Planar chiral palladacycle precatalysts for asymmetric synthesis

Abstract

Chiral non-racemic palladacycles were employed as precatalysts for Pd(0) mediated asymmetric synthesis. Addition of HPAr₂/base to a ferrocenyloxazoline planar chiral palladacycle resulted in ligand synthesis and palladium capture to give a bidentate Phosferrox/Pd(0) complex. A series of these complexes were generated in situ and applied successfully as catalysts for asymmetric allylic alkylation.

---

Introduction

Air-stable palladacycles containing a bidentate carbon-heteroatom ligated palladium(II) core have been employed extensively in catalysis.¹ Following the discovery that high turn-over numbers may be achieved on application of the Herrmann complex to the Heck reaction,² this and other palladacycles have also found application in a number of cross-coupling processes including the Suzuki–Miyaura³ and Buchwald–Hartwig amination reactions.⁴ In addition, related chiral non-racemic palladacycles have been applied successfully to the catalysis of asymmetric reactions.⁵ Examples include the allylic imidate rearrangement,⁶ hydrophosphination,⁷ imine arylation,⁸ the aza-Morita–Baylis–Hillman reaction⁹ and Michael type reactions.¹⁰ Palladium(II) mediated activation of the substrate and retention of the palladacycle C–Pd bond in the chiral catalyst is either demonstrated¹¹ or invoked.^7–10 A recent report on palladacycle catalysed asymmetric Suzuki cross-coupling was explained by the operation of a Pd(II)/Pd(IV) catalytic cycle.¹² In contrast, the use of chiral non-racemic palladacycles as precatalysts for asymmetric reactions proceeding via a Pd(0)/Pd(II) catalytic cycle does not appear to have been demonstrated.¹³

In this paper we describe the use of chiral non-racemic palladacycles as precatalysts for the formation of Pd(0)/chiral ligand coordinated complexes, followed by the application of these as catalysts in asymmetric synthesis. This first report on simultaneous bidentate ligand synthesis/palladium(0) capture provides a new and versatile approach to the generation of Pd/chiral ligand complexes, with potential applications across a range of palladium catalysed asymmetric transformations.¹⁴

Results and discussion

Palladacycles are known to undergo ligand phosphination on reaction with MPPh₂ (M = Li, K), where initial phosphide coordination to palladium followed by reductive elimination results in C–P bond formation.^15,16 In this way a number of N–P, S–P and P–P bidentate ligands have been synthesised from the corresponding S, N or P precursor palladacycles. Many of these product ligands are chiral, including planar chiral examples displaying stereospecific replacement of the C–Pd bond by a new C–P bond.^15a,b Formation of these ligands also results in the generation of Pd(0), but as far as we are aware, isolation of a resulting ligated Pd(0) complex in this way has not been reported, let alone used subsequently in catalysis.

To address this challenge we chose to start with ferrocenyloxazoline (S)-1 ¹⁷ as this may be transformed readily into planar chiral palladacycles (S,S_p)-2a/b (Scheme 1).¹⁸ Furthermore, the Phosferrox ligand (S,S_p)-3 ¹⁹ is also derived from (S)-1, this bidentate ligand having been applied successfully in several palladium catalysed asymmetric reactions.²⁰ It was anticipated that this would aid the identification and comparative application of a corresponding palladacycle derived Pd(0) complex of (S,S_p)-3.

Scheme 1 Palladacycles and a chiral bidentate ligand derived from (S)-1.

We reasoned that identification of in situ formed P–N ligated Pd(0) would be aided by addition of a second ligand to give a readily characterised complex. As maleic anhydride (MA) is known to result in stable low valent complexes of palladium,²¹ we first used this electron deficient alkene with (S,S_p)-3 and Pd₂(dba)₃ for the synthesis of (S,S_p)-4 (Scheme 2). This air-stable Pd(0) complex was purified readily by column chromatography. Its ³¹P NMR spectrum²² contained characteristic major and minor signals at 14.39 and 15.81 ppm respectively due to the two possible (exo/endo) orientations of the cyclic unsaturated ligand.

Scheme 2 Synthesis of Pd(0) complex (S,S_p)-4 by ligand exchange and ligand synthesis/palladium capture.

Phosphine complexes of palladacycles (S,S_p)-2a/b were first investigated using triphenylphosphine. On combination in CH₂Cl₂ at room temperature monomeric adducts 5a/b were formed cleanly (Scheme 2). The phosphine is assigned a trans to nitrogen orientation in the resulting complexes in accord with the geometry of related monomeric phosphine coordinated palladacycles.^15b,23 The ³¹P NMR signals²² observed for 5a (36.5 ppm) and 5b (37.5 ppm), compared to PPh₃ (−5.4 ppm), were used to aid the assignment of the corresponding adducts obtained from HPPh₂ (−40.1 ppm). Addition of this secondary phosphine to chloride ligated 2b gave stable and isolable adduct 6b (8.4 ppm). In contrast, addition to acetate ligated 2a resulted in immediate darkening to brown/black of the initially orange solution, and after stirring for 1 hour, no phosphine adduct was obtained.

To investigate the possibility that this colour change was indicative of reductive elimination from an initially formed phosphine adduct, the experiment was repeated in CD₂Cl₂ and followed by ³¹P NMR spectroscopy. Following addition of HPPh₂ to 2a at room temperature, a major signal observed in the ³¹P NMR spectrum obtained after ca. 10 minutes (7.5 ppm) is indicative of adduct 6a (Fig. 1a). Significantly, following the addition of MA, subsequent NMR analysis revealed the disappearance of this complex and the formation of (S,S_p)-4 (Fig. 1b).

