Preparation of arylphosphonates by palladium(O)-catalyzed cross-coupling in the presence of acetate additives: Synthetic and mechanistic studies

Article abstract of DOI:10.1002/adsc.200900590

An efficient protocol for the synthesis of arylphosphonate diesters via a palladium-catalyzed cross-coupling of H-phosphonate diesters with aryl electrophiles, promoted by acetate ions, was developed. A significant shortening of the cross-coupling time in the presence of the added acetate ions was achieved for bidentate and monodentate supporting ligands, and for different aryl electrophiles (iodo, bromo and triflate derivatives). The reaction conditions were optimized in terms of amount of the catalyst, supporting ligands, and source of the acetate ion used. Various arylphosphonates, including those of potential biological significance, were synthesized using this newly developed protocol. Some mechanistic aspects of the investigated reactions are also discussed.

Full text of DOI:10.1002/adsc.200900590

FULL PAPERS
DOI: 10.1002/adsc.200900590
Preparation of Arylphosphonates by Palladium(0)-Catalyzed
Cross-Coupling in the Presence of Acetate Additives: Synthetic
and Mechanistic Studies
Marcin Kalek,a Martina Jezowska,a and Jacek Stawinskia,b,*
a
Department of Organic Chemistry, Arrhenius Laboratory, Stockholm University, S-106 91 Stockholm, Sweden
Fax : (+46)-8-15-4908; e-mail: js@organ.su.se
Institute of Bioorganic Chemistry, Polish Academy of Sciences, Noskowskiego 12/14, 61-704 Poznan, Poland
b
Received: August 27, 2009; Published online: December 2, 2009
Supporting information for this article is available on the WWW under
http://dx.doi.org/10.1002/adsc.200900590.
Abstract: An efficient protocol for the synthesis of lyst, supporting ligands, and source of the acetate ion
arylphosphonate diesters via a palladium-catalyzed used. Various arylphosphonates, including those of
cross-coupling of H-phosphonate diesters with aryl potential biological significance, were synthesized
electrophiles, promoted by acetate ions, was devel- using this newly developed protocol. Some mecha-
oped. A significant shortening of the cross-coupling nistic aspects of the investigated reactions are also
time in the presence of the added acetate ions was discussed.
achieved for bidentate and monodentate supporting
ligands, and for different aryl electrophiles (iodo, Keywords: acetate additives; arylphosphonates; H-
bromo and triflate derivatives). The reaction condi- phosphonate diesters; nucleotide analogues; palladi-
tions were optimized in terms of amount of the cata- um-catalyzed cross-coupling
Introduction
[for
example,
dppp,[16]
dppb,[17]
PPh2(m-
C6H4SO3M)[18]], and the most efficient was found to
Since its introduction in the early 1980s by Hirao et be a wide-bite-angle ligand,[19] 1,1’-bis(diphenylphos-
al., the cross-coupling reaction of aryl halides with H- phino)ferrocene (dppf).[12,20–22] In this type of cross-
phosphonate diesters catalyzed by PdACTHNUTRGNEUNG
(PPh3)4,[1] has coupling reactions the base has to be used in a stoi-
been used for the preparation of various organophos- chiometric amount. The most commonly used one is
phorus compounds.[2] Relatively mild reaction condi- triethylamine,[23] however, due to possible dealkyla-
tions and the possibility to synthesize arylphospho- tion of sensitive H-phosphonate diesters, it is often re-
nates in a stereospecific manner,[3] contributed to pop- placed by less nucleophilic tertiary amines[22,24] or pro-
ularity of the method.[4,5,6,7] Nevertheless, increasing pylene oxide.[9,12,20] For special reaction conditions, for
applications of aryl- and vinylphosphonates in materi- example, biphasic conditions[25] or for the reactions in-
al sciences[6,8] and biological chemistry,[5,7,9–15] caused a volving microwave heating,[26] using inorganic bases
ꢀ
demand for improved protocols for the C P bond such as K2CO3 and Cs2CO3 can be advantageous. As
forming cross-coupling reactions, to simplify synthesis far as the palladium source is concerned, it was found
of complex organic compounds.
that in contrast to PdACHTUNGTERUNN(G PPh3)4 used in Hiraoꢀs original
To enhance efficiency of the C P bond formation procedure,[1] catalytic systems generated in situ from
ꢀ
(Scheme 1), other supporting ligands than triphenyl- PdACHTNUTRGNEUNG(OAc)2 and appropriate phosphine ligands, were
phosphine used by Hirao et al.,[1] were investigated usually superior in terms of reactivity for promoting
[9,11,20–22,24,27]
ꢀ
the C P bond formation.
All these improvements, for which a mechanistic
basis has been provided in recent years,[19,27,28,29–31]
lend themselves into a modified catalytic cycle for the
ꢀ
palladium-mediated formation of the C P bond as
ꢀ
shown in Scheme 2.
Scheme 1. A palladium-catalyzed C P bond-forming cross-
coupling reaction.
Adv. Synth. Catal. 2009, 351, 3207 – 3216
ꢁ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
3207

Marcin Kalek et al.
FULL PAPERS
On the basis of these mechanistic findings, we set
out to develop an efficient protocol for the synthesis
of arylphosphonates via a palladium-catalyzed cross-
coupling reaction between H-phosphonate diesters
and ArX (X=I, Br, or triflate) in the presence of ex-
ternal acetate ion additives. In this paper we show
that the profound accelerating effect of acetate addi-
tives is a universal phenomenon and is compatible
with a broad spectrum of supporting phosphine li-
gands. The developed protocol employs substoichio-
metric amounts of inorganic acetate and its efficiency
was demonstrated in the synthesis of various complex
arylphosphonate derivatives.
