplete hydrolysis,3c probably due
to non-regioselective hydrolysis pathways, thus low to moderate
yields were observed at the final hydrolysis stage. Bis-benzyl and
bis-tert-butyl phosphates were prepared using their corresponding
phosphorohalidates (Scheme 1, R = t-Bu) but suffered from insta-
bility and low electrophilicity.4
mation of noxious benzyl halides, a toxic by-product, also makes
this approach less attractive for large scale. The trimethylsilyl-eth-
oxy protecting group has been widely used,6c especially for solid
phase synthesis. However, the TFA or TBAF cleavage conditions
makes this option less attractive in the context of large-scale prep-
aration of pharmaceutical phosphate pro-drugs.
The bis-tert-butyl group, since it undergoes facile acidic cleav-
age, may be a superior candidate at circumventing the problems
associated with the above protecting groups. The well known
bis-tert-butyl N,N-dialkyl phosphoramidites have been extensively
used for this transformation3c and the resulting phosphates have
been easily deprotected to their mono-phosphates in high yields
using TFA,7a–c HCl–dioxane,3c TMSCl7d and TMSCl/NaI1c conditions.
However, it has been documented that the bis-tert-butyl phospho-
ramidite gives rise to an undesired phosphonate by-product6,8
(Scheme 3).
The suppression of the phosphonate by-product has been repor-
ted6a by the use of excess phosphoramidite reagent (10–20 equiv),
an unacceptable solution for large scale chemistry. For safety rea-
sons, we modified the mode of addition of the original procedure3
and we experienced increased levels of the phosphonate by-prod-
uct with increasing scale when the phorphoramidite was added
last (not shown). After investigation and isolation of the phospho-
nate by-product, we suspected that the low pKa of 1H-tetrazole
(pKa = 4.89)9 promoted the cleavage of one of the tert-butyl groups
Johns et al.3 reported the phosphorylation using bis-alkyl phor-
phoramidites for the phosphitylation of alcohols. The in situ gener-
ated phosphites were then oxidized in the same pot to the desired
phosphates (Scheme 2).
In the context of large-scale preparation of phosphate pro-drug
development candidates, careful consideration must be given to
the choice of the protecting group. The Pd-catalyzed hydrogenoly-
sis of benzyl groups is less desirable due to the difficulties associ-
ated with the removal of Pd residues and the incompatibility with
halogenated aromatics under standard hydrogenolysis conditions.
O
P
OR'
Hal
O
P
O
P
Hydrolysis
OR'
Base/solvent
O
OR'
OR'
OH
OH
R
H
O
O
R
R
⇑
Corresponding author. Tel.: +1 303 386 1209.
E-mail address: bruno.hache@arraybiopharma.com (B. Haché).
Scheme 1.
0040-4039/$ - see front matter Ó 2011 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetlet.2011.05.020

3626
B. Haché et al. / Tetrahedron Letters 52 (2011) 3625–3629
OtBu
P
N
OtBu
H+
O
P
O
P
OtBu
OtBu
[O]
O
OtBu
OtBu
OH
OH
R
H
P
O
O
O
R
R
R
1H-tetrazole / THF
Scheme 2.
N
N
OtBu
P
OtBu
P
OtBu
N
O
O
OtBu
P
N
H
P
N
R
R
H
H
N
N
OtBu
N
OtBu
N
OtBu
R
R
H
O
O
OtBu
N
1H-tetrazole = H+
pKa = 4.89
H+ = - tBu?
OtBu
H+ = - tBu?
OtBu
1H-tetrazole = H+
1H-tetrazole
P
H
P
H
O
O
N
OtBu
N
N
N
P
H
O
N
Phosphonate
by-product
Scheme 3.
of the reagent and/or the intermediate phosphite.6a Indeed, it was
alarming to see rising levels of undesired phosphonate with
increasing scale. This was probably due to time dilation to control
the exotherm on larger scale during the phosphoramidite addition
(not shown). Furthermore, the activator 1H-tetrazole6 has been re-
cently reclassified as an explosive.9,10 Consequently, accessing this
reagent from commercial sources is now severely restricted.
