The Dichotomy of Mn–H Bond Cleavage and Kinetic Hydricity of Tricarbonyl Manganese Hydride Complexes

Abstract

Acid-base characteristics (acidity, pKa, and hydricity, ΔG°_H− or k_H−) of metal hydride complexes could be a helpful value for forecasting their activity in various catalytic reactions. Polarity of the M–H bond may change radically at the stage of formation of a non-covalent adduct with an acidic/basic partner. This stage is responsible for subsequent hydrogen ion (hydride or proton) transfer. Here, the reaction of tricarbonyl manganese hydrides mer,trans–[L₂Mn(CO)₃H] (1; L = P(OPh)₃, 2; L = PPh₃) and fac–[(L–L′)Mn(CO)₃H] (3, L–L′ = Ph₂PCH₂PPh₂ (dppm); 4, L–L′ = Ph₂PCH₂–NHC) with organic bases and Lewis acid (B(C₆F₅)₃) was explored by spectroscopic (IR, NMR) methods to find the conditions for the Mn–H bond repolarization. Complex 1, bearing phosphite ligands, features acidic properties (pKa 21.3) but can serve also as a hydride donor (ΔG^≠_298K = 19.8 kcal/mol). Complex 3 with pronounced hydride character can be deprotonated with KHMDS at the CH₂–bridge position in THF and at the Mn–H position in MeCN. The kinetic hydricity of manganese complexes 1–4 increases in the order mer,trans–[(P(OPh)₃)₂Mn(CO)₃H] (1) < mer,trans–[(PPh₃)₂Mn(CO)₃H] (2) ≈ fac–[(dppm)Mn(CO)₃H] (3) < fac–[(Ph₂PCH₂NHC)Mn(CO)₃H] (4), corresponding to the gain of the phosphorus ligand electron-donor properties.

1. Introduction

A search for inexpensive and efficient catalysts that give more sustainable alternatives to platinum group metals recently led to remarkable progress in the development of catalytic systems based on organometallic manganese complexes [1,2]. While pincer-type Mn(I) derivatives still dominate in the field of catalytic (de)hydrogenation, it was demonstrated that less elaborated bidentate systems fac–[(L–L′)Mn(CO)₃Br] may also be highly efficient [3]. Generally, the formation of catalytically relevant transition metal hydrides from the corresponding bromide precursors proceeds in situ in the presence of various basic additives. This activation step can be nicely illustrated for Mn(I) complexes fac–[(P–NHC)Mn(CO)₃Br] (P–NHC = κ²P,C–Ph₂PCH₂–NHC) [4] and fac–[(dppm^R)Mn(CO)₃Br] (dppm = κ²P,P–Ph₂PCH(R)PPh₂, R = H, Me, Ph) [5] in the presence of KHMDS leading to the formation of cyclometalated species capable of activating dihydrogen via unconventional metal-ligand cooperation. The resulting hydride products fac–[(L–L′)Mn(CO)₃H] interact with organic substrates to form non-covalent adducts and typically the formation of such intermediates directly precedes hydride/proton transfer steps providing in fine the hydrogenation of polar C=X bonds.

It is generally accepted that transition metal hydrides can be proton, hydride or hydrogen atom donors [6,7,8]. However, the vast majority of hydride complexes exhibit only one reactivity mode being determined by the nature of auxiliary ligands and relative charge on the core metal atom [9]. Nevertheless, there are a few families of hydrides in which the same hydride complex possesses distinct reactivity. Dual reactivity is well documented, for example, for groups 6–8 metal complexes [CpM(CO)₃H] (M = Mo, W), [(CO)₅MH] (M = Mn, Re) and [CpM(CO)₂H] (M = Fe, Ru, Os) [8,10,11,12]. Whether the M–H bond releases a proton or hydride, obviously, depends on the partner reagent. We have shown for the first time that the M–H bond polarity gets adjusted at the stage of a non-covalent complex formation preceding the M–H bond dissociation [13] (Scheme 1).

