Synthesis
DOI: 10.1055/a-2615-1768
paper
Romanian Chemists in Synthesis

Carbon–Carbon Bond Formation Mediated by an Iron(0)–Olefin Pincer Complex

Zachary S. Lincoln
,
Melissa R. Hoffbauer
,

This work was supported by the National Science Foundation grant CAS-2102517 (VMI). Crystallographic data was collected on a diffractometer acquired with the help of the National Science Foundation grant MRI award CHE-2214606.
 


Abstract

C–H functionalization is a highly appealing strategy for accessing complex molecular structures. Herein, we show that π-tethered pincer ligands can engage in C–H activation when coordinated to iron. These reactions result in C(sp2)–C(sp2) bond formation through oxidative coupling and β-hydride elimination/reductive elimination pathways with alkynes and isocyanides.


#

The ability to transform carbon–hydrogen bonds into new carbon–carbon or carbon–heteroatom bonds is an attractive and economical strategy to build complex molecular structures.[1] [2] [3] [4] Typical transition-metal-mediated processes rely on the use of preinstalled directing groups to direct the transition metal to the site to be activated.[5–7] We have previously demonstrated the ability of PCcarbeneP–iron complexes to undergo directed C–H activations with a variety of external π-substrates.[8] Despite the success of this approach, the general requirement to install a directing group into the substrate diminishes its synthetic utility, particularly for substrates such as olefins and alkynes.[9] Metal–ligand cooperative reactions represent a useful alternative:[10] [11] [12] [13] a low-valent metal complex can bind strongly to the π system, allowing for C–H activation, functionalization, and dissociation of the higher oxidation state metal (Scheme [1]).

Zoom Image
Scheme 1 (a) General strategy for C–H functionalization of olefins mediated by transition metals; (b) examples of diphosphine pincer ligands containing internal olefins with late transition metals;[15] [16] [31] (c) this work: iron-promoted C(sp2)–C(sp2) bond formation by coupling of alkynes and isocyanides to the internal olefin of the pincer ligand

Pincer ligands containing a π-tethered moiety have emerged as an interesting strategy to achieve metal–ligand cooperative reactions, and numerous examples of both carbon–carbon and carbon–heteroatom donors have been developed.[14] [15] [16] [17] [18] [19] [20] [21] [22] [23] [24] [25] [26] [27] [28] [29] [30] Our group previously reported the coordination of mid and late transition metals to a pincer ligand containing an internal olefin π-donor, and described the hemilability of the ligand depending on the electronic state of the metal.[15] In the case of d8 nickel and palladium complexes, we observed a reversible C–H activation of one of the olefinic protons, demonstrating the potential for metal–ligand cooperative reactions (Scheme [1b], left).[16] [21] In addition, a number of groups described the ability of metal complexes supported by pincer ligands incorporating terminal olefins to reversibly activate the olefinic C–H bonds,[14] [19] [30] [31] which holds promise as a general strategy to access olefinic C–H functionalization. Moret and coworkers applied this strategy toward nickel-promoted C–C coupling of alkynes with an olefinic pincer ligand (Scheme [1b], right).[31]

Iron-based catalysts are especially attractive to be employed in such a strategy, due to their high abundance and low toxicity.[32] [33] However, no examples of iron centers with pincer ligands containing a terminal olefin π-donor have been reported. Herein, we report the synthesis of a new pincer ligand, its coordination with iron, and its subsequent ability to C–C couple, mediated by the iron center (Scheme [1c]).

To access the desired iron complex, the terminal olefinic pincer pro-ligand 1 was synthesized (Scheme [2]). We have previously obtained 1 by radical addition of dichloromethane to PCcarbeneP–palladium complexes,[34] but its bulk isolation was not reported. Although the diphenylphosphine derivative was previously reported,[29] we were unable to apply the synthesis to the diisopropylphosphine substituents. Thus, an alternative route was devised using the Wittig reaction of bis(2-bromophenyl)methanone and subsequent lithium–halogen exchanges followed by trapping with chlorodiisopropylphosphine, affording 1 as a colorless oil. The 1H NMR spectrum displays the two equivalent olefinic protons as a doublet (J HP = 1.8 Hz) at 5.63 ppm, whereas the phosphorus atoms resonate as a singlet at –3.75 ppm in the 31P{1H} NMR spectrum.

Zoom Image
Scheme 2 Synthesis of 24 and solid-state molecular structure of 3 [most H atoms have been omitted for clarity; atomic displacements are displayed at the 35% probability level; selected distances (Å) and angles (°) for 3: C(1)–C(2), 1.428(2); N–N#, 1.163(1); Fe–C(1), 2.073(1); Fe–C(2), 2.034(1); Fe–C(1)/C(2)centroid, 1.926; Fe–N(1), 1.748(1); Fe–P(1), 2.2717(6); Fe–P(2), 2.2698(7); C(1)/C(2)centroid–Fe–N, 130.39; P(1)–Fe–P(2), 105.60(2); N–Fe–P(1), 112.37(4); N–Fe–P(2), 110.47(4); C(1)/C(2)centroid–Fe–P(1), 99.41; C(1)/C(2)centroid–Fe–P(2), 95.42]

Having synthesized 1, it was coordinated to iron by addition to a suspension of FeCl2 in tetrahydrofuran, affording [{PC(=CH2)P}FeCl2] {2; PC(=CH2)P = 1,1′-bis[2-(diisopropylphosphino)phenyl]ethylene}, which was isolated as a pale yellow solid in 58% yield. Despite extensive attempts, we were unable to obtain single crystals of 2; however, its solution-state magnetic moment corresponds to S = 2, similar to the related internal olefin complex, [(tPCH=CHP)FeBr2] {tPCH=CHP = trans-1,2-bis[2-(diisopropylphosphino)phenyl]ethylene}, which was structurally identified as a tetrahedral Fe(II) center with no olefin interaction.[15] Thus, we propose that 2 does not have an iron–olefin interaction.

