Ylide-Substituted Phosphines with a Cyclic Ylide-Backbone: Angle Dependence of the Donor Strength

Abstract

Ylide-substituted phosphines (YPhos) have been shown to be highly electron-rich and efficient ligands in a variety of palladium catalyzed transformations. Here, the synthesis and characterization of novel YPhos ligands containing a cyclic backbone architecture are reported. The ligands are easily synthesized from a cyclic phosphonium salt and the chlorophosphines Cy₂PCl (L1) and Cy(Flu^Me)PCl (L2, with Flu^Me = 9-methylfluorenyl) and were characterized in both solution and solid states. The smaller PCy₂-substituted ligand, L1, readily formed the biscoordinate L1₂Pd species when treated with Pd₂(dba)₃ and showed no activity in palladium-catalyzed amination reactions even when applied as defined palladium(II) η³-allyl, t-Bu-indenyl, or cinnamyl precursors. Bulkier fluorenyl-substituted ligand L2 similarly was inactive, despite its ability to form the stable monophosphine complex L2·Pd(dba). Assessment of the electronic properties by experimental and computational methods revealed that L1 and L2 are considerably less electron-rich than previously synthesized YPhos ligands. This was shown to be the result of the small P–C–S bond angle, which is sterically enforced due to the cyclic nature of the backbone. Density functional theory calculations revealed that the small angle results in an increased s-character of the lone pair at the ylidic carbon atom and leads to a polarization of the C–P bond toward the carbon atom, thus decreasing the electron density at the phosphorus atom. The results demonstrate the tunability of the donor strength of YPhos ligands by modification of the ligand backbone beyond simple changes of the substitution pattern and are thus important for future ligand design, with a careful balance of many factors to be considered to achieve catalytic activity.

Introduction

Phosphines are a privileged class of ligands ubiquitous in homogeneous catalysis, a vital cornerstone of modern synthetic chemistry, with applications ranging from large-scale pharmaceuticals to fine stereo control of complex natural products.¹⁻³ Although extensive research efforts have focused on N-heterocyclic carbenes (NHCs) as ligand systems in the past few decades due to their in general higher electron-donating properties when compared with phosphines, phosphines remain the ligand systems of choice for many applications. This continuing popularity is due to several reasons, including the simplicity of their synthesis, generally from commercially available starting materials, and the wide breadth of possible substitution patterns, which allows for the precise control of the phosphines electronic and steric properties, enabling the fine-tuning of the phosphine properties toward a desired purpose. Several advances in homogeneous catalysis over previous decades have been driven by this fact.^4,5 For systematic ligand design various metrics are used in order to describe the electronic and steric properties of phosphine ligands, including the Tolman electronic parameter (TEP),^6,7 cone angle, and the buried volume.⁸ These measures facilitate the understanding of structure activity relationships in catalysis and build the foundation to computer-aided catalyst design.⁹

In the past years, many research efforts have focused on the preparation of highly electron-rich phosphines due to their beneficial properties for late transition metal catalysis. For a long time, trialkyl phosphines particularly with tertiary alkyl substituents such as t-Bu or Ad (Ad = adamantyl) groups were the phosphines ligands with the highest donor strength. For example, Carrow and co-workers synthesized tri(1-adamantyl)phosphine, PAd₃, and impressively demonstrated its applicability in palladium catalysis.¹⁰ However, other substituents at phosphorus also have been applied in past years to further increase the donor strength. Thus, Dielmann and co-workers reported on the use of imidazolidin-2-imino substituents (IAPs),¹¹ and Sundermeyer and co-workers on phosphazenyl groups (PAPs) to access super basic phosphines (Figure 1).¹² Their donor capacity can compete with or even surpass that of NHCs, demonstrating the electronic flexibility of specially designed phosphine ligands.

Figure 1.

Examples of highly electron-rich phosphines.

Recently, our group reported on ylide-substituted phosphines (YPhos), which feature an ylide group directly bonded to the phosphorus center.¹³ As a result of this ylide group and its strong electron-donating abilities, these phosphines exhibited strong electron-donating capabilities, with donor capacities surpassing those of simple phosphines and NHCs. These ligands were shown to be ideal scaffolds for catalytic applications, ranging from gold-catalyzed hydroamination and cyclization reactions¹⁴ to palladium catalysis, such as Buchwald–Hartwig aminations as well as Negishi couplings and α-arylation reactions under mild conditions.¹⁵ Recently, it has been shown that YPhos palladium catalysts are even capable of directly coupling aryl chlorides and alkyl lithium reagents, giving very high selectivities while supplementing the need for any additional transmetalation reagents.¹⁶

Due to the efficiency of YPhos ligands in catalysis, we became interested in a more profound understanding of the impact of the ligand architecture on the catalytic ability. In this contribution, we showcase the development of a new class of YPhos ligands with a cyclic ylide functionality and demonstrate that this cyclic structure has a decisive impact on the electronic properties.

Results and Discussion

Ligand and Complex Syntheses and Catalytic Studies

In order to synthesize a cyclic YPhos ligand, we selected easily accessible phosphonium salt 1 as the first test system. We initially set out to synthesize corresponding dicyclohexylphosphine L1 in order to allow for comparison with previously synthesized YPhos ligands. Phosphonium salt [1]I could be obtained in 57% yield in a one-pot reaction of 2 equiv of nBuLi and TMEDA with thiophenol, followed by the slow addition of di-iso-propylchlorophosphine to give an ortho-phosphino thiolate, which was subsequently cyclized via diiodomethane addition (Figure 2). The successful formation of [1]I could be confirmed by a singlet at δ_p = 76.5 ppm in the ³¹P{¹H} NMR spectrum and by XRD analysis (see the Supporting Information for further details). The structure shows a distorted five-membered ring, with a slight deviation of the phosphorus center out of the plane (Figure 2). Starting from [1]I, the corresponding ylide could be isolated in near-quantitative yields by treatment with strong, sterically demanding metal bases such as KHMDS or KOt-Bu. Deprotonation using such bases in THF afforded ylide 2 within 1 h. Attempted deprotonation using other bases such as nBuLi or KH resulted in the formation of a mixture of products, likely due to unwanted deprotonation of the CH group of the isopropyl moiety. The ylide is a dark red oil at room temperature, characterized by a high-field-shifted signal at δ_p = 50.2 ppm in the ³¹P{¹H} NMR spectrum. In keeping with the reactivity of ylide species, exposure of the solution to air resulted in the formation of a decomposition product, in which the phosphonium moiety is oxidized to phosphine oxide, with the concomitant loss of the bridging carbon atom between the sulfur and phosphorus centers (see the Supporting Information). Attempts to obtain the corresponding yldiide of 2 by using a second equivalent of a strong metal base were unsuccessful, even with the use of strong bases such as NaNH₂, KHMDS, or benzyl potassium and at elevated temperatures.

Figure 2.

Synthesis of the YPhos ligands L1 and L2 with a cyclic ylide backbone and molecular structures of the cation of 1. Ellipsoids drawn at 50% probability level; hydrogen atoms omitted for clarity. Selected bond lengths [Å] and angles [°]: P(1)–C(1) 1.803(2), S(1)–C(1) 1.824(2), P(1)–C(1)–S(1) 108.3(1).

With the ylide in hand, desired ligand L1 was synthesized by slow addition of 0.5 equiv of Cy₂PCl to 2. Thereby, the ylide acts as reagent and base, thus leading to the re-formation of the chloride salt of 1 ([1]Cl) and final ligand L1. Within the synthesis pathway, formation of an α-phosphino-substituted phosphonium salt was never observed, suggesting that its deprotonation to L1 by ylide 2 is a fast process. After removal of [1]Cl by filtration ([1]Cl being reobtained in 82% yield), L1 was obtained as orange solid in 79% yield. Analysis via ³¹P{¹H} NMR spectroscopy showed a set of doublets at δ_p = 50.7 ppm for the phosphonium group and at δ_p = −14.9 ppm for the phosphine moiety with a coupling constant of ²J_pp = 108.6 Hz. Recrystallization from hot acetonitrile furnished crystals suitable for XRD analysis, which confirmed the connectivity of the new ligand (Figure 3).

Figure 3.

(A) Molecular structures of L1 and L1·AuCl; ellipsoids are drawn at a 50% probability level. Hydrogen atoms are omitted for clarity. Selected bond lengths [Å] and angles [°]: L1: P1–C1 1.7077(12), S1–C1 1.7920(12), C1–P2 1.7746(12), P1–C1–S1 111.44(6), P1–C1–P2 124.43(7). L1·AuCl: P1–C1 1.717(2), S1–C1 1.793(1), C1–P2 1.754(2), P2–Au1 2.248(1), Au1–Cl1 2.303(1), P1–C1–S1 112.7(1), P1–C1–P2 130.0(1). (B) Steric map of L1 derived from L1·AuCl using the SambVca 2.0 (P–M distance of 2.28 Å including H atoms).¹⁸

Next, the ability of phosphine L1 in palladium catalysis was tested by means of Buchwald–Hartwig aminations. The reactions were carried out using 4-bromotoluene and 4-chlorotoluene as electrophiles and n-butylamine and N-methylaniline as amines with 3 mol % of Pd₂dba₃ and free ligand, both at room temperature and 50 °C.¹⁷ After a period of 24 h, the reaction mixture was analyzed by GC analysis, revealing no detectable conversion for any substrate combination. This contrasts the high activity of the acyclic YPhos ligands which provided high yields in these reactions at room temperature usually within only a couple of minutes.^15,16