The acidity and hydricity of the M–H bond can be quantified in terms of pK_a and ΔG°_H−, respectively [14,15,16]. The quantitative scale of kinetic hydricity (k_H−) has been developed for group 6–7 metal hydrides in the pioneering work of Bullock [11]. However, the data for manganese complexes are still scarce despite their utility as potential catalysts. Experimental values of acidity and hydricity are known only for [(CO)₅MnH] (pK_a = 14.2 in MeCN [8], ΔG°_H− = 59.61 kcal/mol [17], k_H− = 50 M⁻¹s⁻¹ [11]), [(CO)₄(C₆H₆)MnH] (pK_a = 26.8 in MeCN [8]) and cis–[(CO)₄(PPh₃)MnH] (pK_a = 20.4 in MeCN [8], k_H− = 230 M⁻¹s⁻¹ [11]). There is a sole theoretical work devoted to a systematic study of (CO)₅MnH and (CO)_5-n(PH₃)_nMnH complexes [17] demonstrating that the hydricity of the metal hydride bond is greatly amplified when the CO ligand is replaced by a phosphine donor [17].

The research presented herein explores the dichotomy of Mn–H bond cleavage and the potential of its repolarization entailed by intermolecular interactions with Lewis acids and bases. By gaining an understanding of the acid-base characteristics of different manganese hydrides, it should be possible to forecast their catalytic performance and drive the catalyst design.

In this context, we chose two types of octahedral Mn(I) hydride complexes (Scheme 2) in which the Mn–H bond is expected to have different polarity. Electron-deficient phosphite ligands in complex mer,trans–[(P(OPh)₃)₂Mn(CO)₃H] (1) and more donating triarylphosphines in complex mer,trans–[(PPh₃)₂Mn(CO)₃H] (2) should provide them an acidic and basic character, respectively. In agreement with our expectations, complex 1 can be deprotonated by tBuOK [18], and complex 2 exhibits typical reactivity of “hydridic” hydride releasing H₂ [19,20] upon protonation by HBF₄·Et₂O [21]. The incorporation of chelating and more electron-rich ligands such as bidentate phosphine in fac–[(dppm)Mn(CO)₃H] (3) or phosphine–N–Heterocyclic carbene in fac–[(P–NHC)Mn(CO)₃H] (4) ought to increase the Mn–H basicity and hydride donating ability (Scheme 2). The acidity and hydricity of 1–4 have not been evaluated before this work; therefore, the entire set of these complexes could be used to create the scale of their thermodynamic/kinetic hydricity by evaluating the impact of the ligand.

2. Results and Discussion

2.1. Interaction of Tricarbonyl Manganese Hydrides 1–4 with Bases

Hydrogen bonding. Addition of Lewis bases (LB) such as pyridine and hexamethylphosphoramide (HMPA) to complex mer,trans–[(P(OPh)₃)₂Mn(CO)₃H] (1) in methylcyclohexane (MCH) at 190 K leads to weak hydrogen bond formation where Mn–H serves as a proton donor [22,23,24,25]. Hydrogen bonding was evidenced by the appearance of a low-frequency shoulder at the initial ν_CO 1958 cm⁻¹ band (Δν 5–20 cm⁻¹) (Figure S1). Since hydrogen bonding is a reversible process, a temperature increase up to 290 K shifts the equilibrium (Scheme 1, path a) towards free manganese hydride. Complex fac–[(dppm)Mn(CO)₃H] (3) also forms a hydrogen bond with these bases; its three ν_CO bands (1996, 1916, 1909 cm⁻¹) shift to lower frequencies by 3–7 cm⁻¹ in the presence of 70 equiv. HMPA (Figure S2). However, the strength of these bases is insufficient for proton transfer to occur. The change of the nonpolar MCH to polar acetonitrile also does not promote the proton transfer from 3 to pyridine (pK_a = 12.53 in MeCN) or HMPA (pK_a = 6.1 in CH₃NO₂).

Proton transfer. Quantitative deprotonation of 1 was observed in acetonitrile when a stronger base—1,8–diazabicyclo[5.4.0]undec–7–ene (DBU; pK_a = 24.31 in MeCN)—was used. Full conversion of 1 to anionic complex [(P(OPh)₃)₂Mn(CO)₃]⁻[HDBU]⁺ (1⁻) was confirmed by ³¹P NMR in CD₃CN, since the initial signal at δ_P 183.3 ppm converts into downfield resonance δ_P 206.5 ppm after DBU addition (the key spectral parameters for this and all other complexes studied are summarized in

Table 1. IR and ³¹P{¹H} NMR data for neutral, cationic and anionic manganese species 1–4 in different solvents.