In recent reports,9 the crucial activator 1H-tetrazole can be re-
placed by phenanthroline salts,9a anilinium salts,9a benzimidazoli-
um salts,9a,b pyridinium salts9a and imidazolium salts.9 The ease
of activation is reported to be proportional to the degree of acidity
of the activator.9a Of the above salts,9a anilinium salts have a pKa of
4.44–5.34, pyridinium salts have a pKa of 5.23 and imidazolium
salts have a pKa of 6.99. In the case of the activation of bis-tert-butyl
phosphoramidites, we postulated that the use of imidazoleÁhydro-
chloride (imidazoleÁHCl) in combination with imidazole would give
optimal results at suppressing the phosphonate by-product.
(entry 1) with 1H-tetrazole (entry 2). Clearly, imidazoleÁHCl gave
improved product distribution compared to 1H-tetrazole. The role
of imidazoleÁHCl was tested in entry 3. As expected, this experi-
ment confirmed that imidazole alone was not sufficient to activate
the phosphitylation step.
4. Scope
Table 3 shows a variety of alcohols that were tested in our lab-
oratories with our general procedure. Phenol (entry 1), benzyl
alcohol (entry 2) and phenethyl alcohol (entry 3) were smoothly
phosphorylated in good to high yields. Oxidation-sensitive p-
methoxy congeners (entries 4 and 5) did not show over-oxidation
by-products. The secondary benzylic alcohol (entry 7)11 and 2-
hydroxymethyl pyridine (entry 8) were phosphorylated in modest
yields. For the 2-hydroxymethyl pyridine in entry 8, we suspect
the modest yield is due to increased aqueous solubility and/or
N-oxide by-products from exposure to hydrogen peroxide. Phos-
phorylation of more complex alcohols could also be performed.
For example, serine derivative N-CBz serine benzyl ester (entry
9) was chemoselectively phosphorylated in good yield and the
commercial drug Metronidazole (entry 10) was phosphorylated
in good yield.
2. Results
The phosphorylation of phenol (Scheme 4) was optimized to
identify the optimal stoichiometry of the reagents (Table 1). We
quickly identified that it was mandatory to use 1.5 equiv of imidaz-
oleÁHCl and phosphoramidite reagent for complete conversion of
phenol to the phosphate. The synergistic effect of 1.0 equiv of imid-
azole to buffer the reaction media was crucial to nearly eliminate
the occurrence of the phosphonate by-product (Table 3, entry 1).
As can be seen in Table 3, the phosphonate by-product is less
than 2% in most cases (Scheme 5).
We found, however, that the enhanced reactivity of the buffer
activation system led to decreased chemoselectivity. For exam-
ple, phosphorylation of tertiary alcohol (entry 7) proved chal-
lenging, either because of the lack of reactivity due to steric
hindrance or because of in situ cleavage back to the alcohol
through stabilized benzylic carbocation formation. Indeed, when
treating the secondary benzylic alcohol (entry 6), we observed a
subtle limitation to the method. Upon HPLC analysis of the reac-
tion mixture, incomplete conversion of the alcohol to the inter-
mediate phosphite was observed (80–86% conversion). After
addition of hydrogen peroxide, the remaining alcohol had con-
verted to the desired product.
3. Activator effect
Benzyl alcohol was used to study the effect of the activator in
Table 2. We first compared our imidazoleÁHCl/imidazole conditions
1) imidazole.HCl /
2) H2O2
O
P
OtBu
OtBu
imidazole / DMF
OtBu
OtBu
P
PhO
PhO
PhOH
OtBu
P
This observation suggests that the limitation of the method is
the phosphorylation of secondary benzylic alcohol where reaction
completion of the phosphite intermediate should be considered at
the phosphate stage (after hydrogen peroxide oxidation).
N
OtBu
Scheme 4.

