Tervalent phosphorus acid derivatives

This chapter discusses the literature related to the development of tervalent phosphorus acid derivatives, published during 2021. A critical selection of bibliographic material was performed, essentially focused on synthetic strategies for new phosphorus(iii) ligands, metal complexes and some of their most relevant applications in catalysis. Several review articles^1–17  and book chapters^18–20  have been published in the field, with particular emphasis on their catalytic applications.

Several examples regarding the preparation of aminophosphine and their metal complexes were reported during the year 2021, including structural studies,^21–23  and catalytic applications, namely in C–H bond activation,^24,25  cross coupling,²⁶  allylic etherification,²⁷  and annulation²⁸  reactions.

Regarding the synthesis and application of new aminophosphine ligands and catalysts thereof, Staubitz,²⁹  for instance, used [1,1′-biphenyl]-2-ylmagnesium bromide to prepare aminophosphine 1 by reaction with ⁱPr₂NPCl₂, followed by acid treatment (Scheme 1). This synthon was then transformed into the highly unstable phosphine 2, and further in its stabilized counterparts, the borane adduct 3a or the stannane adduct 3b. The authors followed all reactions by ³¹P-NMR, assigning the typical shifts to the corresponding species. Furthermore, all adducts were structurally characterised by single crystal X-ray analysis.

Other authors³⁰  reported the conversion of dichlorophosphinidenoids 4a–b into phosphinidene complexes 5a–b, by reacting hexacarbonyl tungsten with dichlorophosphines, followed by their reaction with 12-crown-4 ether and methylamine (Scheme 2). The formation of a transient complex was confirmed via low-temperature ³¹P-NMR spectroscopy, but strong evidence for its transient formation was also given from the reaction with methylamine in a formal N–H insertion.

Nikonov³¹  reported an efficient method for the preparation of a phosphinoamidinate-supported disilylenes 7a and 7b, starting from phosphinoamidine 6 (Scheme 3).

Using pinacolborane (HBpin), transfer of the ligand from silicon to boron was observed, yielding 8a. Compound 7a was found to be stable in its crystalline form and in dilute solutions under an inert atmosphere for weeks, but, upon heating, it slowly loses its wine-red colour, turning yellow. The authors found that this slow process took about a week, at 70 °C, affording 8b, by intramolecular N–P activation by one of the silylene centres, whose structure was established by single-crystal X-ray diffraction. Furthermore, reaction of the disilylene with diphenylphosphine, HPPh₂, furnished the phosphidosilylene 8c, which was also characterised by single-crystal X-ray diffraction.³¹ 

In parallel, Su and So³²  used phosphinoamidine 9 and promoted its reaction with ⁿBuLi, followed by treatment with B₂Br₄(SMe₂)₂, in toluene, to form the N-phosphinoamidinate-bridged diborane 10 (Scheme 4). It was then reacted with a N-heterocyclic carbene to give the N-phosphinoamidinato NHC-diborene compound 11. The authors found that compound 11 was capable of activating a small molecule, like carbon dioxide (CO₂, 1 atm), in toluene at room temperature, providing compound 12 by [2+2] cycloaddition, isolated as a colourless crystalline solid from the concentrated filtrate, being stable both in solution and the solid state. Further functionalization of 12 with tris(pentafluorophenyl)borane in toluene, at room temperature, afforded compound 13, isolated as colourless crystals from the concentrated filtrate, whose molecular structure was corroborated by X-ray crystallography. The catalytic ability of 11 toward N-formylation of primary and secondary amines by carbon dioxide and pinacolborane was further examined, and N-formamides were obtained in yields ranging from 82 to 99%.³² 

The same group³³  later reported the synthesis of N-phosphinoamidinato digermyne 14, by reacting the same phosphinoamidine 9 with ⁿBuLi, followed by treatment with GeCl₂, followed reaction with excess potassium graphite KC₈ in toluene, to yield the desired compound (Scheme 4). Compound 14 was also able to activate carbon dioxide in toluene at room temperature, yielding compound 15. The authors further showed the reversibility of this reaction, which gave back 14, upon heating in toluene at 90 °C, as confirmed by ¹H NMR spectroscopy.