Complex	νCO (cm−1)	δP (ppm)
1, [(P(OPh)3)2Mn(CO)3H]	2040 w, 2028 w, 1956 s a 2041 w, 2028 w, 1955 s b 2043 w, 2031 w, 1958 s c	183.3
1+, [(P(OPh)3)2Mn(CO)3][HB(C6F5)3]	2011 s, 1969 s a	153.5
1−, [(P(OPh)3)2Mn(CO)3][HDBU]	1815 s c	206.5
2, [(PPh3)2Mn(CO)3H]	1908 s, 1900 s f	80.5
2+, [(PPh3)2Mn(CO)3]+[HB(C6F5)3]−	1980 s, 1930 s f	61.8
2−, [(PPh3)2Mn(CO)3]−[K+]	1770 s, 1741 s e	-
3, [(dppm)Mn(CO)3H]	1996 s, 1916 s, 1909 s a 1993 s, 1914 s, 1903 s b	30.1 a 31.8 b
fac-3+, [(dppm)Mn(CO)3)][HB(C6F5)3] [27]	2040 s, 1973 s, 1935 s d	10.1
mer-3+, [(dppm)Mn(CO)3)][HB(C6F5)3] [27]	2060 s, 2003 s d	10.6, 7.1
3CH−, fac–[(CH−-dppm)Mn(CO)3H][K+]	1957 s, 1871 s, 1876 s e 1956 s, 1870 s b	10.9 a
3Mn−, [(dppm)Mn−(CO)3][K+]	1867 s, 1779 s b	29.9 b
4, fac–[(P–NHC)Mn(CO)3H]	1989 s, 1909 s, 1889 s a 1988 s, 1907 s,1888 s d	95.8
fac-4+, [(P–NHC)Mn(CO)3][HB(C6F5)3]	2032 s, 1949 s, 1921 s d	78.1, 71.1 f
mer-4+, [(P–NHC)Mn(CO)3][HB(C6F5)3]	2038 s, 1968 s, 1942 s d	-

^a toluene, ^b MeCN, ^c MCH, ^d nBuCl, ^e THF, ^f CH₂Cl₂.

). To explore the features of proton transfer equilibrium (Scheme 1, path a), we decided to return into a non-polar solvent that should impede proton transfer process. Indeed, in methylcyclohexane the proton transfer is temperature dependent and starts when the reaction mixture is cooled below 260 K. According to IR spectra, the addition of 1.1 equiv. DBU to a solution of 1 in MCH at low temperatures (190–260 K) leads to the appearance of anionic manganese species [(P(OPh)₃)₂Mn(CO)₃]⁻ (1⁻, ν_CO 1815 cm⁻¹) at the expense of the initial complex (ν_CO 1958 cm⁻¹). Unexpectedly, proton transfer is slow at these temperatures, taking 1.2 h at 230 K to reach the equilibrium (Figure S3). The rate constants analysis gave the activation energies (See Supplementary Materials for more details) for the proton transfer step: ΔH^≠ = 7.5 kcal/mol, ΔS^≠ = −26 cal/(mol·K), ΔG^≠_298K = 15.3 kcal/mol. Since the system does come to equilibrium, the experimental equilibrium constants with thermodynamic parameters of proton transfer in MCH were obtained: ΔHᵒ = −13.1 kcal/mol, ΔSᵒ = −50 cal/(mol·K), ΔGᵒ_298K = 1.9 kcal/mol (See Supplementary Materials for more details). The reaction of 1 with a slightly weaker base [Bu₄N]⁺[4–NO₂C₆H₄O]⁻ (pK_a = 20.7 in MeCN [26]) immediately reaches equilibrium in MeCN with the formation of only ~30% proton transfer product (Figure S4). The known acidity constant of 4–NO₂C₆H₄OH allows an estimation of complex 1 acidity pK_a(1) = 21.3.

