Low-Valent Molybdenum PNP Pincer Complexes as Catalysts for the Semihydrogenation of Alkynes.

Low-valent molybdenum PNP pincer complexes were studied as catalysts for the semihydrogenation of alkynes. For that purpose, tBu-substituted PNP complexes PNP tBuMo(CO)2 (6a) and PNP tBuMo(CO)3 (6c) and the NNP complex NNP iPrMo(CO)2(PPh3) ((rac)-7) were synthesized and characterized. By utilizing the cyclohexyl-substituted complex PNPCyMo(CO)2(CH3CN) (5a), several diphenylacetylene derivatives are transformed to the corresponding (Z)-alkenes with good to very good diastereoselectivities (up to 91:9). Mechanistic experiments indicate an outer-sphere mechanism including metal-ligand cooperativity.

Introduction

The conversion of alkynes to alkenes plays an important role in the bulk and fine chemical industry as well as in basic chemical research.[1,2] Although numerous methods for the synthesis of alkenes have been developed in the past, the catalytic semihydrogenation of alkynes with dihydrogen is arguably one of the most efficient and atom economic approaches in this respect. However, it continues to be challenging due to the required control of stereo- ((Z)/(E)-isomers) and chemoselectivity (alkynes vs alkenes).

Typically, in this transformation heterogeneous systems based on noble metals, in particular Pd,[3,4] are utilized such as the well-known Lindlar’s catalyst (Pb-poisoned Pd/CaCO3).[5] With growing importance of sustainability and resource scarcity, over the past decade many efforts were made to develop catalysts derived from more available and cheaper non-noble metals, especially from the first-row transition metals.[6] More specifically, both heterogeneous and homogeneous systems based on Cr,[7] Fe,[8−11] Mn,[12,13] Co,[14−16] Ni,[17−19] and Cu[20] have been studied for this transformation.

For homogeneous, non-noble metal systems, the first report dates back to 1989, when Bianchini and co-workers discovered that iron(II) complex I with a tetradentate phosphine ligand is capable of selectively hydrogenating terminal alkynes to the corresponding alkenes (Figure ).[8,9] In 2013, the group of Milstein described acridine-based PNP iron(II) pincer complex II with an amidoborane coligand for the (E)-selective semihydrogenation of alkynes.[10] Initially, in this process the (Z)-alkene is formed which is rapidly isomerized to its (E)-isomer. Later, Fout and co-workers reported cobalt(I) dihydrogen complex III bearing a CCC pincer ligand with two NHC moieties for the (E)-selective semihydrogenation of a broad scope of alkynes.[14] As in the case of iron complex II, the (Z)-alkene is formed first and then isomerized. Cationic PNP iron(II) complex IV published by the group of Kirchner efficiently hydrogenates internal alkynes to (Z)-alkenes under mild conditions.[11] Recently, our group published the first homogeneous manganese complex for the semihydrogenation of alkynes.[12] Here, PNP manganese(I) complex V reduces diphenylacetylene derivates under mild conditions with excellent (Z)-selectivity. Mechanistic investigations revealed that hydrogenation proceeds via an outer-sphere mechanism utilizing the amino moiety in the ligand backbone for metal–ligand cooperativity[21] as it is often observed for these systems. No isomerization of the formed alkene takes place under the chosen conditions. Shortly after, Rueping and co-workers reported a similar cationic PNS manganese(I) complex VI.[13] This air-stable complex hydrogenates a variety of alkynes to the (Z)-alkenes under mild conditions.

Figure 1

Selected examples of homogeneous, non-noble metal (pre)catalysts for the semihydrogenation of alkynes.

Besides first-row transition metals, molybdenum also represents an attractive substitute for noble metals due to its low costs and toxicity.[22] Indeed, heterogeneous molybdenum catalysts are widely used for hydrogenation reactions[23] and homogeneous molybdenum complexes were studied for hydrogenations, particularly of N2 and CO2.[24−31] However, reports on catalytic hydrogenations with molybdenum complexes remain scarce in comparison to first-row transition metals.[32−38]

Our group has recently developed a series of structurally related low-valent molybdenum complexes as catalysts for the hydrogenation of ketones,[39] alkenes,[39] formamides,[40] and nitriles.[41] Based on this work, we became interested to explore their potential as catalysts for the semihydrogenation of alkynes.