Sydora³⁴  prepared phosphinoamidines 17a–f (Scheme 5), by reaction of the amine 16 with a variety of chlorocyclophosphines (except 17f, which used chlorodiphenylphosphine). Ligands were reacted with CrCl₃(THF)₃, affording the corresponding Cr(iii) complexes. While complexation of ligands 17a–d were carried out in THF, ligands 17e–f were complexed in MeCN, affording the corresponding MeCN adducts. Chromium complexes with ligands 17a, 17b and 17e were able to provide viable crystals, which could be structurally characterized by single crystal X-ray diffraction, verifying the P,N chelation. The complexes were further evaluated in ethylene oligomerisation. The authors found that 1-octene (C₈) selectivity was highly dependent on phosphine structure, where, at 40 °C, the catalyst derived from the aminophosphine with the smallest steric profile (17a) yielded the highest C₈ selectivity (54%), while the bulkiest PN-ligand (17e) reduced the C₈ selectivity to ca. 37%.³⁴ 

Sousa and Bechtold³⁵  prepared the Cu(i) complexes 19 and 21, from phosphinoamines 18 and 20, respectively (Scheme 6). The complexes were prepared by stirring [Cu(CH₃CN)₄]BF₄ with the corresponding phosphine ligand in dichloromethane under reflux, followed by addition of 1,10-phenanthroline with further reflux. Complexes 19 and 21 were obtained in 82 and 66% yields, respectively. Good stabilities in air under ambient conditions, at least for several months, were observed by the authors. Single crystals were grown for 21, and its molecular structure was determined by single-crystal X-ray diffraction studies. Photophysical studies were performed and the authors found that complex 19 was highly photoluminescent (quantum yield = 17.7%), in comparison with 21 (quantum yield = 4.6%), at 25 °C.³⁵ 

Sukhikh³⁶  reported the synthesis of phenylbenzothiazole derivatives 23–25, bearing phosphine groups (Scheme 7). The compounds were prepared from phenylbenzothiazole 22, by reaction with one equivalent of Ph₂PCl in toluene, in presence of triethylamine (for 23), or Ph₂PCl, ½ equivalents for the synthesis of 24, and ⅔ equivalents for 25. The compounds were evaluated for their photophysical properties and demonstrated dual band emission, whose intensity depended on the excitation wavelength of ultraviolet light both in the solid state and in solution (compound 25 showed the highest intensity).

Several catalytic applications of known phosphoramidites and/or their metal complexes have been reported during 2021, particularly on structural studies,³⁷  addition reactions,^38–45  allylic substitutions,^46–49  carbonylations,^50–52  cyclizations,^53–55  coupling reactions,^56,57  silylations⁵⁸  and alkenylation⁵⁹  reactions.

About the synthesis and application of phosphoramidites, for instance, Gu⁶⁰  reported the synthesis of ligands 26a–b (Scheme 8), through reaction of the p-(α-naphthyl)phenyl derived phosphochloridite with a secondary cyclic amine. The ligands were used in a ring-opening aryl-Narasaka acylation, where 26a showed the best performance, reaching 84% yield with 90% ee, using silafluorenes as substrates.

Cao and Xu⁶¹  prepared ligand 26c (Scheme 8), among other TADDOL phosphoramidites, and evaluated them in palladium-catalysed C–C bond activation of methylenecyclobutanes with aryl boronic acids, where 26c was the best ligand. The ligand was used in the reaction’s scope, where it promoted the synthesis of a family of indanes in yields up to 87%.

Gavrilov⁶²  reported a new biphenylene based P,S-bidentate phosphoramidite ligand 27 (Scheme 9). Its corresponding Pd complex was evaluated in asymmetric allylation reactions, with (E)-1,3-diphenylallyl as electrophile and several dialkyl malonates as C-nucleophiles, reaching conversions up to 100% and 78% ee for S-enantiomeric products.