The coordination of the metal center to the olefin is a key characteristic for C–H activation. On the basis of our previous studies, we found that reduction of the iron center to Fe(I) led to olefin coordination.[15] Attempts to reduce 2 by one electron were not successful. Although a new paramagnetic product was obtained, presumably the iron(I) analogue, a number of unknown impurities were obtained alongside it, and it could not be isolated cleanly. Instead, we reduced 2 by two electrons with KC8 in hexanes, leading to the μ-N2–iron(0) dimer [{PC(=CH2)P}Fe]2N2 (3) as a red solid after two days (Scheme [2]). Solution-state magnetic moment measurements were consistent with S = 1 corresponding to two tetrahedral S = 1/2 centers, as further confirmed by the solid-state molecular structure (Scheme [2]).

The C–C distance of the olefin increases to 1.428(2) Å in 3 (Scheme [2]), compared to 1.34 Å in 1,1-di-p-tolylethylene,[35] indicative of coordination to the iron center, with an Fe–C/Ccentroid distance of 1.926 Å; both parameters are similar to those of other low-valent iron–olefin complexes.[15] , [36] [37] [38] [39] [40] [41] [42] Consistent with solution-state measurements, the iron center is approximately tetrahedral (τ4′ = 0.77), with a bridging dinitrogen ligand. To relieve steric clash, the two ligands are rotated by 75.6°. The Fe–N and N–N distances of 1.748(1) Å and 1.163(1) Å, respectively, are comparable to other low-valent iron complexes with bridging dinitrogen ligands,[43–48] and point toward only a minimal binding of the dinitrogen ligand between the two centers.

Zoom Image
Figure 1 Solid-state molecular structure of 4; [most H atoms have been omitted for clarity; atomic displacements are displayed at the 35% probability level; selected distances (Å) and angles (°) for 4: C(1)–C(2), 1.432(3); Fe–C(1), 2.069(2); Fe–C(2), 2.043(2); Fe–C(1)/C(2)centroid, 1.927; Fe–P(1), 2.2471(6); Fe–P(2), 2.2525(8); Fe–P(3), 2.2337(7); C(1)/C(2)centroid–Fe–P(1), 98.98; C(1)/C(2)centroid–Fe–P(2), 93.58; C(1)/C(2)centroid–Fe–P(3), 117.77; P(1)–Fe–P(2); 108.30(2); P(1)–Fe–P(3), 123.10(2); P(2)–Fe–P(3); 110.77(2)]

We found that 3 was unstable under vacuum, converting to 1 and an insoluble black solid. To circumvent this problem, the bridging dinitrogen was substituted with trimethylphosphine, affording [{PC(=CH2)P}Fe(PMe3)] (4) as a dark red solid (Scheme [2]). As with 3, the solution-state magnetic moment corresponds to S = 1, while the solid-state molecular structure revealed a monomeric complex (Figure [1]). The C–C and Fe–C/Ccentroid distances in 4 change only minimally from those of 3 to 1.432(3) Å and 1.927 Å, respectively, implying a similar bonding situation. Fe complex 4 also maintains a tetrahedral environment (τ4′ = 0.82), with the trimethylphosphine occupying the former dinitrogen coordination site (Figure [1]).

Having obtained a low-valent iron complex with a coordinated olefin, we next sought to explore its reactivity. Several groups have described the reversible insertion of dihydrogen into the tethered olefin,[14] [16] [19] [21] [30] [31] and thus we sought whether our complexes would exhibit this behavior. Unfortunately, no reaction of either 3 or 4 with hydrogen gas was observed, even at higher pressures. Heating either in the presence of hydrogen gas afforded an insoluble black material and clean formation of the free ligand 1, indicating that no hydrogenation occurs at elevated temperatures. This lack of reactivity displays an important divergence from the reactivity observed in other complexes containing olefin-pincer ligands.[14] [16] [19] [21] [30] [31]

Zoom Image
Scheme 3 Synthesis of 5 by C–C coupling of diphenylacetylene with 4 and molecular structure of 5 [most H atoms have been omitted for clarity; atomic displacements are displayed at the 35% probability level; selected distances (Å) and angles (°) for 5: C(1)–C(2), 1.444(10); C(2)–C(3), 1.418(10); C(3)–C(4), 1.441(9); Fe–C(1), 2.046(7); Fe–C(2), 2.022(7); Fe–C(3), 2.086(7); Fe–C(4), 2.163(7); Fe–C(1)/C(2)centroid, 1.902; Fe–C(2)/C(3)centroid, 1.928; Fe–C(3)/C(4)centroid, 1.999; Fe–N(1), 1.817(7); Fe–P(1), 2.225(2); Fe–P(2), 2.228(2); P(1)–Fe–P(2), 108.87(8); P(1)–Fe–N(1), 101.9(2); P(2)–Fe–N(1), 94.8(2); C(1)/C(2)centroid–Fe–N(1), 154.87; C(3)/C(4)centroid–Fe–N(1); C(1)/C(2)centroid–Fe–C(3)/C(4)centroid, 66.23]

Prompted by the reactivity observed with a related nickel complex by Moret[31] as described above (Scheme [1b]), we investigated the reactivity of the newly synthesized iron complexes with alkynes. The addition of diphenylacetylene to a solution of 3 or 4 did not result in any change at ambient temperature as indicated by 1H and 31P NMR spectroscopy. This result was surprising since Holland and coworkers described that alkynes are preferred over both dinitrogen and phosphines as a ligands for low-valent iron complexes.[49] Despite this, slow conversion to a new diamagnetic product was observed by heating a solution of 3 or 4 in the presence of diphenylacetylene, alongside the release of trimethylphosphine in the case of 4. The product of both reactions was identified by its solid-state molecular structure (Scheme [3]) as [{PC(=CH–CPh=CHPh)P}Fe(N2)] (5). C–C coupling of the alkyne and internal olefin is obvious, with a new C(2)–C(3) distance of 1.418(10) Å. Removal of a hydrogen from C(2) and its transfer to the terminus of the alkyne resulted in a butadiene moiety. The C(1) to C(4) chain C–C distances average 1.43 Å; while the Fe–C/Ccentroids average 1.94 Å; this is consistent with 3, 4, and other iron–butadiene complexes.[50] [51] [52] [53] [54] [55] Unlike 3 and 4, 5 is a five-coordinate complex with a terminal dinitrogen ligand, consistent with other literature examples,[52] and corresponding to a reduced electron density at the iron center from π-backbonding to the butadiene.