Results and Discussion
Although by using PdACTHNUTRGNE(NUG OAc)2 as a palladium source
all steps of a cross-coupling reaction are accelerated
due to introduction of an acetate ion, the observed
differences in reactivity are usually small (see Table 1,
entry 1 vs. 2 for aryl halides).[24,27] This phenomenon
we traced back to competition between acetate and
halide anions for a palladium center[27] in intermedi-
ates of type A and B (Scheme 2). Since concentration
of halides gradually increases during the course of the
reaction, less reactive halide-palladium complexes are
Scheme 2. A catalytic cycle for a palladium-promoted cross-
coupling between aryl halides and H-phosphonate diesters
involving acetate ligated palladium complexes.
As it is apparent from the intermediates involved formed, and for this reason, an accelerating effect of
(A–D, Scheme 2), an acetate ligand plays a crucial the acetate is only observed at the initial stages of the
role in all stages of the catalytic cycle. Initially, the cross-coupling reaction, when the concentration of
higher reactivity of catalytic systems generated in situ halides is low. To remedy this problem, we carried out
from PdACHTUNGTRENNUNG(OAc)2 was ascribed to increased reactivity of our screening experiments using stoichiometric
anionic palladium(0) complexes ligated by acetate amounts of acetate, as our previous mechanistic stud-
ions (species A) in the oxidative addition,[28,29,30] how- ies indicated that external addition of acetate increas-
ever, more detailed studies revealed that participation es the overall rate of a cross-coupling reaction.[27]
of acetate ions in the ligand exchange (intermediates
B and C) and the reductive elimination (intermediate
ꢀ
D) steps[27,31] is also of crucial importance.
Effect of Stoichiometric OAc Additive on the C P
ꢀ
As to the accelerating effect of acetate in the ligand Bond Forming Reaction with Different Phosphine
exchange and the reductive elimination, apparently it Ligands
originates from the unique ability of acetate ion to act
as a bidentate k2-ligand (intermediates C and D in We started our study with an examination of how ex-
Scheme 2).[31] Due to this, the rate of incorporation of ternal acetate ions additives influence the rates of the
ꢀ
phosphorus nucleophile into a palladium(II) complex C P bond formation, as a function of a leaving group
(i.e., ligand substitution or transmetallation process) in an electrophilic aromatic substrate and phosphine
is greatly facilitated via intramolecular displacement supporting ligands used. Of particular interest was the
of one of the phosphine ligands by an oxygen atom of behavior of bidentate phosphines, since the postulated
the adjacent acetate. The intermediate formed (C) is mechanisms for ligand substitution and reductive
highly reactive and coordinates an H-phosphonate elimination steps (Scheme 2) implied that, at least at
diester, followed by its deprotonation to form an some stages of the reaction, only one phosphorus
equilibrium mixture of type D intermediates, from atom of the chelating ligand had to be engaged in
which reductive elimination occurs. Due to the pres- complexation (e.g., intermediate C). Although pre-
ence of an acetate ligand these Pd-phosphonate com- liminary experiments proved that it was the case for
plexes are highly fluxional and the reductive elimina- dppp,[31] other bidentate phosphines might not permit
tion is accelerated by a constant supply of species such a mechanistic pathway.
containing aryl and phosphonate moieties in the cis
Thus, to evaluate the compatibility of acetate addi-
tives with common supporting phosphine ligands, we
arrangement, necessary for this process.[31]
3208
asc.wiley-vch.de
ꢁ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2009, 351, 3207 – 3216

Preparation of Arylphosphonates by Palladium(0)-Catalyzed Cross-Coupling
FULL PAPERS
Table 1. Effect of OAcꢀ additives on the cross-coupling using different phosphine ligands.[a,b]
Entry
Pd source/Ligand
OAcꢀ additive[c]
Reaction time [h][d]
1
2
3
4
5
6
7
8
Pd
Pd
Pd
Pd
Pd
Pd
Pd
Pd
Pd
Pd
Pd
N
no
no
yes
no
yes
no
yes
no
8
7
1
10
0.5
21
1
18
16
2.5
25
2
26
3
14
2
23
5
3
3
0.75
2
1
1
0.75
0.75
0.5
6
9
10
11
yes
no
yes
0.75
26
1.5
21
2
[a]
Abbreviations: dppp=1,3-bis(diphenylphosphino)propane, dppb=1,4-bis(diphenylphosphino)butane, dppf=1,1’-bis(di-
phenylphosphino)ferrocene, BINAP=2,2’-bis(diphenylphosphino)-1,1’-binaphthyl (racemic).
[b]
Reaction conditions: 0.30 mmol 1a, 0.33 mmol PhX, 0.36 mmol Et3N, 3 mL THF, 608C; entry 1: 0.03 mmol Pd
tries 2–11: 0.03 mmol Pd(OAc)2, 0.09 mmol monodentate or 0.06 mmol bidentate ligand; 15 min, palladium reduction was
performed before addition of 1a and PhX.
ACHTUNGTERN(NUNG PPh3)3; en-
AHCTUNGTRENNUNG
[c]
[d]
0.30 mmol n-Bu4N
ACHTUNGTRNE(NUNG OAc).
>95% conversion monitored by 31P NMR spectroscopy.
for PhOTf, this effect was significantly stronger
(shortening of the reaction time from 23 to only 5 h).
A completely different picture emerged upon external
addition of acetate to the reaction mixture (entry 3).
In this instance, a significant shortening of the reac-
tion time was observed for the aryl halides, while for
phenyl triflate, the effect was only moderate (from 5
to 3 h). This pattern of reactivity was observed for all
ligands investigated (Table 1, even number entries vs.
odd number ones).