B. Haché et al. / Tetrahedron Letters 52 (2011) 3625–3629
3627
Table 1
Stochiometry optimization with phenola
Entry
Equiv imidazoleÁHCl
Equiv phosphoramidite
Resultsb
Phenol
Phosphonate
Phosphate
1
2
3
4
5
1.4
1.4
1.5
1.7
2.0
1.4
1.4
1.5
1.7
2.0
33.9
1.7
1.3-1.5
1.7
<1
<1
<1
<1
<1
66.1
97.8
97.8–98.0
98.3
97.2
1.9
a
For all entries, 1.0 equiv imidazole was used.
HPLC area % after H2O2 oxidation.
b
Table 2
Activator effect with benzyl alcohola
Entry
Equiv imidazoleÁHCl
Equiv 1H-tetrazole
Equiv imidazole
Resultsb
Benzyl Alcohol
Phosphonate
Phosphate
1
2
3
1.5
1.0
2.5
<1
<1
<1
1.6
No reaction
96.5
87.0
1.5
a
For both entries, 1.5 equiv of phosphoramidite was used.
HPLC area % after H2O2 oxidation.
b
Table 3
Phosphorylation of a variety of alcoholsa
Entry
Substrate
OH
Product
Crude phosphonateb
2.8
Crude productb
95.7
Yield
O
P
OtBu
OtBu
O
1
82
88
63
83
O
P
OtBu
OtBu
2
3
4
<1
96.5
86.2
N/Ac
O
OH
OH
O
OtBu
O
P
1.6
OtBu
O
O
O
O
O
P
N/Ac
OtBu
OtBu
O
OH
OH
O
P
OtBu
OtBu
O
5
6
<1
96.2
98.0
73
82
O
P
OtBu
OtBu
2.4
O
O
O
OH
OH
OH
O
P
OtBu
OtBu
7
8
Phosphate product too labile
2.0
O
P
OtBu
OtBu
92.8
83.4
51
80
N
N
OH
O
P
O
OtBu
OtBu
O
O
O
BenzylO
N
H
9
3.5
1.7
O
BenzylO
OBenzyl
N
H
OBenzyl
O
O2N
OtBu
P
OH
N
N
O2N
O
N
OtBu
10
89.2
69
N
a
b
c
For all entries, 1.5 equiv of phosphoramidite was used.
HPLC area % after H2O2 oxidation.
Starting material and product co-eluted by HPLC. Reaction conditions and isolation were performed as per the general procedure.

3628
B. Haché et al. / Tetrahedron Letters 52 (2011) 3625–3629
OtBu
P
N
OtBu
O
P
H2O2
O
P
OtBu
OtBu
OtBu
OtBu
OtBu
H
H
P
O
O
O
O
R
1.5 equiv imidazole
1.0 equiv imidazole
10 ml/g DMF
·
HCl
R
R
R
Phosphonate < 2%
Scheme 5.
1) Activator / phosphoramidite / DMF
2) H2O2
Ph
O
OH
OH
O
P
OtBu
OtBu
Ph
O
OH
O
Ph
O
O
Ph
O
O
+
+
O
OH
O
P
O
P
OtBu
OtBu
OtBu
imidazole·HCl/
imidazole:
71
80
3
3
24
14
1H-tetrazole:
Scheme 6.
dropwise addition of 30 wt % hydrogen peroxide (3.01 mL,
26.6 mmol) and then warmed to room temperature. Once the oxida-
tion reaction was judged complete by HPLC analysis (<2% phosphite,
usually ꢀ3 h), the reaction was cooled in an ice-water bath and care-
fully quenched with saturated aqueous sodium thiosulphate
(10 mL).14 The product was then partitioned between ethyl acetate
(20 mL) and distilled water (20 mL) in a separatory funnel and the
layers were separated. The top ethyl acetate layer was washed with
brine (2 Â 5 mL) and distilled water (5 mL). The ethyl acetate layer
was concentrated on a rotary evaporator followed by purification
on silica gel using heptane/ethyl acetate (3:1).15 The pure fractions
were concentrated on a rotary evaporator. Further drying under
high vacuum to constant weight gave the phenyl bis-tert-butyl
phosphate product as a colorless oil (2.51 g, 82% yield).