Fig. 1 ³¹P NMR spectra (CD₂Cl₂): (a) following the addition of HPPh₂ (2 eq.) to 2a, (b) ca. 10 minutes after the addition of MA (2 eq.), (c) following the addition of NaCH(CO₂Me)₂ (10 eq.) to 6b, (d) ca. 1 h after the addition of MA (4 eq.).

Two other major signals at 27.3 and 55.2 ppm were observed in the ³¹P NMR spectra obtained before and after the addition of MA. The first corresponds to (S,S_p)-7, a phosferrox derived phosphine oxide for which the identity was confirmed by independent synthesis. The second is assigned to a Pd(II)/PPh₂ complex (vide infra). Formation of 4 was not observed when MA was added to palladacycle 2a before the addition of PHPh₂ due to fast competitive Michael addition.²⁴ The formation of 4 and 7, as a single diastereoisomer in both cases, confirms that replacement of palladium by phosphorus proceeds with no erosion of stereochemical integrity.²⁵

In contrast to 6a, the corresponding experiment with chloride complex 6b (formed in situ from (S,S_p)-2b and HPPh₂) required the addition of a base to promote reductive elimination and formation of the MA adduct 4. Nitrogen bases were employed initially, with triethylamine resulting in the highest conversion (14%) as determined by ³¹P NMR spectroscopy.²⁶ With N,O-bis(trimethylsilyl)acetamide (BSA) the conversion was improved (23%), and the best result (29%) was obtained with sodium dimethyl malonate. Following addition of this base ³¹P NMR spectroscopy revealed two major signals at −18.1 and 23.8 ppm (Fig. 1c). The former corresponds to Phosferrox ligand 3, and on addition of MA this signal disappeared and was replaced predominantly by a corresponding amount of 4 (Fig. 1d). Following isolation of the other complex by column chromatography it was identified by X-ray crystallography as Pd(II)/PPh₂ dimer 8b (Scheme 3 and Fig. 2).²⁷ Addition of silver acetate to 8b gave 8a (³¹P NMR = −55.6 ppm), revealing that this corresponding acetate-bridged dimer is formed from the addition of HPPh₂ to 2a (Fig. 1a and b). Addition of maleic anhydride to a solution of 8b in CD₂Cl₂ did not result in the formation of 4. Dimerisation to give 8a/b is an alternative and apparently irreversible pathway to reductive elimination and formation of ligand 3. It is of note that Pd(0) does not complex with 3 to a significant extent until maleic anhydride is added, an outcome consistent with initial and reversible formation of palladium nanoparticles.²⁸ No oxidation to phosphine oxide (S,S_p)-7 was observed, in contrast to the appreciable formation of this compound from acetate-bridged palladacycle 2a, revealing a significant advantage of starting with chloride-bridged palladacycle 2b.²⁹

Scheme 3 The generation of (S,S_p)-4. by ligand synthesis/palladium capture from palladacycle 2b.

Fig. 2 A representation of the X-ray structure of 8b (hydrogen atoms omitted for clarity). Principal bond lengths [Å] include: Pd(1)–Cl(1) = 2.4308(6), Pd(2)–Cl(1) = 2.4366(6), Pd(1)–P(1) = 2.2447(6), Pd(2)–P(1) = 2.2457(6). Principal bond angles [deg] include: Pd(1)–Cl(1)–Pd(2) = 86.641(18), Pd(1)–P(1)–Pd(2) = 96.09(2). Thermal ellipsoids are drawn at the 50% probability level. Flack parameter = −0.002(5).

The successful formation of (S,S_p)-4 from palladacycle precursors prompted us to investigate the use of these as precatalysts in asymmetric synthesis. The allylic alkylation of trans-1,3-diphenylallyl acetate 9 was chosen as an exemplar reaction due to its experimental simplicity and the potential to employ the basic reaction conditions for in situ ligand/Pd(0) generation (Table 1). Furthermore, ligand (S,S_p)-3 is known to promote high enantioselectivity in this reaction, and in our hands repetition of conditions reported previously^20a using BSA as base (3 eq.) gave (S)-10 in 95% ee (conditions A, Table 1, entry 1). Replacing [Pd(C₃H₅)Cl]₂ with preformed Pd(0) complex (S,S_p)-4 resulted in some loss of enantioselectivity, and a very significant reduction in activity due to inhibition by maleic anhydride (entry 2). This precluded the use of this π-acidic ligand in any subsequent runs. Instead, a catalytic asymmetric allylic alkylation was attempted directly following the addition of HPPh₂ to 2a, and after stirring for 10 minutes, to resulting in situ generated 6a was added sequentially rac-9, dimethyl malonate, BSA and KOAc (entry 3). Although slow, we noted with interest the formation of product 10, subsequently isolated as the S enantiomer in 85% ee. Use in the same way of chloride 2b resulted in a similar outcome (entry 4), and on repetition in the absence of KOAc the product ee increased to 94% (entry 5). No product was formed in the absence of added PHPh₂ (entry 6). These initial results are consistent with the in situ formation of (S,S_p)-3 ligated Pd(0) as the sole catalytically active species, albeit in low concentration. Replacement of CH₂Cl₂ as solvent with THF or toluene resulted in no product formation.