Manganese catalyzed transfer hydrogenation reactions are usually carried out in the presence of strong bases such as potassium tert-butylate or KHMDS [1,3,4]. A possible reaction mechanism includes Mn–H bond deprotonation with formation of anionic complex, which participates in further catalytic transformations [3]. We have tried multiple times to identify the conditions for proton abstraction from mer,trans–[(PPh₃)₂Mn(CO)₃H] (2) and fac–[(dppm)Mn(CO)₃H] (3), where the Mn–H bond is expected to be rather electron-rich due to the presence of phosphine ligands. However, no repolarization of the Mn–H bond and proton transfer was observed for hydride complexes 2 and 3 with various bases (pyridine, HMPA, DBU, TBD (1,5,7–triazabicyclo[4.4.0]dec–5–ene, pK_a = 26.03 in MeCN [28])) even in polar solvents (MeCN, THF). Fortunately, treatment of complex 2 at ambient temperature in THF with KHMDS excess (2 equiv., pK_a = 26 in THF [29]) led to partial proton transfer indicated by the appearance of two new low-frequency ν_CO bands at 1770 and 1741 cm⁻¹ in IR spectrum that correspond to anionic complex 2⁻ (Figure S5) [30]. From these spectral data the pK_a value 27.3 for manganese hydride 2 in THF was estimated.

The reaction of hydride complex 3 with KHMDS (5 equiv.) in THF leads to full proton transfer and the appearance of three new low-frequency-shifted CO bands (1957, 1871, 1876 cm⁻¹) in the IR spectra (Figure 1). The shift of ν_CO bands by ca. 40 cm⁻¹ in the IR spectra is not consistent with the presence of the negative charge at the metal atom in the case of Mn–H bond deprotonation. Moreover, ¹H NMR spectrum shows the presence of a triplet signal at δ_H −5.54 ppm in the hydride region, while for the initial complex 3 the hydride signal (δ_H −5.53 ppm) is a triplet of doublets due to the ²J_PH and an additional ⁴J_HH spin-spin interaction with one of the CH₂ bridge H-atoms. The phosphorus resonance observed in this 3/KHMDS mixture by ³¹P{¹H} NMR spectroscopy is shifted to stronger field (δ_P 10.9) relative to that (δ_P 30.1) of the initial manganese hydride 3 (Figure S6). These spectral data allow the proposal that deprotonation occurs at the bridging CH₂ group of dppm ligand with the formation of [3^CH−][K⁺]. In the ¹³C spectrum, the triplet (δ_C 20.8, t, ¹J_CP = 51.4 Hz, PCHP) of corresponding anionic carbon CH⁻ is also shifted to stronger field compared to the CH₂ signal of 3 (δ_C 48.0, t, ¹J_CP = 22.4 Hz, PCH₂P). The estimated pK_a value of methylene CH-proton in complex 3 equals 26 in THF. A more polar acetonitrile deprotonation of 3 by the same amount of KHMDS goes through [3^CH−][K⁺] formation yielding ultimately the anionic manganese complex [3^Mn−][K⁺] (Scheme 3). In the IR and NMR spectra, measured immediately after mixing, three manganese species could be observed; however, in 15 min, the intermediate [3^CH−][K⁺] (ν_CO 1956, 1870 cm⁻¹; δ_P 10.4 ppm) completely transforms into the final species [3^Mn−][K⁺] ((ν_CO 1867, 1779 cm⁻¹; δ_P 29.9 ppm) (Figure 1 and Figure S6) that has no hydride signal in the ¹H NMR spectrum. According to our DFT calculations, the intramolecular proton transfer from manganese to the anionic bridge CH⁻ is explained by thermodynamic preference of the metal deprotonated form (ΔG_298K = −0.2 kcal/mol) in acetonitrile compared with THF (

Table 2. Relative energies (in kcal/mol) of the complex [3^Mn−][K⁺] formation computed at DFT/ωB97XD/def2-TZVP level. Complex [3^Mn−][K⁺] is taken as a zero.

	Toluene	THF	MeCN
Eel	+2.8	−0.7	−3.1
ΔH	+5.7	+2.0	−0.2
ΔG	+7.6	+2.5	−0.2

). That implies that even manganese complexes with pronounced hydride character [11,17] can be deprotonated by KHMDS, thus allowing the corresponding anionic intermediates to participate in hydrogenation reactions.