Scheme 3. A model reaction for an acetate-promoted cross-
coupling.
performed a series of model reactions in which dieth-
On the whole, the data in Table 1 are consistent
yl H-phosphonate (1a) was coupled with iodobenzene, with the catalytic cycle in Scheme 2 and point to the
bromobenzene, and phenyl triflate (Scheme 3). importance of the acetate-mediated ligand exchange
In all cases 10 mol% of the palladium pre-catalyst, process that seems to operate for all the bidentate
in the form of Pd(OAc)2 [except for the benchmark supporting phosphine ligands. Crucial to the final out-
reaction with Pd(PPh3)4 – Table 1, entry 1], was used, come of these reactions is probably an equilibrium
AHCTUNGTRENNUNG
ACHTUNGTRENNUNG
together with an appropriate phosphine that acted system shown in Scheme 4 that controls amounts of
both as a reducing agent for palladium and a support-
ing ligand for the catalyst formed.[30,32]
Each reaction was carried out separately in the ab-
sence and in the presence of n-Bu4NACTHNUTRGNE(NUG OAc) (1 equiv.),
and the progress of the reactions was monitored by
31P NMR spectroscopy. The completion times for
these cross-coupling reactions (>95% conversion into
phenylphosphonate 2a) are presented in Table 1.
As it is apparent from the data in Table 1, the reac-
tivities of PhI, PhBr and PhOTf in a cross-coupling
reaction catalyzed by Pd
ACHTUNGTREN(NNUG PPh3)4 (entry 1) paralleled
their relative rates of oxidative addition.[33] By using
Scheme 4. Equilibria between palladium(II) complexes that
undergo ligand exchange with H-phosphonate diesters. L*
stands for a neutral ligand, for example, solvent molecule or
diphosphine monoxide.[30] For monodentate ligands the pre-
PdACHTUNGTRENNUNG(OAc)2 as the palladium source (entry 2; 0.2 equiv.
of OAcꢀ is introduced to the reaction mixture with
this pre-catalyst), a slight acceleration of the product
formation was observed for PhI and PhBr, whereas sented Pd(II) complexes have the trans configuration.
Adv. Synth. Catal. 2009, 351, 3207 – 3216
ꢁ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
3209

Marcin Kalek et al.
FULL PAPERS
different Pd(II) complexes in the reaction mixture
To conclude this part, it seems that a mechanism
that undergo ligand exchange with phosphorus nucle- presented in Scheme 2 operates for various bidentate
ophiles (H-phosphonate diesters). supporting phosphine ligands. Relative rates of the
A dramatic accelerating effect of acetate observed cross-coupling reactions observed for different aryl
in the reactions involving phenyl triflate can be as- electrophiles are understandable. The most affected
cribed to a weak bonding of the triflate anion to a by the external addition of acetate are the reactions
palladium(II) center.[33] Due to this, even in presence in which the ligand substitution could be enhanced to
of a substoichiometric amount of acetate ions [intro- the highest degree, that is, those involving PhBr and
duced to the reaction mixture along with PdACTHNUTRGNEUNG(OAc)2], PhI as electrophiles. For PhOTf, since the reaction al-
the equilibrium is apparently shifted towards the reac- ready proceeds mainly via the acetate-ligated species
tive acetate-ligated species (Scheme 4). For PhI and even in the presence substoichiometric amounts of
PhBr, however, relatively high affinity of halides vs. OAcꢀ introduced with Pd
ACHTUNGTRENNUNG
acetate-ligated Pd(II) species were formed in signifi-
Finally, since for the cross-coupling reactions in the
cant amounts only at the beginning of the reaction, presence of acetate all phosphine ligands investigated
when the concentration of halide ions was low. As the showed similar efficiency, it means that differences
reactions progressed, the dominant Pd(II) species observed in the absence of this additive reflect the
became those with ligated halides, and this slowed ability of the supporting ligands to facilitate a ligand
down the ligand exchange process, and in conse- exchange process. In this respect, the most efficient
quence, the whole cross-coupling. In contrast to this, seemed to be monodentate PPh3 and the wide-bite-
when a stoichiometric amount of acetate was present angle phosphine dppf (Table 1, entries 2 and 8), and
in the reaction mixture (external addition), then the latter was used in our further studies.
throughout the reaction the dominant Pd(II) species
contained ligated acetate. This secured high rates of
the individual steps and made the differences between Effect of Amount and Source of the Acetate
reactions involving various aryl substrates (aryl hal- Additives on the Cross-Coupling Reactions
ides vs. aryl triflate), small.
By comparing the reactions carried out in the ab- Since preliminary experiments with external acetate
sence of external acetate additives, for most ligands additives indicated a complex relationship between
(Table 1, entries 2, 4, 6 and 8) iodobenzene reacted the accelerating effect and the amount and the source
faster than its bromo counterpart. The exception was, of the acetate ion used, more detail investigations
however, BINAP (entry 10) for which a faster cross- were carried out to find optimal reaction conditions.
coupling reaction was observed with bromo- vs. iodo-
benzene. A possible explanation for this could be that ethyl H-phosphonate 1a and bromobenzene, catalyzed
palladium(II) complexes containing bromide undergo by the Pd(OAc)2/dppf was carried out using different
To this end a cross-coupling reaction between di-
AHCTUNGTRENNUNG
faster ligand substitution (ca. 2 times) by H-phospho- amounts of the acetate additive and its various sour-
nate diesters than those in with iodide,[31] and thus, in ces. The results are summarized in Table 2.