Table 1 Comparative catalysis of asymmetric allylic alkylation with preformed and palladacycle generated Phosferrox/Pd(0) complexes^a

Entry	Pd source (mol%)	Conditions	Time (h)	Con. (%)	ee (%)
a 1 mmol rac-9 (0.6 M) unless otherwise stated. b Lit. value 90% ee.^20a c 0.33 mmol scale. d No KOAc. e Reaction essentially complete within 1 h (tlc).
1	(S,Sp)-3 (2.5) + [Pd(C3H5)Cl]2 (1)	A	<1	100	95^b
2	(S,Sp)-4 (1)	A	96	6	86
3	(S,Sp)-2a (1) + HPPh2 (2)	A	96	51	85
4	(S,Sp)-2b (1) + HPPh2 (2)	A	96	77	85
5	(S,Sp)-2b (1) + HPPh2 (2)	A ^c^,^d	96	20	94
6	(S,Sp)-2a (1) or (S,Sp)-2b (1)	A	96	0	na
7	(S,Sp)-3 (2.5) + [Pd(C3H5)Cl]2 (1)	B	3.5	100	88
8	(S,Sp)-2b (1) + HPPh2 (2)	B	2.5	100	77
9	(S,Sp)-2b (1) + HPPh2 (2)	C	1	99	86
10	8b (1)	A	24	0	na
11	(S,Sp)-2b (1) + 11a (2)	C	24	18	85
12	(S,Sp)-2b (1) + 11b (2)	C	24^e	100	84
13	(S,Sp)-2b (1) + 11c (2)	C	24^e	100	83

---

As sodium dimethyl malonate gave the highest and cleanest conversion to (S,S_p)-4, this was employed next as both a base and a nucleophile, and the process referenced initially by use of (S,S_p)-3/[Pd(C₃H₅)Cl]₂ (3 fold excess, conditions B). The lower product ee, compared to that obtained from the use as base of BSA (95% – entry 1 vs. 88% – entry 7) is consistent with the superiority of the former conditions noted previously.³⁰ Significantly, addition of a three-fold excess of sodium dimethyl malonate to rac-9 and 2 mol% of in situ generated 6b resulted in complete conversion within 2.5 h to allylic alkylation product (S)-10 formed in 77% ee (entry 8). Modifying these conditions further (conditions C) to the use sodium dimethyl malonate (10 mol%) as the base for in situ catalyst formation, and then the addition of BSA/dimethyl malonate (3 equivalents of each) further decreased the reaction time to 1 hour and increased the ee of (S)-10 to 86% (entry 9). The use of conditions C provided an in situ generated catalyst with comparable activity and selectivity to that obtained using a preformed ligand, and consistent with this outcome is the absence of any catalysis with Pd(II) complex 8b (entry 10).

Finally, this led us to explore the use of asymmetric allylic alkylation to test the viability of using electronically contrasting diarylphosphines in the ligand synthesis/Pd(0) capture protocol (Scheme 4). For this reaction previous work with related PHOX ligands containing PAr₂ units with various para substituents revealed little difference in enantioselectivity compared to the parent PPh₂ containing ligand,³¹ albeit with faster conversion with electron-withdrawing substituents.^31a Accordingly, secondary phosphines 11a–c were added separately to palladacycle (S,S_p)-2b followed by sodium dimethyl malonate, and to the resulting solutions were added rac-9 (50 equivalents w.r.t. Pd) and dimethyl malonate/BSA (i.e. conditions C). Use of electron-deficient phosphine 11a resulted in a similar level of product enantioselectivity albeit in reduced yield (Table 1, entry 11). Further examination by ³¹P NMR spectroscopy (CD₂Cl₂) of the addition of 11a (−40.8 ppm) to 2b confirmed the formation of the phosphine coordinated adduct (10.9 ppm), but little Pd⁰ complex was observed following the addition of sodium dimethyl malonate and MA. In this instance catalyst formation appears to be limited by the inefficiency of reductive elimination.³² In contrast, the catalysts derived from phosphines 11b (entry 12) and 11c (entry 13) gave essentially complete conversion after 1 h and ee values very similar to that obtained from diphenylphosphine.

Scheme 4 Phosphine variation for in situ ligand synthesis/Pd(0) capture from palladacycle (S,S_p)-b2.

Conclusion

We have established that ferrocenyloxazoline-based planar chiral palladacycles 2a/b, on combination with diphenylphosphine (and for 2b a base), are viable precatalysts for in situ catalyst generation using a new ligand synthesis/palladium(0) capture approach. This was demonstrated by formation of a stable Phosferrox/maleic anhydride coordinated Pd(0) complex (S,S_p)-4a, and by application of Phosferrox/Pd(0) as a catalyst for asymmetric allylic alkylation in the absence of maleic anhydride. The other palladium(II) complex formed in this process is not a competitive catalyst or catalyst precursor. Product configuration and enantioselectivity using a palladacycle derived catalyst are similar to that obtained using a preformed Phosferrox ligand and a source of Pd(0). In addition to being a precursor to an active and selective catalyst generated in one step, palladacycle 2b may also be used for multiple catalyst generation with other diarylphosphines. It is anticipated that this approach will aid significantly the generation and identification of palladium based catalysts for application in other areas of asymmetric synthesis.

Experimental

General information

Tetrahydrofuran was distilled over sodium and benzophenone ketyl and dichloromethane distilled over calcium hydride. All reactions were carried out under an inert atmosphere of either nitrogen or argon. Silica gel (60 Å pore size, 40–63 μm technical grade) was used for chromatography. All starting materials not commercially available are specifically referenced.