To understand the origin of the inactivity of L1, we set out to examine its steric and electronic properties. Reaction of the new ligand with (tht)AuCl allowed us to obtain the corresponding gold complex in 73% yield, which was fully characterized by various analytical methods, including XRD analysis (see the Experimental Section for full details). The obtained molecular structure of L1·AuCl shows a slight shortening of the C1–P2 bond [1.754(2) Å] upon coordination of the metal [1.775(1) Å in L1], as expected. The P1–C1–P2 angle increases from 124.43(7) to 130.0(1)°, thus reducing steric repulsion between the iPr groups and the gold center (Figure 3). From this structure, the buried volume was calculated using SambVca.¹⁸ The obtained buried volume of 42.5% is lower than that of other acyclic YPhos ligands with PCy₃ or PPh₃ moieties, which usually showed volumes of approximately 50%.¹³⁻¹⁵ Furthermore, the steric map of L1 shows areas of steric bulk only on one side of the of the ligand, namely, the phosphonium moiety. The cyclohexyl groups are unable to reach toward the metal center. This observation suggests, that L1 might be too small to efficiently stabilize the monoligated LPd species and instead leads to the formation of the corresponding bisphosphine complex L₂Pd, which have been shown to be inactive in catalysis.^16,19

To prove this hypothesis, we attempted the isolation of palladium complexes with L1. Reaction of L1 with Pd₂(dba)₃ resulted in instantaneous formation of a new species, detected via ³¹P{¹H} NMR spectroscopy. The NMR spectrum showed a set of two doublets at δ_p = 53.6 and 13.1 ppm. The reaction reached full conversion within 10 min, and the resulting bisphosphine ligated complex could be isolated in high yields of 81%. Crystallization from pentane and THF confirmed the formation of the L₂Pd species (Figure 4). While the formation of such complexes was found to be hampered with the acyclic YPhos ligands reported earlier,¹⁶L1 readily forms the bisphosphine species because of its low steric bulk. In the molecular structure, the palladium lies on the inversion center between the two phosphine atoms, with the P(2)–Pd(1) bond of 2.300(1) Å being shorter than the one found for the acyclic keYPhos ligand, Cy₃PC(Me)PCy₂, featuring a PCy₃ moiety (2.332(6) Å).¹⁶ Also of interest is the observation of a short Pd(1)–H(11) distance of only 2.47(3) Å between the palladium and the hydrogen atom of an iPr group. Thus, the palladium center adopts a pseudo-square-planar geometry. The bonding nature of the Pd–H interaction was also confirmed by DFT calculations. QT/AIM studies showed the presence of a bond critical point between the Pd and H atoms, suggesting the presence of a stabilizing interaction. Such interactions have been previously observed with the keYPhos ligand; however, in the latter case a slightly larger distance of 2.52(3) Å was observed.^15a,16

Figure 4.

(Top) Molecular structure of L1₂Pd; ellipsoids are drawn at a 50% probability level. Hydrogen atoms are omitted for clarity. Selected bond lengths [Å] and angles [°]: P(1)–C(1) 1.706(3), S(1)–C(1) 1.796(3), C(1)–P(2) 1.766(3), P(1)–Pd(1) 2.300(1), Pd(1)–H(11) 2.44(3), P(1)–C(1)–S(1) 110.9(1), P(1)–C(1)–P(2) 126.4(1), P(2)–Pd(1)–H(11) 88.59(2). (Bottom left) Contour-line diagram of the Laplacian distribution of L1₂Pd. Solid lines indicate charge depletions [∇²ρ(r) > 0]; dashed lines indicate charge concentrations [∇²ρ(r) < 0]. (Bottom right) Section of the molecular structure of L1₂Pd.

As previously shown via mechanistic studies, monoligated PdL species are the active catalysts in a plethora of palladium-catalyzed reactions involving bulky monophosphines, while in contrast the PdL₂ species are often considered as catalytically inactive.^16,19 To prevent the formation of PdL₂ defined precatalysts featuring a ligand to metal ratio of 1:1 have been designed to facilitate the formation of catalytically active Pd species and inhibit the formation of PdL₂.²⁰ Prominent examples of such monophosphine systems include η³-allyl, t-Bu-indenyl, and cinnamyl complexes.²¹ In this vein, precatalysts of L1 based on the η³-allyl, cinnamyl, and t-Bu-indenyl complexes were synthesized (Scheme 1) by mixing of the free ligand and the dimeric palladium precursors in pentane giving the three desired complexes (Pd_al1, Pd_cin1, and Pd_ind1) in good yields directly from solution. In all three cases, NMR spectroscopic analysis shows one set of doublets with similar shifts in the ³¹P{¹H} NMR spectrum, with the resonances for the phosphonium moiety appearing at approximately δ_p = 56 ppm and for the phosphine moiety between δ_p = 20.3 and 27.0 ppm. Structural analysis was carried out on Pd_cin1 and Pd_ind1, with both showing shortened carbon–phosphorus bonds [1.7563(3) and 1.720(2) Å] and larger P(1)–C(1)–P(2) angles [130.0(2) and 130.1(1)°] compared to the free ligand. The isolated palladium complexes were not only applied in BHA but also showed no activity under the reaction conditions mentioned above.

Scheme 1. Synthesis of the Palladium Complexes with L1 and L2.

We also attempted to crystallize the Pd–allyl complex; however, we were instead only able to obtain a decomposition product (Figure 5), in which a second phosphonium moiety is formed through substitution of the allyl group from the Pd–allyl precursor together with [Pd(allyl)Cl₂]⁻ as counterion. Formation of this species is very slow, with only small quantities formed after several days in solution, thus excluding that this decomposition reaction is the reason for the catalytic activity of L1.

Figure 5.

(Top) Molecular structures of (left) Pd_cin1 and (right) Pd_ind2 with ellipsoids drawn at a 50% probability level. Hydrogen atoms are omitted for clarity. Selected bond lengths [Å] and angles [°]: Pd_cin1: P1–C1 1.715(3), S1–C1 1.791(3), C1–P2 1.756(3), P2–Pd1 2.306(1), Pd1–Cl1 2.365(1), P1–C1–S1 111.1(1), P1–C1–P2 130.0(2). Pd_ind1: P1–C1 1.720(2), S1–C1 1.796(2), C1–P2 1.756(2), P2–Pd1 2.283(1), Pd1–Cl1 2.376(1), P1–C1–S1 110.4(1), P1–C1–P2 130.1(1). (Bottom) Molecular structure of decomposition product of Pd_al1.

We then assessed the electronic properties of the dicyclohexyl ligand by measurement of the CO stretching vibration in [L1·Rh(acac)(CO)]. Correlation with stretching frequencies in LNi(CO)₃ complexes yielded a TEP of 2057.1 cm^–1. This value is surprisingly high compared with other YPhos ligands, which usually featured high donor strengths similar to carbenes. For example, the acyclic YPhos ligands keYPhos with a PCy₃ moiety and a methyl group in the ylide backbone exhibited a TEP value of 2050.1 cm^–1.¹² The high TEP value of L1 was further confirmed by DFT calculations. To this end, the minimum of the molecular electrostatic potential (MESP) at the phosphorus atom was calculated and correlated with the TEP, a method previously described by Suresh and Koga.²² This method yielded a TEP_ESP of 2057.1 cm^–1, thus corroborating with the experimentally obtained value.

To understand whether the catalytic inactivity of the palladium complexes of L1 is due to the low steric demand or the weak donor strength, we accessed an analogues ligand to L1 with a bulkier phosphine moiety containing a 9-methylfluorenyl substituent. The fluorenyl group was expected to further expand toward the coordination sphere of the metal and thus impede the formation of the detrimental L₂Pd species. New ligand L2 was synthesized from in situ formed ylide 2 and CIPCyFlu, which was made from 9-methylfluorene (Flu^MeH, see the Supporting Information)²³ by deprotonation with nBuLi and slow addition to a solution of cyclohexyldichlorophosphine at −60 °C. Similar to L1, slow addition of 0.5 equiv of the chlorophosphine to 2 directly gave way to final ligand L2, which was obtained as a yellow solid in 65% yield after purification. L2 is characterized by a set of doublets at δ_p = 52.0 and 12.2 ppm with a coupling constant of ²J_pp = 119.5 Hz in the ³¹P{¹H} NMR spectrum. Recrystallizing from hot acetonitrile furnished single crystals of L2 suitable for single-crystal XRD analysis (Figure 6). The ligand adopts the desired conformation, whereby the fluorenyl substituent is directed toward the coordination site of the phosphine. Thus, L2 appears to be bulkier than L1. The P1–C20 bond length, which is that between the fluorenyl and the phosphine, amounts to 1.927(1) Å and is thus relatively long when compared to typical bonds between phosphorus and an alkyl group (e.g., P2–C14 1.847(1) Å). In contrast, the P1–C1–P2 bond angle of 124.5(1)° is similar to that in L1 and thus seems to be unaffected by the increased bulk of the fluorenyl group.

Figure 6.

Molecular structure of L2. Selected bond lengths [Å] and angles [°]: P1–C1 1.713(1), S1–C1 1.794(1), C1–P2 1.761(1), P2–C20 1.927(1), P1–C1–S1 109.5(1), P1–C1–P2 124.5(1).

To gain further insights into the steric properties of L2 we attempted the synthesis of the gold chloride complex in order to determine the buried volume, %V_bur. Reaction with (tht)AuCl resulted in the formation of a new species, as detected by NMR spectroscopic studies. We were able to observe a set of doublets in the ³¹P{¹H} NMR spectrum at δ_p = 58.4 and 44.7 ppm, with a coupling constant of ²J_pp = 57.4 Hz. However, due to rapid decomposition in solution, isolation of this species was not possible. Thus, we right away focused on the synthesis of palladium complexes with L2. The reaction of L2 with Pd₂(dba)₃ gave way to two species. The first of these was identified as bisphosphine complex L2₂Pd, which showed the characteristic coupling pattern with a set of two doublets at δ_p = 52.0 and 38.8 ppm in the ³¹P{¹H} NMR spectrum, similar to L1₂Pd. Unambiguous identification came from XRD analysis (Figure 7). Unfortunately, disorder within the structure prevented a detailed discussion of bond lengths and angles, but the molecular connectivity is unambiguous. Of note is the change in conformation of the fluorenyl group when compared to the free ligand. Presumably, this change is instrumental to allow two ligands to coordinate to the Pd(0) center.

Figure 7.

Molecular structure of L2₂Pd with ellipsoids drawn at 50% probability level. Hydrogen atoms are omitted for clarity.