Wang and Gong⁶³  also prepared a small family of biphenylene based phosphoramidite ligands (28a–c, Scheme 9), which was used in regioselective Pd-catalysed allylic amination reaction of dienyl allylic carbonates and aromatic amines. The authors found that 28b and 28c were very selective, the former providing the linear amines exclusively, while the latter gave the branched amines, always in quite high yields.

Trapp⁶⁴  reported ligand 31, which was prepared by reaction of biphenylene amine 29 with pivaloylglycine through amide coupling, using HOBt and EDCI, followed by deprotection. The obtained diol 30 was then rreacted with dichloro(diethyl)phosphine to yield the desired phosphoramidite ligand 31 (Scheme 10). The ligand was then evaluated in the Rh-catalysed asymmetric hydrogenation reactions of 2-acetamidoacrylate and methyl 2-acetamido-3-phenylacrylate, achieving enantiomeric excesses of up to 94%.

Chung and Wang⁶⁵  prepared ligand 32 (Scheme 11), based on H8-BINOL, and used it, among other known ligands, to promote the Cu(i)-catalysed asymmetric 1,3-dipolar cycloaddition reactions of azomethine ylides with several electron-deficient alkenes, providing a straightforward protocol for the preparation of enantioenriched pyrrolidines. Ligand 32, having a triple homoaxial (S,S,S)-chirality, played a pivotal role in the reaction, allowing the preparation of pyrrolidines in yields up to 99% and 99% ee.

Wang and Gong⁶⁶  reported the synthesis of new phosphoramidite 33a (Scheme 11) and evaluated it, along with other known phosphoramidite ligands, in the palladium-catalysed asymmetric allylic C–H alkylation of 1,4-pentadienes with α-angelica lactones. Ligand 33a showed the best performance, reaching 87% yield, with a Z/E 9 : 1 ratio and 87% ee. Under the optimized reaction conditions, the protocol was then extended to several other α-angelica lactones, reaching enantiomeric excesses up to 90%.

The same group⁶⁷  further synthesised phosphoramidite ligands 33b–c (Scheme 11), tested in the kinetic resolution of racemic thioanilide atropisomers by palladium catalysed C–H arylation, in presence of an anionic chiral Co(iii)-resolution complex, leading to both enantioenriched atropisomeric arylation thioanilides (up to 99% ee) and N–Me atropisomeric thioanilides (up to 99% ee), simultaneously.

Fürstner⁶⁸  developed a new class of phosphoramidites 35a–c and 36, prepared by reacting substituted BINOL derivatives 34 with N-benzylated dichlorocyclophosphazanes (Scheme 12). These and other ligands were evaluated in the asymmetric nickel-catalysed coupling of aldehydes with silyloxydienes, with 35c showing the best performance, reaching 35% ee for the 1,2 diol-product. Using other silyloxydienes, the scope of the reaction was tested, where 35c reached very high ee (up to 94%), in good yields (up to 85%).

Geng and Chang⁶⁹  reported the synthesis of BINOL based ligand 37a and, among other ligands, evaluated it on iridium-catalysed asymmetric reductive amination reactions (Scheme 13). Ligand 37a revealed the highest selectivity in the reaction of acetophenone with ammonium acetate, reaching 76% conversion and 95% ee for the S,S-diastereoisomer. The authors even managed to improve yield and selectivity, by using p-toluenesulfonic acid as additive, to 91% yield and 97% ee. Further scope using a family of aryl acetophenones produced the corresponding amines in quite high yields (79–91%) and ee > 94%.

Xu⁷⁰  also synthesised phosphoramidite 37b (Scheme 13), and compared it with other known ligands in the asymmetric Pd-catalysed hydrosilylation of ynones with varied substitution patterns (e.g., Me, OMe, F, Cl, Br, or ^tBu). Catalyst 37b showed the highest selectivity for the corresponding silylenones, in moderate to good yields (up to 94%), with high enantioselectivities (up to 98 : 2 er).

You’s group⁷¹  prepared ligands 38a–b (Scheme 13), which were evaluated along with other ligands in Ir-catalysed asymmetric allylic substitution reactions of indole derivatives, with 38a showing very high selectivity for the corresponding Z-product with up to 94% ee, in up to 93% yields.