2.2. Interaction of Tricarbonyl Manganese Hydrides with Lewis Acids

Non-covalent adducts. Interaction of manganese hydride complexes (1–4) with Lewis acids (LA) leads to Mn–H bond polarization, making the hydride atom more negatively charged. Recently, we have proven that the formation of non-covalent adducts between Lewis acid and hydride complex fac–[(dppm)Mn(CO)₃H] (3) at 180 K precedes the hydride ion abstraction [27] (Scheme 4). Noteworthy, while the initial hydride 3 has facial geometry, the most stable form of the non-covalent adduct is the isomer with the meridional geometry. The latter then undergoes additional transformations, including intramolecular hydride transfer with the formation of meridional cation stabilized by the solvent molecule mer–[(dppm)Mn(CO)₃(κ¹Cl–CH₂Cl₂)](HBAr₃) and its isomerization to a more stable facial isomer fac–[(dppm)Mn(CO)₃(κ¹Cl–CH₂Cl₂)](HBAr₃).

The analogous chemical behavior was observed upon addition of B(C₆F₅)₃ to complex fac–[(P–NHC)Mn(CO)₃H] (4) at 180 K in methylene chloride. The Lewis acid coordination to the hydride ligand and formation of the non-covalent adduct 4···LA gives in ¹H NMR spectrum a broadened resonance at δ_H −7.98 ppm shifted high-field relative to the initial signal of 4 (δ_H −7.42) (Figure 2). A weak BrØnsted acid mer,trans–[(P(OPh)₃)₂Mn(CO)₃H] (1) also interacts with Lewis acids as a hydride-ion donor. Addition of B(C₆F₅)₃ (10 equiv.) to the toluene solution of 1 (ν_CO 2040 w, 2028 w, 1956 s cm⁻¹) at 190 K leads to the formation of a non-covalent complex 1···B(C₆F₅)₃ (ν_CO 2035 s, 1973 s cm⁻¹). We assume that this adduct has a facial configuration, since its ν_CO bands are strongly shifted to high frequencies (cf. 2046 s, 1980 s cm⁻¹ for fac-1 and 2026 w, 1942 s cm⁻¹ for mer,trans-1 in nujol) [31]. Formation of non-covalent adduct with B(C₆F₅)₃ stabilizes fac-1 and in the 190–220 K temperature range the equilibrium between two species, mer,trans-1 and fac-1···B(C₆F₅)₃, is observed in the IR spectra (Figure S7). In ¹H NMR spectra of the 1 + B(C₆F₅)₃ mixture, measured under the same conditions, the formation of the intermediate complex fac-1⋯B(C₆F₅)₃ is evidenced by new broad hydride resonance shifted to the higher field (δ_H −9.36 ppm) (Figure S8, left). In ³¹P{¹H} NMR spectra (Figure S8, right), the resonance at δ_P 163 ppm exhibits the same temperature behavior and can be attributed to fac-1⋯B(C₆F₅)₃. Thus, complex 1 is the least reactive toward hydride transfer (see below) and its non-covalent adduct with Lewis acid exists in a wider temperature range.

Hydride transfer. The non-covalent adducts 1⋯LA–4⋯LA precede the intramolecular hydride transfer that yields the cationic complexes with different geometries, which are stabilized by coordination of solvent molecule or [H-LA]⁻ anion. Thus, for fac–[(P–NHC)Mn(CO)₃H] (4) interacting with B(C₆F₅)₃ at 180 K, new high-frequency ν_CO bands at 2038, 1968 and 1942 cm⁻¹ appear in the IR spectrum assigned to the cationic mer-intermediate; at 190 K they already transform into fac-product with ν_CO bands at 2032, 1949 and 1921 cm⁻¹ (Figure S9). In contrast to isostructural fac–[(dppm)Mn(CO)₃H] (3), in the case of complex 4 two configurations are possible when Mn(CO)₃ moiety is in meridional arrangement: hydride ligand being trans to the phosphorus atom (mer-P-4) or trans to the NHC carbon atom (mer-C-4) (Figure S8). Since complex mer-P-4 should feature more hydridic H-atom than mer-C-4 due to the higher trans-effect of phosphorus [32], we assumed that the hydride transfer occurs through the formation of the meridional trans-to-P form (mer-P-4⋯B(C₆F₅)₃).