As a consequence of the increased bulk of the fluorenyl group relative to the Cy group and the required conformational changes to form L2₂Pd, the diphosphine complex was not formed selectively from the reaction with Pd₂(dba)₃. The second species showed a pair of doublets at δ_p = 55.2 and 29.0 ppm, with the former being in the region expected for the phosphonium moiety and the latter in the region previously observed for PCy₂ moiety of the LPd(dba) complex with an acyclic YPhos ligands. Thus, this species presumably is L2·Pd(dba), which we would expect to be a catalytically competent species according to previous studies by our as well as other groups.^16,19

To further understanding how the ligand forms the two complexes, L2·Pd(dba) and L2₂Pd, we endeavored to follow the reaction of stoichiometric quantities of L2 with Pd₂(dba)₃ via ³¹P{¹H} NMR spectroscopy using triphenylphosphine as an internal standard. The reaction plot (Figure 8) shows that the beginning of the reaction is characterized by the rapid formation of L2·Pd(dba). Within the first 30 min, the rate of formation slows considerably, while simultaneously there is an increase in the formation of L2₂Pd and a decrease in the concentration of free ligand. Within 3 h, an equilibrium state is reached. Throughout the course of the reaction, L2₂Pd is observed to slowly precipitate. These observations suggest that the combination of free ligand and L2·Pd(dba) leads to the formation of PdL₂. Nonetheless, a significant amount of the proposed active catalyst L2·Pd(dba) remains available for entering the catalytic cycle.

Figure 8.

Schematic showing the formation of L2Pd(dba) and L2₂Pd complexes from free ligand L2 and Pd₂(dba)₃ as monitored by ³¹P(¹H) NMR spectroscopy with Ph3PO as internal standard.

Analogous to L1, we also synthesized palladium precatalysts Pd_al2, Pd_cin2, and Pd_ind2, which were obtained as pure solids in yields of 66–76% (Scheme 1). Analysis by ³¹P{¹H} NMR spectroscopy showed two sets of doublets in the case of all complexes; all three had similar chemical shift and coupling constant values. These two sets of signals are probably due to the formation of diastereomers connected with the stereocenter at phosphorus and were also seen in the ¹H and ¹³C NMR spectra. In the molecular structures of Pd_al2 and Pd_ind2, however, only one isomer was observed (Figure 9). In contrast to the structure of L2₂Pd, the fluorenyl groups in the Pd(II) complexes remained in the desired orientation pointing toward the metal center. Nonetheless, one fluorenyl group seems not to be sufficient to fully shield one site of the metal, especially in comparison to biaryl phosphine ligands.²⁴ This can be seen from the steric map of 2 constructed based on the molecular structure of Pd_al2. Whereas the map clearly reflects the increased steric bulk of L2 compared to L1 exerted from both the ylide and fluorenyl substituents, the steric demand remains moderate.

Figure 9.

(Top) Molecular structures of (left) Pd_al2 and (right) Pd_ind2. Selected bond lengths [Å] and angles [°]: Pd_al2: P1–C1 1.724(2), S1–C1 1.780(3), C1–P2 1.754(2), P2–C20 1.935(2), P2–Pd1 2.349(1), Pd1–Cl1 2.389(1), P1–C1–S1 110.3(1), P1–C1–P2 129.9(1). Pd_ind2: P1–C1 1.730(2), S1–C1 1.769(2), C1–P2 1.763(2), P2–C20 1.955(2), P2–Pd1 2.319(1), Pd1–Cl1 2.373(1), P1–C1–S1 110.8(1), P1–C1–P2 131.1(1). (Bottom) Steric map of Y_Cycl2 derived from the complex Pd_al2 using SambVca 2.0 (P–M distance of 2.28 Å including H-atoms).¹⁸

Applications of L2 in Buchwald–Hartwig aminations analogous to those described above remained unsuccessful independently of whether the active species was formed from L2 and Pd₂(dba)₃ or one of the isolated precatalysts. Since the reaction monitoring of L2 and Pd₂(dba)₃ has revealed that in contrast to L1 at least some active catalyst [L2·Pd(dba)] should be present, we reasoned that the donor strength of L2 would be too low to efficiently enable oxidative addition under the reaction conditions. Measurement of the CO stretching frequency in the in situ formed [L2·Rh(acac)(CO)] complex yielded a TEP of 2059.4 cm^–1. Surprisingly, this value is even higher than that of L1, suggesting that L2 is an even weaker donor ligand. The high value was again confirmed by calculation of the minimum of the electrostatic potential (MESP) at the phosphorus (TEP_ESP = 2059.8 cm^–1). As both ligands display extraordinarily high TEP values and both feature the cyclic ylide moiety, we concluded that the cyclic ylide itself does not behave as a strongly electron-donating substituent.

Impact of the Cyclic Backbone on the Donor Properties

Two possible reasons can be theorized for the low electron-donating properties of the cyclic ylide: (i) the anion-stabilizing effects of the sulfur bound to the ylidic carbon or (ii) the cyclic nature of the ylide. To probe these theories further, a series of phosphines with similar cyclic ylide backbones but differing substituents Z in the backbone were investigated via DFT calculations and compared with related previously synthesized acyclic YPhos ligands featuring similar backbones. The MESP at the phosphine was determined, and the corresponding TEP calculated. The results are summarized in Table 1. As clearly shown by the obtained results, all cyclic phosphines examined exhibit high TEP values independent of the Z substituent. In contrast, the related acyclic analogues feature much lower TEP values. This observation impressively demonstrates that the presence of the sulfur atom does not play the decisive role in the observed electronic properties, but instead it is the cyclic nature of the ylide, that exerts the biggest influence on the electron-donating abilities.

Table 1. Calculated TEP Values of YPhos Ligands with Cyclic Ylide Backbones and Comparable Acyclic YPhos Ligands Previously Reported.

Z (cyclic YPhos)	MESP [eV]	TEPESP [cm–1]	acyclic YPhos ligand	TEPexp [cm–1]
S	–54.15914	2057.1
SO	–54.14821	2060.3
SO2	–54.14466	2061.3	YSPCy213	2055.1
NMe	–54.15479	2058.4
SiMe2	–54.15951	2057.0	YSiPCy213	2048.8
CMe2	–54.16339	2055.9	YMePCy214a	2050.1

To further understand the impact of the cyclic geometry on the electronic properties of the YPhos ligands, we analyzed the electronic structure of the ligands by natural bond orbital (NBO) analyses (Table 2). These NBO studies reinforced our assumption by showing significant changes in the character of the P–C bond. Thus, the orbital of the ylidic carbon atom involved in the P–C bond to the phosphine moiety of L1 and L2 features a higher s-character in comparison to what would be expected for an ideal sp²-hybridized carbon center (values of 39.4 vs 33.3%). This increased s-character is due to the small P–C–S angle of 111.4° which is enforced by the geometric restraints of the ring structure and results in a higher p-character in the corresponding bonds. For comparison reason, the electron-rich, acyclic keYPhos and joYPhos ligands were also investigated. As expected, keyPhos showed a much lower s-character in the P–C bond due to the larger P–C–C_Me angle of 117.2°, and joYPhos, which is most electron-rich donor in this series, features the largest angle of 123.0° and thus displays the lowest s-character within the orbital with only 30.8%. As a result of this high s-character within the bonding orbital of L1, the C–P bond is much more polarized toward the carbon atom, thus explaining why the cyclic ylide in L1 and L2 donates less electron density toward the phosphine. This explanation is also in agreement with the observation that keYPhos and joYPhos are much more electron-rich, with these species having a lower s-character and hence a less polarized C–P bond. This observation impressively demonstrates that the angle between the phosphonium moiety, the ylidic carbon atom, and Z substituent has a significant influence on the overall donor ability of the ligand system. It is noteworthy that studies on the angle dependency of metal ligand interactions have been reported before, particularly focusing on the impact of the angle at the donor site (i.e., the C–P–C angles) on the donor properties.²⁴

Table 2. Results of the NBO Analysis of YPhos Ligands, Observed P–C–Z Angles, and Their Measured TEP Values^a.

ligand	α-P–C–Z angle [deg]	σ(C) s-character	σ(C) p-character	bond polarization % P	bond polarization % C	TEP [cm–1]
L1	111.44	39.4%	60.5%	34.8%	65.2%	2057.1
keYPhos	117.17	33.5%	66.4%	36.5%	63.5%	2050.1
joYPhos	122.98	30.8%	69.1%	41.2%	58.8%	2049.3
L3	111.14	38.2%	61.8%	33.6%	66.3%	2061.3b
YSPCy2b	121.50	36.5%	63.5%	34.9%	65.1%	2057.0

^a

^bCalculated TEP from MESP.

To obtain a quantitative estimation of the impact of the cyclic structure on the donor properties, cyclic ligand L3 with Z = SO₂ was calculated and compared to its acyclic counterpart Y_SPCy₂. Similar to the observations mentioned above, the acyclic ligand features a much larger P–C–S angle, resulting in a higher p-character on the σ(C) orbital and therefore in a higher bond polarization toward the phosphorus atom and a lower TEP value.

We further evaluated the angle-dependency of the donor strength by computationally increasing the angle at the ylidic carbon atom on the example of the acyclic YPhos ligand, keYPhos and monitoring the impact on the electronic structure (Figure 10). Increasing the Cy₃P–C–C_Me angle from 90 to 120° resulted in an increase of s-character in the respective bonds and hence in an increase of p-character of σ(C) in the P–C bond to the phosphine moiety. This simultaneously leads to a lower bond polarization toward the carbon atom and hence to a lower (more negative) MESP value. Overall, the widening of the P–C–C-angle in keYPhos by 30° leads to a decrease of the TEP value by 1.4 cm^–1. This change is similar to the expected change when replacing one n-butyl group (χ_i = 1.4 cm^–1) such as in P(n-Bu)₃ by a tert-butyl substituent (χ_i = 0 cm^–1).^6b

Figure 10.