NMR spectroscopy of 5 supported the solid-state structure, with the inequivalent phosphorus atoms resonating as doublets (J PP = 5.5 Hz) at 82.5 and 79.9 ppm. In the 1H NMR spectrum, two-dimensional NMR identified a single olefinic proton at 4.95 ppm. The terminal butadiene proton was identified upfield at 1.0 ppm, representing a significant shift from related nickel complexes.[31] This might be due to shielding from one of the ligand’s aryl rings, however, a significant upfield shifting of olefinic protons to ca. 2.0 ppm was also observed with [(tPCH=CHP)CoCl], the effect of which was reduced when a strong π-acceptor ligand was added.[15]

We hypothesized that we could also take advantage of the presence of a π-acceptor, and insert a second equivalent of alkyne, similar to Moret’s results. As with 3, adding diphenylacetylene did not displace the dinitrogen ligand, and heating of the reaction mixture in a sealed tube did not afford any noticeable reaction until 130 °C, at which point decomposition of 5 occurred. Thus, unlike the nickel-based Moret example, a second alkyne coupling was not observed in our system, likely impeded by the substitution of the dinitrogen ligand due to steric crowding.

To extend this strategy to other π-systems, we also explored the reactivity with olefins. As with diphenylacetylene, adding stilbene to 3 or 4 did not lead to any noticeable reactivity at ambient temperature. Furthermore, no reaction was observed at 80 °C, likely a result of the lower binding constants for olefins.[49] At 120 °C, a small amount of a new product, with similar 31P chemical shifts to 5, was observed; however, after 10 days at this temperature, only ca. 5% of the starting materials had reacted. Increasing the temperature to 130 °C resulted in decomposition, and, thus, we could not isolate the product to confirm that olefins also undergo this coupling reaction.

However, we did observe a similar reactivity with other π-systems. The addition of two equivalents of phenyl isocyanide (CNPh) resulted in the immediate formation of [{PC(=CH2)P}Fe(CNPh)2] (6) as evidenced by the 1H and 31P NMR spectra, where the phosphorus atoms resonate equivalently at 87.2 ppm (Scheme [4]). However, single crystals obtained of 6 at –35 °C revealed the formation of a new compound, [{PC(=CH–CH=NPh)P}Fe(CNPh)] (7; Scheme [4]). Monitoring a solution of 6 reveals the slow formation of 7 at ambient temperature over the course of a day, and the reaction can be completed in several hours by heating of 6 at 60 °C.

The solid-state molecular structure of 7 (Scheme [4]) revealed a pseudo-square pyramidal iron center. The coupling of one of the isocyanide ligands to the olefin is evident, with a new C(2)–C(3) distance of 1.409(3) Å. As with 5, migration of one of the olefinic C–H bonds coincides with this; however, instead of being at the terminus, the proton is located on the erstwhile isocyanide carbon. Correspondingly, an increase in the C(3)–N(3) distance to 1.347(3) Å occurs, indicating the formation of an imine. Overall, the formation of an azabutadiene, analogous to the butadiene observed in 5, is evident in the structure.[56] [57]

Zoom Image
Scheme 4 Synthesis of 6 by ligand displacement from 4 and subsequent C–C coupling to 7, with molecular structure of 7 below [most H atoms have been omitted for clarity; atomic displacements are displayed at the 35% probability level; selected distances (Å) and angles (°) for 7: C(1)–C(2), 1.457(3); C(2)–C(3), 1.409(3); C(3)–N(3), 1.347(3); C(4)–N(4), 1.195(3); Fe–C(1), 2.077(2); Fe–C(2), 2.032(2); Fe–C(3), 2.094(2); Fe–N(3), 2.205(2); Fe–C(1)/C(2)centroid, 1.921; Fe–C(2)/C(3)centroid, 1.940; Fe–C(3)/N(3)centroid, 2.042; Fe–P(1), 2.2339(6); Fe–P(2), 2.1766(8); P(1)–Fe–P(2), 105.66(3); C(1)/C(2)centroid–Fe–C(4), 152.95; C(3)/N(3)centroid–Fe–C(4), 102.10; C(1)/C(2)centroid–Fe–C(3)/N(3)centroid, 63.10; C(4)–N(4)–C(41), 157.7(2)]

The NMR characterization of 7 (Scheme [4]) differs significantly from that of 6. The phosphines of 7 are inequivalent, with the phosphorus atoms resonating as singlets at 102.3 and 89.6 ppm. A single olefinic proton is observed at 4.73 ppm as a doublet of triplets (J = 4.9, 2.5 Hz) due to coupling with the inequivalent phosphine and the nearby olefinic proton. In contrast to 5, the migrated proton resonates at 6.73 ppm as a doublet of doublets (J = 2.8, 1.0 Hz); we hypothesize that this change in chemical shift is attributed not only to the imine, but also the significant π-accepting nature of the isocyanide ligand.

Though insertions of isocyanides into olefins are well established,[58] [59] [60] to the best of our knowledge, our results represent the first example where azabutadienes are formed by C–H activation and migration.

The formation of 5 likely proceeds through an η2-alkyne complex, followed by an oxidative coupling between the olefin and alkyne, leading to a metallacyclopentene intermediate (Scheme [5]). From here, a β-hydride elimination can occur, followed by reductive elimination, affording 5. Although Moret and coworkers propose a ligand-to-ligand hydrogen transfer from nickel metallacyclopentenes,[31] to the best of our knowledge, this type of process has not been established with iron, whereas β-hydride elimination is well known.[61] Furthermore, the proposed formation of 5 is supported by analogous ruthenium(II) examples that were proposed to proceed through a ruthenacyclopentene intermediate that underwent β-hydride elimination and reductive elimination.[62] [63] [64]

Zoom Image
Scheme 5 Proposed mechanism for the formation of 5 by oxidative coupling of the η2-alkyne and internal olefin, followed by a β-hydride elimination and reductive elimination from a metallacyclopentene intermediate

We propose a similar mechanism for 7 (Scheme [6]), whereby oxidative coupling of the olefin and isocyanide would furnish an iminoacyl complex;[65] [66] β-hydride elimination follows, and reductive elimination to the more nucleophilic carbon center occurs. In this case, only one stereoisomer was observed in both solution and solid state, in agreement with the proposed iron-mediated mechanism.