Above 230 K, the equilibrium is completely shifted to cationic products, the ³¹P{¹H} spectrum contains only two resonances δ_P 78.0 and 71.0 ppm in 1:4 ratio (Figure 2). After 4 days at room temperature, this ratio inverts and becomes 2:1. By selectively decoupling ³¹P, the CH₂-bridge H-atoms δ_H 5.07 (dd, ²J_HH = 14.0, ²J_PH = 5.1 Hz) and 4.94 (dd, ²J_HH = 14.1, ²J_PH = 6.6 Hz) can be attributed to complex 4a with δ_P 78.1 ppm and δ_H 5.48 (vt, ²J_HH = 14.4 Hz, ²J_PH = 13.7 Hz) and 4.99 (d, ²J_HH = 14.4 Hz) to complex 4b δ_P 71.1. While the ³¹P{¹H} resonance of 4a shifts (between 74.1 and 78.1 ppm) in response to changing the solvent (toluene, nBuCl, C₆H₅Cl, CD₂Cl₂), the chemical shift of 4b keeps its position (Table S1). Thus, we suggest that the two cationic products are [(P–NHC)Mn(CO)₃(Solvent)][HB(C₆F₅)₃)] (4a), which has a coordinated solvent molecule, and [(P–NHC)Mn(CO)₃][HB(C₆F₅)₃)] (4b), which is a contact ionic pair stabilized by B–H bond interaction with the cationic metal center (4b) remains in tranquility. The sensitivity of the ³¹P{¹H} resonance of 4a to the media is indicative of the solvent’s influence on its structure. To obtain an independent proof of its structure, we reacted the manganese bromide fac–[(P–NHC)Mn(CO)₃Br] with AgBF₄ in MeCN as a solvent. The cationic complex 4^MeCN obtained features acetonitrile molecule coordinated to the metal center as it was confirmed by X-ray diffraction (Figure 3), and its ³¹P{¹H} resonance (77.5 ppm in CD₂Cl₂) is in the range of 4a-type complexes.

For the non-covalent complex fac-1⋯B(C₆F₅)₃ the hydride transfer occurs above 230 K (Scheme 5). The single product of the hydride transfer is a cationic manganese complex mer,trans–[1]⁺[HB(C₆F₅)₃]. The signal of the cationic product [1]⁺[HB(C₆F₅)₃]⁻ becomes clearly seen in ³¹P spectra at 250 K (δ_P 153.0 ppm).

Structurally similar complex mer,trans–[(PPh₃)₂Mn(CO)₃H] (2) bearing more donating phosphine ligands was found to be more reactive. Indeed, it transforms into mer,trans–[(PPh₃)₂Mn(CO)₃(solv)][HB(C₆F₅)₃] (2⁺) already at 180 K (Figure S11). IR spectra in the 180–250 K range show the transformation of hydride 2 directly into cationic complex [2⁺][HB(C₆F₅)₃] (ν_CO 1980, 1930 cm⁻¹) without detectable non-covalent adduct formation and mer-to-fac isomerization. At temperatures higher than 250 K, partial decomposition of this cationic species was observed, resulting in the formation of known tetracarbonyl complex trans–[(PPh₃)₂Mn(CO)₄]⁺ exhibiting a ν_CO band at 2002 cm⁻¹ [33]. ³¹P{¹H} NMR spectra confirm the formation of sole cationic product (δ_P 61.8 ppm) and its further decomposition. Direct synthesis of mer,trans–[(PPh₃)₂Mn(CO)₃(CH₃CN)]⁺[BF₄]⁻ 2^MeCN allows the formation of a stable cationic product with ν_CO at 2055 w, 1977 s, 1946 s cm⁻¹ characterized by X-ray diffraction (Figure 4).

2.3. Kinetic Hydricity of Manganese Tricarbonyl Complexes

The hydride transfer from complexes 1–4 to B(C₆F₅)₃ is relatively slow at low temperatures that allow the estimation of their kinetic hydricities (See Supplementary Materials for more details; Figures S12 and S13). The low-temperature IR monitoring of hydride transfer kinetics in the presence of 1.1 equiv. B(C₆F₅)₃ in nBuCl gave effective rate constants (k_eff). The value of free energy ΔG^≠_298K characterizes the kinetic hydricity of manganese complexes but depends on the Lewis acidity of the given hydride abstractor. As expected, the hydride complex 1, which is a weak acid due to electron-withdrawing phosphite ligands, has the lowest hydricity (ΔG^≠_298K = 19.8 kcal/mol), while complex 4 bearing phosphine-carbene ligand, the most electron-rich in this series, possesses the highest hydricity (ΔG^≠_298K = 16.5 kcal/mol) among the complexes studied (

Table 3. Effective rate constants for the interaction between 1–4 and B(C₆F₅)₃ in nBuCl at 220 K and activation parameters of hydride transfer.