5. Phosphorylation of 1,2-diols
In the context of the phosphorylation of 1,2-diols with 1H-tet-
razole, details from the literature on product distribution are lim-
ited.12 The phosphorylation of diols with our imidazoleÁHCl/
imidazole buffer system proved challenging and gave three main
products. LC–MS of the reaction mixture suggested predominant
phosphorylation of the primary alcohol over the secondary alcohol
and the 1,2-cyclic phosphate as shown in Scheme 6.
Intrigued by this observation, we tested the phosphorylation
with both the imidazoleÁHCl/imidazole system and 1H-tetrazole.13
Indeed, modest selectivity was achieved with both systems. How-
ever, a slightly better selectivity was observed with 1H-tetrazole.
This also demonstrates that the enhanced activation reactivity of
the imidazoleÁHCl/imidazole buffer system led to decreased che-
moselectivity compared to the well established 1H-tetrazole.
In conclusion, we discovered an improved activation system for
the phosphorylation of alcohols using bis-tert-butyl phosphorami-
dites. The improved conditions using imidazoleÁHCl/imidazole as
the activator gave less than 2% of phosphonate by-products in most
cases. The methodology is applicable to a variety of alcohols with a
few limitations. This improved the imidazoleÁHCl/imidazole activa-
tion method which is safer to use than the commonly used activa-
References and notes
1. (a) Cohen, P. Nature 1982, 296, 613; (b) Jadhav, P. K.; Woerner, F. J.; Aungst, B.
Bioorg. Med. Chem. Lett. 1996, 6, 2259; (c) Pettit, G. R.; Rhodes, M. R. Anti-cancer
Drug Design 1998, 13, 183; (d) Bentley, A.; Butters, M.; Green, S. P.; Learmonth,
W. J.; MacRae, J. A.; Morland, M. C.; O’Connor, G.; Skuse, J. Org. Process Res. Dev.
2002, 6, 109.
2. (a) Lang, P.; Gerez, C.; Tritsch, D.; Fontecave, M.; Biellmann, J.-F.; Burger, A.
Tetrahedron 2003, 59, 7315; (b) Ingall, A. H.; Bailey, A.; Coomba, M. E.; Cos, D.;
McInally, J. I.; Hunt, S. F.; Kindon, N. D.; Teobald, B. J.; Willis, P. A.; Humphries,
R. G.; Leff, P.; Clegg, J. A.; Smith, J. A.; Tomlinson, W. J. Med. Chem. 1999, 42, 213;
(c) Sekine, M.; Tsuruoka, H.; Iimura, S.; Kusuoku, H.; Wada, T. J. Org. Chem.
1996, 61, 4087.
3. (a) Perich, J. W.; Johns, R. B. Tetrahedron Lett. 1987, 28, 101; (b) Bannwarth, W.;
Trzeciak, A. Helv. Chim. Acta 1987, 70, 175; (c) Perich, J. W.; Johns, R. B. Synthesis
1988, 142.
4. (a) Gajda, T.; Zwierzak, A. Synthesis 1977, 623; (b) Gajda, T.; Zwierzak, A.
Synthesis 1976, 243.
5. Greene, T. W.; Wuts, P. G. M. Protective Groups Org. Synth. 1991, 47–53.
6. (a) Perich, J. W. Lett. Peptide Sci. 1998, 5, 49. and references therein; (b) Poteur,
L.; Trifilieff, E. Lett. Peptide Sci. 1996, 2, 271; (c) Chao, H. G.; Bernatowicz, M. S.;
Klimas, C. E.; Matsueda, G. R. Tetrahedron Lett. 1993, 34, 3377.