Zoom Image
Scheme 6 Proposed mechanism for the formation of 7 by oxidative coupling of one of the isocyanide ligands of 6 and subsequent β-hydride elimination and reductive elimination from the intermediate

In summary, our study illustrated the utility and reactivity of iron complexes supported by pincer ligands bearing a terminal olefin. Unlike typical low-valent iron complexes, the presence of an alkyne or olefin does not immediately displace coordinated dinitrogen or phosphine ligands, as a result of the tethered π-ligand on the pincer ligand. At elevated temperatures, alkynes undergo an oxidative C(sp2)–C(sp2) coupling with the internal olefin, followed by a β-hydride elimination and reductive elimination from the erstwhile metallacyclopentene intermediate to give a tethered η4-butadiene. Unlike in related Ni(0) systems, we could not extend the observed reactivity to further additions, likely because the iron center becomes increasingly deactivated toward additional π-bonding. In line with this, weaker π-ligands, such as olefins, do not undergo any coupling under reasonable conditions, whereas the strong π-acceptor isocyanides undergo this facile C–C coupling to afford η4-1-azabutadienes even at low temperatures.

All experiments were performed under an inert atmosphere of N2 using standard glovebox techniques unless otherwise specified. The solvents hexanes, n-pentane, Et2O, toluene, and THF were dried by passing through a column of activated alumina and stored in the glovebox over 3 Å molecular sieves. C6D6 was dried over CaH2 followed by vacuum transfer, stored in the glovebox over 3 Å molecular sieves, and filtered through a Celite pad before use. Bis(2-bromophenyl)methanone,[67] methyltriphenylphosphonium iodide,[68] and potassium graphite (KC8)[69] were synthesized according to literature procedures. All other chemicals were commercially available and used without further purification. 1H, 13C{1H}, 31P{1H}, COSY, HSQC, and HMBC spectra were recorded on a Bruker DRX 400/500/800 or Varian DirectDrive 600 spectrometer. All chemical shifts are referenced to the residual solvent resonance of the deuterated solvents for 1H and 13C chemical shifts. 31P chemical shifts were referenced to external 85% H3PO4. The crystal structures of 35 and 7 were determined by X-ray diffraction using single crystals grown by their respective methods (see below). The crystals were mounted on a Nylon cyroloop coated in Paratone N oil and were kept at 120 K during data collection. X-ray diffraction data were collected on a Bruker D8 Venture diffractometer with a Photon III CPAD detector equipped with Helios high-brilliance multilayer mirror optics or a Bruker Quest diffractometer with a Photon III CMOS area detector equipped with Helios high-brilliance multilayer mirror optics. The structures were solved and refined using the SHELXle software suite.[70] [71] Structures were visualized using Mercury.[72] CCDC 2434845–2434848 (35 and 7) contain the supplementary crystallographic data for this paper. The data can be obtained free of charge from The Cambridge Crystallographic Data Centre via www.ccdc.cam.ac.uk/structures.


#

2,2′-(Ethene-1,1-diyl)bis(bromobenzene)

To a suspension of methyltriphenylphosphonium iodide (424 mg, 1.05 mmol, 1.2 equiv) in THF (50 mL), 1.6 M n-BuLi (0.65 mL, 1.04 mmol, 1.2 equiv) was added dropwise. The mixture was stirred at ambient temperature for 1 h before bis(2-bromophenyl)methanone (334 mg, 0.98 mmol, 1.0 equiv) in THF (10 mL) was added dropwise. The dark brown solution was stirred at 60 °C for 24 h, after which 10% aq NH4Cl (10 mL) was added. The mixture was extracted with Et2O and dried over Na2SO4, and the volatile components were removed under reduced pressure to yield a dark brown residue, which was purified by column chromatography (silica gel, hexanes/EtOAc 99:1) to provide a colorless oil.

Yield: 281 mg (85%).

1H NMR (400 MHz, C6D6, 295 K): δ = 7.35 (dd, J = 8.0, 1.2 Hz, 2 H, ArH), 7.09 (dd, J = 7.7, 1.7 Hz, 2 H, ArH), 6.82 (td, J = 7.5, 1.3 Hz, 2 H, ArH), 6.64 (ddd, J = 8.0, 7.5, 1.7 Hz, 2 H, ArH), 5.51 (s, 2 H, =CH 2).

13C{1H} NMR (126 MHz, C6D6, 295 K): δ = 148.2 (s, -C=CH2), 142.4 (s, ArC), 133.8 (s, ArC), 132.0 (s, ArC), 129.0 (s, ArC), 127.3 (s, ArC), 123.1 (s, C=CH2), 122.9 (s, ArC).


#

PC(=CH2)P (1)

A 100 mL round-bottom flask equipped with a stir bar was charged with 2,2′-(ethene-1,1-diyl)bis(bromobenzene) (271 mg, 0.8 mmol, 1.0 equiv) and Et2O (20 mL), and the solution was cooled to –35 °C. To the cooled solution was added 1.6 M n-BuLi (1.1 mL, 1.76 mmol, 2.1 equiv), and the resulting solution was stirred at ambient temperature for 30 min. Chlorodiisopropylphosphine (0.30 mL, 1.88 mmol, 2.1 equiv) was added dropwise to the solution, and the mixture was stirred overnight at ambient temperature to yield a light yellow solution with copious precipitate. To the solution, 10% aq NH4Cl (10 mL) was added, and the mixture was stirred for 10 min before the aqueous layer was eliminated and the organic fraction was dried over Na2SO4. Upon filtration through a Celite pad and removal of the volatile components, a pale-orange oil was isolated and further purified by passing through a silica plug (hexanes); this gave 1 as a colorless oil. (Compound 1 is typically contaminated with small amounts of (n-Bu)P(iPr)2 that are difficult to separate from the product and so 1 is generally used without further purification.)