Zhou⁷²  reported a family of spiro phosphoramidite ligands (39a–f and 40a–b, Scheme 14) and evaluated them in nickel-catalysed desymmetrising cyclization of 1,6-dienes. Ligand 40a showed the best performance, providing the desired products in yields ranging from 80% to 96%, with ee up to 97%.

Wang and Wang reported the utilisation of spirobiphenoxasilin-diol (SPOSiOL) 41, a Si-centered spirocyclic skeleton, to prepare its derived chiral phosphoramidite ligands 42a–c (Scheme 15). The ligands were then evaluated in Rh-catalysed asymmetric hydrogenation of α-dehydroamino acid derivatives, where 42a showed the best performance, reaching full conversions with ee in the range 91–96%.

Diamidophosphites have been used in hydroformylation,⁷³  allylic alkylation⁷⁴  and cycloaddition^75,76  reactions. Regarding the synthesis and application of diamidophosphites, Gavrilov’s groups have been particularly active in 2021.^77–80 

They prepared the family of diamidophosphites 45a–f, 47a–d, 48a–e and 49a–c, using chlorodiamidophosphites 43a–c (Scheme 16).^77,78  The ligands 45a–f were prepared by reacting 43a–c with alcohol 44, while 47a–d, 48a–e and 49a–c were obtained by reaction of chlorodiamidophosphites 43a–c with alcohol 46. These ligands were used in Pd-catalysed asymmetric allylic alkylation of (E)-1,3-diphenylallyl acetate with dialkyl malonates and the best results were obtained with P*,S-diamidophosphites 47a–d containing the 1,3-diaza-2-phosphabicyclo[3.3.0]octane unit with a stereogenic phosphorus atom. Ligands 47a,b provided an enantioselectivity as high as 91% ee, regardless of the type of thioether substituent.⁷⁷ 

The same group extended their ligand portfolio to the synthesis of bis-diamidophosphite ligand 50,⁷⁹  bearing stereogenic phosphorus atoms in the 1,3,2-diazaphospholidine rings, prepared from reaction of the corresponding chlorodiamidophosphite with (1R,2R)-[N,N′-bis(3-hydroxybenzylidene)]-1,2-diaminocyclohexane (Scheme 17).⁷⁹  This ligand was tested in Pd-catalysed asymmetric allylic alkylation of (E)-1,3-diphenylallyl acetate with dimethyl malonate, providing up to 73% ee and up to 80% ee in its amination with pyrrolidine. In addition, the same group also prepared ligands 51a,b, using the same chlorodiamidophosphite and reacting it with 5,10,15,20-tetraphenylporphyrin 51a and its zinc complex 51b (Scheme 17).⁸⁰  The ligands were further applied in asymmetric Pd-catalysed allylic alkylation of (E)-1,3-diphenylallyl acetate with dimethylmalonate, providing up to 63% ee.

During 2021, several reports were described regarding the use of known phosphites in a broad range of applications. Phosphite compounds have been studied by computational methods,^81–83  applied as promoters in photocatalytic hydroamination of unactivated alkenes,⁸⁴  studied in complexation reactions⁸⁵  and used as ligands in catalysis, namely in Rh-catalysed hydroformylation,^86,87  Ir-catalysed hydrogenation,⁸⁸  Ni-catalysed hydrocyanation,^89,90  and Ru-catalysed olefin methathesis.⁹¹ 

Regarding the development of new phosphites, Trofimov described the synthesis of polyfluoroalkyl phosphites 53a–b, through reaction of dichlorophosphites 52a–b with propargyl alcohol in Et₃N/hexane (−25 °C/−20 °C) (Scheme 18). The authors found that the phosphites 53a,b, obtained as transparent yellowish liquids, are both highly unstable and, upon storage at 5–7 °C, they easily undergo phosphite–phosphonate rearrangement and prototropic isomerisation, which lead to the formation of the corresponding allenylphosphonates and 1-propenyl phosphonates. The structure of the polyfluoroalkyl phosphite compounds 53a–b was confirmed by ¹H, ¹³C and ³¹P NMR, which presented a single peak at ca. 140 ppm, typical of monophosphite resonance.⁹² 

Clarke developed a series of phospholane–phosphite ligands, in order to study the effect of the ligand backbone structure on the selectivity, rate and catalyst stability of hydroformylation rhodium catalysts (Schemes 19 and 20).