this particular case, not oxidative addition but ligand
exchange, became a turnover limiting step of the cata- affects efficiency of our cross-coupling reaction, we
lytic cycle. Upon addition of external acetate to such used the well soluble in organic solvent n-Bu4N(OAc)
To assess how the amount of the external acetates
ACHTUNGTRENNUNG
a BINAP-mediated reaction (Table 1, entry 11), how- as an acetate source. Entry 1 in Table 4 shows the re-
ever, a significant accelerating effect was observed sults for our reference reaction where no external
and the cross-coupling reaction with iodobenzene acetate additive was introduced, and entry 2, for the
became faster (slightly) than that with bromobenzene. reaction in the presence of 1 equivalent of
The above results show that the ligand exchange n-Bu4NACHTUNTRGNEUNG(OAc). By carrying out experiments with dif-
process has the largest impact on the overall reaction ferent amounts of acetate ions we noticed that the re-
ꢀ
kinetics in the C P bond forming cross-coupling with actions were fastest with 1 equivalent of n-Bu4N-
H-phosphonate diesters, although, the rate of oxida-
ACHTUNGTRENNUNG
tive addition also matters. Slow cross-couplings in the of the added n-Bu4NACHTUNGTRENNUNG
absence of acetate indicate that H-phosphonates ex- could be observed (Table 2, entry 3). The same phe-
hibit relatively weak nucleophilicity towards the palla- nomenon we observed for other acetate salt soluble
dium(II) center. The most probable reason for that is in organic solvents, Et3NH
ACHTUNGERTN(NUNG OAc) (Table 2, entries 4
that, unlike amines, alcohols or thiols, these com- and 5). A smaller accelerating effect exerted by this
pounds do not bear a lone electron pair on the heter- salt vs. n-Bu4NACTHNUTRGNEU(GN OAc) (Table 2 entry 4 vs. 2) originated
oatom forming bond to the metal. Hence, a preceding probably from a partial hydrogen bonding of AcOꢀ
association must take place first, most probably with with Et3NH+, that lowered its effective activity.
participation of the phosphoryl oxygen.[31]
3210
asc.wiley-vch.de
ꢁ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2009, 351, 3207 – 3216

Preparation of Arylphosphonates by Palladium(0)-Catalyzed Cross-Coupling
FULL PAPERS
Table 2. Effect of different acetate additives on the cross-
coupling of diethyl H-phosphonate 1a with PhBr, using Pd-
Although using n-Bu4NACTHNUTRGNE(UNG OAc) secured homogenous
reaction conditions, its high price, hydroscopic proper-
ties, and problems with its removal during work-up,
prompted us to look for other sources of acetate
anions. To this end we investigated various alkali
metal acetates (Li, Na, Cs, and K) (Table 2, entries 6–
9). Interestingly, although these salts were only spar-
ingly soluble in THF under reflux (<1 mgmLꢀ1), they
gave a reasonable shortening of the reaction time.
Since the reactions occurred under heterogeneous
conditions and the concentration of acetate ions was
controlled by the solubility factor, using smaller
amounts of these salts (e.g., entries 9–12 for KOAc),
did not affect the reaction time (within the experi-
mental error).
ACHTUNGTRENNUNG
(OAc)2 dppf.[a]
Entry
Pd loading
(mol%)
Acetate additive
(equiv.)
Reaction time
[h][b]
1
2
3
4
5
6
7
8
10
10
10
10
10
10
10
10
10
10
10
10
5.0
2.5
1.0
none
n-Bu4N
n-Bu4N
Et3NH
Et3NH
LiOAc (1)
NaOAc (1)
CsOAc (1)
KOAc (1)
KOAc (0.5)
KOAc (0.2)
KOAc (0.1)
KOAc (0.1)
KOAc (0.1)
KOAc (0.1)
14
2
G
E
no reaction
3.5
no reaction
5
G
E
4
3.5
3.5
3
4
3
4
2.5
4.5
9
10
11
12
13
14
15
What was somewhat unclear about using alkali
metal acetates as external additives, however, was a
significant shortening of the reaction time achieved
with a small amount of acetate ions present in the re-
action mixtures (e.g., entry 12 vs. 2 in Table 2). If we
[a]
Reaction conditions: 0.30 mmol 1a, 0.33 mmol PhBr, assume that concentration of Pd(II) complexes ligated
0.36 mmol Et3N, 3 mL THF, 608C; 15 min, palladium re-
duction was performed before addition of 1a and PhX.
>95% conversion monitored by 31P NMR spectroscopy.
with acetate is controlled by Eq. (1) (L=mono- or bi-
dentate ligand, X=halide),[29] the expressions for the
corresponding equilibrium constant [Eqs. (2) and (3)]
indicate that a favorable ratio of acetate vs. halide li-
gated complexes is controlled by the ratio of [OAcꢀ]/
[b]
The observed inhibition of the cross-coupling reac- [Xꢀ] in the reaction media. Under homogenous reac-
tion by a high concentration of acetate ions was not tion conditions, this can be increased only by higher
completely unexpected, in light of the involvement of concentrations of the added acetate, while under het-
acetate ions in various reactions steps (see, erogeneous conditions, also by removal of the halides
Scheme 2). The 31P NMR experiments revealed that from the reaction solution.
under such reaction conditions acetate ions apparent-
A postulated model for such a scenario is depicted
ly interfered with the oxidative addition step as no ar- in Scheme 5. At the beginning of a cross-coupling re-
ylpalladium(II) complex formation could be observed
upon addition of bromobenzene to a model reaction
mixture containing a palladium(0) complex Pd
ACHTUGNRTEN(NUGN PPh3)4
and 50 equivalents of n-Bu4N
AHCTUNGTRENNUNG
hand, a high concentration of the acetate probably
did not affect significantly the reduction step of
PdACHTUNGTRENNUNG(OAc)2 with PPh3 as judged from the formation of
the corresponding phosphine oxide (31P NMR experi-
ment).
ACTHNUTRGNEUNG
Scheme 5. A mechanism for maintaining a favorable ratio [OAcꢀ]/[Brꢀ] in the reaction mixture when using KOAc as an ace-
tate source.
Adv. Synth. Catal. 2009, 351, 3207 – 3216
ꢁ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
3211

Marcin Kalek et al.