Yield: 224 mg (68%).

1H NMR (500 MHz, C6D6, 295 K): δ = 7.58–7.35 (m, 2 H, ArH), 7.35 (dd, J = 7.4, 1.4 Hz, 2 H, ArH), 7.15–7.04 (m, 4 H, ArH), 5.63 (d, J = 1.8 Hz, 2 H, C(=CH 2)), 1.86 (dt, J = 13.9, 7.0 Hz, 4 H, -CH(CH3)2), 1.08 (ddd, J = 13.2, 6.9, 1.5 Hz, 12 H, -CH(CH 3)2), 0.85 (dd, J = 10.2, 7.1 Hz, 12 H, -CH(CH 3)2).

13C{1H} NMR (101 MHz, C6D6, 295 K): δ = 151.9 (s, ArC), 151.6 (s, ArC), 151.4 (t, J = 5.6 Hz, ArC), 137.0–136.6 (m, ArC), 132.8 (s, ArC), 132.6 (t, J = 6.5 Hz, ArC), 126.6 (s, ArC), 122.3 (t, J = 5.4 Hz, C(=CH2)), 25.9 (t, J = 4.4 Hz, -P(CH(CH3)2)), 25.8 (t, J = 4.4 Hz, -P(CH(CH3)2)), 21.0–20.7 (m, -P(CH(CH3)2)), 20.7–20.4 (m, -P(CH(CH3)2)).

31P{1H} NMR (202 MHz, C6D6, 295 K): δ = –3.75 (s).


#

[{PC(=CH2)P}FeCl2] (2)

In a 20 mL scintillation vial with a stir bar, FeCl2 (131 mg, 1.0 mmol, 1.1 equiv) was suspended in THF (10 mL). Compound 1 (369 mg, 0.90 mmol, 1.0 equiv) in THF (2 mL) was added to the suspension, which was then stirred at ambient temperature for 2 h; over time, the suspended FeCl2 slowly dissolved, and the solution became yellow. The volatile components were removed under reduced pressure, and the yellow residue was triturated with hexanes (10 mL) and filtered over a Celite pad. The remaining yellow solid was dissolved in CH2Cl2 (10 mL), and the solution was filtered through a Celite pad and then concentrated under reduced pressure. Analytically pure 2 was obtained by layering this concentrated solution with hexanes at –35 °C.

Yield: 281 mg (58%).

1H NMR (400 MHz, C6D6, 295 K): δ = 80.7 (υ1/2 = 745 Hz), 40.2 (υ1/2 = 895 Hz), 38.6 (υ1/2 = 100 Hz), 32.6 (υ1/2 = 134 Hz), 24.2 (υ1/2 = 234 Hz), 23.6 (υ1/2 = 78 Hz), 13.8 (υ1/2 = 58 Hz), 12.4 (υ1/2 = 99 Hz), 10.6 (υ1/2 = 314 Hz), 8.4 (υ1/2 = 49 Hz), 6.1 (υ1/2 = 156 Hz), 5.7 (υ1/2 = 116 Hz), 5.2 (υ1/2 = 178 Hz), –1.1 (υ1/2 = 175 Hz), –3.5 (υ1/2 = 90 Hz), –4.1 (υ1/2 = 85 Hz), –6.4 (υ1/2 = 397 Hz), –15.6 (υ1/2 = 236 Hz), –66.2 (υ1/2 = 365 Hz), –82.1 (υ1/2 = 132 Hz).

Anal. Calcd for C26H38FeP2Cl2: C, 57.91; H, 7.10. Found: C, 57.83, H, 8.14.

μeff = 4.79 μB.


#

[{PC(=CH2)P}Fe]2N2 (3)

In a 20 mL scintillation vial equipped with a stir bar, 2 (47 mg, 0.09 mmol, 1.0 equiv) was suspended in hexanes (10 mL) and freshly prepared potassium graphite (41 mg, 0.3 mmol, 2.2 equiv) was added to the suspension. The resulting solution was stirred at ambient temperature for 2 d, over which time the solution became green followed by dark red. The resulting solution was filtered through a Celite pad and concentrated under reduced pressure. The concentrated solution was left at –35 °C to yield dark-red crystals of 3.

Yield: 15 mg (36%).

1H NMR (400 MHz, C6D6, 295 K): δ = 45.9 (υ1/2 = 700 Hz), 36.2 (υ1/2 = 136 Hz), 19.0 (113 Hz), 4.5 (υ1/2 = 509 Hz), –5.0 (υ1/2 = 299 Hz), –14.2 (υ1/2 = 91 Hz).

Anal. Calcd for C52H76N2Fe2P4: C, 64.74; H, 7.94; N, 2.90. Found: C, 63.91; H, 7.25; N, 1.98.

μeff = 5.01 μB.

X-ray crystal structure of 3: Single crystals were obtained as dark-red rods grown from a saturated n-pentane solution at ambient temperature in the glovebox. Crystal data for C37H56FeN2P2: Mr = 646.62; monoclinic; space group C2/c; a = 15.964(4) Å; b = 22.881(4) Å; c = 17.507(3) Å; α = 90°; β = 104.484(4)°; γ = 90°; V = 6192(2) Å3; Z = 8; T = 120(2) K; λ = 0.71073 Å; μ = 0.621 mm–1; d calc = 1.387 g.cm–3; 72570 reflections collected; 7711 unique (R int = 0.0346); giving R 1 = 0.0549, wR 2 = 0.1907 for 6553 data with [I > 2σ(I)] and R 1 = 0.0659, wR 2 = 0.2041 for all 7711 data. Residual electron density (e·Å–3) max/min: 2.055/–0.405. The SQUEEZE function in Platon was used to remove density associated with a highly disordered molecule of solvent.


#

[{PC(=CH2)P}Fe(PMe3)] (4)

In a 20 mL scintillation vial equipped with a stir bar, 3 (59 mg, 0.12 mmol, 1.0 equiv) was dissolved in n-pentane (3 mL) and 2 M trimethylphosphine in Et2O (150 μL, 0.3 mmol, 2.5 equiv) was added to the stirring solution. The resulting solution was stirred at ambient temperature for 2 h, and the volatile components were then removed under reduced pressure. The crude residue was dissolved in n-pentane (2 mL) and filtered through a Celite pad, concentrated under reduced pressure, and stored at –35 °C to yield dark-red crystals of 4.

