Chemistry of palladium phosphinite (PPh₂(OR)) and phosphonite (P(OPh)₂(OH)) complexes: catalytic activity in methoxycarbonylation and Heck coupling reactions

Abstract

The new phosphinite and phosphonite complexes (1–8) are very efficient catalysts for the methoxycarbonylation of iodobenzene and Heck cross-coupling of bromobenzene with butyl acrylate. High catalytic activity of these complexes can be explained by their in situ transformations during the reaction, stimulated by the presence of water, acid (HCl) or base (NEt₃). Hydrolysis of phosphinite palladium complexes of the form trans-PdCl₂[PPh₂(OR)]₂ (R = C₆F₅2, ^tBu 3, or O-menthyl 4) results in the formation of the dimeric complex [μ-ClPd(PPh₂OH)(PPh₂O)]₂5, which is deprotonated by NEt₃, producing a polymeric complex of formula [Pd(P(O)PPh₂)₂]_n8. The reverse reaction, protonolysis of 8 with HCl, leads back to 5 and the monomeric complex 5a. The phosphinite complex PdCl₂[PPh₂(OBu)]₂1 with a more lipophilic ligand, PPh₂(OBu), does not undergo hydrolysis under the same conditions. In the reaction of PdCl₂(cod) with P(OPh)₂(OH), the new dimer [μ-ClPd(P(OPh)₂OH)(P(OPh)₂O)]₂6 was obtained, whereas reaction of Pd(OAc)₂ with P(OPh)₂(OH) leads to the polymeric complex [Pd[P(O)(OPh)₂]₂]_n7. Protonolysis of 7 with HCl results in the formation of 6.

---

Introduction

Phosphinites (PPh₂(OR)) and hydrogen phosphonates (HP(O)(OR)₂) represent an interesting alternative to phosphines as ligands for metal ions used as catalysts in organic synthesis. Their practical application in catalytic systems is however complicated by sensitivity towards hydrolysis, leading to a change in the ligand and catalyst structure. This report demonstrates an improvement in catalytic activity as a result of such transformations initiated by the presence of water as well as HCl or NEt₃.

Phosphinites as free ligands can be easily prepared in the reaction of PPh₂Cl with alcohols or diols (eqn (1)).^1,2

PPh₂Cl + ROH → PPh₂(OR) + HCl (1)

Alternatively, phosphinites, specifically vicinal diphenylphosphinites, can be synthesized through metal template synthesis, as has been demonstrated for reactions of MCl₂(PPh₂Cl)₂ complexes (M = Pd, Pt) with diols, leading to square planar complexes containing chelating diphosphinite (eqn (2)).^3,4

Metal template synthesis is particularly useful for the preparation of optically active derivatives which can then be used as catalysts in chiral synthesis. Recently, several applications of such catalysts have been reported, mainly for the enantioselective formation of C–C and C–heteroatom bonds through allylic substitution reactions.^5–9 A palladium complex with phosphinite-oxazoline ligands derived from D-glucosamine worked efficiently in asymmetric allylic substitution.⁵ Good results, represented by high ee values, have also been obtained in this reaction with a palladium catalyst containing furanoside thioether-phosphinite,⁶ amino-phosphonites,⁷ as well as N,P-donor oxazoline phosphinites.^8,9 Surprisingly, in spite of their generally higher air-stability compared to phosphines, phosphinites have seldom been used in palladium-catalyzed C–C bond forming reactions, such as methoxycarbonylation of aryl halides or Heck or Suzuki coupling. Catalytic systems containing Pd(dba)₂ and PPh₂(O-menthyl) or aminophosphine phosphinite have been used in the carbonylation of α-methylbenzyl bromide under phase-transfer conditions but exhibited low activity and poor optical yield compared to aminophosphine or 2-substituted-3,1,2-oxazaphospholanes.¹⁰ Another example of the application of phosphinite palladium complexes in carbonylation reactions, although a non-asymmetric one, has been published by us.¹¹ It was found that the yield of phenylacetic acid methyl ester formed by methoxycarbonylation of benzyl bromide depends on the structure of the phosphinite ligand coordinated to palladium in the PdCl₂P₂ complex, and generally better results were obtained with ligands with less steric hindrance (P = PPh₂(OMe), PPh₂(OBu)).¹¹ The activity of palladium phosphinite complexes in Heck and Suzuki reactions has been illustrated mostly by complexes with phosphinite PCP pincer ligands. A phosphinito palladium PCP pincer complex efficiently catalyzed Heck coupling of aryl chlorides with styrene at 120–180 °C.¹² Suzuki coupling of aryl bromides with a bis(phosphinite) PCP pincer palladium complex¹³ and with P,C-bidentate phosphinite palladacycles¹⁴ have been performed successfully. A palladium complex with a hemilabile amino- and sulfur-containing phosphinite ligand has been found to catalyze the Heck reaction of aryl bromides with styrene at 130 °C, with high turnover numbers and a good selectivity towards trans-stilbenes.¹⁵ An air-stable phosphinite ligand, P(^tBu)₂(OH), has been found as an ideal ligand for palladium catalysts for a variety of C–C, C–N, and C–S bond forming reactions,¹⁶ also with aryl chlorides as substrates.^17,18 Arylation of 2,3-dihydrofuran using a palladium complex with a phosphinite-oxazoline ligand originating from D-glucosamine proceeded smoothly to give 78–96% ee; however, it gave lower ee in enantioselective phenylation of cis- and trans-crotyl alcohols.¹⁹ A supported phosphinite ligand obtained from a polymer-bound aminoalcohol template reacted with Pd(OAc)₂ and the obtained complex was used in the Heck reaction.²⁰