Calculated MESP values at the lone pair at the phosphorus, p-character of σ(C), and polarization of the C–PCy₂ bond toward the P atom depending on the P–C–C angle in the keYPhos ligand.

Conclusion

In summary, we have reported on the synthesis and characterization of two novel YPhos ligands with a cyclic ylide and a PCy₂ (L1) and P(Cy)(Flu^Me) substituent (L2, Flu^Me = 9-methylfluorenyl), respectively. The unexpected complete inactivity of various palladium complexes of these ligands in Buchwald–Hartwig aminations led us to a detailed investigation of the spatial and electronic properties of the novel phosphines. Analysis from a spatial viewpoint revealed that the cyclic ylide offers very little steric protection toward the metal center in complexes with L1, thus leading to the facile formation of the corresponding catalytically inactive L₂Pd species. Formation of this bisphosphine complex can be hampered by the bulkier fluorenyl substituent in L2, but this still does not allow for any catalytic activity. Experimental determination of the donor strength revealed a remarkably decreased donor capacity of L1 and L2 compared with previously reported YPhos ligands featuring acyclic ylide substituents. Computational studies showed that the decreased donor strength results from the cyclic backbone. The tight P–C–S bond angle in the cyclic ylide increases the s-character of the bonding orbital of the ylidic carbon atom in the C–P bond, thus resulting in a strong polarization of the C–P bond toward the carbon and hence in a decreased electron density at the phosphorus atom. Overall, these results demonstrate a marked angle-dependency of the donor strength of ylide-substituted phosphines, which may be used as further tool in the design of new YPhos ligands in the future.

Experimental Section

General Procedures

All experiments unless otherwise stated were performed under a dry, oxygen-free argon atmosphere using standard Schlenk techniques. Solvents involved were dried with an MBraun solvent-purification system (SPS 800) or with standard methods and stored under an argon atmosphere over 3 or 4 Å molecular sieves. ¹H, ¹³C{¹H}, and ³¹P{¹H} NMR spectra were recorded on an Avance III 400 spectrometer from Bruker at 25 °C. Chemical shifts are given in ppm regarding the δ scale and are referenced to the residual protons of the deuterated solvent. All spin–spin coupling constants are given in Hz. Multiplicities are indicated with the following abbreviations: s = singlet, d = doublet, t = triplet, q = quartet, dt = doublet of triplets, and so on. The assignment of signals was supported by HSQC, HMBC, APT, and COSY experiments. Elemental analyses were performed by Manuela Winter and Dr. Harish Parala on an Elementar vario MICRO-cube from Elementar. All chemicals were obtained from commercial sources and used without further purification. Dicyclohexylchlorophosphine and (tht)AuCl were made according to literature methods.^25,26

Synthesis of [1]I

In a flask fitted with an addition funnel, 60 mL (99.8 mmol) of a 1.66 M nBuLi solution in hexane was added to 50 mL of cyclohexane. An additional 30 mL of cyclohexane and 15 mL (11.6 g, 99.8 mmol) of TMEDA were added afterward slowly, followed by a further 20 mL of cyclohexane. The solution was cooled to 0 °C and 4.66 mL (5 g, 45.4 mmol) of thiophenol dissolved in 10 mL of cyclohexane was added slowly. The solution was stirred for 30 min at 0 °C, warmed to room temperature, and stirred overnight. A colorless solid precipitated that was subsequently filtered off and washed with three 30 mL portions of hexane. The solid was dried in vacuo (10.0 g, 44.8 mmol). The dilithiated product was dissolved in 100 mL of THF and cooled to −80 °C. Next, 7.15 mL (6.84 g, 44.8 mmol) of diisopropylchlorophosphine was slowly added via an addition funnel at −80 °C over a period of 1 h. The solution was stirred and allowed to warm to room temperature overnight. The solvent was removed, and the remaining solid was dried in vacuo (15 g, 38.4 mmol). The obtained solid was dissolved in 200 mL of diethyl ether and 3.09 mL (10.2 g, 38.4 mmol) of diiodomethane was added slowly before being left to stir overnight. The solvent was removed, and the residue redissolved in 50 mL of DCM. The undissolved salts were filtered off. The solvent was removed, and the remaining solid was suspended in 50 mL of THF, filtered, and washed (2 × 20 mL) with THF. The colorless solid was dried in vacuo. Phosphonium salt [1]I could be obtained as a colorless solid (9.4 g, 25.7 mmol, 56.5% yield). ¹H NMR (400 MHz, DCM-d₂): δ [ppm] = 1.33 (d, 3H, ³J = 7.15 Hz, iPr CH₃), 1.38 (d, 3H, ³J = 7.15 Hz, iPr CH₃), 1.41 (d, 3H, ³J = 7.12 Hz, iPr CH₃), 1.46 (d, 3H, ³J = 7.12 Hz, iPr CH₃), 3.61 (ds, 2H, ³J = 7.15, 7.15, 7.12, 7.12 Hz, ²J_HP = 10.57 Hz, iPr CH(CH₃)₂), 4.30 (d, 2H, ²J = 9.45 Hz, PCHS), 7.43 (m, 1H, CH Pos. 4), 7.49 (m, 1H, CH Pos. 2), 7.68 (m, 1H, CH Pos. 1), 7.89 (m, 1H, CH Pos. 3). ¹³C{¹H} NMR (100.6 MHz, DCM-d₂): δ [ppm] = 16.2 (d, ²J_CP = 15.2 Hz, iPr CH₃), 16.3 (d, ²J_CP = 15.2 Hz, iPr CH₃), 17.5 (d, ¹J_CP = 42.7 Hz, PCH₂S), 23.9 (d, ¹J_CP = 40.4, iPr CH(CH₃)₂), 113.2 (d, ¹J_CP = 85.6, C Pos. 6), 124.8 (d, ³J_CP = 9.2 Hz, CH Pos. 2/4), 126.9 (d, ³J_CP = 10.3 Hz, CH Pos. 2/4), 132.4 (d, ²J_CP = 10.6 Hz, CH Pos. 1), 136.1 (d, ⁴J_CP = 2.6 Hz, CH Pos. 3), 150.3 (d, ²J_CP = 16.9 Hz, CH Pos. 5). ³¹P{¹H} NMR (162.1 MHz, DCM-d₂): δ [ppm] = 76.5 (s, PCS). IR (ATR) [cm^–1]: 3051 (w), 2979 (w), 2859 (m), 1574 (s), 1452 (s), 1427 (s), 1271 (m), 1140 (s), 1116 (m), 1036 (w), 884 (m), 783 (s), 145 (m), 676 (s), 632 (m), 482 (s), 461 (s), 450 (s), 430 (m). CHNS for C₁₃H₂₀IPS: Calcd: C: 42.63, H: 5.50, S: 8.75. Measured: C: 42.57, H: 5.38, S: 8.43. Melting point: 160.1 °C.

Synthesis of 2

In a Schlenk tube, 2.50 g (6.83 mmol) of phosphonium salt [1]I and 930.0 mg (8.29 mmol) of KOt-Bu were suspended in 15 mL of THF. The suspension immediately turned orange. The suspension was stirred for 2 h at room temperature before removal of the solvent under reduced pressure, with the residue redissolved in 10 mL of pentane and stirred for 15 min. The solution was filtered, and the white residue was washed (2 × 5 mL) with pentane. The solvent of the combined solutions was removed and 1.55 g (6.47 mmol, 95%) of ylide 2 was obtained as a dark red liquid. ¹H NMR (400 MHz, C₆D₆): δ [ppm] = 0.76 (d, 3H, ³J = 7.06 Hz, iPr CH₃), 0.79 (d, 3H, ³J = 6.76 Hz, iPr CH₃), 0.83 (d, 3H, ³J = 6.76 Hz, iPr CH₃), 0.86 (d, 3H, ³J = 7.11 Hz, iPr CH₃), 1.42 (d, 1H, ²J_HP = 23.0 Hz, PCHS), 1.81 (dq, 1H, ³J = 7.12, 7.12, 7.12 Hz, ²J_HP = 14.2 Hz, iPr CH(CH₃)₂), 6.83 (m, 1H, CH Pos. 4), 6.94 (m, 1H, CH Pos. 2), 7.14(m, 1H, CH Pos. 1), 7.40 (m, 1H, CH Pos. 3). ¹³C{¹H} NMR (100.6 MHz, C₆D₆): δ [ppm] = −3.4 (d, ¹J_CP = 103.0 Hz, PCHS), 16.5 (dd, ²J_CP = 154.2 Hz, ²J = 2.42 Hz, iPr CH₃), 29.1 (d, ¹J_CP = 48.6, iPr CH(CH₃)₂), 115.2 (d, ¹J_CP = 94.8, C Pos. 6), 121.4 (d, ⁴J_CP = 9.6 Hz, CH Pos. 4), 124.2 (d, ⁴J_CP = 7.48 Hz, CH Pos. 2), 127.8 (d, ⁴J_CP = 2.35 Hz, CH Pos. 3), 129.2 (d, ²J_CP = 9.84 Hz, CH Pos. 1), 153.6 (d, ²J_CP = 21.4 Hz, CH Pos. 5). ³¹P{¹H} NMR (162.1 MHz, C₆D₆): δ [ppm] = 50.2 (s, PCS).