Their synthesis involved, first, the preparation of the hydroxylated phospholanes 56 and 57a–b (Scheme 19). The hydroxylated phospholane aduct 56 was prepared through deprotonation of secondary phospholane 54 with n-BuLi, and subsequent reaction with ethane-1,2-diyl bis(4-methylbenzenesulfonate), followed by BH₃ deprotection of 55. On the other hand, deprotonation of phospholane 54 with n-BuLi, followed by a regioselective attack of the resulting anion at the less substituted carbon of the racemic 2-(trifluoromethyl)oxirane afforded a mixture of hydroxylated phospholane aducts 57a and 57b, in a 58 : 42 diastereomeric ratio, determined by ³¹P NMR spectroscopy. These diastereomers could be separated by column chromatography in silica gel. Then, the synthesis of phospholane–phosphite 58 was accomplished via the bromophosphite intermediate, obtained from reaction of the tropos biphenol with PBr₃, and direct reaction with precursor 56, in the presence of 1,4-diazabicyclo-[2,2,2]-octane (DABCO). Similarly, the preparation of phospholane–phosphite 60a–b, bearing a strongly electron-withdrawing CF₃ group in the backbone, involved the synthesis of the chlorophosphite intermediate 59, which was reacted directly with the precursors 57a or 57b, in the presence of DABCO (Scheme 19). The crystal structure of 57b was elucidated by X-ray diffraction studies, which allowed determination of the absolute configuration of each diastereomer.⁹³ 

The authors also reported the synthesis and catalytic evaluation of enantiomerically pure ligand (R _ax,R,R)-61, a variant of ligand 58, as well as the racemic ligands 62 and 63, variants of electron-deficient phospholane–phosphite ligand 60a (Scheme 20). A significant effect of the ligand structure on the catalytic activity and selectivity was observed in the Rh-catalysed hydroformylation of propene, where the best iso-regioselectivity (ca. 80%) was obtained with Rh/62, and the highest activity was achieved with Rh/63 (TON = 1148). This result was attributed to the presence of electron withdrawing groups at the ligand backbone, which turn the phosphorus atom less basic, leading to a decrease of the electron density in the metal center, with a consequent increased activity of the catalyst. Furthermore, the [–CH₂] extended chain provide the phospholane–phosphite ligands with high chemical stability, even at high temperatures (105 °C).⁹³ 

Thioether–phosphite ligands based on axially chiral (S)-BINOL and (S)-H₈-BINOL were developed by Bakos and Farkas. Their synthesis was performed by reaction of the corresponding chlorophosphites with thioether alcohols in basic medium (Scheme 21). After purification by column chromatography on silica gel, under an inert atmosphere, the thioether–phosphites 64–66 were obtained as waxy solids, and were shown to be highly stable at room temperature. In order to study their coordination chemistry, the reaction of one molar equivalent of chiral ligand with [Rh(COD)₂]BF₄ was carried out, leading to the formation of rhodium complexes [Rh(COD)(phosphite)]BF₄, isolated as air-stable yellow solids in good to excellent yields (84–92%). In the ³¹P NMR spectra, a doublet with a typical ¹ J (¹⁰³Rh,³¹P) coupling constant of ca. 250 Hz was observed. The coordination ability of Rh/thioether–phosphite complexes was investigated by 1D and 2D NMR techniques, using 1 molar equiv. or 1–5 molar equiv. of ligand, which confirmed their coordination to the metal centre in a bidentate hemilabile fashion. In the presence of excess of ligand, firm evidence was found for a dynamic behaviour, which involves the dissociation of the Rh–sulfur bond. Moreover, the thioether–phosphite ligands 64–66 and the corresponding Rh complexes were further applied in the catalytic asymmetric hydrogenation of benchmark substrates (methyl acetamidocinnamate, methyl acetamidoacrylate and dimethyl itaconate), leading to high conversions (up to 99% in 24 h) and moderate to good enantioselectivities (up to 91% for dimethyl itaconate, using [Rh(COD)₂]BF₄/66, in a 1 : 2 molar ratio).⁹⁴ 