Synthesis of (S,S_p)-4

(S,S_p)-3 ¹⁹ (0.087 g, 0.18 mmol), tris(dibenzylideneacetone)-dipalladium(0) (0.076 g, 0.08 mmol) and maleic anhydride (0.024 g, 0.24 mmol) were added to a flame dried Schlenk tube and dissolved in toluene (15 mL). The resulting deep purple solution was stirred vigorously overnight resulting in the formation of a dark suspension. The reaction was filtered through Celite™ using toluene as the eluent and the solvent removed in vacuo. Purification by column chromatography (SiO₂, 40% EtOAc/hexane) yielded an orange solid (0.07 g, 61%); mp decomposed ∼200 °C (under argon); HRMS (AS) [M + H]⁺ anal. calcd for C₃₂H₃₁FeNO₄PPd 686.0389, found 686.0389; [α]^{20.5 °C}_D = −538 (c = 0.46, CHCl₃); ν_max (film)/cm⁻¹ 3056, 3016, 2960, 2923, 2873, 1790, 1723, 1626; ¹H NMR (500 MHz, CDCl₃) major isomer (ratio of isomers = 1.6 : 1) δ 7.80 (2H, ddd, J = 11.6, 7.8, 1.5 Hz), 7.56–7.46 (3H, m), 7.33–7.27 (3H, m), 7.08 (2H, ddd, J = 11.0, 6.6, 3.2 Hz), 5.04 (1H, brs), 4.70 (1H, t, J = 2.6 Hz), 4.41–4.38 (1H, m), 4.35 (1H, t, J = 3.7 Hz), 4.33–4.21 (3H, m), 4.07 (1H, dd, J = 10.5, 3.8 Hz), 3.96 (5H, s), 2.80 (1H, dtd, J = 13.8, 6.9, 3.3 Hz), 1.07 (3H, d, J = 7.1 Hz), 0.99 (3H, d, J = 6.9 Hz); minor isomer 7.76–7.68 (2H, m), 7.56–7.46 (3H, m), 7.33–7.27 (3H, m), 7.21–7.14 (2H, m), 5.04 (1H, brs), 4.72 (1H, t, J = 2.6 Hz), 4.50–4.46 (1H, m), 4.33–4.21 (3H, m), 3.86 (5H, s), 3.87–3.79 (2H, m), 2.67 (1H, dtd, J = 13.7, 6.9, 3.4), 1.04 (3H, d, J = 7.0 Hz), 0.95 (3H, d, J = 6.8 Hz); ¹³C NMR (125 MHz, CDCl₃) major isomer δ 172.6 (d, J = 2.0 Hz), 171.5 (d, J = 5.2 Hz), 169.1, 135.5 (d, J = 36.7 Hz), 134.9 (d, J = 16.7 Hz), 134.7 (d, J = 40.1 Hz), 131.6 (d, J = 13.3 Hz), 131.2 (d, J = 2.0 Hz), 129.4 (d, J = 1.6 Hz), 128.6 (d, J = 10.8 Hz), 128.5 (d, J = 9.7 Hz), 75.5 (d, J = 2.6 Hz), 74.0 (d, J = 2.3 Hz), 73.5 (d, J = 46.2 Hz), 73.4 (d, J = 1.5 Hz), 73.1, 71.8 (d, J = 0.5 Hz), 71.4, 67.7, 47.4 (d, J = 31.7 Hz), 46.4 (d, J = 1.8 Hz), 29.2, 19.2, 14.7; minor isomer 172.9 (d, J = 1.5 Hz), 171.2 (d, J = 5.0 Hz), 169.1, 137.3 (d, J = 37.8 Hz), 135.1 (d, J = 16.6 Hz), 133.7 (d, J = 40.2 Hz), 131.4 (d, J = 13.1 Hz), 131.4 (d, J = 2.1 Hz), 129.1 (d, J = 1.5 Hz), 128.8 (d, J = 11.2 Hz), 128.5 (d, J = 9.7 Hz), 75.9 (d, J = 2.7 Hz), 75.1 (d, J = 1.8 Hz), 74.0 (d, J = 33.0 Hz), 74.0 (d, J = 2.1 Hz), 73.4 (d, J = 1.3 Hz), 73.1, 71.6, 67.7, 47.8 (d, J = 1.7 Hz), 46.7 (d, J = 32.7 Hz), 29.2, 19.2, 14.9; ³¹P NMR (202 MHz, CDCl₃) major isomer δ 14.39; minor isomer 15.81.

General procedure for the synthesis of 5a, 5b and 6b

(S,S_p)-2a ^18a,b or (S,S_p)-2b ^18b (0.029 mmol) and either triphenylphosphine or diphenylphosphine (0.057 mmol) were added to a flame dried Schlenk tube and dissolved in dichloromethane (1 mL) and stirred at room temperature for 1 hour. The solvent was removed in vacuo yielding the product.