Synthesis of L1

First, 1.55 g (6.47 mmol) of ylide 2 was dissolved in 15 mL of THF and cooled to 0 °C. To this, 0.79 mL (865.8 mg, 3.56 mmol) of PCy₂Cl was slowly added over 10 min. Upon addition of the chlorophosphine, the phosphonium salt precipitated as a colorless solid. It was stirred for 2 h at room temperature and filtered, and the solid was washed with (2 × 5 mL) THF. The solvent was removed under reduced pressure, and the red-orange residue was suspended in 5 mL of acetonitrile and stirred for 15 h. Overnight, an orange solid precipitated that was filtered off, washed (2 × 5 mL) with acetonitrile, and dried in vacuo. Ylide-substituted phosphine L1 was obtained in 79% yield (2.54 mmol, 1.1 g). ¹H NMR (400 MHz, THF-d₈): δ [ppm] = 1.12 (d, 3H, ³J = 7.12 Hz, iPr CH₃), 1.16 (d, 3H, ³J = 7.12 Hz, iPr CH₃), 1.21 (d, 3H, ³J = 7.12 Hz, iPr CH₃), 1.25 (d, 3H, ³J = 7.12 Hz, iPr CH₃), 1.20–1.35 (m, 10H, PCy₂ H₂₊₃₊₄), 1.70–1.85 (m, 2H, PCy₂ H₂₊₃₊₄), 1.92 (m, 2H, PCy₂ H₁), 2.49 (dp, 2H, ³J = 7.12, 7.12, 7.12, 7.12 Hz, ¹J_HP = 10.1 Hz, iPr CH(CH₃)₂), 7.03 (m, 1H, CH Pos. 4), 7.20 (m, 1H, CH Pos. 2), 7.29 (m, 1H, CH Pos. 1), 7.48 (m, 1H, CH Pos. 3). ¹³C{¹H} NMR (100.6 MHz, THF-d₈): δ [ppm] = 5.7 (dd, ¹J_CP = 34.7, 105.32 Hz, PCS), 16.75 (dd, ²J_CP = 1.99 Hz, iPr CH₃), 17.8 (d, ²J_CP = 2.22 Hz, iPr CH₃), 28.0 (s, CH₂ PCy₂ C_3/4), 28.7 (m, CH₂ PCy₂ C_3/4), 30.3 (d, ¹J_CP = 49.36 Hz, iPr CH(CH₃)₂), 31.5 (m, CH₂ PCy₂ C_3/4), 32.8 (d, ²J_CP = 18.7 Hz, CH₂ PCy₂ C₂), 36.6 (dd, ²J = 4.8 Hz, ²J_CP = 9.43, CH₂ PCy₂ C₁), 118.4 (d, ¹J_CP = 91.0 Hz, C Pos. 6), 122.8 (d, ⁴J_CP = 9.2, CH Pos. 3), 123.8 (d, ³J_CP = 6.9, CH Pos. 2/4), 129.5 (d, ³J_CP = 1.3 Hz, CH Pos. 2/4), 129.9 (d, ²J_CP = 2.6 Hz, CH Pos. 1), 155.7 (dd, ²J_CP = 18.9 Hz, C Pos. 5). ³¹P{¹H} NMR (162.1 MHz, THF-d₈): δ [ppm] = −14.9 (d, ²J_PP = 108.6 Hz, PCy₂), 50.67 (d, ²J_PP = 108.6 Hz, PCS). IR (ATR) [cm^–1]: 2917 (s), 2851 (m), 1439 (m), 1108 (s), 1048 (m), 874 (m), 849 (w), 711 (m), 690 (s), 653 (m), 538 (w), 471 (m), 450 (w). CHNS for C₂₅H₄₀P₂S: Calcd: C: 69.09, H: 9.28, S: 7.38. Measured: C: 68.89, H: 9.14, S: 7.38. Melting point: 93.7 °C.

Synthesis of L1·AuCl

A Schlenk tube was charged with L1 (200 mg, 459 μmol, 1 equiv), (tht)AuCl (147 mg, 459 μmol, 1 equiv), and 4 mL of THF, and the mixture was stirred overnight. The solvent was removed in vacuo, and the residue dissolved in 2 mL of DCM. The solution was layered with 20 mL of pentane and left for 1 day. A yellow solid precipitated. The solution was then cooled to −30 °C for 2 more days. The solid was filtered off and washed (2 × 5 mL) with pentane. 224 mg (336 μmol, 73%) of the yellow crystalline solid was obtained. Slow diffusion of pentane into a solution of L1·AuCl in C₆D₆ allowed crystals suitable for single-crystal XRD to be obtained. ¹H NMR (400 MHz, CD₂Cl₂): δ [ppm] = 1.14–1.34 (4 ×rd, 12 H, ³J_PH = 7.0 Hz, iPr CH₃), 1.14–1.34 (m, 6H, PCy₂ CH₂), 1.34–1.57 (m, 6H, PCy₂ CH₂), 1.63–1.71 (m, 2H, PCy₂ CH₂), 1.77–1.88 (m, 4H, PCy₂ CH₂), 1.95–2.06 (m, 4H, PCy₂ CH₂), 2.15 (tdd, 2H, ²J_HP = 12.1 Hz, ³J = 7.0, 3.3 Hz), 3.13 (dp, 2H, ²J_HP = 8.9 Hz, ³J = 7.1, 7.1, 7.0, 7.0, iPr CH), 7.12–7.18 (m, 1H, Pos. 4), 7.30–7.39 (m, 2H, Pos. 2 + Pos. 3), 7.46–7.51 (m, 1H, Pos. 1). ¹³C{¹H} NMR (101 MHz, CD₂Cl₂): δ [ppm] = 0.8 (dd, ¹J_CP = 103.9, 66.1 Hz, PCS), 17.0 (dd, ²J_CP = 42.6 Hz, ²J = 2.0 Hz, iPr CH₃), 26.6 (d, ²J = 1.7 Hz, PCy₂CH₂), 27.1–27.5 (m, PCy₂CH₂), 29.2 (d,¹J_CP = 48.9 Hz, iPr CH), 30.3 (d, ²J = 1.9 Hz, PCy₂CH₂), 32.0 (d, ²J = 2.8 Hz, PCy₂CH₂), 37.7 (dd, ¹J_CP = 41.5 Hz, ²J = 1.8 Hz, PCy₂CH), 118.2 (dd, ¹J_CP = 92.7 Hz, ²J = 6.7 Hz, C Pos. 6), 122.9 (d, ³J_CP = 7.1 Hz, CH Pos. 2), 123.7 (d, ³J_CP = 9.6 Hz, CH Pos. 4), 129.6 (d, ²J_CP = 10.1 Hz, CH Pos. 1), 130.8 (d, ³J = 2.2 Hz, CH Pos. 3), 151.5 (dd, ²J_CP = 15.8 Hz, ²J = 5.1 Hz, C Pos. 5). ³¹P{¹H} NMR (162.1 MHz, CD₂Cl₂): δ [ppm] = 27.1 (d, ²J_PP = 51.8 Hz, PCy₂), 57.2 (d, ²J_PP = 50.5 Hz, PCS). IR (ATR) [cm^–1]: 2919 (m), 1444 (m), 1110 (m), 1066 (s), 897 (m), 741 (s), 711 (s), 687 (m), 667 (m), 636 (w), 520 (m), 445 (m), 408 (m). CHNS for C₂₅H₄₀AuClP₂S: Calcd: C: 45.02, H: 6.04, S: 4.81. Measured: C: 44.79, H: 6.10, S: 4.43. Melting point: 207.3 °C (decomposition).

Synthesis of L1₂Pd

A Schlenk tube charged with L1 (200 mg, 460 μmol, 1 equiv), Pd₂dba₃ (155 mg, 230 μmol, 1 equiv), and 4 mL of THF was stirred at room temperature for 1 h. The solution was subsequently filtered, over-laid with 10 mL of pentane, and left for 3 days. A yellow solid precipitated from the solution. The solid was filtered off and washed (2 × 5 mL) with THF; 181 mg (185 μmol, 81%) of the yellow solid was obtained. Slow diffusion of pentane into a solution of L1₂Pd in C₆D₆ allowed for crystals suitable for single-crystal XRD to be obtained. ¹H NMR (400 MHz, THF-d₈): δ [ppm] = 1.12 (dd, 6H, ³J = 7.16, 16.71 Hz, iPr CH₃), 1.41 (dd, 6H, ³J = 7.16, 16.71 Hz, iPr CH₃), 1.3–1.5 (m, 6H, PCy2 H_1+2+3+4), 1.76 (m, 2H, PCy2 H_1+2+3+4), 1.8–2.0 (m, 8 H, PCy₂ H₂₊₃₊₄), 2.1–2.4 (m, 6H, PCy₂ H₂₊₃₊₄), 3.72 (dp, 2H, ³J = 7.05, 7.05, 7.05, 7.05 Hz, ²J_HP = 10.1 Hz, iPr CH(CH₃)₂), 6.72 (m, 1H, CH Pos. 4), 6.88 (m, 1H, CH Pos. 2), 6.92 (m, 1H, CH Pos. 1), 7.02 (m, 1H, CH Pos. 3). ¹³C{¹H} NMR: Not possible due to poor solubility of the pure compound in deuterated solvent. ³¹P{¹H} NMR (162.1 MHz, THF-d₈): δ [ppm] = 13.11 (dd, ²J_PP = 44.9, 34.2 Hz PCy₂), 50.7 (dd, ²J_PP = 45.0, 34.2 Hz, PCS). IR (ATR) [cm^–1]: 2914 (m), 2843 (m), 1441 (m), 1257 (w), 1109 (m), 1058 (s), 1000 (w), 887 (m), 868 (s), 851 (w), 741 (s), 712 (m), 366 (m), 543 (w), 514 (m), 471 (m), 441 (m), 406 (m). CHNS for C₅₀H₈₀P₄PdS₂: Calcd: C: 61.56, H: 8.27, S: 6.57. Measured: C: 61.88, H: 7.87, S: 6.49. Melting point: 195.8 °C (decomposition).