Results and Discussion

Synthesis and Characterization of Mo Complexes

Initially, a series of low-valent molybdenum complexes have been prepared starting from Mo(CH3CN)2(CO)2(PPh3)2 (1) and the corresponding PNP ligands (Scheme , Figure ) according to previous protocols.[39−41] Sterically low demanding substituents (Ph, Et) on the phosphines of the pincer ligand lead to complexes with a facial coordination mode of the PNP ligand. In these complexes a PPh3 ligand remains besides the two strongly bound CO ligands on the Mo atom.

Scheme 1

General Synthesis of Low-Valent PNP Molybdenum Complexes Used in This Work

Figure 2

Series of molybdenum complexes tested in this work as catalysts for the semihydrogenation of alkynes.

Increased steric demand on the phosphines (iPr, Cy) causes a meridional arrangement of the PNP ligand, leading to the formation of more classical pincer complexes. Due to the structure of 1, in which the weak-field CH3CN ligands are located trans to the strong-field CO ligands, a meridional coordination of the PNP ligand inevitably results in the preservation of one CH3CN ligand in the complex sphere. If the synthesis of these complexes is performed in DCM as a solvent, chlorinated molybdenum(I) complexes are obtained, in which the neutral CH3CN ligand is replaced by a formally anionic Cl ligand.[39,41] The facially coordinated Ph-substituted complex 3 is more stable to chlorination and therefore can be prepared in DCM at room temperature. However, refluxing in DCM yields a heptacoordinated Mo(II) complex bearing two Cl ligands.[41] The reaction of the PNPEt ligand and 1 in DCM leads to product mixtures even at −20 °C.[41]

The new complex 6a was prepared according to Scheme . In contrast to the known iPr- and Cy-substituted congeners, the conversion is not complete after stirring a solution of 1 and the PNP ligand at room temperature overnight, but increased temperature and/or longer reaction times were necessary (e.g., 40 °C for 24 h). Compound 6a is sparingly soluble in common solvents (toluene, THF, acetonitrile, methanol, DMSO), and therefore, no crystals of the complex nor a meaningful NMR spectrum was obtained. Hence, 6a could only be analyzed by IR spectroscopy and elemental analysis. The IR spectrum of 6a features strong CO absorption bands at similar energies as observed for its iPr- and Cy-substituted congeners 4a and 5a. However, no CN-stretch of a potential CH3CN ligand was detected. Elemental analysis of the obtained material also pointed toward the absence of a CH3CN ligand. This might be rationalized by the high steric demand of the tBu-groups leaving little space in the coordination sphere.

Scheme 2

Synthesis of tBu-Substituted PNP Molybdenum Complexes 6a, 6b, and 6c

If compound 1 is reacted with the PNP ligand in DCM or 6a is dissolved in this solvent, formation of another complex was observed by IR spectroscopy, which proofed as a useful tool due to the strong CO absorption bands that all these complexes exhibit (see Supporting Information (SI) for details). However, attempts to isolate this compound in analytical purity remained unsuccessful. Most likely, molybdenum(I) complex 6b is formed, as can be concluded from the observed reactivity of the analogous complexes 4a and 5a in DCM and the similarity of the IR spectrum of the formed complex with these of complexes 4b and 5b. In solution (THF, DCM, DMSO) 6a and 6b slowly form the tri(carbonyl)complex 6c and other unidentified species.

Complex 6c could be separately prepared by refluxing Mo(CO)6 with a slight excess of PNP in toluene for 19 h. Furthermore, 6c is obtained by the exposure of a suspension of complex 6a in THF to CO gas, visible by a color change from brown red to yellow in a couple of minutes (see SI for images). The complex crystallizes as fine yellow needles from DCM/heptane, which were suitable for single crystal X-ray diffraction analysis. In the solid state of 6c (Figure ) the PNP ligand coordinates meridionally to the molybdenum center, forcing two CO ligands in a trans position. This leaves the complex in a strongly distorted octahedral geometry with a C21–Mo1–C23 angle of 153.79(8)°, far from linearity.

Figure 3

Molecular structure of 6c in the solid-state. Displacement ellipsoids set at 30% probability level, carbon-bound hydrogen atoms and lower occupied atoms of the disordered tBu-group are omitted for clarity. Selected bond lengths [Å] and angles [deg]: Mo1–N1 2.3582(14), Mo1–P1 2.5182(4), Mo1–C21 1.9963(19), Mo1–C22 1.9289(18), Mo1–C23 2.0042(19), C21–O1 1.166(2), C22–O2 1.179(2), C23–O3 1.161(2), N1–Mo1–C21 93.88(7), N1–Mo1–C22 168.91(7), N1–Mo1–C23 112.31(6), P1–Mo1–P2 154.23(2), C21–Mo1–C23 153.79(8).