5a . Obtained as a deep orange solid (0.034 g, 83%): mp 136–137 °C (under argon), MS (EI) [M]⁺ anal. calcd for C₃₆H₃₆FeNO₃PPd 723.1, found 723.1; [α]^{23.5 °C}_D = −572 (c = 0.55, MeCN); ν_max (film)/cm⁻¹ 3052, 2961, 2925, 2871, 1615, 1582, 1499; ¹H NMR (500 MHz, MeCN-d³) δ 7.78–7.72, (6H, m), 7.52–7.48 (3H, m), 7.46–7.41 (6H, m), 4.63 (1H, t, J = 9.5 Hz), 4.59–4.55 (1H, m), 4.45 (1H, d, J = 1.9 Hz), 4.07 (1H, ddd, J 10.0, 7.5, 4.2 Hz), 4.02 (1H, t, J = 2.2 Hz), 3.90 (5H, s), 3.30 (1H, d, J = 2.2 Hz), 2.34–2.26 (1H, m), 1.25 (3H, s), 1.03 (3H, d, J = 6.9 Hz), 0.88 (3H, d, J = 7.2 Hz); ¹³C NMR (125 MHz, MeCN-d³) δ 179.8, 176.3, 135.7 (d, J = 12.0 Hz), 132.1 (d, J = 50.3 Hz), 131.5 (d, J = 2.4 Hz), 129.1 (d, J = 10.7 Hz), 90.2 (d, J = 8.6 Hz), 77.3 (d, J = 9.0 Hz), 73.8 (d, J = 1.3 Hz), 73.1 (d, J = 3.0 Hz), 70.9, 69.7 (d, J = 2.6 Hz), 68.3 (d, J = 3.3 Hz), 65.6, 29.3, 23.9, 19.3, 15.7; ³¹P NMR (202 MHz, MeCN-d³) δ 36.48.

5b . Purified by column chromatography (SiO₂, 20% EtOAc/hexane) to give the product as a deep orange solid (0.038 g, 95%): mp decomposed ∼230 °C (under argon); HRMS (AS) [M-Cl]⁺ anal. calcd for C₃₄H₃₃FeNOPPd 664.0699, found 664.0711; [α]^{22.8 °C}_D = −825 (c = 0.26, MeCN); ν_max (film)/cm⁻¹ 3077, 3052, 2958, 2925, 2867, 1618, 1503; ¹H NMR (500 MHz, MeCN-d³) δ 7.78–7.72 (6H, m), 7.53–7.49 (3H, m), 7.48–7.43 (6H, m), 4.67–4.57 (2H, m), 4.47 (1H, d, J = 1.9 Hz), 4.18 (1H, ddd, J = 10.2, 6.3, 4.2 Hz), 4.02 (1H, t, J = 2.1 Hz), 3.97 (5H, s), 3.04 (1H, brs), 2.87 (1H, brs), 1.06 (3H, d, J = 6.9 Hz), 0.92 (3H, d, J = 7.2 Hz); ¹³C NMR (125 MHz, MeCN-d³) δ 179.9, 135.6 (d, J = 11.7 Hz), 132.9 (d, J = 51.3 Hz), 131.6, 129.1 (d, J = 10.8 Hz), 94.6, 75.9 (d, J = 8.7 Hz), 74.2, 73.3, 70.9, 69.4, 68.0 (d, J = 3.1 Hz), 65.6, 29.4, 19.2, 15.5; ³¹P NMR (202 MHz, MeCN-d³) δ 37.49.

6b . Obtained as a deep orange solid (0.025 g, 73%): mp 110–112 °C (under argon), MS (EI) [M]⁺ anal. calcd for C₂₈H₂₉ClFeNOPPd 623.0, found 623.0; [α]^{22.8 °C}_D = −882 (c = 0.20, MeCN); ν_max (film)/cm⁻¹ 3077, 3055, 2961, 2929, 2871, 1615, 1503; ¹H NMR (500 MHz, MeCN-d³) δ 8.05 (2H, dd, J = 12.0, 7.4 Hz), 7.88 (2H, dd, J = 12.7, 7.3 Hz), 7.58–7.49 (4H, m), 7.47 (2H, td, J = 7.3, 2.0 Hz), 6.56 (1H, d, J = 388.9 Hz), 4.64 (1H, dd, J = 9.0, 6.2 Hz), 4.57 (1H, t, J = 9.7 Hz), 4.54 (1H, d, J = 2.2 Hz), 4.45 (1H, brs), 4.37 (1H, t, J = 2.4 Hz), 4.17–4.12 (1H, m), 4.07 (5H, s), 2.90 (1H, brs), 1.04 (3H, d, J = 5.9 Hz), 0.90 (3H, d, J = 7.0 Hz); ¹³C NMR (125 MHz, MeCN-d³) δ 179.9, 135.2 (d, J = 11.1 Hz), 134.8 (d, J = 11.4 Hz), 132.2 (d, J = 2.6 Hz), 132.0 (d, J = 2.6 Hz), 129.9 (d, J = 10.8 Hz), 129.7 (d, J = 11.0 Hz), 129.5 (d, J = 51.2 Hz), 129.5 (d, J = 52.8 Hz), 91.6 (d, J = 8.0 Hz), 74.8 (d, J = 1.0 Hz), 73.8 (d, J = 15.7 Hz), 73.2 (d, J = 3.2 Hz), 70.7, 70.2 (d, J = 4.5 Hz), 68.0 (d, J = 3.4 Hz), 66.8, 29.2, 19.1, 15.1; ³¹P NMR (202 MHz, MeCN-d³) δ 8.44.