MnH	keff220K, M−1·s−1	∆H≠, kcal/mol	∆S≠, cal/(mol·K)	ΔG≠220K, kcal/mol	ΔG≠298K, kcal/mol
1	0.00008	8.6 ± 0.2	−37 ± 1	16.9 ± 0.1	19.8 ± 0.1
2	0.147	4.1 ± 0.5	−43 ± 3	13.6 ± 0.3	17.0 ± 0.3
3	0.006	9.4 ± 0.6	−26 ± 3	15.0 ± 0.2	17.0 ± 0.2
4	0.706	3.8 ± 0.2	−43 ± 1	13.2 ± 0.1	16.5 ± 0.1

). The kinetic hydricity of the manganese complexes of interest increases in the order mer,trans–[(P(OPh)₃)₂Mn(CO)₃H] (1) < mer,trans–[(PPh₃)₂Mn(CO)₃H] (2) ≈ fac–[(dppm)Mn(CO)₃H] (3) < fac–[(P-NHC)Mn(CO)₃H] (4), reflecting the gain of the phosphorus ligand electron donor properties (basicity). On the activation free energy scale (ΔG^≠_298K) there is only small by 0.5 kcal/mol difference between complexes 2–4 bearing electron-donating ligands. However, the distinction between their properties becomes more pronounced at lower temperatures due to the impact of highly negative activation entropy ∆S^≠ (cf. k_{eff 220K}, Table 3).

3. Materials and Methods

All reactions were performed using standard Schlenk procedures under a dry argon atmosphere. Dry and oxygen-free organic solvents (toluene, THF) were obtained using a solvent purification system from M. Braun (Garching, Germany). Methylcyclohexane was dried over Na and distilled under an argon atmosphere. nBuCl, CH₂Cl₂ and acetonitrile were stored over CaH₂ and distilled under an argon atmosphere before use. Deuterated solvents for NMR were degassed before use by three freeze–pump–thaw cycles and kept over 3Å molecular sieves. A liquid nitrogen/ethanol or nitrogen/isopropanol slush bath was used to maintain samples at the desired low temperature.

Variable-temperature (VT) NMR spectra were recorded on Bruker Avance 300, Bruker Avance 400 (Bruker, Billerica, MA, USA) and Varian Inova 400 (Varian, Palo Alto, CA, USA) spectrometers operating at 300 and 400 MHz in the 180–300 K temperature range. ¹H and ¹³C{¹H}, chemical shifts are reported in parts per million (ppm) downfield of tetramethylsilane (TMS) and were calibrated against the residual resonance of the deuterated solvent, while ³¹P{¹H} chemical shifts were referenced to 85% H₃PO₄ with downfield shift taken as positive. The IR spectra were recorded at different temperatures (160–293 K) using a home-modified cryostat (Carl Zeiss Jena) with a Nicolet iS50 FTIR (Thermo Scientific, Waltham, MA, USA) spectrometer using 0.05 cm CaF₂ cells. The accuracy of the experimental temperature adjustment was ±0.5 °C. The cryostat modification allowed the transfer of the reagents (premixed at either low or room temperature) under an inert atmosphere directly into the cells.

Manganese hydride complexes mer,trans–[(P(OPh)₃)₂Mn(CO)₃H] (1) [31], mer,trans–[(PPh₃)₂Mn(CO)₃H] (2) [34], fac–[(dppm)Mn(CO)₃H] (3) [27] and fac–[(P-NHC)Mn(CO)₃H] (4) [4] (Scheme 2) were prepared according to the literature methods. Commercially available tris(pentafluorophenyl)boron was purified by sublimation before use. All other reagent-grade chemicals purchased from commercial sources were used as received.