FULL PAPERS
action bromides are removed from the reaction media
First, aryl substrates containing different leaving
via precipitation of insoluble in organic solvents po- groups, namely iodobenzene, bromobenzene and
tassium bromide, and at some point (e.g., 10% con- phenyl triflate, were coupled with diethyl H-phospho-
version for entry 12, or 20% conversion for entry 11), nate.
the whole amount of AcOꢀ ions is in solution in the
As is apparent form entries 1–3 (Table 3), all of
ACHTUNGTRENNUNG(OAc). As the reaction progresses, the them showed comparable reactivity in the cross-cou-
form of Et3NH
bromides produced are still removed from the solu- pling reactions in the presence of acetate ions. Di-
tion, this time in the form of an Et3NHBr precipitate. methyl H-phosphonate (Table 3, entry 4) reacted with
ACTHNUTRGNEUNG
In consequence, the ratio [OAcꢀ]/[Brꢀ] remained bromobenzene similarly to diethyl H-phosphonate
steady for the most part of the reaction and secured a (3 h for the completion), but the presence of two iso-
fast cross-coupling, even when only 10 mol% KOAc propyl groups (entry 5) in the H-phosphonate moiety
additive was used for the reaction.
Further experiments showed that for a given (10 h), most likely due to the steric hindrance.
amount of the added acetate it was possible to reduce Then, cross-couplings with aryl bromides bearing
visibly slowed down the cross-coupling reactions
the amount of the catalyst even 10-fold, without af- polycyclic aromatic moieties were investigated
fecting significantly the efficiency of the reaction (Table 3, entries 6–9). 2-Bromonaphthalene (entry 6)
(Table 2, entries 13–15). We found that for synthetic reacted similarly to bromobenzene (ca. 2 h), but its
purposes 2.5 mol% of Pd in combination with positional isomer, 1-bromonaphthalene (entry 7), 9-
0.1 equivalent of KOAc (Table 2, entry 14) was the bromophenanthrene (entry 8), and 1-bromopyrene
best choice.
(entry 9), underwent cross-coupling with diethyl H-
To sum up this part, we demonstrated that a suc- phosphonate significantly slower (23–32 h), probably
cessful implementation of the mechanistic findings due to highly unfavorable interactions with the peri-
concerning the acetate-promoted catalytic cycle in hydrogen atoms. This slow kinetics, however, did not
Scheme 2 can be achieved in two different ways: via affect the efficiency of the reactions, and the isolated
the addition of stoichiometric amounts of soluble in yields of the corresponding arylphosphonates formed
organic solvents acetates [for example, n-Bu4N
ACHTUNGTREN(NUGN OAc) were rather high. In line with the importance of steric
or Et3NH(OAc)], or by using sub-stoichiometric factors in this type of reactions, a very slow cross-cou-
ACHTUNGTRENNUNG
amounts of the acetate additives with a simultaneous pling was observed for o-bromotoluene (Table 3,
removal of halides from the reaction media. Both entry 13).
methods secured a favorable ratio [OAcꢀ]/[Xꢀ] and
Concerning the influence of electronic factors, the
provided comparable acceleration of the Pd-catalyzed cross-coupling was rather insensitive to the presence
cross-coupling reactions of H-phosphonate diesters of electron-withdrawing substituents in the aromatic
with aryl electrophiles. The latter approach, however, ring of the electrophilic substrates (e.g., p-nitro- and
seems to be preferable for preparative purposes as it p-fluorobromobenzene, entries 10 and 11, respective-
permits one to carry out the reaction with a lower cat- ly), but the introduction of a strongly electron-donat-
alyst load.
ing p-methoxy group (entry 12), significantly slowed
down the reaction (10 h).
Since pyridylphosphonates emerged recently as po-
tential, biologically important nucleotide ana-
logues,[7,15] we carried out two cross-coupling reactions
involving 3-bromopyridine (entry 14) and 4-bromo-
Synthesis of Arylphosphonate Diesters Promoted by
the Acetate Additive
The above studies led us to the reaction conditions pyridine (entry 15). The reactions proceeded smooth-
shown in Scheme 6 that we successfully used in the ly and the corresponding diethyl pyridylphosphonates
synthesis of various arylphosphonates with diverse were isolated in high yields (>90%).
structural features both in the aryl and in the ester
moieties (Table 3, Table 4, and Table 5).
Looking for a possible chemoselectivity in this type
of cross-coupling reactions, we investigated aromatic
All the reactions in Table 3, Table 4, and Table 5 dihalides as electrophilic substrates (Table 4).
were run to >95% conversion (31P NMR spectrosco-
To this end we reacted p-bromoiodobenzene
py) and the products were isolated by silica gel chro- (Table 4, entry 1) and 2,5-dibromopyridine (Table 4,
matography.
entry 2) with diethyl H-phosphonate under the devel-
oped reaction conditions. In both instances the reac-
tions were high yielding and completely chemoselec-
tive. For bromoiodobenzene, it was the iodine atom
that was replaced by the phosphorus nucleophile
forming diethyl 4-bromophenylphosphonate, and for
dibromopyridine, the bromine at the C-2 position of
Scheme 6. Optimized reaction conditions for the synthesis of
arylphosphonate diesters.
3212
asc.wiley-vch.de
ꢁ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2009, 351, 3207 – 3216

Preparation of Arylphosphonates by Palladium(0)-Catalyzed Cross-Coupling
FULL PAPERS
Table 3. Acetate-promoted cross-coupling of aryl halides with H-phosphonate diesters.[a]
Entry Aryl halide
H-phospho-
nate (1)
Reaction
time [h][b]
Product
[%][c]
Entry Aryl halide H-phospho-
Reaction Product
nate (1)
time[b]
[%][c]
1
3
2a (81)
9
32
2g (82)
2
3
4
5
6
7
2.5
3
2a (75)
2a (79)
2b (72)
2c (95)
2d (85)
2e (97)
2f (92)
10
11
12
13
14
15
4
3
2h (98)
2i (91)
2j (90)
2k (63)
2l (92)
2m (94)
3
10
32
1.5
5
10
2
36
23
8
[a]
Reaction conditions: 0.03 mmol PdACHTNUGTRNEG(UN OAc)2, 0.06 mmol dppf, 1.5 mmol Et3N, and 0.13 mmol KOAc in 5 mL THF, 688C,
15 min; then 1.25 mmol 1 and 1.38 mmol ArX were added.