Synthesis of (S,S_p)-7

(S)-1 ¹⁷ (0.059 g, 0.2 mmol) was added to a flame dried Schlenk tube under an atmosphere of argon and dissolved in dry THF (5 mL). The resulting orange solution was cooled to −78 °C and stirred for 5 min after which n-butyl lithium (2.5 M in hexanes) (0.11 mL, 0.28 mmol) was added slowly. After stirring for 2 hours the mixture was warmed to 0 °C and chlorodiphenylphosphine oxide (0.076 mL, 0.4 mmol) was added and the reaction was allowed to warm to room temperature. After an additional 15 min, the reaction was diluted with diethyl ether and then quenched with saturated sodium hydrogen carbonate solution. The organics were separated with H₂O, washed with brine, dried over MgSO₄ and the solvent removed in vacuo giving a crude product containing a 2 : 1 mixture of diastereoisomers. The diastereoisomers were separated by column chromatography (SiO₂, 2% MeOH/Et₂O) yielding the desired product as a yellow brown solid (0.0255 g, 26%): mp 197–198 °C; HRMS (ES) [M + H]⁺ anal. calcd for C₂₈H₂₉FeNO₂P 498.1280, found 498.1263; [α]^{19.4 °C}_D = +114 (c = 0.2, CHCl₃); ν_max (film)/cm⁻¹ 3080, 3062, 2963, 2930, 2871, 1648, 1120; ¹H NMR (500 MHz, CDCl₃) δ 7.76–7.71 (2H, m), 7.69–7.64 (2H, m), 7.50–7.46 (1H, m), 7.45–7.36 (5H, m), 5.06 (1H, dt, J = 2.6, 1.4 Hz), 4.49 (5H, s), 4.45 (1H, dd, J = 4.4, 2.4 Hz), 4.22–4.18 (1H, m), 3.93 (1H, dd, J = 3.9, 2.5 Hz), 3.72 (1H, td, J = 9.3, 6.3 Hz), 3.13 (1H, t, J = 8.8 Hz), 1.53–1.47 (1H, m), 0.81 (3H, d, J = 6.8 Hz), 0.61 (3H, d, J = 6.8 Hz); ¹³C NMR (125 MHz, CDCl₃) δ 164.6, 134.86 (d, J = 107 Hz), 134.8 (d, J = 110.4 Hz), 131.6 (d, J = 9.3 Hz), 131.4 (d, J = 9.7 Hz), 131.3 (d, J = 2.6 Hz), 131.0 (d, J = 2.9 Hz), 128.2 (d, J = 4.3 Hz), 128.1 (d, J = 4.8 Hz), 78.5 (d, J = 14.6 Hz), 75.7 (d, J = 109.9 Hz), 74.5 (d, J = 9.1 Hz), 73.5 (d, J = 8.2 Hz), 72.4, 71.5 (d, J = 10.9 Hz), 71.4, 70.1, 32.3, 19.2, 18.0; ³¹P NMR (202 MHz, CDCl₃) δ 26.40.

Synthesis of (S,S,S_p,S_p)-8b

To a flame dried Schlenk tube (tube-1) was added sodium tert-butoxide solution (2 M in THF) (0.05 mL, 0.10 mmol) followed by a solution of dimethyl malonate (22.6 μl, 0.20 mmol) in THF (1 mL) at 0 °C. After stirring for 30 min the solvent was removed under high vacuum to give a beige residue. In a separate flame dried Schlenk tube (tube-2), (S,S_p)-6b (0.0125 g, 0.020 mmol) was dissolved in CH₂Cl₂ (1.7 mL) and then the contents of Schlenk tube (2) were transferred to Schlenk tube (1) and stirred at room temperature for 10 min. The solvent was then removed in vacuo and the residue purified by column chromatography (SiO₂, 10% EtOAc/hexane) to give an orange solid (0.007 g, 70%): mp > 230 °C (under argon); HRMS (AS) [M − Cl]⁺ anal. calcd for C₄₄H₄₆Fe₂N₂O₂PPd₂ 991.0103, found 991.0093; [α]^{20.1 °C}_D = −2000 (c = 0.23, CHCl₃); ν_max (film)/cm⁻¹ 3096, 2958, 2923, 2967, 1617, 1495; ¹H NMR (500 MHz, CDCl₃) δ 8.12–8.05 (4H, m), 7.38–7.34 (6H, m), 4.49–4.47 (4H, m), 4.36 (2H, d, J = 1.8 Hz), 4.08 (2H, td, J = 8.0, 4.6 Hz), 4.04 (10H, s), 4.02 (2H, t, J = 2.3 Hz), 3.22 (2H, d, J = 1.6 Hz), 2.51–2.43 (2H, m), 1.06 (6H, d, J = 6.8 Hz), 0.97 (6H, d, J = 7.0 Hz); ¹³C NMR (125 MHz, CDCl₃) δ 177.9, 137.2 (d, J = 32.8 Hz), 135.4 (d, J = 12.7 Hz), 128.8, 127.8 (d, J = 10.5 Hz), 91.4 (d, J = 2.9 Hz), 75.1 (d, J = 8.6 Hz), 73.3 (d, J = 1.6 Hz), 72.2, 70.0, 69.0, 66.9, 64.5, 29.3, 19.6, 16.0; ³¹P NMR (202 MHz, CDCl₃) δ 23.40.

Synthesis of (S,S,S_p,S_p)-8a

Addition of excess AgOAc to an NMR tube containing (S,S,S_p,S_p)-8b showed the formation of acetate bridged palladacycle (S,S,S_p,S_p)-8a; ν_max (film)/cm⁻¹ 3096, 2958, 2919, 2849, 1621, 1556, 1495; ¹H NMR (500 MHz, CDCl₃) δ 8.12–7.97 (4H, m), 7.33–7.27 (6H, m), 4.51–4.46 (4H, m), 4.34 (2H, brs), 4.09–4.03 (2H, m), 4.00 (10H, s), 3.97 (2H, brs), 3.00 (2H, brs), 2.70–2.60 (2H, m), 2.01 (3H, brs), 1.04 (6H, d, J = 6.8 Hz), 0.95 (6H, d, J = 7.1 Hz); ³¹P NMR (202 MHz, CDCl₃) δ 55.63.