During 2021, several reports were described regarding the use of known phosphonites and phosphinites in a range of applications. Particularly, phosphonites have presented interesting electroluminescent properties,⁹⁵  and have been used as ligands in cycloaddition reactions,⁹⁶  intramolecular cyclisations⁹⁷  hydoborations,⁹⁸  and coupling reactions.⁹⁹  On the other hand, phosphinites have been used as ligands in hydrosilylation,¹⁰⁰  coupling reactions¹⁰¹  and asymmetric hydrogenations.⁸⁸ 

Regarding the synthesis and application of new phosphonite ligands, Blokhin¹⁰²  reported the synthesis of binuclear non-symmetrical phosphor(iii)arene macrocyclic phosphonite 69, bearing different aromatic fragments, synthesised using two different bisphenols (4,4′-iminodiphenol and 4,4′-methylenediphenol) as the starting materials (Scheme 22). A two-step synthetic procedure was used, since the one-step three-component reaction between diols 67a, 67b and phenylphosphonic acid tetraethyldiamide failed to give the desired product 69. The first step of the two-step procedure involved the parallel bisphosphorylation of both 4,4′-iminodiphenol (67a) and 4,4′-methylenediphenol (67b) with PhP(Net)₂ to afford the corresponding arenebisamidophenylphosphonites 68a and 68b, respectively. The second step involved the reaction of either 68a with 67b or 68b with 67a in 50-fold excess of toluene at 100–105 °C. Ligand 69 was spectroscopically characterised by IR and ³¹P-NMR and its reactivity towards oxidation and sulfurization was evaluated by the authors.

Barbazanges reported the synthesis of helical chiral bisphosphinite ligands 72a–b (Scheme 23). The bis-methoxy derivative 70b, prepared via Sonogashira cross-coupling, was treated with BBr₃ leading to helixol 71b. However, in the case of 70a, a partial cleavage of the benzylic ethers has occured and acyclic adducts were obtained. Thus, basic conditions were employed (K₂CO₃, acetone, reflux) in order to obtain the desired heterohelixol 71a. The authors have developed preparative chiral HPLC methods, which allowed efficient separation of the enantiomers of each helixol 71a and 71b (ee > 99%). Finally, the subsequent phosphorylation reactions, carried out by treatment of 71a–b with chlorodiphenylphosphine PPh₂Cl, in basic medium, afforded the desired bisphosphinites 72a–b (Scheme 23). These ligands were applied in enantioselective Pd-catalyzed Tsuji–Trost allylation of N,O-bis(trimethylsilyl)acetamide with dimethyl malonate, leading to moderate yields (up to 62%) and moderate enantioselectivity (up to 66% ee). Moreover, the ligand to metal ratio was shown to be crucial, with the (R) and (S) product being selectively formed, by changing the P : Pd ratio from 4 : 1 to 1 : 1.¹⁰³ 

Sundermeyer developed a reductive aromatization strategy to convert peropyrenequinone 73 into peropyrene tetra-phosphinite 74 (Scheme 24). After treatment of 73 with n-butyllithium, the resulting tetra-lithoxy intermediate was directly reacted with di-tert-butylchlorophosphine, in THF, to afford air sensitive phosphinite 74 in 45% isolated yield (Scheme 24). The influence of the phosphinite substituents on the optoelectronic properties of the π-system-extended peropyrene was investigated, in the solid state, by X-ray crystallography and, in solution, by UV-Vis and fluorescence spectroscopy. The fluorescence quantum yield of 74, measured in CH₂Cl₂, was 72%.¹⁰⁴ 