>95% conversion monitored by 31P NMR spectroscopy.
Isolated yield.
[b]
[c]
Table 4. Selective acetate-promoted cross-coupling of aryl
dihalides with 1a.[a]
the pyridine ring reacted selectively to produce dieth-
yl 5-bromopyridin-2-ylphosphonate.
Entry Aryl
halide
Reaction
time [h][b]
Product
Yield
[%][c]
As a final part of these investigations we tested the
efficacy of the developed cross-coupling method in
the synthesis of more complex arylphosphonate dies-
1
2
3
96
83
ACHTUNGTRENNUNGters (Table 5) that are of potential importance to me-
dicinal chemistry[10,35] and nucleic acids-based thera-
peutics.[13,15,36]
In entry 1 (Table 5), a protected tyrosine triflate de-
rivative 4 in the reaction with diethyl H-phosphonate
was successfully converted into the corresponding ar-
ylphosphonate 9 in high yield. Entry 2 (Table 5)
shows the transformation of cholesteryl H-phospho-
nate diester 5 into the highly lipophilic 2-naphthyl-
phosphonate derivative 10. Also various dinucleoside
H-phosphonates 6–8 (Table 5, entries 3–5) reacted ef-
ficiently with bromobenzene to produce the corre-
sponding dinucleoside phenylphosphonates. The reac-
2
[a]
Reaction conditions: 0.03 mmol PdACTHNUTRGENUG(N OAc)2, 0.06 mmol
dppf, 1.5 mmol Et3N, and 0.13 mmol KOAc in 5 mL THF,
688C, 15 min; then 1.25 mmol 1 and 1.25 mmol ArX
were added.
[b]
[c]
>95% conversion monitored by 31P NMR spectroscopy.
Isolated yield.
Adv. Synth. Catal. 2009, 351, 3207 – 3216
ꢁ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
3213

Marcin Kalek et al.
FULL PAPERS
Table 5. Synthesis of some biologically relevant arylphosphonate derivatives.[a,b]
Entry
Aryl
derivative
H-phosphonate
diester
Reaction
time
Product
Yield [%][d]
[h][c]
1
2.5
86
71
2[e]
5
3[e]
1.5
54
4[e]
7
60
63
56
ACHTUNGTRENNUNG
5
5.5
ACHTUNGTRENNUNG
[a]
Abbreviations: TBDPS=tert-butyldiphenylsilyl, TBDMS=tert-butyldimethylsilyl, DMT=4,4’-dimethoxytrityl.
Reaction conditions: 0.01 mmol PdACHTUNTGRNEUNG(OAc)2, 0.02 mmol dppf, 0.48 mmol Et3N, and 0.04 mmol KOAc in 4 mL THF, 688C, 15 min;
[b]
then 0.40 mmol H-phosphonate and 0.44 mmol ArX were added.
>95% conversion monitored by 31P NMR spectroscopy.
Isolated yield.
[c]
[d]
[e]
Mixture of the H-phosphonate diastereoisomers (ca. 1:1) was used.
3214
asc.wiley-vch.de
ꢁ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2009, 351, 3207 – 3216

Preparation of Arylphosphonates by Palladium(0)-Catalyzed Cross-Coupling
FULL PAPERS
vessel. The vessel was sealed and filled with N2, by applying
2 cycles of vacuum, followed by N2. THF was introduced via
a septum (5 mL), followed by triethylamine (1.50 mmol,
152 mg, 208 mL), and the mixture was stirred and heated at
688C After 15 min, H-phosphonate (1.25 mmol) and aryl
halide (1.38 mmol) were added (in the case of solid reagents,
they were dissolved in a small amount of THF). After heat-
ing at 688C for the time indicated in Table 3, the solvent
was evaporated and the product purified by silica gel chro-
matography (depending on the case, pentane-AcOEt mix-
ture from 1:1 to 1:9 was used as an eluent).
tion time varied depending on nucleosidic composi-
tion and the protecting groups present, but it did not
exceed 7 h. Although in all instances the 31P NMR
spectroscopy indicated a complete conversion into the
corresponding phenylphosphonates 11–13, the isolat-
ed yields were lower than for other arylphosphonates,
due to loses during purifications.
To find out if the involvement of acetates in a cata-
lytic cycle (Scheme 2) does not erode stereochemistry
at the phosphorus center, separate diastereomers of
dithymidine H-phosphonate 8 (8-RP and 8-SP) were
subjected to a cross-coupling with bromobenzene Supporting Information
(Table 5, entry 13). The 31P NMR spectra revealed
Characterization data for the synthesized arylphosphonates
that the reactions were completely stereospecific (8-
RP!13-RP and 8-SP!13-SP) and thus confirmed that
the developed procedure did not compromise the
usual stereospecificity of this type of cross-coupling
reactions.[3,5,7,37]
(Table 3, Table 4, and Table 5) and their 1H, 13C, and
31P NMR spectra are available as Supporting Information.
Acknowledgements
Conclusions
Financial support from the Swedish Research Council is
gratefully acknowledged.
We have developed a mild, and efficient procedure
for the synthesis of arylphosphonates that makes use
of an acetate-promoted, palladium-catalyzed cross-
coupling reaction of aryl halides (or triflates) with H-
phosphonate diesters. The reaction conditions of the
new protocol were optimized in terms of the support-
ing ligands, kinds and amounts of the acetate additive,
and the catalyst load. The method seems to be rather
general, accepts a wide range of electrophilic aryl
(iodide, bromide and triflate) derivatives, and H-phos-
phonate substrates (simple alkyl, cholesteryl, dinu-
cleosides), and may provide a convenient entry to
complex arylphosphonate diesters of biological impor-
tance.