The meridional structure of 6c contrasts with the solid-state structure of the iPr-substituted tri(carbonyl)molybdenum complex 4c, which was found to have a facial ligand geometry.[39] The NMR spectra of complexes 4c, 5c, and 6c show the presence of two different species whose ratio is solvent dependent.[42] Likely, in solution an equilibrium exists between the facial and the meridional complexes.

The resonance of the tBu-groups of 6c in the 1H NMR spectrum shows a complex coupling pattern due to 1H–31P coupling. Decoupling of 31P simplifies the resonances to two singlets in accordance with the C symmetry of the complex.

Next to Mo PNP pincer compounds, the coordination of NNP ligands to molybdenum was studied. Complex (rac)-7 can easily be prepared by stirring 1 with a slight excess of the NNP ligand in THF at room temperature. In contrast to the symmetric PNP pincer complexes, which all feature a mirror plane, (rac)-7 exhibits helical chirality due to the unsymmetric NNP ligand. In the solid state, the NNP ligand coordinates facially to the molybdenum center with the phosphine moiety being located trans to the one remaining PPh3 ligand (Figure ). The facial structure can be explained by the low steric demand of the N-methylimidazolyl group (compare Scheme ). Interestingly, in (rac)-7 the PPh3 ligand is located trans to the phosphine of the NNP ligand, whereas in 2 and 3 it opposes the amino functionality of the PNP ligand. However, heating a solution of 3 in toluene-d8 to 80 °C leads to a mixture of complexes as indicated by NMR experiments. One of the formed complexes likely is a complex (iso)-3 in which the PPh3 ligand is opposing a phosphine of the PNP ligand in analogy to the solid-state structure of (rac)-7 (see SI for details). Also, resonances that can be assigned to meridional complexes were observed, indicating that fac-mer isomerization takes places at elevated temperatures. Due to the helical chirality of (rac)-7, protons bound to the same carbon atom in the methylene and ethylene linker in the ligand backbone as well as the two iPr-groups are diastereotopic to each other. Together with the limited solubility of (rac)-7, this leads to a low intensity of the resonances of these protons in the 1H NMR spectrum. As expected, the 31P NMR spectrum of (rac)-7 in CD3CN exhibits two doublets at 67.24 and 55.54 ppm, respectively. Furthermore, some free PPh3 is observed which originates from the complex, since no resonance for free PPh3 is found for a suspension of (rac)-7 in toluene-d8, in which (rac)-7 itself is insoluble. This implies that PPh3 is not strongly bound in (rac)-7.

Figure 4

Molecular structure of (rac)-7 in the solid-state. Displacement ellipsoids set at 30% probability level; carbon-bound hydrogen atoms omitted for clarity. Selected bond lengths [Å] and angles [deg]: Mo1–N1 2.2619(15) Mo1–N3 2.3839(14), Mo1–P1 2.4519(5), Mo1–P2 2.4347(5), Mo1–C14 1.9203(18), Mo1–C15 1.9202(17), C14–O1 1.190(2), C15–O2 1.185(2), N1–Mo1–N3 72.92(5), P1–Mo1–P2 168.83(2), N1–Mo1–C14 176.41(6), N3–Mo1–C15 169.17(6).

The IR spectrum of (rac)-7 exhibits two strong CO absorption bands at 1778 and 1697 cm–1. The very low energy of the absorptions—being in the area typically observed for C=O double bonds—displays the strong π-backbonding to the CO ligands and the electron richness of the complex. For comparison, complexes 2 and 3 show CO absorptions at higher energies (1812, 1730 cm–1 and 1827, 1743 cm–1, respectively)[41] which illustrates the weaker electron-donating ability of the phosphines compared with the N-methylimidazolyl group. Liu and co-workers observed similar ligand properties with analogous manganese complexes.[43]

Catalytic Hydrogenation of Diphenylacetylene with Low-Valent Molybdenum Complexes

We started our catalytic experiments by investigating the different molybdenum(0) and molybdenum(I) complexes as catalysts for the hydrogenation of alkynes (Table ). Diphenylacetylene was reacted as a model substrate under 30 bar of H2 pressure at 80 °C using 5 mol % catalyst and 5 mol % NaHBEt3 as base, since NaHBEt3 has significantly affected catalyst activity in previous works.[39−41]

Table 1

Molybdenum-Catalyzed Hydrogenation of Diphenylacetylenea

Entry	[Mo]	Convb	Yield 9a (%)	(Z)/(E)	Yield 10a (%)
1	2	17	15	36:64	2
2	3	54	51	57:43	3
3	4a	68	65	75:25	3
4	4b	14	7	76:24	7
5	5a	93	89	85:15	4
6	5b	7	2	n.d.	5
7	6a	1	1	n.d.	0
8	6c	3	2	n.d.	1
9	(rac)-7	19	15	77:23	4
10	1	21	13	33:67	8
11	none	1	1	n.d.	0

Reaction conditions: 0.5 mmol of 8a, a 0.5 M solution of NaHBEt3 in toluene and 2 mL of toluene were used.