Morales-Morales reported the synthesis and characterization of a series of non-symmetric Ni(ii)–POCOP pincer complexes 78a–b and 79a–b, functionalized with benzothiazole or benzimidazole moieties at the meta position of the aryl ring (Scheme 25). First, the POCOP-pincer ligands 76a–77a (benzothiazole) and 76b–77b (benzimidazole) were obtained through a procedure that involved the reaction of the dihydroxylated proligands 75a and 75b, respectively, with a base (Et₃N or DMAP) in THF, followed by treatment with different chlorophosphines (R = ⁱPr or ^tBu), under nitrogen atmosphere. The resulting POCOP ligands were directly used, without any purification procedure, for metalation with NiCl₂ in refluxing toluene, generating air-stable Ni–POCOP pincer complexes 78a, 78b, 79a and 79b. All complexes were fully characterised by spectroscopic techniques and demonstrated to be air, water and thermally stable. In addition, single crystal X-ray diffraction studies confirmed the tridentate coordination mode of the pincer ligands, leading to a distorted square planar geometry around the Ni(ii) center. Furthermore, the catalytic activity of the complexes was explored in cross-coupling reactions of functionalized benzaldehydes and boronic acid derivatives to produce diarylketones, and the best performance was achieved with nickel complexes bearing a less sterically hindered donor group 78a–b (R = ⁱPr). On the other hand, the benzothiazole and benzimidazole moieties did not show any noteworthy influence on the catalytic activity of the Ni pincer complexes.¹⁰⁵ 

Mixed phosphorus compounds, bearing at least two different tervalent phosphorus moieties, have also been reported during 2021, including in asymmetric hydrogenation¹⁰⁶  and asymmetric hydroformylation/intramolecular cyclisation¹⁰⁷  reactions.

Ling and Zhong prepared the mixed phosphine–phosphoramidite ligands 82, by reacting the ferrocene-based phosphine 80 with the phosphochloridite 81 (Scheme 26).¹⁰⁸  The ligands were evaluated in the Ru-catalysed enantioselective amination of diols, where phosphine–phosphoramidite (Sc, Rp, S)-82 presented the best performance, providing amino alcohols in high yields and ee (up to 98%).

Hu also reported the highly stable structurally rigid bicyclic bridgehead phosphine–phosphoramidites 84a–e (Scheme 27),¹⁰⁹  which were synthesised from the reaction of aminophosphine 83 with PBr₃, leading to the target ligands. These were then tested in Rh-catalysed asymmetric hydrogenation of 2-vinylanilides for the synthesis of optically active anilines bearing an ortho-tertiary benzylic stereocenter in up to 99% yields, with up to 99% ee.

Jayaprakash synthesised mixed ligands 87a–b, along with their corresponding Mn(i) complexes 88a–b (Scheme 28).¹¹⁰  The synthesis started with the condensation between aldehydes 85a–b and (1S,2S)-2-aminocyclohexan-1-ol in methanol at room temperature, followed by their in situ reduction with sodium borohydride, affording phosphines 86a–b. Deprotonation of 86a and 86b, followed by the addition of PPh₂Cl, provided mixed phosphine–phosphinite ligands 87a and 87b with 95% purity, determined by ³¹P-NMR. Further complexation with MnBr(CO)₅ afforded catalyst 88a and 88b, which were then evaluated in the asymmetric transfer hydrogenation of aromatic ketones, with 88b being the best catalyst, giving ee around 85% for electron-withdrawing substituted ketones and >90% ee for electron-donating substituted ketones.

Habib¹¹¹  described the synthesis of phosphine–phosphinite ligands 91a–e (Scheme 29), which were prepared by derivatization of phenols by introducing a PR₂ moiety (R = ⁱPr₂, aryl) at the oxygen atom, and then a second PR₂ moiety on the ortho position. The second phosphine moiety was introduced by rearrangement of the phosphine moiety bound to oxygen by ortho lithiation with n-butyl lithium, followed by the reaction of 90a–d with chlorophosphine ClPR₂. The phosphine–phosphinite ligands 91a–e were further used to study the potential of the phosphine moiety to affect the C–Ni bond functionality of dimeric nickel complexes.