References
[1] T. Hirao, T. Masunaga, Y. Ohshiro, T. Agawa, Tetrahe-
dron Lett. 1980, 21, 3595–3598; T. Hirao, T. Masunaga,
Y. Ohshiro, T. Agawa, Synthesis 1981, 56–57; T. Hirao,
T. Masunaga, N. Yamada, Y. Ohshiro, T. Agawa, Bull.
Chem. Soc. Jpn. 1982, 55, 909–913.
[2] Y. Xu, Z. Li, J. Xia, H. Guo, Y. Huang, Synthesis 1983,
377–378; Y. Xu, J. Zhang, Synthesis 1984, 778–780; Y.
Xu, Z. Li, J. Xia, H. Guo, Y. Huang, Synthesis 1984,
781–782; Y. Xu, J. Zhang, Tetrahedron Lett. 1985, 26,
4771–4774.
[3] Y. Xu, J. Zhang, J. Chem. Soc. Chem. Commun. 1986,
1606–1606.
[4] G. Guerrero, P. H. Mutin, F. Dahan, A. Vioux, J. Orga-
nomet. Chem. 2002, 649, 113–120; M. Terinek, A. Va-
sella, Helv. Chim. Acta 2004, 87, 719–734; M. A. Ka-
zankova, I. G. Trostyanskaya, S. V. Lutsenko, I. P. Be-
letskaya, Tetrahedron Lett. 1999, 40, 569–572; C. D.
Edlin, D. Parker, Tetrahedron Lett. 1998, 39, 2797–
2800; V. Penicaud, F. Odobel, B. Bujoli, Tetrahedron
Lett. 1998, 39, 3689–3692; W. M. Sharman, S. V. Kudre-
vich, J. E. Lier, Tetrahedron Lett. 1996, 37, 5831–5834;
H.-J. Cristau, A. Herve, F. Loiseau, D. Virieux, Synthe-
sis 2003, 2216–2220; X. Lu, J. Zhu, Synthesis 1987,
726–727; K. S. Petrakis, T. L. Nagabhushan, J. Am.
Chem. Soc. 1987, 109, 2831–2833; D. A. Holt, J. M.
Erb, Tetrahedron Lett. 1989, 30, 5393–5396; J. T. Suri,
D. D. Steiner, C. F. Barbas III, Org. Lett. 2005, 7, 3885–
3888; H. L. Ngo, A. Hu, W. Lin, J. Mol. Catal. A:
Chem. 2004, 215, 177–186; S. Oishi, S.-U. Kang, H.
Liu, M. Zhang, D. Yang, J. R. Deschamps, T. R. Burke,
Tetrahedron 2004, 60, 2971–2997; K. Muthukumaran,
R. S. Loewe, A. Ambroise, S. Tamaru, Q. Li, G.
Mathur, D. F. Bocian, V. Misra, J. S. Lindsey, J. Org.
Chem. 2004, 69, 1444–1452; W.-Q. Liu, C. Olszowy, L.
Experimental Section
General
All reagents and solvents were of analytical grade, obtained
from commercial suppliers and used without further purifi-
cation. THF was dried using a VAC solvent purifier system.
All reactions were carried out using standard Schlenk tech-
niques. Column chromatography was preformed on silica gel
(Grace Davison, Davsil, 0.035–0.070 mm). The NMR spectra
were registered using a Bruker Avance II 400 MHz instru-
ment. The chemical shifts are reported in ppm, relative to
solvent peaks (1H, 13C) or 2% H3PO4 solution in D2O
(31P NMR). Assignment of the NMR signals was done on
the basis of 2D correlation experiments (COSY, HSQC).
Experimental Procedure for the Acetate-Promoted
Preparation of Arylphosphonates
Pd
ACHTUNGTRENNUNG(OAc)2 (7 mg, 0.03 mmol), dppf (33 mg, 0.06 mmol) and
KOAc (13 mg, 0.13 mmol) were placed in the reaction
Adv. Synth. Catal. 2009, 351, 3207 – 3216
ꢁ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
3215

Marcin Kalek et al.
FULL PAPERS
Bischoff, C. Garbay, Tetrahedron Lett. 2002, 43, 1417–
1419; M. Kant, S. Bischoff, R. Siefken, E. Grꢂnde-
mann, A. Kçckritz, Eur. J. Org. Chem. 2001, 477–481;
P. Machnitzki, T. Nickel, O. Stelzer, C. Landgrafe, Eur.
J. Org. Chem. 1998, 1029–1034; S. S. Chauhan, A. Var-
shney, B. Verma, M. W. Pennington, Tetrahedron Lett.
2007, 48, 4051–4054.
[21] M. Kalek, J. Stawinski, Collection Symp.Series 2008, 10,
214–218.
[22] Y. Belabassi, S. Alzghari, J. L. Montchamp, J. Organo-
met. Chem. 2008, 693, 3171–3178.
[23] D. Prim, J.-M. Campagne, D. Joseph, B. Andrioletti,
Tetrahedron 2002, 58, 2041–2075; A. L. Schwan, Chem.
Soc. Rev. 2004, 33, 218–224.
[24] L. J. Goossen, M. K. Dezfuli, Synlett 2005, 445–448.
[25] I. P. Beletskaya, E. G. Neganova, Y. A. Veits, Russ. J.
Org. Chem. 2004, 40, 1782–1786.
[26] M. Kalek, A. Ziadi, J. Stawinski, Org. Lett. 2008, 10,
4637–4640.
[27] M. Kalek, J. Stawinski, Organometallics 2007, 26, 5840–
5848.