Conversion, yield, and (Z)/(E) ratio were determined by GC analysis using hexadecane as internal standard.

The facial complexes 2 and 3 showed mediocre conversions and diastereoselectivities (Table , entries 1–2), whereas the meridional complexes 4a and 5a performed better especially regarding the diastereoselectivity (Table , entries 3 and 5). Complex 5a gave the best results with nearly full conversion (93%) and a good diastereoselectivity of 85:15 in favor of the (Z)-diastereomer. In contrast, the chlorinated molybdenum(I) complexes 4b and 5b only showed low conversions.

Interestingly, the observed difference in catalytic activity between 4a and 5a was surprisingly large (93% to 68% conversion). Since the electronic properties of the iPr- and the Cy-substituents are quite similar, we suggest that the improved catalytic activity might be caused by the increased steric demand.

To prove this assumption the tBu-substituted PNP complexes 6a and 6c were tested in the semihydrogenation of diphenylacetylene. Unfortunately, both complexes, 6a and 6c, showed nearly no catalytic activity (Table , entries 7–8).

For related Mn-complexes, NNP-type ligands showed an improved catalytic activity in hydrogenation reactions in comparison to the PNP complexes.[43,44] Therefore, the respective Mo NNP pincer complex (rac)-7 (vide supra) was tested under our standard conditions. However, complex (rac)-7 showed moderate conversion and good diastereoselectivity (Table , entry 9). By applying complex 1, moderate conversion and diastereoselectivity were obtained (Table , entry 10), and by just using NaHBEt3 without a molybdenum complex, no conversion of diphenylacetylene was observed (Table , entry 11).

Variation of the Reaction Conditions and Substrate Scope of the Semihydrogenation

After identifying complex 5a as the most promising precatalyst, we investigated the influence of different reaction parameters such as solvents and bases on the outcome of the reaction (Table ). Polar solvents like THF, acetonitrile, or DCM led to poor conversions (<10%), whereas nonpolar, aprotic solvents like toluene, cyclohexane, and heptane are suitable for the reaction. Both toluene and cyclohexane provided high conversions and yields around 90% (Table , entries 1 and 3). In the base screening, NaHBEt3 presented the best results under the tested conditions (Table , entry 1). For NaHMDS a slightly lower conversion and yield were observed (Table , entry 7). The weaker bases NaHBH3, NaOH, and NaOtBu gave only low conversions (≤10%) (Table , entries 8–10), which are comparable with the results obtained without addition of base (Table , entry 11).

Table 2

Influence of Solvent and Base on the Molybdenum-Catalyzed Hydrogenation of Diphenylacetylenea

Entry	Deviation	Convb	Yield 9a (%)b	(Z)/(E)	Yield 10a (%)b
1	–	93	89	85:15	4
2	THF	9	4	68:32	5
3	heptane	46	44	73:27	2
4	acetonitrile	1	1	n.d.	0
5	DCM	3	1	n.d.	2
6	cyclohexane	92	87	86:14	5
7	NaHMDS	79	75	87:13	4
8	NaHBH3	9	5	58:42	4
9	NaOH	9	5	52:48	4
10	NaOtBu	10	4	45:55	6
11	no base	9	4	43:57	5

Reaction conditions: 0.5 mmol of 8a, a 0.5 M solution of NaHBEt3 in toluene and 2 mL of toluene were used.

Conversion, yield, and (Z)/(E) ratio were determined by GC analysis using hexadecane as internal standard.

Thus, a strong base seems to be needed to form the active catalyst species, likely an anionic amido complex which is generated by the deprotonation of the amine in the ligand backbone.[40] Interestingly, variation of the cation of the additive has a significant influence on the catalytic performance. While LiHBEt3 showed lower activity (32% conv, 23% yield, 84:16 (Z)/(E)) in comparison to NaHBEt3, KHBEt3 produced comparable conversion and yield but lower diastereoselectivity (84% conv, 79% yield, 73:27 (Z)/(E)). The influence of the cation is further supported by experiments in the presence of 15-crown-5, which is known to selectively bind potassium ions (see SI for details). A similar effect was found for the addition of THF.