Synthesis of Pd_al1

250 mg (0.58 mmol) of L1 and 108 mg of [Pd(allyl)Cl]₂ (56.4 m% Pd, 0.29 mmol) were suspended in 10 mL of pentane and stirred for 30 min. A yellow suspension formed, which was filtered, and the remaining solid was washed (2 × 5 mL) with pentane. Complex Pd_al1was obtained as a yellow solid in 74% yield (255 mg, 0.56 mmol). By slow diffusion of pentane into a solution of Pd_al1in C₆D₆ crystals of a decomposition product suitable for single-crystal XRD could be obtained. ¹H NMR (400 MHz, C₆D₆): δ [ppm] = 0.88–1.10 (m, 6H, iPr CH₃), 1.10–1.29 (m, 8H, iPr CH₃, PCy₂ CH₂ H₂₊₃₊₄), 1.30–1.92 (m, 13H, PCy₂ CH₂ H₂₊₃₊₄), 2.21–2.31 (d, 2H, J³ = 12.5 Hz, PCy₂ CH₂ H₂₊₃₊₄), 2.32 (d, 2H, J³ = 11.2 Hz, Allyl CH₂ H₅), 2.38–2.53 (dt, 2H, J³ = 9.0, 9.0, 22.2 Hz, PCy₂ CH₂ H₂₊₃₊₄), 2.53–2.71 (p, 2H, J³ = 15.1, 15.1, 17.1, 17.1 Hz, PCy₂ CH₂ H₁), 3.47 (dd, 1H, J³ = 8.7, 13.8 Hz, Allyl CH₂ H₂), 5.53 (m, 1H, Allyl CH₂ H₄), 3.6 (dp, 1H, J³ = 5.24, 5.24, 5.25, 5.25 Hz, J²_HP = 14.2 Hz, iPr CH), 3.74 (dp, 1H, J³ = 7.04, 7.04, 7.05, 7.05 Hz, J²_HP = 14.2 Hz, iPr CH), 4.52 (t, 1H, J³ = 7.19, 7.19 Hz, Allyl CH H₃), 4.89 (ddd, 1H, J³ = 7.96, 13.7, 17.9 Hz, Allyl CH H₁), 6.83–6.91 (m, 1H, CH Pos. 4), 6.83–6.91 (m, 1H, CH Pos. 2), 7.01–7.12 (m, 1H, CH Pos. 1), 7.13–7.23 (m, 1H, CH Pos. 3). ¹³C{¹H} NMR (101 MHz, C₆D₆) δ [ppm] = 6.01 (dd, ¹J_CP = 37.9, 106.5 Hz, PCS), 18.1 (d, J = 2.2 Hz, iPr CH₃), 26.9 (s, CH₂ PCy₂ C₂₊₃₊₄), 27.7 (d, ²J = 12.1 Hz, CH₂ PCy₂ C₂₊₃₊₄), 28.1 (d, ²J = 12.6 Hz, CH₂ PCy₂ C₂₊₃₊₄), 30.0 (s, CH₂ PCy₂ C₂₊₃₊₄), 30.1 (s, CH₂ PCy₂ C₂₊₃₊₄), 30.4(s, CH₂ PCy₂ C₂₊₃₊₄), 30.6 (d, ¹J_CP = 6.5 Hz, iPr CH(CH₃)₂), 40.4 (dd, ¹J_CP = 74.1 Hz, ²J = 26.7 Hz, CH PCy₂ C₁), 50.7 (d, ²J = 2.3 Hz, CH₂ allyl), 79.6 (d, ²J = 29.1 Hz, CH₂ allyl), 114.3 (d, ²J = 4.6 Hz, CH allyl), 118.0 (dd, ¹J_CP = 90.0 Hz,²J = 5.9 Hz, Pos. 6), 122.2 (d, ²J = 9.2 Hz, Pos. 4), 122.9 (d, ²J = 6.7 Hz, Pos. 3), 129.5 (d, ²J = 2.2 Hz, Pos. 2), 130.2 (d, ²J = 9.3 Hz, Pos. 1), 152.8 (dd, ²J = 15.5 Hz, ²J = 3.1 Hz, Pos. 5). ³¹P{¹H} NMR (162.1 MHz, C₆D₆): δ [ppm] = 20.3 (d, ²J_PP = 60.8 Hz, PCy₂), 55.5 (d, ²J_PP = 61.1 Hz, PCS). IR (ATR) [cm^–1]: 2924 (m), 2845 (w), 1442 (m), 1262 (w), 1106 (m), 1064 (s), 864 (m), 742 (s), 657 (m), 459 (m), 405 (w). CHNS for C₂₈H₄₅ClP₂PdS: Calcd: C: 54.37, H: 7.50, S: 5.18. Measured: C: 54.22, H: 7.79, S: 5.18. Melting point: 146.7 °C (decomposition).

Synthesis of Pd_cin1

First, 200 mg (0.46 mmol) of L1 and 118 mg of [Pd(Cinnamyl)Cl]₂ (40.2 m% Pd, 0.23 mmol) were suspended in 8 mL of pentane and stirred for 2 h. A red suspension formed. This was filtered, and the solid was washed (2 × 5 mL) with pentane. Complex Pd_cin1 was obtained as a red-orange solid in 55.2% yield (176 mg, 0.25 mmol). By slow diffusion of pentane into a solution in THF-d₈, crystals suitable for X-ray diffraction could be obtained. ¹H NMR (400 MHz, THF-d₈): δ [ppm] = 1.07–1.85 (m, 27H, iPr CH₃ + PCy₂ CH₂ H₂₊₃₊₄), 1.97–2.26 (m, 5H, PCy₂ CH₂ H₂₊₃₊₄), 2.30–2.52 (m, 2H, PCy₂ CH H₁), 2.69–2.77 (d, 1H, ³J = 11.9 Hz, cin CH₂), 3.58 (m, 1H, iPr CH), 3.71 (m, 1H, iPr CH), 3.80 (d, 1H, ³J = 6.7 Hz, iPr CH), 5.06 (dd, 1H, ³J = 13.1, 9.3 Hz, cin CH), 5.83 (m, 1H, cin CHPh), 7.02–7.10 (m, 1H, Pos. 4), 7.19–7.29 (m, 4H, cin CH Ph + Pos. 2), 7.31–7.37 (m, 1H, Pos. 1), 7.42–7.46 (m, 2H, cin CH Ph), 7.49–7.55 (m, 1H, Pos.3). ¹³C NMR (101 MHz, THF-d₈): δ [ppm] = 7.2 (dd, ¹J_CP = 107.2, 39.1 Hz, PCS), 18.6 (d, ²J_CP = 2.2 Hz, iPr CH₃), 18.8 (d, ²J_CP = 2.0 Hz, iPr CH₃), 23.4 (s, PCy₂CH₂), 27.6 (s, PCy₂CH₂), 28.0–31.3 (m, PCy₂CH₂), 32.0 (d, ¹J_CP = 46.4 Hz, iPr CH), 40.0 (d, ¹J_CP = 27.2 Hz, PCy₂CH), 42.5 (d, ¹J_CP = 25.6 Hz, PCy₂CH), 48.1 (d, ²J = 2.6 Hz, cin CH₂), 98.8 (d, ²J = 26.8 Hz, cin CH), 109.6 (d, ²J = 5.2 Hz, cin CHPh), 118.7 (dd, ¹J_CP = 89.7 Hz, ²J = 5.6 Hz, Pos. 6), 123.0 (d, ³J_CP = 9.2 Hz, Pos. 4), 123.4 (d, ³J_CP = 6.7 Hz, Pos.2), 127.9 (d, ²J_CP = 2.4 Hz, Pos. 1), 128.7 (d, ²J = 3.3 Hz, cin CH Ph), 129.1 (d, ²J = 2.0 Hz, cin CH Ph), 130.2 (d, ²J = 2.2 Hz, cin CH Ph), 131.5 (d, ²J = 9.5 Hz, Pos. 3), 138.7 (d, ²J = 6.1 Hz, cin C Ph), 153.5 (dd, ²J_CP = 15.3 Hz, ²J = 3.1 Hz, Pos. 5). ³¹P NMR (162 MHz, THF-d₈) δ = 24.8 (d, ²J_PP = 61.3 Hz, PCy₂), 55.53 (d, ²J_PP = 60.9 Hz, PCS). IR (ATR) [cm^–1]: 2918 (m), 1442 (m), 1107 (m), 1059 (s), 872 (m), 741 (s), 688 (s), 657 (w), 513 (w), 458 (m), 411 (m). CHNS for C₃₄H₄₉ClP₂PdS: Calcd: C: 58.87, H: 7.12, S: 4.62. Measured: C: 59.18, H: 7.22, S: 4.39. Melting point: 115.2 °C (decomposition).