Yield: 47 mg (69%).

1H NMR (400 MHz, C6D6, 295 K): δ = 90.6 (υ1/2 = 376 Hz), 28.2 (υ1/2 = 587 Hz), 19.98 (υ1/2 = 132), 15.0 (υ1/2 = 569 Hz), 12.3 (υ1/2 = 266 Hz), 9.6 (υ1/2 = 33 Hz), 0.52 (υ1/2 = 222 Hz), –1.40 (υ1/2 = 517 Hz), –6.6 (υ1/2 = 470 Hz), –8.3 (υ1/2 = 48 Hz).

X-ray crystal structure of 4: Single crystals were obtained as red/brown blocks from a concentrated solution of n-pentane/Et2O (1:1) at –35 °C in the glovebox. Crystal and refinement data for C120H198Fe4OP12: Mr = 2251.81; triclinic; space group P1; a = 10.2754(15) Å; b = 17.934(3) Å; c = 18.672(3) Å; α = 102.557(2)°; β = 104.537(2)°, γ = 104.307(2) Å; V = 3080.9(8) Å3; Z = 1; T = 120(2) K; λ = 0.71073 Å; μ = 0.663 mm–1; d calc = 1.214 g·cm−3; 85192 reflections collected; 10821 unique (R int = 0.0207); giving R 1 = 0.0300, wR 2 = 0.0726 for 10257 data with [I > 2σ(I)] and R 1 = 0.0319, wR 2 = 0.0749 for all 10821 data. Residual electron density (e·Å–3) max/min: 0.663/–0.659.

Anal. Calcd for C29H47FeP3Cl2: C, 63.97; H, 8.70. Found: C, 63.85, H, 8.21.

μeff = 3.1 μB.


#

[{PC(=CH-CPh=CHPh)P}Fe(N2)] (5)

A J-Young NMR tube was charged with a solution of 4 (32 mg, 0.06 mmol, 1.0 equiv) and diphenylacetylene (11 mg, 0.06 mmol, 1.0 equiv) in C6D6 (0.7 mL). The tube was sealed and placed in an oil bath at 80 °C for 48 h, over which the dark-red solution became progressively lighter. The tube was brought back into the glovebox, and the contents were transferred to a 20 mL scintillation vial. The volatile components were removed under reduced pressure, followed by the addition of hexanes (10 mL) and the removal of the volatile components under reduced pressure. The orange residue was extracted with n-pentane (3 mL) and filtered through a pad of Celite. The solution was concentrated and stored at –35 °C, from which analytically pure 5 was obtained after 1 d.

Yield: 24 mg (62%).

1H NMR (400 MHz, C6D6, 295 K): δ = 8.03–7.89 (m, 2 H, ArH), 7.54–7.48 (m, 2 H, ArH), 7.40–7.33 (m, 1 H), 7.28–7.17 (m, 4 H), 7.14–7.03 (m, 4 H), 6.98 (s, 2 H), 6.92–6.84 (m, 2 H), 6.82–6.75 (m, 3 H), 4.95 (d, J = 6.5 Hz, 1 H, -C=CH-CPh=CHPh), 2.80 (ddt, J = 22.0, 14.8, 7.3 Hz, 1 H, -CH(CH3)2), 2.47 (dd, J = 11.9, 7.0 Hz, 1 H, -CH(CH3)2), 2.08–1.88 (m, 3 H, -CH(CH3)2), 1.60 (ddd, J = 13.7, 12.2, 7.4 Hz, 3 H, -CH(CH 3)2), 1.45 (dd, J = 11.1, 7.2 Hz, 6 H, -CH(CH 3)2), 1.30 (dd, J = 9.3, 7.1 Hz, 3 H, -CH(CH 3)2), 1.21 (dd, J = 13.6, 7.3 Hz, 3 H, -CH(CH 3)2), 1.11 (ddd, J = 11.0, 6.9, 3.7 Hz, 3 H, -CH(CH 3)2), 1.01–0.82 (m, 7 H, -CH(CH 3)2 { -C=CH-CPh=CHPh), 0.56 (dd, J = 13.7, 7.1 Hz, 3 H, -CH(CH 3)2).

13C{1H} NMR (101 MHz, C6D6, 295 K): δ = 159.3 (d, J = 27.9 Hz, ArC), 152.6 (d, J = 29.5 Hz, ArC), 147.1 (d, J = 36.8 Hz, ArC), 143.8 (s, ArC), 141.6 (s, ArC), 136.7 (d, J = 32.2 Hz, ArC), 132.1–131.5 (m, ArC), 130.9–130.6 (m, ArC), 130.1 (s, ArC), 130.0 (s, ArC), 129.0 (s, ArC), 128.7 (s, ArC), 128.5 (s, ArC), 128.3 (s, ArC), 127.3 (s, ArC), 125.5 (s, ArC), 124.7 (d, J = 4.5 Hz, ArC), 123.0 (s, ArC), 105.0 (d, J = 2.3 Hz, -C=CH-CPh=CHPh), 87.3 (s, -C=CH-CPh=CHPh), 58.4 (s, -C=CH-CPh=CHPh), 33.5 (d, J = 11.1 Hz, -CH(CH3)2), 31.5 (d, J = 12.4 Hz, -CH(CH3)2), 30.5 (d, J = 11.2 Hz, -CH(CH3)2), 29.9 (dd, J = 18.1, 4.3 Hz, -CH(CH3)2), 24.4 (d, J = 13.8 Hz, -CH(CH3)2), 23.3 (d, J = 8.8 Hz, -CH(CH3)2), 21.0 (dd, J = 10.7, 4.9 Hz, -CH(CH3)2), 20.8–20.1 (m, -CH(CH3)2), 19.7–19.4 (m, -CH(CH3)2), 19.3 (d, J = 7.0 Hz, -CH(CH3)2).