The role of hydrolytic transformations of phosphinites under Suzuki reaction conditions with orthometallated palladium phosphinito complexes as catalysts has been found to be important and dependent on the kind of ligand and palladium precursor.¹³

It has been discovered that PPh₂(OH) (in the form HP(O)Ph₂) can be coupled with alkynes in hydrophosphinylation reactions catalyzed by palladium complexes.^21,22 In a similar application, hydrophosphorylation of alkenes and alkynes leading to phosphonate derivatives, hydrogen phosphonates existing in two tautomeric forms have been found (eqn (3)).^23–25

Simple substrates such as HP(O)(OR)₂ (R = alkyl, aryl) are often unreactive, but the addition of α-pinacol-derived hydrogen phosphonate to a variety of olefins in the presence of palladium complexes has been successful.^24,25 Palladium-catalyzed hydrophosphorylation of alkynes with HP(O)(OEt)₂ and HP(O)(OMe)₂ has also been reported.²³ The formation of MeP(O)(OPh)₂ by reductive elimination from palladium complexes of the type L₂PdMe(P(O)(OPh)₂) (L = PMePh₂, PMe₂Ph, PEt₃etc.), a key step in P(O)–C(sp³) bond forming reactions, has been investigated, and the effect of steric properties of the ligand L on the reaction rate has been demonstrated.^26,27

During our earlier studies on the catalytic activity of palladium phosphinite complexes in the methoxycarbonylation of benzyl bromide, we observed that some of these complexes were unstable in solution, and their recrystallization was seriously complicated by unidentified structural changes.¹¹ It seemed interesting to investigate and explain the nature of those transformations and to determine their influence on the catalytic activity of the systems. Literature data concerning the catalytic activity of palladium complexes with monodentate phosphinites are rather scarce, and even less is known about the catalytic applications of palladium phosphonate complexes, which have been tested only in phosphorylation reactions.^23–25 Therefore, we chose to study not only methoxycarbonylation but also the Heck reaction (Scheme 1). Both reactions proceed in the presence of base, acting as scavengers of HX formed as a side product and therefore, the catalyst should suffer basic as well as acidic conditions without decomposition. The studies of reactions of palladium phosphinito and phosphonato complexes with NEt₃ and HCl has allowed the determination of the stability of catalysts.

Scheme 1 Methoxycarbonylation and Heck reaction patterns.

The important part of these studies is related to the elementary steps of the catalytic processes, common to both reactions (Scheme 1), namely reactions of Pd-catalyst precursors with NEt₃, representing the base present in both catalytic systems. The eventual reduction of Pd(II) to Pd(0) with contributions from amine was expected, just as has been reported for PdCl₂[P(OPh)₃]₂ (eqn (4)).²⁸

PdCl₂P₂ + N(C₂H₅)₃ + H₂O → [Pd⁰] + 2 HCl + CH₃CHO +HN(C₂H₅)₂ + 2 P (4)

P = P(OPh)₃

PdCl₂[P(OPh)₃]₂ reacted with NEt₃ in the presence of H₂O with the formation of a Pd(0) complex according to reaction (4).²⁸

Results and discussion

Hydrolysis of Pd(II) phosphinite complexes of the type PdCl₂[PPh₂(OR)]₂ (2–4)

The phosphinito palladium(II) complexes selected for these studies, of the PdCl₂P₂ type (P = PPh₂(OR)), can be divided into cis and trans isomers depending on the steric hindrance of the phosphinite ligand¹¹ (Scheme 2). Only PdCl₂[PPh₂(OⁿBu)]₂1 forms a cis isomer, and for a phosphorus ligand P = PPh₂(C₆F₅) 2, PPh₂(O^tBu) 3, or PPh₂(O-menthyl) 4, the formation of trans isomers was confirmed.¹¹ The structure of the complexes was determined from spectroscopic data, and in the case of 4, preliminary X-ray data indicated trans coordination of phosphinite ligands, but the quality of the crystals was not sufficient to resolve the final structure.

Scheme 2

During the measurements of ³¹P NMR spectra of palladium(II) phosphinite complexes 2, 3, and 4 in solution, it was observed that all the complexes were transformed to the same compound 5 with a characteristic signal at 78 ppm.

The new compound was detected after one hour in a solution of incompletely dried CDCl₃. The observed position of the signal in the ³¹P NMR spectrum (78 ppm) is similar to that reported in the literature (78.1 ppm) for the dimer 5 of formula [μ-ClPd(PPh₂OH)(PPh₂O)]₂;^3,29 however, very similar data (79.6 ppm) have also been published for monomeric cis-PdCl₂[PPh₂(OH)]₂,¹³ complex 5a.