Synthesis of Pd_in1

First, 250 mg (0.58 mmol) of L1 and 178 mg of [Pd(^tBu-Indenyl)Cl]₂ (33.2 m% Pd, 0.29 mmol) were suspended in 10 mL of pentane and stirred for 2 h. The dark gray suspension formed was filtered, and the solid was washed (2 × 5 mL) with pentane. Complex Pd_in1 was obtained as a gray solid in 84.2% yield (352 mg, 0.47 mmol). By slow diffusion of pentane into a solution of Pd_in1 in C₆D₆, crystals suitable for single-crystal XRD could be obtained. ¹H NMR (400 MHz, C₆D₆): δ [ppm] = 0.84–0.98 (m, 4H, iPr CH₃), 1.02–1.13 (m, 6H, iPr CH₃), 1.18 (m, 2H, iPr CH₃), 1.22–1.88 (m, 18H, PCy₂ CH₂ H₂₊₃₊₄), 1.68 (s, 9H, ^tBu CH₃), 2.21 (m, 3H, PCy₂ CH₂ H₂₊₃₊₄), 2.42 (m, 1H, PCy₂ CH₂ H₂₊₃₊₄), 2.60 (dqd, 2H, ³J = 3.17, 9.35, 9.47, 12.1 Hz, PCy₂ CH H₁), 3.55 (dp, 1H, ²J_PH = 14.2 Hz, ³J = 7.2, 7.2, 7.24, 7.24 Hz, iPr CH), 4.18 (dp, 1H, ²J_PH = 14.2 Hz, ³J = 7.2, 7.2, 7.24, 7.24 Hz, iPr CH), 5.43 (dd, 1H, ²J = 1.37, 2.89 Hz, Ind), 6.32 (dd, 1H, ²J = 0.92, 2.93 Hz, CH, Ind), 6.71 (m, 1H, CH Pos. 4), 6.79 (m, 1H, CH, Ind), 6.85 (m, 1H, CH Pos. 2), 6.93 (m, 1H, CH, Ind), 7.03 (m, 2H, CH Pos 1 + Ind), 7.19 (m, 1H, CH Pos 3), 7.47 (m, 1H, CH, Ind). ¹³C{¹H} NMR (101 MHz, C₆D₆) δ = 7.24 (dd, ²J = 106.6, 42.8 Hz, PCS), 17.9 (d, ²J = 1.9 Hz, iPr CH₃), 18.1 (d, ²J = 2.2 Hz, iPr CH₃), 18.2 (d, ²J = 2.1 Hz, iPr CH₃), 18.6 (d, ²J = 2.2 Hz, iPr CH₃), 26.5 (d, ²J = 20.5 Hz, CH₂ PCy₂ C₂₊₃₊₄), 27.2 (dd, ²J = 12.0, 4.0 Hz, CH₂ PCy₂ C₂₊₃₊₄), 27.7 (dd, ²J = 15.0, 12.9 Hz, CH₂ PCy₂ C₂₊₃₊₄), 29.1 (s, CH₂ PCy₂ C₂₊₃₊₄), 29.5 (s, ^tBu C(CH₃)₃), 29.5 (s, ^tBu C(CH₃)₃), 29.8 (d, ²J = 41.8 Hz, CH Ind), 31.0 (d, ²J = 47.0 Hz, CH Ind), 34.6 (d, ²J = 4.5 Hz, ^tBu C(CH₃)₃), 41.2 (dd, ¹J_CP = 52.2 Hz, ²J = 26.6 Hz, CH PCy₂ C₁), 64.8 (d, ²J = 3.5 Hz, CH Ind), 107.9 (d, ²J = 5.8 Hz, CH Ind), 117.8 (dd, ¹J_CP = 88.9 Hz,²J = 5.3 Hz, C Pos. 6), 118.0(s, CH Ind), 120.5 (s, CH Ind), 121.7 (d, ²J = 9.1 Hz, CH Pos. 4), 122.4 (d, ²J = 6.6 Hz, CH Pos. 3), 124.71 (d, ²J = 63.4 Hz, CH Ind), 125.9 (d, ²J = 23.2 Hz, C Ind), 129.0 (d, ²J = 2.2 Hz, CH Pos. 2), 130.4 (d, ²J = 9.2 Hz, CH Pos. 1), 138.7 (d, ²J = 5.4 Hz, C Ind), 140.3 (s, C Ind), 152.1 (dd, ³J_CP = 14.8 Hz, ²J = 3.5 Hz, C Pos. 5). ³¹P NMR (162 MHz, C₆D₆) δ = 27.0 (d, ²J_PP = 55.8 Hz, PCy₂), 56.7 (d, ²J_PP = 56.0 Hz, PCS). IR (ATR) [cm^–1]: 2926 (m), 2843 (w), 1444 (m), 1362 (w), 1275 (w), 1196 (s), 1104 (m), 1060 (s), 1046 (m), 873 (m), 743 (s), 687 (m), 628 (w), 511 (m), 461 (s), 426 (s). CHNS for C₃₈H₅₅ClP₂PdS: Calcd: C: 61.04, H: 7.66, S: 4.74. Measured: C: 61.27, H: 7.89, S: 4.49. Melting point: 155.3 °C (decomposition).

Synthesis of 9-Methyl-fluorene (9-Me-Fluorene)

Following a slightly modified literature procedure,²⁷ 4.99 g (30 mmol) of fluorene was dissolved in 60 mL of THF and cooled to −60 °C. To this was slowly added (over a period of 15 min) 26 mL of a 1.55 M solution of nBuLi in hexane. The solution was warmed to room temperature and stirred for 1.5 h. The solution was cooled to −60 °C, and 2.8 mL (6.39 mg, 45 mmol) of MeI was added slowly over 5 min. The solution became clear and colorless and was then stirred for 15 min at −60 °C and 2 h at room temperature. Excess MeI was quenched with 100 mL of aqueous KOH solution. The aqueous phase was separated and extracted (2 × 5 mL) with Et₂O. The combined organic phases were dried with MgSO₄, and the solvent was removed. The crude product was filtered over a short silica column with hexane as eluent. The product was obtained as a yellow waxlike solid in 91% yield (5.0 g, 27.4 mmol). ¹H NMR (400 MHz, THF-d₈): δ [ppm] = 1.44 (d, 3H, ³J = 7.34 Hz, CH₃), 3.85 (q, 1H, ³J = 7.30 Hz, 9HFlu), 7.26 (m, 4H, Ar), 7.42 (d, 2H, ³J = 7.15 Hz, Ar), 7.67 (m, 2H, Ar).²⁷

Synthesis of CIPCyFlu

First, 5.00 g (27.4 mmol) of 9-Me-Fluorene was dissolved in 110 mL of THF and cooled to −60 °C. 17.7 mL of a 1.55 M solution of nBuLi in hexane was added slowly over 15 min. The solution turned red and was stirred for 30 min at 60 °C, before being cooled and stirred for 2 h at room temperature. A second solution of 4.2 mL (29 mmol, 5.07 g) of PCyCl₂ in 40 mL of THF was prepared and cooled to −60 °C. The lithiated 9-Me-Fluorene solution was cooled to −60 °C and slowly added to the chlorophosphine solution over a period of 30 min. This was then stirred for 30 min at −60 °C and allowed to come to room temperature overnight. The solvent was removed, and the residue was suspended in 20 mL of hexane and filtered over Celite. The solvent was removed in vacuo, and the crude product purified by distillation (1 × 10^–3 bar at 147 °C). Yield was not determined. ¹H NMR (400 MHz, DCM-d₂): δ [ppm] = 0.40 (m, 1H, PCy₂ CH₂ H₂₊₃₊₄), 0.5–1.0 (m, 6H, PCy₂ CH₂ + CH H_1+2+3+4), 1.2–1.4 (m, 3H, PCy₂ CH₂ H₂₊₃₊₄), 1.48 (s, 1H, PCy₂ CH₂ H₂₊₃₊₄), 1.86 (d, ³J_HP = 13.3 Hz, CH₃). ¹³C{¹H} NMR (100.6 MHz, DCM-d₂): δ [ppm] = 23.5 (d, ³J_CP = 21.6 Hz, CH₃ Flu), 26.1 (d, ²J = 1.43 Hz, CH₂ Cy C₃₊₄), 26.7 (d, ²J_CP = 8.98 Hz, CH₂ Cy C₂), 26.8 (d, ²J = 1.37 Hz, CH₂ Cy C₃₊₄), 28.8 (d, 6.26 Hz, CH₂ Cy H₃₊₄), 30.0 (d, ²J_CP = 26.5 Hz, CH₂ Cy C₂), 38.5 (d, ¹J_CP = 38.5 Hz, CH Cy C₁), 53.0 (d, ¹J_CP = 41.1 Hz, C Flu₉). IR (ATR) [cm^–1]: 2917 (m), 2849 (s), 1474 (s), 1437 (m), 1026 (s), 763 (s), 742 (s), 729 (s), 603 (w), 565 (w), 473 (s), 453 (s), 423 (w). CHNS for C₂₀H₂₂ClP: Calcd: C: 73.06, H: 6.74. Measured: C: 73.05, H: 6.81. Melting point: 69.2 °C.

Synthesis of L2

First, 1g (2.73 mmol) of phosphonium salt 1 and 306 mg (2.73 mmol) of KOt-Bu were suspended in 8 mL of THF and stirred for 4 h, forming a red solution. The solvent was removed in vacuo, and the residue was redissolved in 8 mL of pentane. The solution was subsequently filtered, and the solvent was removed. The residue was dissolved in 8 mL of THF and cooled to 0 °C. A solution of 449 mg of ClPCyFlu (0.5 equiv, 1.37 mmol) in 1 mL of THF was added slowly. A beige solid immediately precipitated from the solution. The solution was stirred for 2 h at room temperature and filtered, and the solvent was removed in vacuo. The residue was suspended in 15 mL of hot acetonitrile and stirred until dissolution. The solution was filtered while hot and allowed to slowly cool room temperature, precipitating a yellow solid. The solution was filtered, and the remaining solid was washed (3 × 1.5 mL) with acetonitrile. Phosphine L2 was obtained as this yellow solid in 65% yield (940 g, 1.8 mmol). ¹H NMR (400 MHz, THF-d₈) δ = 0.27–0.55 (m, 3H, Cy CH₂ H₂₊₃₊₄), 0.67 (m, 3H, Cy CH₂ H₂₊₃₊₄), 0.84–1.08 (m, 9H, iPr CH₃), 1.14 (m, 1H, Cy CH H₂₊₃₊₄), 1.23 (m, 2H, Cy CH₂ H₂₊₃₊₄ + CH H₁), 1.32 (dd, ³J = 16.1, 7.2 Hz, 3H, iPr CH₃), 1.37 (s, 1H, Cy CH₂ H₂₊₃₊₄), 1.48 (d, ³J = 11.8 Hz, 3H, Flu CH₃HH), 1.51 (s, 1H, Cy CH₂ H₂₊₃₊₄), 1.97 (dp, ³J = 10.8, 7.1 Hz, 1H, iPr CH), 2.61 (dp, ³J = 10.6, 7.1 Hz, 1H, iPr CH), 6.85 (m, 1H, CH Pos. 4), 6.98–7.08 (m, 5H, CH Pos. 2, CH Flu), 7.10 (m, 1H, CH Pos. 1h), 7.27 (m, 1H, CH Pos. 3), 7.33–7.39 (m, 1H, CH Flu), 7.43–7.50 (m, 1H, CH Flu), 7.51 (d, ³J = 7.4 Hz, 1H, CH Flu), 7.75 (d, ³J = 7.4 Hz, 1H, CH Flu). ¹³C{¹H} NMR (100.6 MHz, DCM-d₂): δ [ppm] = 9.4 (dd, ¹J_CP = 42.8, 103.6 Hz, PCS), 16.7 (t, ²J_CP = 2.51 Hz, iPr CH₃), 17.6 (d, ²J_CP = 2.6 Hz, iPr CH₃), 17.8 (dd, ²J_CP = 2.32, 12.7 Hz, iPr CH₃), 18.1 (t, ²J_CP = 2.56 Hz, iPr CH₃), 25.9 (m, CH₂ Cy C₂₊₃₊₄), 27.6 (s, CH₃ Flu), 28.6 (d, ²J = 7.78 Hz, CH₂ Cy C₂₊₃₊₄), 27.4 (d, ²J_CP = 17.8 Hz, CH₂ Cy C₂₊₃₊₄), 32.3 (d, ²J = 6.95 Hz, CH Cy C₂₊₃₊₄), 32.7 (d, ²J_CP = 22.8 Hz, CH₂ Cy C₂₊₃₊₄), 27.9 (d, ²J_CP = 47.9 Hz, CH iPr), 33.9 (d, ²J_CP = 46.4 Hz, CH iPr), 37.3 (dd, ²J_CP = 6.55, 10.4 Hz, CH₂ Cy C₁), 55.6 (dd, ²J_CP = 3.27 Hz, ²J = 30.4 Hz, CCH₃ Flu₉), 118.7 (dd, ¹J_CP = 90.5 Hz, ²J = 7.24 Hz, C Pos. 6), 120.5 (d, ²J = 18.8 Hz, CH Flu), 123.3 (d, ²J_CP = 9.01 Hz, CH Pos. 4), 124.0 (d, ²J = 6.97 Hz, CH Flu), 125.5 (d, ²J_CP = 10.5 Hz, CH Pos. 1), 126.9 (dd, ²J = 8.45, 1.77 Hz, CH Flu), 127.4 (d, ²J = 4.11 Hz, CH Flu), 129.6 (d, ²J = 9.32 Hz, CH Pos. 3), 130.2 (d, ²J = 2.18 Hz, CH Pos. 2), 141.1 (d, ²J_CP = 3.44 Hz, C Flu₁₀₊₁₁), 141.3 (s, C Flu₁₂₊₁₃), 155.8 (dd, ²J_CP = 18.3 Hz, ²J = 1.10 Hz, C Pos. 5). ³¹P{¹H} NMR (162.1 MHz, THF-d₈): δ [ppm] = 12.2 (d, ²J_PP = 119.5 Hz, PCyFlu), 52.0 (d, ²J_PP = 119.5 Hz, PCS). IR (ATR) [cm^–1]: 2912 (w), 1439 (m), 1255 (w), 1103 (m), 1056 (m), 1023 (m), 868 (w), 808 (w), 757 (s), 739 (s), 729 (s), 666 (w), 636 (m), 571 (w), 447s, 426 (m). CHNS for C₃₃H₄₀P₂S: Calcd: C: 74.69, H: 7.60, S: 6.04. Measured: C: 74.23, H: 7.51, S: 5.69. Melting point: 76.2 °C (degradation).