Furthermore, catalyst loading, reaction temperature, and the dihydrogen pressure were investigated (Table ). Even a slight decrease of the catalyst loading resulted in a major loss in conversion and yield (Table , entries 2–3). At 60 °C the reaction proceeded slower, but by applying longer reaction times, full conversion could be reached (Table , entries 4–5). Under 20 bar of dihydrogen pressure, a slight decrease in conversion and yield is obtained, and lowering the pressure to 10 bar results in much lower conversion and yield (Table , entries 6–7). Conversion and yield could be enhanced by running the reaction at higher concentrations (Table , entry 8). During all these variations of parameters, the diastereoselectivity was not influenced.

Table 3

Influence of Catalyst Loading, Temperature and Dihydrogen Pressure on the Molybdenum-Catalyzed Hydrogenation of Diphenylacetylenea

Entry	x [mol %]	T [°C]	p [bar]	t [h]	Convb	Yield 9ab (%)	(Z)/(E)	Yield 10ab (%)
1	5.0	80	30	2	93	89	85:15	4
2	4.0	80	30	2	64	61	86:14	3
3	2.5	80	30	2	11	7	51:49	4
4	5.0	60	30	2	71	68	85:15	3
5	5.0	60	30	6	98	96	86:14	3
6	5.0	80	20	2	81	79	86:14	2
7	5.0	80	10	2	39	32	84:16	7
8c	5.0	80	30	2	98	96	86:14	4

Reaction conditions: 0.5 mmol of 8a, a 0.5 M solution of NaHBEt3 in toluene and 2 mL of toluene were used.

Conversion, yield, and (Z)/(E) ratio were determined by GC analysis using hexadecane as internal standard.

1 mL instead of 2 mL of toluene as solvent.

Next, various substituted diphenylacetylenes were reacted under the optimized reaction conditions (Scheme ). While a methyl substitution in the para- (8b) or meta-position (8c) on one of the phenyl rings was well tolerated and showed no significant difference compared to diphenylacetylene 8a, methylation in the ortho-position (8d) resulted in drastically lowered conversion but similar diastereoselectivity. Fluorine-containing substrate 8e was smoothly converted to the corresponding alkene 9e with good diastereoselectivity, whereas for the more reactive chloride 8f and bromide 8g, small amounts of hydrodehalogenation products were observed. Likely, the catalyst is deactivated by halogenation in these cases, resulting in lower conversions and yields.

Scheme 3

Molybdenum Catalyzed Semihydrogenation of Selected Alkyne Substrates

Reaction conditions: 0.5 mmol of 8, a 0.5 M solution of NaHBEt3 in toluene and 1 mL of toluene were used.

Conversion and (Z)/(E) ratio were determined by GC analysis.

Yield determined by GC analysis; isolated yield given in brackets.

60 °C, 16 h.

GC yield for all diastereomers, isolated yield for (Z, Z) diastereomer, diastereomeric ratio of (Z, Z) to other diastereomers.

Electron-rich substituted diphenylacetylenes (8h–k) including ethers, thioethers, and silanes were readily converted to the corresponding alkenes (9h–k). Under standard reaction conditions (80 °C, 2 h) in all cases small amounts (>10%) of alkane were observed. By applying lower temperatures at an increased reaction time (60 °C, 16 h), reduction to the alkane could mostly be suppressed and high yields (91–86%) as well as good diastereoselectivities (91:9–79:21) were obtained.

For diphenylacetylenes featuring electron-withdrawing substituents like nitriles and ketones as well as nitro, ester, or trifluoromethyl groups, only low conversions (≥10%) were observed (see SI for detailed results). Even when a strongly electron-donating methoxy group is present, a trifluoromethyl group on the other phenyl ring, like in substrate 8m, resulted in a low conversion. A possible explanation might be that the catalyst is inhibited by the substrates through the formation of stable alkyne complexes. A similar effect was observed for the pyridine derivative 8q (see SI). Diyne 8p featuring two directly connected alkyne functionalities resulted in poor conversion and yield (see SI), whereas substrate 8n could be easily converted to the corresponding diene with a selectivity of the (Z, Z)-diastereomer 9n to the other diastereomers of 86:14. Although sulfur-containing compounds often completely block hydrogenation catalysts, the thiophene-containing substrate 8o showed moderate conversion.