[28] C. Amatore, E. Carre, A. Jutand, M. A. M’Barki, G.
Meyer, Organometallics 1995, 14, 5605–5614; C. Ama-
tore, A. Jutand, Acc. Chem. Res. 2000, 33, 314–321.
[29] C. Amatore, A. Jutand, J. Organomet. Chem. 1999, 576,
254–278.
[30] C. Amatore, A. Jutand, A. Thuilliez, Organometallics
2001, 20, 3241–3249.
[5] G. Lavꢃn, J. Stawinski, Collection Symp. Series 2005, 7,
195–199.
[6] S. M. Zakeeruddin, M. K. Nazeeruddin, P. Pechy, F. P.
Rotzinger, R. Humphry-Baker, K. Kalyanasundaram,
M. Gratzel, V. Shklover, T. Haibach, Inorg. Chem.
1997, 36, 5937–5946; O. R. Evans, D. R. Manke, W.
Lin, Chem. Mater. 2002, 14, 3866–3874.
[7] T. Johansson, J. Stawinski, Chem. Commun. 2001,
2564–2565.
[8] T. Bock, H. Mçhwald, R. Mꢂlhaupt, Macromol. Chem.
Phys. 2007, 208, 1324–1340; K. D. Belfield, C. Chinna,
K. J. Schafer, Tetrahedron Lett. 1997, 38, 6131–6134; T.
Ogawa, N. Usuki, N. Ono, J. Chem. Soc. Perkin Trans.
1 1998, 2953–2958; S. Jin, K. E. Gonsalves, Macromole-
cules 1998, 31, 1010–1015.
[31] M. Kalek, J. Stawinski, Organometallics 2008, 27, 5876–
5888.
[32] C. Amatore, A. Jutand, M. A. M’Barki, Organometal-
lics 1992, 11, 3009–3013; C. Amatore, E. Carre, A.
Jutand, M. A. M’Barki, Organometallics 1995, 14,
1818–1826.
[9] B. Whittaker, M. Lera Ruiz, C. J. Hayes, Tetrahedron
Lett. 2008, 49, 6984–6987.
[10] K. J. Ullrich, G. Rumrich, T. R. Burke, S. P. Shirazie-
Beechey, G.-L. Lang, J. Pharmacol. Exp. Ther. 1997,
283, 1223–1229.
[11] S. Abbas, C. J. Hayes, Synlett 1999, 7, 1124–1126.
[12] S. Abbas, R. D. Bertram, C. J. Hayes, Org. Lett. 2001, 3,
3365–3367.
[13] M. R. Harnden, A. Parkin, M. J. Parratt, R. M. Perkins,
J. Med. Chem. 1993, 36, 1343–1355.
[14] T. Johansson, A. Kers, J. Stawinski, Tetrahedron Lett.
[33] A. Jutand, A. Mosleh, Organometallics 1995, 14, 1810–
1817.
[34] Since there was a remote possibility that lack of the re-
action between bromobenzene and diethyl H-phospho-
nate in the presence of high acetate concentration
(Table 2, entries 3 and 5) could be due to consumption
of bromobenzene in the reaction with acetate ions (see,
for example, T. Forngren, Y. Andersson, B. Lamm, B.
Langstrçm, Acta Chem. Scand. 1998, 52, 475–479), we
checked the reaction mixtures for a possible presence
2001, 42, 2217–2220.
[15] K. Zmudzka, T. Johansson, M. Wojcik, M. Janicka, M.
Nowak, J. Stawinski, B. Nawrot, New J. Chem. 2003, 27,
1698–1705.
1
of phenyl acetate. Since neither H NMR spectroscopy
[16] L. Kurz, G. Lee, D. Morgans, M. J. Waldyke, T. Ward,
nor TLC analysis revealed the presence of this com-
pound, such a possibility seemed to be less likely.
[35] S. A. Holstein, D. M. Cermak, D. F. Wiemer, K. Lewis,
R. J. Hohl, Bioorg. Med. Chem. 1998, 6, 687–694; W.
Jiang, G. Allan, J. J. Fiordeliso, O. Linton, P. Tannen-
baum, J. Xu, P. Zhu, J. Gunnet, K. Demarest, S.
Lundeen, Z. Sui, Bioorg. Med. Chem. 2006, 14, 6726–
6732.
Tetrahedron Lett. 1990, 31, 6321–6324.
[17] Y.-Y. Yan, T. V. RajanBabu, J. Org. Chem. 2000, 65,
900–906; G. Bringmann, A. Wuzik, M. Breuning, P.
Henschel, K. Peters, E.-M. Paters, Tetrahedron: Asym-
metry 1999, 10, 3025–3031.
[18] A. L. Casalnuovo, J. C. Calabrese, J. Am. Chem. Soc.
1990, 112, 4324–4330.
[19] R. A. Stockland Jr. , A. M. Levine, M. T. Giovine, I. A.
Guzei, J. C. Cannistra, Organometallics 2004, 23, 647–
656; M. C. Kohler, T. V. Grimes, X. Wang, T. R. Cun-
dari, R. A. Stockland, Organometallics 2009, 28, 1193–
1201.
[36] K. L. Agarwal, F. Riftina, Nucleic Acids Res. 1979, 6,
3009–3024; J. W. Engels, J. Parsch, Nucleic Acid Drugs,
in: Molecular Biology in Medicinal Chemistry, (Eds.: T.
Dingermann, D. Steinhilber, G. Folkers), Wiley-VCH,
Weinheim, 2005.
[20] S. Abbas, C. J. Hayes, Tetrahedron Lett. 2000, 41, 4513–
4517.
[37] J. Zhang, Y. Xu, G. Huang, H. Guo, Tetrahedron Lett.
1988, 29, 1955–1958.
3216
asc.wiley-vch.de
ꢁ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2009, 351, 3207 – 3216