To determine which structure, dimeric 5 or monomeric 5a, is correct, complexes 2 and 4 were recrystallized respectively from a mixture of CHCl₃ with EtOH and from only C₆H₆, and the obtained crystals were analyzed by X-ray crystallography. The complex, which recrystallized from a CHCl₃ and EtOH mixture, was identified as the dimer 5 with 2 molecules of CHCl₃ in the lattice, identical with that described in ref. 29 (Scheme 3). The process of the formation of 5 in ref. 29 was based on the reaction of the phosphinite complex cis-PdCl₂[PPh₂OCH₂)₂·ⁿBuCl_x (where ⁿBuCl_x is tetra-n-butoxycalix[4]arene) with HCl. The dimer 5 was also been obtained earlier from the reaction of PdCl₄²⁻ with PPh₂(OH).³⁰ The recrystallization of 4 from C₆H₆ produced the same dimer 5 with one molecule of benzene in the crystal lattice (Fig. 1, Tables 1, 2). It is interesting to note that palladium phosphinite complexes undergo hydrolysis even with traces of water but not solvolysis in ethanol solution, and in consequence the formation of a palladium complex with the expected solvolysis product, PPh₂(OEt), was not proved experimentally.

Table 1 Bond lengths (Å) and angles (°) in 5^a

a Symmetry transformation #1 −x, −y, −z defines center of symmetry located in position 0,0,0.
Pd(1)–P(1)	2.2461(5)	Pd(1)–Cl(1)#1	2.4216(4)
Pd(1)–P(2)	2.2499(5)	Cl(1)–Pd(1)#1	2.4216(4)
Pd(1)–Cl(1)	2.4166(4)
P(1)–Pd(1)–P(2)	91.789(17)	P(2)–Pd(1)–Cl(1)#1	90.615(16)
P(1)–Pd(1)–Cl(1)	92.973(16)	Cl(1)–Pd(1)–Cl(1)#1	84.701(16)
P(2)–Pd(1)–Cl(1)	174.768(16)	Pd(1)–Cl(1)–Pd(1)#1	95.299(16)
P(1)–Pd(1)–Cl(1)#1	176.917(17)

---

Table 2 Hydrogen-bonds for 5 [(Å) and angles (°)]

Symmetry transformations used to generate equivalent atoms: #1 −x, −y, −z.
D–H⋯A	d(D–H)	d(H⋯A)	d(D⋯A)	<(DHA)
O(1)–H(1)⋯O(2)	1.04(3)	1.39(3)	2.4058(17)	164(3)

---

Fig. 1 Molecular view of [μ-ClPd(PPh₂OH)(PPh₂O)]₂, 5.

Scheme 3 Hydrolysis of phosphinite palladium complexes and their reactions with NEt₃ and HCl.

By analogy to the pathway proposed for the hydrolysis of phosphites,³⁰ the reaction of a phosphinite complex with water is presented in Scheme 4. Such a reaction of complexes 2, 3, or 4 would provide complex 5a, which eventually formed the dimer 5 after the release of two HCl molecules per 1 mol of the dimer 5.

Scheme 4 Hydrolysis of Pd–phosphinite complex.

Transformation of complex 4 to the dimer 5 was examined with great care to elucidate the reaction mechanism. Comparison of ¹H NMR spectra (in the region 0–2 ppm) of complex 4 with those of menthyl alcohol and free phosphinite PPh₂(O-menthyl) led to the conclusion that the formation of menthyl alcohol, identified by the doublet of the CH₃ group at 0.73 ppm, J(H–H) 7 Hz, versus a doublet at 0.4 ppm, J(H–H) 7 Hz, in 4, was simultaneous with the formation of the dimer 5. The formation of the monomeric complex 5a as well as dissociation of the phosphinite ligand were not observed. The lack of phosphinite ligand dissociation made it possible to assume that hydrolysis proceeded without Pd–P bond breaking.

To better understand the equilibrium between complexes 5 and 5a, an experiment described in ref. 3 was repeated, and reaction of in situ prepared complex cis-PdCl₂(PPh₂Cl)₂ (obtained in the reaction of PdCl₂(cod) with PPh₂Cl) with water was studied. The ³¹P NMR spectrum of the isolated crude product showed only one signal at 76.8 ppm (close to that observed earlier at 78 ppm) confirming the presence of the dimer 5 only.

It is worth noting that, in contrast to the phosphinite complexes (2, 3, 4), complex 1 does not undergo hydrolysis, which may be explained by the difficult access for water molecules caused by the lipophilic character of the butoxy group. In the case of the remaining complexes, the attack of water molecules is facilitated by attraction of electron density from phosphorus by the R group in the PPh₂(OR) ligand.

Reactions of phosphinite palladium complexes (2–5) with NEt₃ and HCl

The experiments were performed in NMR tubes using a ten-fold excess of NEt₃versus the palladium complex. A few minutes after the addition of NEt₃ to the solution of one of the complexes, 2, 3, 4, or 5, a precipitate of a new compound appeared in the tubes and was subsequently identified as the polymeric complex 8 containing P(O)Ph₂⁻ anionic ligands coordinated to palladium (Scheme 3). This compound had been described earlier as the product of deprotonation of the dimer 5 after the addition of a base.³⁰ In our case NEt₃ played the role of a deprotonating agent that fixed HCl formed during the reaction. This was confirmed by the presence in the IR spectrum of frequencies characteristic for NEt₃·HCl (2500, 2620, 2700 cm⁻¹). In the IR spectrum of 8, a band at 1100 cm⁻¹, assigned to ν(P=O), was observed, whereas no bands in the region 200–400 cm⁻¹, corresponding to ν(Pd–Cl), were found. The very low solubility of 8 in common solvents does not make it possible to record its ³¹P NMR spectrum. However, when a suspension of 8 in CDCl₃ was treated with a few drops of concentrated HCl in water, the mixture became clear, and its ³¹P NMR spectrum presented two signals, at 76.8 ppm (the higher intensity) and at 84.9 ppm, due to the dimer 5 and the monomer 5a respectively. It was the only case when we were able to see together both complexes, 5 and 5a, in solution.