31P{1H} NMR (162 MHz, C6D6, 295 K): δ = 82.5 (br s, 1 P), 79.9 (d, J = 5.5 Hz, 1 P).

Anal. Calcd for C45H60FeN2P2·C5H12: C, 72.38; H, 8.10; N, 3.75. Found: C, 72.59; H, 8.15; N, 3.21.

X-ray crystal structure of 5: Single crystals were obtained as orange blocks from a concentrated n-pentane solution at –35 °C. Crystal data for C40H48FeN2P2: M r = 674.59; monoclinic; space group P2 1 ; a = 11.5209(11) Å; b = 15.2386(15) Å; c = 19.4373(19) Å; α = 90°; β = 92.877(2)°; γ = 90°; V = 3408.2(6) Å3; Z = 4; T = 120(2) K; λ = 0.71073 Å; μ = 0.568 mm–1; d calc = 1.315 g·cm–3; 15162 reflections collected; 14951 unique (R int = 0.0735); giving R 1 = 0.0739, wR 2 = 0.1134 for 11851 data with [I > 2σ(I)] and R 1 = 0.1106, wR 2 = 0.1230 for all 14951 data. Residual electron density (e·Å–3) max/min: 1.220/–0.804.


#

In-Situ Synthesis of [{PC(=CH2)P}Fe(CNPh)2] (6)

A J-Young NMR tube was charged with 4 (28 mg, 0.05 mmol, 1.0 equiv) in C6D6 (0.6 mL). A solution of phenyl isocyanide (10 mg, 0.10 mmol, 2.0 equiv) in C6D6 (0.2 mL) was added to the tube and then the tube was sealed and gently rotated for 5 min. 31P NMR analysis indicated full conversion to 6.

31P{1H} NMR (162 MHz, C6D6, 295 K): δ = 87.2 (s).


#

[{PC(=CH-CH=NPh)P}Fe(CNPh)] (7)

A J-Young NMR tube containing 4 (52 mg, 0.10 mmol, 1.0 equiv), phenyl isocyanide (20 mg, 0.194 mmol, 2.0 equiv), and C6D6 was sealed and heated in an oil bath at 60 °C for 12 h, over which time the yellow solution slowly became dark orange. After the reaction was complete (by NMR analysis), the tube contents were transferred to a 20 mL scintillation vial, and the volatile components were removed under reduced pressure. The dark orange residue was taken up in hexanes (10 mL), filtered through a Celite plug, and concentrated to ca. 1 mL under reduced pressure. The concentrated solution was stored at –35 °C, from which analytically pure 7 was isolated after 1 d.

Yield: 26 mg (39%).

1H NMR (400 MHz, C6D6, 295 K): δ = 7.60–7.57 (m, 1 H, ArH), 7.51–7.47 (m, 2 H, ArH), 7.23–7.16 (m, 2 H, ArH), 7.07–6.96 (m, 10 H, ArH), 6.93 (t, J = 7.3 Hz, 1 H, ArH), 6.87 (tt, J = 6.9, 1.4 Hz, 1 H, ArH), 6.82 (tt, J = 7.3, 1.5 Hz, 1 H, ArH), 6.73 (dd, J = 2.8, 1.0 Hz, 1 H, -C=CH-CH=NPh), 4.73 (dt, J = 4.9, 2.5 Hz, 1 H, -C=CH-CH=NPh), 2.81–2.70 (m, 1 H, -CH(CH3)2), 2.44 (dp, J = 13.9, 7.1 Hz, 1 H, -CH(CH3)2), 2.30–2.17 (m, 1 H, -CH(CH3)2), 1.54–1.46 (m, 8 H, overlapping signals: -CH(CH3)2 { -CH(CH 3)2), 1.44–1.39 (m, 3 H, -CH(CH 3)2), 1.29–1.21 (m, 6 H, -CH(CH 3)2), 1.09–1.03 (m, 6 H, -CH(CH 3)2), 0.38 (dd, J = 13.9, 7.1 Hz, 3 H, -CH(CH 3)2).

13C{1H} NMR (101 MHz, C6D6, 295 K): δ = 194.5 (dd, J = 34.4, 4.5 Hz, Fe-CNPh), 160.6 (d, J = 25.4 Hz, ArC), 156.6 (s, ArC), 145.2 (d, J = 35.7 Hz, ArC), 136.8 (d, J = 29.3 Hz, ArC), 134.4 (s, ArC), 131.9 (d, J = 15.7 Hz, ArC), 130.7 (s, ArC), 129.3 (s, ArC), 128.4 (s, ArC), 125.8 (s, ArC), 125.4 (d, J = 10.5 Hz, ArC), 125.1 (d, J = 4.9 Hz, ArC), 124.8 (s, ArC), 124.3 (d, J = 4.2 Hz, ArC), 123.7 (s, ArC), 121.6 (s, ArC), 107.7 (s, -C=CH-CH=NPh), 80.5 (s, -C=CH-CH=NPh), 32.0 (d, J = 14.4 Hz, -CH(CH3)2), 29.8 (dd, J = 12.4, 5.7 Hz, -CH(CH3)2), 28.1 (d, J = 23.0 Hz, -CH(CH3)2), 22.9 (d, J = 7.2 Hz, -CH(CH3)2), 21.6 (s, -CH(CH3)2), 21.0–20.8 (m, -CH(CH3)2), 20.7 (s, -CH(CH3)2), 20.3 (d, J = 5.1 Hz, -CH(CH3)2), 18.6 (d, J = 7.1 Hz, -CH(CH3)2).

31P{1H} NMR (162 MHz, C6D6, 295 K): δ = 102.3 (s, 1 P), 89.6 (s, 1 P).

Anal. Calcd for C46H62FeN2P2·C6H14: C, 72.62; H, 8.21; N, 3.68. Found: C, 72.35; H, 8.25; N, 3.75.