Besides aryl–aryl alkynes also terminal alkynes, as well as aryl-alkyl or alkyl–alkyl alkynes, were tested. Unfortunately, all these substrate classes showed only low conversions under the applied conditions in the presence of either 4a or 5a as catalyst (see SI).

Finally, a series of mechanistic control experiments were performed. To investigate if isomerization of the formed alkenes takes place during the reaction, (Z)- and (E)-stilbene were used as starting material, respectively (Scheme ). When (E)-stilbene was used, no significant isomerization and only traces of alkane were observed. In the case of (Z)-stilbene, formation of 2% of the more stable (E)-isomer and of the alkane were detected, respectively. This clearly shows that isomerization reactions are no major pathway under these conditions. Next, to examine whether metal–ligand cooperativity (MLC) is involved, the free NH-moiety in the ligand backbone of 4a was blocked by substitution. When the N-methylated derivative 4a-Me was applied under standard conditions, no significant hydrogenation occurred indicating an outer-sphere mechanism including MLC. Also, when 5a was used with 5 mol % of PPh3 as an additive under standard conditions (see SI for details), no decrease in catalytic activity was found, as would be expected for an inner-sphere mechanism due to the blocking of a vacant coordination site. This further supports an outer-sphere mechanism for the described hydrogenation reactions.

Scheme 4

Mechanistic Experiments Performed for the Molybdenum Catalyzed Semihydrogenation of Alkynes

Reaction conditions: 0.5 mmol of diphenylacetylene, (Z)-stilbene or (E)-stilbene as starting material, 5 mol % of molybdenum catalyst, a 0.5 M solution of NaHBEt3 in toluene, 30 bar of dihydrogen pressure and 1 mL of toluene as a solvent were used at 80 °C for 2 h.

Yields were determined by GC analysis using hexadecane as an internal standard.

Summary and Conclusions

Here, we describe novel homogeneous catalytic hydrogenation of alkynes using molybdenum complexes. The new tBu substituted Mo PNP pincer complexes 6a and 6c as well as the NNP pincer complex (rac)-7 have been prepared and characterized. These complexes as well as related complexes were tested as catalysts in the semihydrogenation of diphenylacetylene. The best performance was obtained in the presence of the cyclohexyl-substituted complex 5a PNPCyMo(CO)2(CH3CN). Utilizing this catalyst, various internal diaryl alkynes are hydrogenated to the corresponding alkenes with good to very good chemo- and diastereoselectivity for the (Z)-alkene and no significant isomerization taking place. However, the tolerance of this catalytic system toward substrates with electron-withdrawing substituents is limited, allowing for further improvement. Mechanistic experiments pointed toward an outer-sphere mechanism including MLC.

Experimental Section

General Information

All manipulations, except when indicated otherwise, were carried out under an argon atmosphere with exclusion of air and moisture using standard Schlenk and glovebox techniques. Solvents were dried over activated alumina columns using a solvent purification system (Innovative Technology PS-MD-6) and stored over molecular sieves (4 Å) under an argon atmosphere. Deuterated solvents were purchased from eurisotop, degassed by three successive freeze–pump–thaw cycles, and stored over molecular sieves (4 Å) under an argon atmosphere. NMR spectra were recorded on a Bruker Avance (300 MHz, 400 MHz) or Bruker Fourier (300 MHz) instrument. Chemical shifts (δ) are reported in parts per million (ppm) and are referenced to residual proton solvent signals or carbon resonances.[45,46] Infrared spectra were recorded in the solid state on a Nicolet iS 5 FT-IR spectrometer equipped with a PIKE Technologies GladiATR ATR. Elemental analyses were carried out in the Microanalysis Laboratory of the institute on a Leco TruSpec Micro CHNS device. GC analyses were carried out on an Agilent 7890A chromatograph using an HP-5 column (30 m × 0.25 m × 0.25 m). Bis[2-(diphenylphosphino)ethyl]ammonium chloride (15-7306), bis[2-(di-isopropylphosphino)ethyl]amine (15-7304), and bis[2-(dicyclohexylphosphino)ethyl]amine (15-7310) were purchased from Strem Chemicals and used without further purification. Bis[2-(di-tert-butylphosphino)ethyl]amine (AB393139) was purchased from abcr and used without further purification. Bis[2-(diethylphosphino)ethyl]amine[47] and 2-(di-isopropylphosphino)-N-((1-methyl-1H-imidazol-2-yl)methyl)ethylamine[48] were synthesized according to literature. NaHBEt3 was purchased from Sigma-Aldrich (227307). Mo(CO)6 was purchased from Sigma-Aldrich (577766) with a metal purity of ≥99.9%. The complexes 1,[49−51]2,[41]3,[41]4a,[39]4a-Me,[42]4b,[39]5a,[41] and 5b(41) were synthesized according to literature procedures.