Formation of Pd–phosphonate complexes [μ-ClPd(P(OPh)₂OH)(P(OPh)₂O)]₂6 and [Pd[P(O)(OPh)₂]₂]_n7

The reactions of diphenyl phosphite P(OPh)₂(OH) with PdCl₂(cod) and Pd(OAc)₂ were studied in solution using ³¹P NMR spectrum measurements, and different products were found in each system (Scheme 5). When the PdCl₂(cod) complex reacted with a 2-fold excess of diphenyl phosphite, the evolution of the cod ligand was confirmed by the appearance of two singlets in ¹H NMR (2.36 and 5.59 ppm). In the ³¹P NMR spectrum of the reacting solution recorded in situ, an intensive signal at 65.5 ppm and a much less intensive one at 78.1 ppm were observed. Further addition of P(OPh)₂(OH) did not change the signal ratio, and only a free phosphite signal appeared at −2 ppm. In the ³¹P NMR spectrum of the isolated solid palladium complex obtained in reaction of PdCl₂(cod) with P(OPh)₂(OH) (Scheme 5), only one signal at 65.5 ppm was observed which was tentatively assigned to the dimer 6. In the IR spectrum of complex 6, two bands at 380.7 and 339.0 cm⁻¹, assigned to ν(Pd–Cl) stretching frequencies, were observed, and the dimeric structure 6 was supported by the presence of a peak at m/z 1184 in ESI-MS attributed to the [M − Cl]⁺ fragment.

Scheme 5 Reactions of palladium complexes with phosphonate ligands.

Reaction of Pd(OAc)₂ with P(OPh)₂(OH) leads selectively to the formation of complex 7, which has a polymeric structure with phosphonate ligands coordinating as P(O)(OPh)₂⁻ anions. Such a structure of 7 is supported by the characteristic position of a ³¹P NMR signal at 86.8 ppm due to the P(O)(OPh)₂⁻ moiety²⁶ and the presence of an intensive IR band at 1100 cm⁻¹ assigned to ν(P=O), which was not observed in the IR spectrum of 6. The ³¹P NMR spectrum of 7 was measured only in situ, in solution containing Pd(OAc)₂ and P(OPh)₂(OH) ligands. It is worth noting that complex 7, after having been isolated in the solid state, is insoluble in any available solvents used for NMR measurements and therefore its NMR spectra cannot be measured.

Attempts to transform the dimer 6 to the polymeric complex 7 by deprotonation with NEt₃, similar to the phosphinite complexes 5 and 8, failed. When complex 6 in CDCl₃ was treated with NEt₃, some new signals appeared in the ³¹P NMR spectrum of the solution in the region of 54–102 ppm, but a signal originating from complex 7 could not be found. Moreover, 7 can be easily transformed into 6 in a reaction with concentrated HCl, as was confirmed by the appearance of only one signal at 65.5 ppm in the ³¹P NMR spectrum of a CDCl₃ solution obtained by treatment of the suspension of 7 with HCl.

The new route for the synthesis of palladium phosphonate complexes from Pd(OAc)₂ and P(OPh)₂(OH) as substrates makes it possible to obtain products containing the P(O)(OPh)₂⁻ anionic ligand with high selectivity. For comparison, recently a number of monomeric palladium complexes of the Pd(Me)L₂(P(O)(OPh)₂) type have been described^26,27 which were obtained with the application of AgP(O)(OPh)₂ as the source of the phosphonate ligand.

Catalytic properties of palladium phosphinite and phosphonite complexes 1–8

Methoxycarbonylation of α-methylbenzyl bromide, benzyl bromide, and benzyl chloride catalyzed by 4. The obtained results confirmed the rather low catalytic activity of 4 in carbonylation of α-methylbenzyl bromide observed earlier,¹⁰ despite different reaction conditions and different co-catalysts used. The yield of ester did not exceed 3% (at 60 °C, 5 atm CO), even in the presence of n-tetrabutyloammonium iodide as co-catalyst. Instead of ester, the main products were styrene and ether.

In spite of the very low activity of 4 in the carbonylation of α-methylbenzyl bromide, we checked its reactivity in the catalytic methoxycarbonylation of benzyl chloride and benzyl bromide (Scheme 6), and the yield of methyl phenylacetate obtained in these reactions is given in Table 3.

Table 3 Methoxycarbonylation of benzyl chloride and bromide catalyzed by 4

Substrate PhCH2X (X =)	T/°C	p(CO)/atm	Co-catalyst	Yield of ester (%)
Reaction conditions: [Pd] = 5 × 10^−5 mol; [PhCH2X] = 9 × 10^−3 mol; [[^nBu4N]I] : [PhCH2X] = 1; [MeOH] : [PhCH2X] = 8.3; [NEt3] : [PhCH2X] = 2.2; 3 h.
Cl	60	5	—	15
Br	40	1	—	27
Br	60	5	[^nBu4N]I	41

---

Scheme 6 Methoxycarbonylation of benzyl halides.