Synthesis of Pd_al2

First, 300 mg (0.565 mmol) of L2 and 106 mg of [Pd(allyl)Cl]₂ (56.4 m% Pd, 0.283 mmol) were suspended in 8 mL of pentane and stirred for 2 h. After this time, the yellow suspension formed was filtered and the solid washed (2 × 5 mL) with pentane. Complex Pd_al2 was obtained as a yellow solid in 73% yield (295 mg, 0.414 mmol). By slow diffusion of pentane into a solution of Pd_al2 in THF-d₈, crystals suitable for single-crystal XRD could be obtained. (The obtained molecular structures and NMR data suggest two different conformers in solid state and solution.) ³¹P{¹H} NMR (162.1 MHz, THF-d₈): δ [ppm] = 34.5 (d, ²J_PP = 63.4 Hz, PCyFlu), 35.4 (d, ²J_PP = 63.3 Hz, PCyFlu), 59.0 (d, ²J_PP = 62.8 Hz, PCS), 59.1 (d, ²J_PP = 63.4 Hz, PCS). ¹H NMR and ¹³C NMR could not be evaluated due to the presence of two different conformers in solution. IR (ATR) [cm^–1]: 2916 (w), 1441 (m), 1105 (m), 1048 (m), 1013 (m), 866 (m), 739 (s), 727 (s), 687 (s), 631 (m), 571 (w), 448 (s), 425 (m). Melting point: 133.7 °C (degradation). CHNS for C₃₆H₄₅ClP₂PdS: Calcd: C: 60.59, H: 6.36, S: 4.49. Measured: C: 59.74, H: 6.20, S: 4.19

Synthesis of Pd_cin2

First, 300 mg (0.565 mmol) of L2 and 149 mg of [Pd(cinnamyl)Cl]₂ (40.2 m% Pd, 0.283 mmol) were suspended in 8 mL of pentane and stirred for 2 h. The orange suspension formed was filtered, and the solid was washed (2 × 5 mL) with pentane. Complex Pd_cin2was obtained as an orange solid in 76% yield (340 mg, 0.431 mmol). By slow diffusion of pentane into a solution of Pd_cin2 in THF-d₈, crystals suitable for X-ray diffraction could be obtained. (The obtained molecular structures and NMR data suggest two different conformers in solid state and solution.) ³¹P{¹H} NMR (162.1 MHz, THF-d₈): δ [ppm] = 37.3 (d, ²J_PP = 62.7 Hz, PCyFlu), 39.2 (d, ²J_PP = 62.57 Hz, PCyFlu), 59.10 (d, ²J_PP = 63.13 Hz, PCS), 59.24 (d, ²J_PP = 62.49 Hz, PCS). ¹H NMR and ¹³C NMR could not be evaluated due to the presence of two different conformers in solution. IR (ATR) [cm^–1]: 2922 (w), 1443 (m), 1110 (w), 1037 (m), 869 (m), 763 (w), 728 (s), 689 (s), 653 (m), 623 (w), 511 (w), 452 (m). Melting point: 159.2 °C (degradation).

Synthesis of Pd_ind2

First, 120 mg (0.23 mmol) of L2 and 72.4 mg of [Pd(^tBu-Indenyl)Cl]₂ (33.2 m% Pd, 0.12 mmol) were suspended in 4 mL of pentane and stirred for 2 h. A dark brown suspension formed, which was subsequently filtered, and the solid was washed (2 × 5 mL) with pentane. Complex Pd_ind2 was obtained as brown solid in 66% yield (127 mg). By slow diffusion of pentane into a solution in THF-d₈, crystals suitable for single-crystal XRD could be obtained. (The obtained molecular structures and NMR data suggest two different conformers in solid state and solution.) ³¹P{¹H} NMR (162.1 MHz, THF-d₈): δ [ppm] = 40.80 (d, ²J_PP = 56.7 Hz, PCyFlu), 44.5 (d, ²J_PP = 56.6 Hz, PCyFlu), 58.4 (d, ²J_PP = 56.8 Hz, PCS), 60.3 (d, ²J_PP = 56.7 Hz, PCS). ¹H NMR and ¹³C NMR could not be evaluated due to the presence of two different conformers in solution. IR (ATR) [cm^–1]: 2916 (m), 1444 (m), 1363 (w), 1106 (m), 1025 (s), 861 (m), 737 (s), 654 (m), 621 (w), 601 (w), 463 (m), 448 (m). Melting point: 148.8 °C (degradation).

Computational Details

All calculations were performed without symmetry restrictions. Starting coordinates obtained with GaussView 6.0²⁸ or directly from the crystal structure analyses for synthesized ligands. The geometry optimization and NBO²⁹ analysis were carried out with the Gaussian16 (revision C.01)³⁰ program package, and the AIM analysis was done with the Multiwfn program version 3.6.³¹ The geometry optimization and AIM analysis of the palladium complex was performed using density functional theory (DFT)^32,33 with the PBE0 functional,³⁴ through the LANL2TZ(f) basis set and the corresponding ECP for palladium³⁵⁻³⁷ and the def2svp basis set³⁸ for all other atoms, in conjunction with Grimme’s D3 dispersion correction with Becke–Johnson damping.³⁹⁻⁴¹ For the examined ligands, geometry optimizations were performed using DFT with B3LYP-D3/6-31+G*.⁴²⁻⁴⁷ The metrical parameters of the energy-optimized geometries are in good agreement with those determined by X-ray diffraction. Harmonic vibrational frequency analyses were performed at the same level of theory as the optimizations to determine the nature of the structure.⁴⁸ The vibrational frequency analysis showed no imaginary frequencies. The NBO analysis was performed with the B3LYP functional, the LANL2TZ(f) basis set, and the corresponding ECP for palladium and the 6-31+G* basis set for all other atoms. The charges were fit to the electrostatic potential at points selected according to the CHelp scheme and the electrostatic property for the phosphorus atom of the phosphine was taken from that fit.⁴⁹

Crystallographic Details

Data collection of all compounds was carried out on either an Oxford SuperNova diffractometer (Cu microsource) or an XtaLAB Synergy (HyPix detector). Suitable crystals of all compounds were mounted in an inert oil (perfluoropoly aklylether) and directly transferred into a cold nitrogen stream. All crystal structure determinations were carried out at 100 K. The structures were solved using direct methods, refined using full-matrix least-squares techniques on F² with the SHELX software package^50,51 and expanded using Fourier techniques. Data collection parameters are given in Tables S1–S4. Crystallographic data (including structural factors) have been deposited with the Cambridge Crystallographic Data Centre as supplementary publication CCDC numbers 2089207–2089217.

Supporting Information Available

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/acs.organomet.1c00349.

• NMR spectra, crystallographic and computational details (PDF)

• Coordinates of the energy-optimized structures (XYZ)

Accession Codes

CCDC 2089207–2089217 contain the supplementary crystallographic data for this paper. These data can be obtained free of charge via www.ccdc.cam.ac.uk/data_request/cif, or by emailing data_request@ccdc.cam.ac.uk, or by contacting The Cambridge Crystallographic Data Centre, 12 Union Road, Cambridge CB2 1EZ, UK; fax: +44 1223 336033.

The authors declare the following competing financial interest(s): The authors have filed patent WO2019030304 covering the YPhos ligands and precatalysts discussed, which is held by UMICORE, and products will be made commercially available.