Phosphine insertion studies

Using a glove box, 0.01 mmol of either (S,S_p)-2a or (S,S_p)-2b was dissolved in CD₂Cl₂ (0.7 mL) in an NMR tube fitted with a J. Young valve under an inert atmosphere. After addition of the reagents approximately 10 minutes elapsed before the first NMR spectrum was recorded.

In situ catalyst preparation and asymmetric allylic alkylation (conditions C)

To a flame dried Schlenk tube (tube-1) was added sodium tert-butoxide solution (2 M in THF) (0.05 mL, 0.10 mmol) followed by a solution of dimethyl malonate (22.6 μl, 0.20 mmol) in THF (1 mL) at 0 °C. After stirring for 30 min the solvent was removed under high vacuum to give a beige residue. In a separate flame dried Schlenk tube (tube-2), (S,S_p)-2b (0.0088 g, 0.010 mmol) was dissolved in CH₂Cl₂ (1.7 mL) and phosphine (0.020 mmol) was added and stirred for 10 min. 1,3-Diphenylallyl acetate (0.252 g, 1.0 mmol) was dissolved in CH₂Cl₂ (1.7 mL) and then added to Schlenk tube (2) followed by dimethyl malonate (0.34 mL, 3.0 mmol). The contents of Schlenk tube (2) were transferred to Schlenk tube (1) and stirred for 5 min after which N,O-bis(trimethylsilyl)acetamide (0.74 mL, 3.0 mmol) was added and the reaction allowed to stir at room temperature. After completion or 24 hours (whichever came first), the reaction was quenched with saturated ammonium chloride, separated with diethyl ether, dried with magnesium sulphate and the solvent removed in vacuo. A ¹H NMR was performed at this point to determine the conversion. Purification by column chromatography (SiO₂, 5% EtOAc/hexane) gave the product (S)-10 as a colourless oil for which the ¹H/¹³C NMR spectra matched previously reported data.³³ The enantiomeric excess was determined by HPLC analysis using a CHIRALCEL OD-H column; eluent = 98 : 2 (hexane : IPA); flow rate = 0.5 mL min⁻¹; concentration = 0.0015 g mL⁻¹, injection volume = 5 μl, wavelength = 254 nm. (S)-Enantiomer RT = 20.5 min, (R)-enantiomer RT = 18.9 min.

Conflicts of interest

There are no conflicts to declare.

Acknowledgements

The EPSRC (EP/N019393/1) (R. A. A.) is thanked for financial support. We also thank the EPSRC National Mass Spectrometry Centre (University of Wales, Swansea).