Methoxycarbonylation of benzyl chloride compared with that of benzyl bromide required more stringent conditions. The yield of ester obtained from benzyl chloride in reaction catalyzed only by complex 4 at 60 °C, 5 atm CO for 3 h was 15%, whereas methoxycarbonylation of benzyl bromide gave 27% of ester at 40 °C under 1 atm CO. The addition of n-tetrabutyloammonium iodide made it possible to obtain 41% of ester. The positive effect of ammonium halides on the yield of methoxycarbonylation was also observed in other systems, and therefore further catalytic tests were performed in the presence of [ⁿBu₄N]Cl.

Methoxycarbonylation of iodobenzene catalyzed by complexes 1–8. To compare the catalytic activity of all of the studied palladium complexes (1–8), phosphinite and phosphonite ones, methoxycarbonylation reactions of iodobenzene were carried out under the same reaction conditions (90 °C, 5 atm CO) with ammonium salts as co-catalysts (Scheme 7).

Scheme 7 Methoxycarbonylation of iodobenzene.

It was expected that comparison of the obtained results would enable the determination of the impact of hydrolysis on the catalytic activity of the complexes under study and the relation between the molecular structure of the palladium complexes and their catalytic activity. Additionally, a comparison of the catalytic activity of phosphinite and phosphonite complexes could help in the estimation of the electronic effect of the respective phosphorus ligands on their reactivity.

Almost all of the palladium complexes under study presented high or very high catalytic activity in the test reaction: methoxycarbonylation of iodobenzene (the yield of methyl benzoate amounting to 66–100%, which corresponds to a TON of 400–600) (Table 4). The lowest yield of ester was obtained for complex 8 (46%), in contrast to 92% produced by the dimer 5. This indicates that the transformation of 8 into 5, which is quite fast in the presence of HCl, is not complete during the catalytic process.

Table 4 Methoxycarbonylation of iodobenzene catalyzed by palladium phosphinite (1–5 and 8) and phosphonite (6, 7) complexes

Catalyst	Yield of ester (%)	Catalyst	Yield of ester (%)
Reaction conditions: [Pd] = 1.41 × 10^−5 mol; [PhI] = 9.4 ×10^−3 mol; [[^nBu4N]Cl] : [PhI] = 1; [NEt3] : [PhI] = 2.2; 90 °C, 3 h
1	81	5	92
2	66	6	100
3	70	7	100
4	95	8	46

---

Generally the catalytic activity of palladium phosphinite complexes 2 and 3, as represented by the yield of ester, is high, though slightly lower than that of complex 5, the final product of their hydrolysis. One can conclude that hydrolysis is not very efficient under carbonylation reaction conditions, because of the presence of methanol as well as some water in the hygroscopic ammonium salt. Similar yields of ester obtained in reactions catalyzed by 4 and 5 indicate probable total hydrolysis of 4 during the reaction.

The high catalytic activity of both phosphonate complexes, 6 and 7, suggests that they may be perfect methoxycarbonylation catalysts. Their additional advantage is high resistance to pH changes, as was demonstrated in reactions with NEt₃ and HCl.

Heck coupling of bromobenzene with butyl acrylate catalyzed by complexes 1–8. Heck reactions of bromobenzene with n-butyl acrylate were performed at 130 °C with [ⁿBu₄N]Br as co-catalyst. The application of a 2.3-fold excess of PhBr versus n-butyl acrylate facilitated the formation of not only a trans-monoarylated (a) but also a diarylated product (b) (Scheme 8).

Scheme 8 Heck cross-coupling.

The activities of the catalysts under study were compared on the basis of the reaction product ratio as well as the stability of the catalytic systems in repetitive experiments (runs 1–3). Higher-activity catalysts produced more of the diarylated product and were successfully recycled several times without a significant lowering of the reaction yield (Table 5).

Table 5 The yield of Heck coupling reactions catalyzed by palladium phosphinite (1–5, 8) and phosphonite complexes (6, 7) (Scheme 8)

Catalyst	Run 1		Run 2		Run 3
Product	Mono (a)	Di (b)	Mono (a)	Di (b)	Mono (a)	Di (b)
Reaction conditions: [Pd] = 1.41 × 10^−5 mol; [PhBr] = 9.4 × 10^−3 mol; [CH2=CHC(O)OBu] = 4.36 × 10^−3 mol; [[^nBu4N]Br] = 2.3 × 10^−3 mol; [NaHCO3] = 4.4 × 10^−3 mol; 140 °C, 4 h.
1	64	36	68	27	63	37
2	58	42	86	10	39	61
3	75	25	86	11	91	5
4	50	50	63	37	20	3
5	75	25	29	71	40	60
6	45	55	80	20	81	7
7	62	38	65	35	76	24
8	41	59	68	30	61	0

---

The results presented in Table 5 indicate very good catalytic performance with all the catalysts studied. In all reactions, even after catalyst recycling, the conversion of butyl acrylate was 95–100%. In most cases the diarylated product was formed with reasonable yield even in the third run, confirming the high activity and stability of the systems. Comparison of results obtained with phosphinite complexes 2–4 and with the dimer 5 leads to the conclusion that the dimer is more active and hydrolysis of complexes 2–4 is a relatively slow process under the Heck reaction conditions, like in the case of methoxycarbonylation. Complex 8, which presented the lowest productivity in methoxycarbonylation, appeared to be quite a good catalyst in the Heck coupling. Both phosphonate complexes, 6 and 7, also gave very good results, 7 being slightly better than 6 if the yield of the diarylated product is considered.