Synthesis of Mo(CH3CN)2(CO)2(PPh3)2 (1)

The procedure was adapted from literature.[49−51] In a 500 mL three-necked round-bottom equipped with a large stirring bar and a reflux condenser, Mo(CO)6 (5.07 g, 19.2 mmol, 1.0 equiv) was suspended in 40 mL of acetonitrile and 40 mL of benzene and heated to reflux (95 °C oil bath temperature) for 4 h. The reaction mixture was allowed to cool to room temperature, and allylbromide (2.32 g, 19.2 mmol, 1.0 equiv) was added dropwise resulting in a color change from yellow to red. The mixture was heated to reflux for another 19 h, and an orange solid precipitated upon cooling to room temperature. The solution was filtered off at 0 °C, and the residue was suspended in 60 mL of acetonitrile. Triphenylphosphine (15.11 g, 57.61 mmol, 3.0 equiv) was added in one portion resulting in a color change to red. Upon heating to reflux for 2 h, a yellow solid precipitated. After cooling to room temperature, the solution was filtered off and the residue was thoroughly washed with acetonitrile (3 × 20 mL) and dried in vacuo, yielding 1 as a light-yellow solid (9.95 g, 68%). 31P{1H} NMR (161.99 MHz, DMSO-d6, 295 K): δ [ppm] = 54.75 (s), 52.04 (s), 50.29 (s), 36.77 (s) 25.46 (s). Multiple resonances are observed because of ligand exchange with DMSO. Therefore, a resonance at −6.9 ppm is observed which is assigned to free PPh3. A clear assignment of the resonances to the different formed complexes was not performed. Data are still provided for comparison. IR (ATR): ν [cm–1] = 1806 (CO), 1734 (CO).

Synthesis of PNPCyMo(CO)3 (5c)

In a 25 mL Schlenk tube Mo(CO)6 (150.0 mg, 568 μmol, 1.0 equiv) and bis[2-(dicyclohexylphosphino)ethyl]amine (277.8 mg, 596.6 μmol, 1.05 equiv) were dissolved in 10 mL of toluene. The reaction mixture was heated to reflux for 19 h. The orange suspension was allowed to cool to room temperature. The solvent was filtered off, and the residue was washed with toluene (2 × 2 mL) and heptane (2 × 2 mL). The light-yellow solid was dried in vacuo (311.0 mg, 89%). 1H NMR (400.13 MHz, CD2Cl2, 295 K): δ [ppm] = 3.22–3.05 (m, 2H), 2.77–1.10 (m, 49H), 0.84–0.68 (m, 2H). 31P{1H} NMR (161.99 MHz, CD2Cl2, 295 K): δ [ppm] = 65.67 (s), 41.43 (s). Elemental Analysis Calcd for C31H53MoNO3P2: C, 57.67; H, 8.27; N, 2.17. Found: C, 57.57; H, 8.64; N, 1.99. IR (ATR): ν [cm–1] = 1897 (CO), 1795 (CO), 1761 (CO).

Synthesis of PNPMo(CO)2 (6a)

In a 25 mL Schlenk tube Mo(CH3CN)2(CO)2(PPh3)2 (199.9 mg, 263 μmol, 1.0 equiv) was suspended in 5 mL of THF, bis[2-(di-tert-butylphosphino)ethyl]amine (10 wt % in toluene, 100.0 mg, 277 μmol, 1.05 equiv) was added, and the reaction mixture was heated at 40 °C for 20 h. The reaction mixture was allowed to cool to room temperature, the solvent was filtered off, and the red-brown residue was washed with THF (4 × 2 mL). Drying in vacuo yielded 6a as a red solid (145.1 mg, quant.). Elemental Analysis Calcd for C22H45MoNO2P2: C, 51.46; H, 8.83; N, 2.73. Found: C, 52.48; H, 8.59; N, 1.78. Although these results are outside the range viewed as establishing analytical purity, they are provided to illustrate the best values obtained to date. IR (ATR): ν [cm–1] = 1769 (CO), 1677 (CO).