X-ray crystal structure of 7: Single crystals were obtained as orange blocks from a concentrated n-pentane solution at –35 °C. Crystal data for C40H48FeN2P2: M r = 674.59; triclinic; space group P-1; a = 9.1320(3) Å; b = 18.5053(7) Å; c = 21.8809(7) Å; α = 107.630(2)°; β = 90.2350(10)°; γ = 102.0260(10)°; V = 3437.7(2) Å3; Z = 4; T = 120(2) K; λ = 0.71073 Å; μ = 0.563 mm–1; d calc = 1.303 g·cm–3; 106315 reflections collected; 20862 unique (R int = 0.0588); giving R 1 = 0.0576, wR 2 = 0.1184 for 16200 data with [I > 2σ(I)] and R 1 = 0.0812, wR 2 = 0.1269 for all 20862 data. Residual electron density (e·Å–3) max/min: 1.664/–0.592.


#
#

Conflict of Interest

The authors declare no conflict of interest.

Acknowledgment

We thank Dr. Allen Oliver for helpful suggestions for solving the crystal structures.

Supporting Information


Corresponding Author

Vlad M. Iluc
Department of Chemistry and Biochemistry, University of Notre Dame
Notre Dame, Indiana 46556
USA   

Publication History

Received: 28 March 2025

Accepted after revision: 16 May 2025

Accepted Manuscript online:
16 May 2025

Article published online:
17 June 2025

© 2025. Thieme. All rights reserved

Georg Thieme Verlag KG
Oswald-Hesse-Straße 50, 70469 Stuttgart, Germany


Zoom Image
Scheme 1 (a) General strategy for C–H functionalization of olefins mediated by transition metals; (b) examples of diphosphine pincer ligands containing internal olefins with late transition metals;[15] [16] [31] (c) this work: iron-promoted C(sp2)–C(sp2) bond formation by coupling of alkynes and isocyanides to the internal olefin of the pincer ligand
Zoom Image
Scheme 2 Synthesis of 24 and solid-state molecular structure of 3 [most H atoms have been omitted for clarity; atomic displacements are displayed at the 35% probability level; selected distances (Å) and angles (°) for 3: C(1)–C(2), 1.428(2); N–N#, 1.163(1); Fe–C(1), 2.073(1); Fe–C(2), 2.034(1); Fe–C(1)/C(2)centroid, 1.926; Fe–N(1), 1.748(1); Fe–P(1), 2.2717(6); Fe–P(2), 2.2698(7); C(1)/C(2)centroid–Fe–N, 130.39; P(1)–Fe–P(2), 105.60(2); N–Fe–P(1), 112.37(4); N–Fe–P(2), 110.47(4); C(1)/C(2)centroid–Fe–P(1), 99.41; C(1)/C(2)centroid–Fe–P(2), 95.42]
Zoom Image
Figure 1 Solid-state molecular structure of 4; [most H atoms have been omitted for clarity; atomic displacements are displayed at the 35% probability level; selected distances (Å) and angles (°) for 4: C(1)–C(2), 1.432(3); Fe–C(1), 2.069(2); Fe–C(2), 2.043(2); Fe–C(1)/C(2)centroid, 1.927; Fe–P(1), 2.2471(6); Fe–P(2), 2.2525(8); Fe–P(3), 2.2337(7); C(1)/C(2)centroid–Fe–P(1), 98.98; C(1)/C(2)centroid–Fe–P(2), 93.58; C(1)/C(2)centroid–Fe–P(3), 117.77; P(1)–Fe–P(2); 108.30(2); P(1)–Fe–P(3), 123.10(2); P(2)–Fe–P(3); 110.77(2)]
Zoom Image
Scheme 3 Synthesis of 5 by C–C coupling of diphenylacetylene with 4 and molecular structure of 5 [most H atoms have been omitted for clarity; atomic displacements are displayed at the 35% probability level; selected distances (Å) and angles (°) for 5: C(1)–C(2), 1.444(10); C(2)–C(3), 1.418(10); C(3)–C(4), 1.441(9); Fe–C(1), 2.046(7); Fe–C(2), 2.022(7); Fe–C(3), 2.086(7); Fe–C(4), 2.163(7); Fe–C(1)/C(2)centroid, 1.902; Fe–C(2)/C(3)centroid, 1.928; Fe–C(3)/C(4)centroid, 1.999; Fe–N(1), 1.817(7); Fe–P(1), 2.225(2); Fe–P(2), 2.228(2); P(1)–Fe–P(2), 108.87(8); P(1)–Fe–N(1), 101.9(2); P(2)–Fe–N(1), 94.8(2); C(1)/C(2)centroid–Fe–N(1), 154.87; C(3)/C(4)centroid–Fe–N(1); C(1)/C(2)centroid–Fe–C(3)/C(4)centroid, 66.23]
Zoom Image
Scheme 4 Synthesis of 6 by ligand displacement from 4 and subsequent C–C coupling to 7, with molecular structure of 7 below [most H atoms have been omitted for clarity; atomic displacements are displayed at the 35% probability level; selected distances (Å) and angles (°) for 7: C(1)–C(2), 1.457(3); C(2)–C(3), 1.409(3); C(3)–N(3), 1.347(3); C(4)–N(4), 1.195(3); Fe–C(1), 2.077(2); Fe–C(2), 2.032(2); Fe–C(3), 2.094(2); Fe–N(3), 2.205(2); Fe–C(1)/C(2)centroid, 1.921; Fe–C(2)/C(3)centroid, 1.940; Fe–C(3)/N(3)centroid, 2.042; Fe–P(1), 2.2339(6); Fe–P(2), 2.1766(8); P(1)–Fe–P(2), 105.66(3); C(1)/C(2)centroid–Fe–C(4), 152.95; C(3)/N(3)centroid–Fe–C(4), 102.10; C(1)/C(2)centroid–Fe–C(3)/N(3)centroid, 63.10; C(4)–N(4)–C(41), 157.7(2)]
Zoom Image
Scheme 5 Proposed mechanism for the formation of 5 by oxidative coupling of the η2-alkyne and internal olefin, followed by a β-hydride elimination and reductive elimination from a metallacyclopentene intermediate
Zoom Image
Scheme 6 Proposed mechanism for the formation of 7 by oxidative coupling of one of the isocyanide ligands of 6 and subsequent β-hydride elimination and reductive elimination from the intermediate