Conclusions

The reported results demonstrate that phosphonite and phosphonato palladium complexes are ideal catalysts of reactions of aryl halides, representing high activity, stability and recyclability.

Phosphonite palladium complexes of formula PdCl₂(PPh₂(OR))₂, where R = C₆F₅2, ^tBu 3, or O-menthyl 4, undergo hydrolysis in the presence of water and form the dimeric complex [μ-ClPd(PPh₂OH)(PPh₂O)]₂5. The catalytic activity of 5 in methoxycarbonylation and in the Heck cross-coupling is higher than that of complexes 2–4, which means that hydrolysis of phosphinite complexes improves the catalytic systems.

The phosphinite complex PdCl₂(PPh₂(OBu))₂1, which is not hydrolyzed in the presence of water, exhibited catalytic properties similar to those of 5.

Deprotonation of 5 by NEt₃ led to the polymeric species 8, containing P(O)Ph₂⁻ moieties. The reaction is reversible, and the addition of HCl led to the recovery of 5. These reactions are very important in catalytic systems in which a base is always present and HX is formed as a side product during methoxycarbonylation as well as during the Heck reaction.

The phosphonato complexes, [μ-ClPd(P(OPh)₂OH)(P(OPh)₂O)]₂6 and [Pd[P(O)(OPh)₂]₂]_n7, represent a new class of highly active and stable catalysts of both methoxycarbonylation and Heck coupling reactions. The slightly soluble polymeric species 7 can be easily transformed to the soluble dimer 6 in acidic conditions, as was demonstrated by its reaction with HCl. This transformation, very probable also under catalytic reaction conditions, can stabilize palladium in its active form.

Experimental

Syntheses

Palladium complexes 1, 2 and3 were obtained according to the method described in ref. 11.

PPh₂(O-menthyl). 10 mmol of L-menthol was dissolved in THF under an atmosphere of N₂ and then 10 mmol of n-butyllithium in hexane was added dropwise with efficient cooling. The reaction mixture was vigorously stirred for 1 h, cooled to −78 °C and then 10 mM of the phosphine PPh₂Cl dissolved in a few cm³ of THF was added dropwise. The cooling bath was removed and the mixture was stirred for 24 h. After that time THF was removed and the viscous oil obtained was dissolved in ca. 20 cm³ of diethyl ether and filtered through a pad of Celite to remove LiCl. The ether was next removed under vacuum or the phosphinite was used for further syntheses in ether solution. MS: 340 (M⁺), 201 (OPPh₂), 183 (PPh₂), 155 (menthyl). ³¹P NMR (CDCl₃): 107.3 ppm.

PdCl₂[PPh₂(O-menthyl)]₂, 4. To 0.11 g of PdCl₂(cod) in 2 cm³ of CH₂Cl₂ 0.5 cm³ of diethyl ether containing 1 mmol of PPh₂(O-menthyl) was added. The yellow solution was stirred for 15 min and evaporated to dryness. The residue was washed with hexane giving a pale yellow precipitate. Yield: 76% Anal. Calc. for PdC₄₄H₅₈O₂P₂Cl₂: C, 61.6; H, 3.38. Found: C, 61.1; H, 3.5%. ¹H NMR (CDCl₃): 0.4 d (J(H–H) 7 Hz), 0.81 d (J(H–H) 7 Hz), 0.87 d (J(H–H) 7 Hz), 1.5 m, 2.53 m, 3.27 m, 5.26 m, 7.06 m, 7.92 m; ³¹P NMR (CDCl₃): 100.2 ppm; UV-Vis (CHCl₃): 264, 340 nm; IR/KBr (cm⁻¹): 2900vs, 1420s, 1100s, 1000vs, 680s, 530s, 500s; IR/Nujol (cm⁻¹): 316.

[μ-ClPd(PPh₂OH)(PPh₂O)]₂, 5. To 0.066 g PdCl₂(cod) 0.2 cm³ of PPh₂Cl in 2 cm³ of benzene was added and the resultant mixture was stirred for 2 h. Next H₂O (0.05 cm³) was added and after 5 min 2 cm³ of EtOH. Stirring was continued for 1 h and the resultant white precipitate was filtered off, washed with ethyl ether and dried. Yield: 0.17 g. Anal. Calc. For Pd₂Cl₂C₄₈H₄₂O₄P₄: C, 52.89; H, 3.70 Found: C, 52.64; H, 3.50%. IR/KBr (cm⁻¹): 1480vs, 1100vs, 1015m, 730m, 710m, 680vs, 500m; IR/Nujol (cm⁻¹): 333.3, 275.0 (ν(Pd–Cl)); UV-Vis (CHCl₃): 350 nm; ¹H NMR(CDCl₃): 7.15m (Ph), 10.0 br (OH) ppm. ³¹P NMR(CDCl₃): 76.6 ppm (at 500 MHz with methylenediphosphonic acid as external standard), 78.0 ppm (at 300 MHz with 85% H₃PO₄ as external standard).