Notes and references

1. (a) J. Dupont, C. S. Consorti and J. Spencer, Chem. Rev., 2005, 105, 2527; (b) Palladacycles: Synthesis Characterization and Applications, ed. J. Dupont and M. Pfeffer, Wiley-VCH, Weinheim, 2008; (c) Palladacycles: Catalysis and Beyond, ed. A. R. Kapdi and D. Maiti, Elsevier, 2019.
2. W. A. Herrmann, C. Brossmer, C.-P. Reisinger, T. H. Riermeier, K. Ó¦fele and M. Beller, Chem. – Eur. J., 1997, 3, 1357.
3. F. Bellina, A. Carpita and R. Rossi, Synthesis, 2004, 2419.
4. B. P. Fors and S. L. Buchwald, J. Am. Chem. Soc., 2010, 132, 15914.
5. V. V. Dunina, O. N. Gorunova, P. A. Zykov and K. A. Kochetkov, Russ. Chem. Rev., 2011, 80, 51.
6. (a) H. Nomura and C. J. Richards, Chem. – Asian J., 2010, 5, 1726; (b) J. S. Cannon and L. E. Overman, Acc. Chem. Res., 2016, 49, 2220.
7. R. J. Chew and P.-H. Leung, Chem. Rec., 2016, 16, 141.
8. C. Schrapel and R. Peters, Angew. Chem., Int. Ed., 2015, 54, 10289.
9. K. Hyodo, S. Nakamura and N. Shibata, Angew. Chem., Int. Ed., 2012, 51, 10337.
10. (a) M. Weber, S. Jautze, W. Frey and R. Peters, J. Am. Chem. Soc., 2010, 132, 12222; (b) M. Weber and R. Peters, J. Org. Chem., 2012, 77, 10846.
11. (a) M. P. Watson, L. E. Overman and R. G. Bergman, J. Am. Chem. Soc., 2007, 129, 5031; (b) J. S. Cannon, S. F. Kirsch, L. E. Overman and H. F. Sneddon, J. Am. Chem. Soc., 2010, 132, 15192.
12. O. N. Gorunova, Y. K. Grishin, M. M. Ilyin (Jr.), K. A. Kochetkov, A. V. Churakov, L. G. Kuz’mina and V. V. Dunina, Russ. Chem. Bull., Int. Ed., 2017, 66, 282.
13. A number of chiral non-racemic Pd(II)-based pincer complexes have also been applied to asymmetric catalysis, but again there appears to be no evidence that these proceed via a Pd(0)/Pd(II) catalytic cycle. For a recent review see: J.-K. Liu, J.-F. Gong and M.-P. Song, Org. Biomol. Chem., 2019, 17, 6069.
14. (a) L. F. Tietze, H. Ila and H. P. Bell, Chem. Rev., 2004, 104, 3453; (b) Q.-A. Chen, Z.-S. Ye, Y. Duan and Y.-G. Zhou, Chem. Soc. Rev., 2013, 42, 497.
15. (a) V. I. Sokolov, L. L. Troitskaya and O. A. Reutov, J. Organomet. Chem., 1980, 202, C58; (b) C. Bolm, K. Wenz and G. Raabe, J. Organomet. Chem., 2002, 662, 23; (c) V. A. Stepanova, V. V. Dunina and I. P. Smoliakova, Organometallics, 2009, 28, 6546; (d) V. A. Stepanova, V. V. Dunina and I. P. Smoliakova, J. Organomet. Chem., 2011, 696, 871; (e) N. J. Korte, V. A. Stepanova and I. P. Smoliakova, J. Organomet. Chem., 2013, 745–746, 356; (f) G. C. Dickmu, N. J. Korte and I. P. Smoliakova, J. Organomet. Chem., 2015, 797, 13; (g) J. E. Kukowski, V. A. Stepanova and I. P. Smoliakova, J. Organomet. Chem., 2017, 830, 155.
16. In addition, C–P bonds have also been formed by palladium catalysed ortho-directed C–H activation via an intermediate palladacycle. See: L. Bonnafoux, R. Gramage-Doria, F. Colobert and F. R. Leroux, Chem. – Eur. J., 2011, 17, 11008.
17. (a) T. Sammakia, H. A. Latham and D. R. Schaad, J. Org. Chem., 1995, 60, 10.
18. (a) J.-B. Xia and S.-L. You, Organometallics, 2007, 26, 4869; (b) R. A. Arthurs, D. L. Hughes and C. J. Richards, Organometallics, 2019, 38, 4271.
19. (a) C. J. Richards, T. Damalidis, D. E. Hibbs and M. B. Hursthouse, Synlett, 1995, 74; (b) Y. Nishibayashi and S. Uemura, Synlett, 1995, 79.
20. (a) K. H. Ahn, C.-W. Cho, J. Park and S. Lee, Tetrahedron: Asymmetry, 1997, 8, 1179; (b) A. J. Hennessy, Y. M. Malone and P. J. Guiry, Tetrahedron Lett., 2000, 41, 2261; (c) M. Lautens, J.-L. Renaud and S. Hiebert, J. Am. Chem. Soc., 2000, 122, 1804; (d) S.-L. You, X.-L. Hou, L.-X. Dai, Y.-H. Yu and W. Xia, J. Org. Chem., 2002, 67, 4684; (e) C. Ma, Y. Huang and Y. Zhao, ACS Catal., 2016, 6, 6408; (f) C. Zhu, D. Wang, Y. Zhao, W.-Y. Sun and Z. Shi, J. Am. Chem. Soc., 2017, 139, 16486; (g) M. Okumura, A. S. Shved and D. Sarlah, J. Am. Chem. Soc., 2017, 139, 17787.
21. H. Minematsu, Y. Nonaka, S. Takahashi and N. Hagihara, J. Organomet. Chem., 1973, 59, 395.
22. ³¹P NMR spectral data from the use of CD₂Cl₂ as solvent excepting (S,S_p)-4, 8b, PPh₃ and HPPh₂ (CDCl₃); (S,S_p)-5a/b and (S,S_p)-6b (MeCN-d³).
23. D. S. Black, G. B. Deacon and G. L. Edwards, Aust. J. Chem., 1994, 47, 217.
24. J. A. Van Doorn and R. L. Wife, Phosphorus, Sulfur Silicon Relat. Elem., 1990, 47, 253.
25. The (S,R_p) epimer of (S,S_p)-7 is known (³¹P NMR = 36.10 ppm/CDCl₃). R. A. Arthurs and C. J. Richards, Org. Lett., 2017, 19, 702.
26. Full details including ³¹P NMR spectra are available in the ESI.†.
27. CCDC 1983850 contains the supplementary crystallographic data for this paper.†.
28. A. J. Reay and I. J. S. Fairlamb, Chem. Commun., 2015, 51, 16289.
29. Acetate-mediated phosphine oxidation with concomitant Pd(II) reduction is known, but in this instance Pd(0) formation is consistent with reductive elimination. See: C. Amatore, A. Jutand and M. A. M'Barki, Organometallics, 1992, 11, 3009.
30. B. M. Trost and D. L. Van Vranken, Chem. Rev., 1996, 96, 395.
31. (a) K. Tani, D. C. Behenna, R. M. McFadden and B. M. Stoltz, Org. Lett., 2007, 9, 2529; (b) D. C. Behenna, J. T. Mohr, N. H. Sherden, S. C. Marinescu, A. M. Harned, K. Tani, M. Seto, S. Ma, Z. Novák, M. R. Krout, R. M. McFadden, J. L. Roizen, J. A. Enquist Jr., D. E. White, S. R. Levine, K. V. Petrova, A. Iwashita, S. C. Virgil and B. M. Stoltz, Chem. – Eur. J., 2011, 17, 14199.
32. J. F. Hartwig, Inorg. Chem., 2007, 46, 1936.
33. (a) Y. Tanaka, T. Mino, K. Akita, M. Sakamoto and T. Fujita, J. Org. Chem., 2004, 69, 6679; (b) D. Liu, F. Xie and W. Zhang, J. Org. Chem., 2007, 72, 6992.

---

Footnote

† Electronic supplementary information (ESI) available: Additional experimental detail, copies of the ¹H, ¹³C and ³¹P NMR spectra, HPLC traces and details of the X-ray crystal structure determination. CCDC 1983850. For ESI and crystallographic data in CIF or other electronic format see DOI: 10.1039/d0ob01331e

---