Synthesis of PNPMo(CO)3 (6c)

From 6a: In a 25 mL Schlenk tube 6a (110 mg, 214 μmol, 1.0 equiv) was suspended in 5 mL of THF. CO gas was introduced into the suspension, and a color change from brown red to yellow was observed within minutes. After 2 h, all volatiles were removed in vacuo yielding 6c as a yellow solid (116 mg, quant.). From Mo(CO): In a 25 mL Schlenk tube Mo(CO)6 (148.3 mg, 562 μmol, 1.0 equiv) was dissolved in 10 mL of toluene and bis[2-(di-tert-butylphosphino)ethyl]amine (10 wt % in THF, 213.0 mg, 590 μmol, 1.05 equiv) was added. The reaction mixture was heated to reflux for 19 h. The red-brown suspension was allowed to cool to room temperature. The solvent was filtered off, and the residue was washed with toluene (2 × 1 mL). The residue was dissolved in DCM, and a yellow solid precipitated by adding heptane. The solvent was filtered off, and the residue was washed with heptane (2 × 2 mL). Drying in vacuo yielded 6c as a yellow solid (190 mg, 63%). Yellow needles suitable for single-crystal X-ray diffraction were obtained by slow diffusion of heptane into a solution of 6c in DCM at 0 °C. 1H NMR (300.20 MHz, CD2Cl2, 295 K): δ [ppm] = 3.40–3.20 (m, 2H), 2.46–2.19 (m, 3H), 2.18–2.05 (m, 2H), 1.60–1.41 (m, 2H), 1.39–1.30 (m, 36H). 1H{31P} NMR (400.13 MHz, CD2Cl2, 295 K): δ [ppm] = 3.34–3.27 (m, 2H), 2.43–2.21 (m, 3H), 2.15–2.07 (m, 2H), 1.55–1.44 (m, 2H), 1.35 (s, 18H), 1.34 (s, 18H). 31P{1H} NMR (121.52 MHz, CD2Cl2, 295 K): δ [ppm] = 101.96 (s). Elemental Analysis Calcd for C23H45MoNO3P2: C, 51.01; H, 8.38; N, 2.59. Found: C, 51.08; H, 8.45; N, 2.41. IR (ATR): ν [cm–1] = 1904 (CO), 1782 (CO), 1761 (CO).

Synthesis of NNPMo(CO)2(PPh3) ((rac)-7)

In a 25 mL Schlenk tube Mo(CH3CN)2(CO)2(PPh3)2 (569.3 mg, 750 μmol, 1.0 equiv) was suspended in 15 mL of THF, 2-(di-isopropylphosphino)-N-((1-methyl-1H-imidazol-2-yl)methyl)ethylamine (10 wt % in toluene, 210.8 mg, 826 μmol, 1.1 equiv) was added, and the reaction mixture was stirred at room temperature for 22 h. The solvent was filtered off, and the yellow residue was washed with hexane (5 × 5 mL). Drying in vacuo yielded (rac)-7 as a yellow solid (403 mg, 80%). Red crystals suitable for single-crystal X-ray diffraction were obtained by slow diffusion of Et2O into a solution of (rac)-7 in acetonitrile at 0 °C. 1H NMR (300.20 MHz, CD3CN, 295 K): δ [ppm] = 7.61–7.52 (m, 6H), 7.33–7.21 (m, 9H), 6.62–6.58 (m, 2H), 3.32 (bs, 1H), 3.23 (s, 3H), 3.17–3.09 (m 1H), 2.86–2.61 (m, 3H), 2.40–2.56 (m, 1H), 1.82–1.68 (m, 1H), 1.39–1.17 (m, 10H), 1.09–0.99 (m, 4H). 31P NMR (121.52 MHz, CD3CN, 295 K): δ [ppm] = 67.24 (d, J = 141 Hz), 55.54 (d, J = 141 Hz). IR (ATR): ν [cm–1] = 1778 (CO), 1697 (CO).

General Procedure for Hydrogenation Experiments

All hydrogenation reactions were carried out in a 300 mL autoclave (Parr Instrument Company). In a glovebox a 4 mL glass vial was charged with the corresponding molybdenum catalyst and a stirring bar. Solvent and NaHBEt3 (0.5 M in toluene) were subsequently added, and the reaction mixture was stirred for approximately 10 min. The corresponding alkyne was added, and the vial was closed with a screw cap containing a septum. The septum of the vial was punctured with a needle to allow for the exchange of atmosphere, and the vial was transferred into an autoclave. The sealed autoclave was purged ten times with 10 bar of pressure of dihydrogen gas before the desired pressure was set. The autoclave was heated in a preheated aluminum block for the desired reaction time. Afterward, the autoclave was cooled in an ice bath and carefully depressurized. The reaction mixture was diluted with ethyl acetate, a known amount of hexadecane was added as an internal standard, and the mixture was filtered through a pad of Celite.