[μ-ClPd(P(OPh)₂OH)(P(OPh)₂O)]₂, 6. To 0.07 g of PdCl₂(cod) in 1 cm³ of benzene 0.12 cm³ of P(OH)(OPh)₂ was added and a white precipitate formed immediately. The mixture was stirred for 1 h and after that time the product was filtered off, washed with diethyl ether and dried in vacuo. Yield: 0.09 g. Anal. Calc. For Pd₂Cl₂C₄₈H₄₂O₁₂P₄: C, 47.4; H, 3.3 Found: C, 47.1; H, 3.19%. IR/KBr (cm⁻¹): 1600s, 1480vs, 1200vs, 900vs, 750m, 700m, 500m; IR/Nujol (cm⁻¹): 381.7, 338.0 (ν(Pd–Cl)); ¹H NMR(CDCl₃): 7.15m (Ph), 10.0 br (OH) ppm; ³¹P NMR(CDCl₃: 65.5 ppm.

[Pd[P(O)(OPh)₂]₂]_n, 7. To 0.22 g of Pd(OAc)₂ in 2 cm³ of benzene 0.4 cm³ of P(OH)(OPh)₂ was added and the resultant mixture was stirred at ambient temperature. After a few minutes the solution became homogeneous and colourless, then a white precipitate started to form. Stirring was continued for 2 h and next diethyl ether was added to complete precipitation of product which was filtered off and dried in vacuo. Yield: 0.38 g. IR/KBr (cm⁻¹): 1600s, 1490vs, 1200vs, 1100vs, 920vs, 760m, 700m, 500m; IR/Nujol(cm⁻¹): 3400 (ν(OH)); ³¹P NMR (CDCl₃, measured in situ): 86.8 ppm.

[Pd[P(O)Ph₂)]₂]_n, 8. To a solution of complex 5 (0.05 g) in 1 cm³ of CHCl₃ triethylamine (0.05 cm³) was added dropwise. After a few minutes a pale yellow precipitate was formed, which was filtered off, washed with CHCl₃ and dried under vacuum. Yield; 0.038 g. IR/KBr (cm⁻¹): 1437s, 1108s, 1022vs, 742m, 713m, 690s, 535s.

Measurements

¹H, and ³¹P NMR spectra were recorded on Bruker 300 and Bruker 500 spectrometers. For ³¹P NMR at 300 MHz 85% H₃PO₄ was used as an external standard and at 500 MHz methylenediphosphonic acid was used as an external standard.

FTIR spectra were measured on Nicolet Impact 400 spectrometer.

GC-MS an GC analyses were made on a Hewlett Packard 8452A instrument.

Catalytic reactions

The Heck reactions were carried out in a Schlenk tube with magnetic stirring. The catalyst (1.41 × 10⁻⁵ mol) was weighed in a small Teflon vessel, [ⁿBu₄N]Br (2.3 × 10⁻³ mol) and NaHCO₃ (4.4 × 10⁻³ mol) were introduced to the Schlenk tube which was next evacuated and filled with N₂. Reagents: 1.89 × 10⁻³ mol of butyl acrylate, 4.36 × 10⁻³ mol of bromobenzene and 7.15 × 10⁻⁴ mol of mesitylene (internal standard) were introduced under an N₂ atmosphere. Next the Schlenk tube was sealed with a rubber tap and introduced into an oil bath pre-heated to 140 °C for the reaction time 4 h. After that time the Schlenk tube was cooled and any organic products were separated by extraction with diethyl ether (3 times with 10 cm³) and GC-FID analyzed (Hewlett Packard 8452A). The products were identified by GC-MS (Hewlett Packard 8452A). ¹H NMR spectroscopy was used as a complementary method of product identification as well as determination of the reaction yield. Samples for ¹H NMR measurements were prepared in CDCl₃ after vacuum removal of diethyl ether from the extract.

For recycling experiments the solid residue obtained after the extraction of the liquid products was dried in vacuum, next the liquid reagents and NaHCO₃ in the amounts given above were added under a N₂ atmosphere.

The methoxycarbonylation reactions were carried out in a 130 cm³ thermostated steel autoclave with magnetic stirring. Reagents: PhI (9.4 × 10⁻³ mol), NEt₃ (2.15 × 10⁻² mol), methanol (2.5 × 10⁻² mol), mesitylene-internal standard (4.6 × 10⁻³ mol), and palladium catalyst (1.41 × 10⁻⁵ mol), were introduced into the autoclave under an N₂ atmosphere. Next, the N₂ atmosphere was replaced by CO (5 atm). The reaction was carried out at 90 °C for 3 h. Afterwards, the autoclave was cooled and any organic products were separated by extraction with diethyl ether (3 times with 3 cm³) and GC-FID analyzed (Hewlett Packard 8452A) or condensed under vacuum to remove diethyl ether and analyzed using ¹H NMR spectroscopy (Bruker 500 MHz) after CDCl₃ addition.

X-Ray crystallography

A suitable crystal was glued on top of a glass fiber and transferred into the cold nitrogen stream on a Kuma KM4CCD κ-axis diffractometer (graphite-monochromated Mo-Kα radiation, λ = 0.71073 Å). The crystal was positioned at 65 mm from the CCD camera. 612 frames were measured at 0.75° intervals with a counting time of 20 s. Accurate cell parameters were determined and refined by least-squares fit of 4800 of the strongest reflections. Data were corrected for Lorentz and polarization effects. Analytical absorption correction was also applied. Data reduction and analysis were carried out with the Oxford Diffraction (Poland) programs. The structure was solved by the heavy method (program SHELXS97³¹) and refined by the full-matrix least-squares method on all F² data using the SHELXL97 programs.³² Non-hydrogen atoms were refined with anisotropic thermal parameters; hydrogen atoms were included from the geometry of the molecules and the ΔF map but were not refined (except the H1 atom). Neutral atom scattering factors were taken from Cromer and Waber.³³ Final refinement details are: Chemical formula: C₆₀H₅₄O₄Cl₂P₄Pd₂, formula weight (M) 1246.65; crystal system: monoclinic; unit-cell dimentions (Å, °): a = 8.6925(6); b = 16.1096(7); c = 20.1450(10); β = 102.338(5); V/ų = 2755.8(3); T/K = 100(2); space group: P2₁/c; Z = 2; μ (mm⁻¹) = 0.912; reflections measured: 17881; independent reflections: 5996 (0.0266); Final R₁/wR₂ indices (I > 2σ_I): 0.0246/0.0559.