7. Selected references: (a) Lindberg, J.; Johan, E.; Konradsson, P. J. Org. Chem. 2002,
67, 194; (b Burke, T. R., Jr.; Barchi, J. J., Jr.; George, C.; Wolf, G.; Showlson, S. E.;
Yan, X. J. Med. Chem. 1995, 38, 1386; (c) Wang, X.-Z.; Yao, Z.-J.; Liu, H.; Zhang,
M.; Yang, D.; George, C.; Burke, T. R., Jr. Tetrahedron 2003, 59, 6087; (d) Sekine,
M.; Iimura, S.; Nakanishi, T. Tetrahedron Lett. 1991, 32, 395.
tor 1H-tetrazole,13
a
recently classified explosive.9,10Large
quantities of imidazoleÁHCl and imidazole are both easy to obtain
and inexpensive. The improved activation procedure will further
expand the usefulness of the phosphorylation of alcohols using
bis-tert-butyl phorphoramidites.
6. General procedure (using phenol as an example)
To a 50 mL round-bottom flask was charged phenol (1.00 g,
10.6 mmol), imidazoleÁHCl (1.67 g, 15.9 mmol) and imidazole
(0.723 g, 10.6 mmol). The mixture was dissolved in DMF (10 mL)
under magnetic stirring and nitrogen atmosphere. Once complete
dissolution was achieved, N,N-diisopropyl bis-tert-butyl phospho-
ramidite (5.03 mL,15.9 mmol) was charged dropwise over 1–
3 min. The reaction was stirred at room temperature until complete
consumption of phenol was observed by HPLC (less than 1 h). The
reaction mixture was cooled in an ice-water bath and treated with
8. For mechanistic considerations: With tetrazole, DPPA was also used as
a
phosphoramidite equivalent: (a) Chen, S.-B.; Li, Y.-M.; Luo, X.-Z.; Zhao, G.; Tan,
B.; Zhao, Y.-F. Phosphorus, Sulfur Silicon Relat. Elem. 2000, 164, 277; Nurminen,
E. J.; Mattinen, J. K.; Lonnberg, H. J. Chem. Soc., Perkin Trans. 2 1998, 1621. see
also Ref. 9.

B. Haché et al. / Tetrahedron Letters 52 (2011) 3625–3629
3629
9. (a) Sanghvi, Y. S.; Guo, Z.; Pfundheller, H. M.; Converso, A. Org. Process Res. Dev.
2000, 4, 175; (b) Hayakawa, Y.; Kataoka, M.; Noyori, R. J. Org. Chem. 1996, 61,
7996.
10. (a) Stull, D. R. Fundamentals of Fire and Explosion, AIChE Monograph, Series No.
10; American Institute of Chemical Engineers: New York, 1977. Vol. 72, p 22;
(b) See MSDS for T 7641 in Sigma database.
644); PyridineÁHCl and PyridineÁHCl/imidazole. (no phosphonate discussion,
rather O,N selectivity): Kato, Y.; Oka, N.; Wada, T. Tetrahedron Lett. 2006, 47,
2501; Imidazolium Trifluoroacetate (no phosphonate discussion): Gomez, C.;
Chen. J.; Wang, S. Tetrahedron Lett. 2009, 50, 6691–6692.
14. In some cases, less sodium thiosulphate is necessary to prevent decomposition
of the crude phosphate product.
11. The commercial alcohol was 47% pure by HPLC.
15. In some cases, multiple chromatographies were required to remove the N,N-
diisopropyl bis-tert-butyl phosphoramidate:
12. Literature precedents, mostly in patents often purify by prep-HPLC and no
product distribution data are available.
OtBu
P
O
P
H2O2
OtBu
OtBu
13. ScifinderÒ search of the N,N-di-alkyl-bis-tert-butyl phosphoramidite as
a
N
OtBu
N
reagent returned >100 references, all of the references were using 1H-
tetrazole as the activator. Three exceptions, pyridineÁHCl (WO 2001040236,
no phosphonate discussion); 1,2,3-triazole (WO 2001074369 and WO
2001074368, no phosphonate discussion) and pyridineÁHCl + pyridine
(phosphonate discussion with vague product distribution details);
pyridineÁHCl giving phosphonate by-product (Barnwell, N.; Cheema, L;
Cherryman, J.; Dubiez, J; Howells, G.; Wells, A. Org. Process Res. Dev. 2006, 10,
0.5 equiv
excess