CCDC reference number 287627.

For crystallographic data in CIF or other electronic format see DOI: 10.1039/b512835h

Acknowledgements

The financial support by the Ministry of Research and Informatization with the Grant PBZ is gratefully acknowledged.

References

1. D. E. C. Corbridge, Phosphorus 2000. Chemistry, Biochemistry & Technology, Elsevier, Amsterdam, 2000.
2. M. S. Balakrishna, P. P. George and S. M. Mobin, Polyhedron, 2005, 24, 475.
3. P. Bergamini, V. Bertolasi, M. Cattabriga, V. Ferretti, V. Loprieno, N. Mantovani and L. Marrelli, Eur. J. Inorg. Chem., 2003, 5, 918.
4. P. Bergamini, V. Bertolasi and F. Milani, Eur. J. Inorg. Chem., 2004, 6, 1277.
5. K. Yonehara, T. Hashizume, K. Mori, K. Ohe and S. Uemura, J. Org. Chem., 1999, 64, 9374.
6. E. Guimet, M. Dieguez, A. Ruiz and C. Claver, Tetrahedron: Asymmetry, 2005, 16, 959.
7. G. Chen, X. Li, H. Zhang, L. Gong, A. Mi, X. Ciu, Y. Jiang, M. K. C. Choi and A. S. C. Chan, Tetrahedron: Asymmetry, 2002, 13, 809.
8. D. Franco, M. Gomez, F. Jimenez, G. Muller, M. Rocamora, M. A. Maestro and J. Mahia, Organometallics, 2004, 23, 3197.
9. G. Jones and G. J. Richards, Tetrahedron Lett., 2001, 42, 5553.
10. H. Arzoumanian, G. Buono, M. Barek Choukrad and J.-F. Petrignani, Organometallics, 1988, 7, 59.
11. A. M. Trzeciak, H. Bartosz-Bechowski, Z. Ciunik, K. Niesyty and J. J. Ziółkowski, Can. J. Chem., 2001, 79, 752.
12. D. Morales-Morales, R. Redon, C. Yung and C. M. Jensen, Chem. Commun., 2000, 1619.
13. R. B. Bedford, S. L. Hazelwood and M. E. Limmert, Organometallics, 2003, 22, 1364.
14. R. B. Bedford and S. L. Welch, Chem. Commun., 2001, 129.
15. I. D. Kostas, B. R. Steele, A. Terzis and S. V. Amosova, Tetrahedron, 2003, 59, 3467.
16. G. Y. Li, Angew. Chem., Int. Ed., 2001, 40, 1513.
17. G. Y. Li, G. Zheng and A. F. Noonan, J. Org. Chem., 2001, 66, 8677.
18. G. Y. Li, J. Org. Chem., 2002, 67, 3643.
19. K. Yonehara, K. Mori, T. Hashizume, K.-G. Chung, K. Ohe and S. Uemura, J. Organomet. Chem., 2000, 603, 40.
20. A. Mansour and M. Portnoy, Tetrahedron Lett., 2003, 44, 2195.
21. L.-B. Han, N. Choi and M. Tanaka, Organometallics, 1996, 15, 3259.
22. L.-B. Han, R. Hua and M. Tanaka, Angew. Chem., Int. Ed., 1998, 37, 94.
23. L.-B. Han and M. Tanaka, J. Am. Chem. Soc., 1996, 118, 1571.
24. L.-B. Han, F. Mirzaei, C.-Q. Zhao and M. Tanaka, J. Am. Chem. Soc., 2000, 122, 5407.
25. C.-Q. Zhao, L.-B. Han and M. Tanaka, Organometallics, 2000, 19, 4196.
26. R. A. Stockland, Jr., A. M. Levine, M. T. Giovine, L. A. Guzei and J. C. Cannistra, Organometallics, 2004, 23, 647.
27. A. M. Levine, R. A. Stockland, Jr., R. Clark and I. Guzei, Organometallics, 2002, 21, 3278.
28. A. M. Trzeciak, Z. Ciunik and J. J. Ziółkowski, Organometallics, 2002, 21, 132.
29. D. R. Evans, M. Huang, J. C. Fettinger and T. L. Williams, Inorg. Chem., 2002, 41, 5986.
30. B. M. Roundhill, R. P. Sperline and W. B. Beaulieu, Coord. Chem. Rev., 1978, 26, 263.
31. G. M. Sheldrick, SHELXS97: Program for Solution of Crystal Structures, University of Göttingen, Germany, 1997.
32. G. M. Sheldrick, SHELXL97: Program for Crystal Structure Refinement, University of Göttingen, Germany, 1997.
33. D. T. Cromer, J. T. Waber, International Tables for X-ray Crystallography, Kynoch Press, Brimingham, UK, 1974, vol. 4.

---