• Synthesis of fac–[(P–NHC)Mn(CO)₃(MeCN)]BF₄ (4^MeCN)

AgBF₄ (0.375 mmol, 73 mg) was placed into the Schlenk flask containing fac-(P-NHC)Mn(CO)₃Br (0.375 mmol, 226 mg) under inert atmosphere. Then 5 mL of CH₃CN was added upon stirring at room temperature. The obtained suspension was sonicated for 5 min and left stirring overnight till complete product formation that was controlled by the IR spectroscopy. After that, the CH₃CN was removed under reduced pressure, the yellow oil residue was dissolved in 5 mL of CH₃CN again to avoid colloid solution formation. The resulting solution was separated from the precipitate via filtration through a Pasteur pipette with Celite, concentrated to 1/10 of initial volume, and dry Et₂0 (80 mL) was added dropwise to induce precipitation of the product. The precipitate was left under supernatant overnight at −20 °C. The next day, the supernatant was removed, and the precipitate was washed with 5 mL two times and dried in a vacuum yielding (211 mg, 86%) pale-yellow powder.

• Synthesis of mer,trans–[(PPh₃)₂Mn(CO)₃(CH₃CN)][BF₄] (2^MeCN)

A suspension of (PPh₃)₂Mn(CO)₃H (0.008 mmol, 5.0 mg) and [Ph₃C][BF₄] (0.008 mmol, 2.6 mg) in CH₃CN (5 mL) was placed in the ultrasonic bath for 5 min at room temperature and left stirring till complete product formation that was controlled by the IR spectroscopy. The solvent was removed under reduced pressure. The resulting pale-yellow solid was dissolved in CH₂Cl₂ (1 mL) and then filtered through a Pasteur pipette with Celite. Hexane (5 mL) was added to the solution to induce crystallization. The supernatant was removed by decantation, the crystalline precipitate was washed with hexane (2 × 5 mL) and then dried in the vacuum to yield [(PPh₃)₂Mn(CO)₃(CH₃CN)][BF₄] (4.6 mg, 76%) as pale-yellow crystals.

Structures of reactants and complexes were optimized at the ωB97-XD level [35], applying def2-TZVP basis set [36] by Gaussian 09 [37]. Optimizations were done in toluene, THF and CH₃CN introduced by a SMD solvent model [38].

Cationic complexes mer,trans–[(PPh₃)₂Mn(CO)₃(MeCN)][BF₄] 2^MeCN and fac–[(P–NHC)Mn(CO)₃(MeCN)][BF₄] 4^MeCN were crystalized from the CH₂Cl₂/hexane system. X-ray diffraction data were collected at 100 K with a Bruker Quest D8 CMOS diffractometer (Bruker, MA, USA), using graphite monochromated Mo-Kα radiation (λ = 0.71073 Å). Using Olex2 [39], the structures were solved with the ShelXT [40] structure solution program using Intrinsic Phasing and refined with the XL [41] refinement package using Least-Squares minimization against F² in anisotropic approximation for non-hydrogen atoms. Positions of hydrogen atoms were calculated, and they were refined in the isotropic approximation within the riding model. Crystal data and structure refinement parameters are given in Table S2. CCDC 2241621 and 2241622 contain the supplementary crystallographic data for 2^MeCN and 4^MeCN, respectively.

3.1. General Procedure for the Interaction with Bases

• For Variable Temperature IR Studies

The solution of mer,trans–[(P(OPh)₃)₂Mn(CO)₃H] (1, c = 0.003 M) was prepared at room temperature in methylcyclohexane. Then itthe was placed into a cryostat and cooled to 190 K. After the spectrum of the initial complex was acquired, the solution from the cryostat was added to the solution of corresponding base (pyridine, HMPA, DBU; 1–70 eq., c = 0.003–0.21 M dissolved in a small amount of solvent kept at 190 K in liquid nitrogen/iPrOH slush bath). The mixture obtained was quickly returned into the cryostat, and the IR spectra were monitored in the 190–290 K temperature range.

Solid fac–[(dppm)Mn(CO)₃H] (3, m = 10.0 mg, n = 0.02 mmol) and KHMDS (m = 11.5 mg, n = 0.10 mmol) were placed in separate Schlenk tubes under inert atmosphere. The complex was dissolved in THF or acetonitrile at room temperature, and a small aliquot was taken to IR cell. After the spectrum of the initial complex was acquired, the Schlenk tubes with solution and solid KHMDS were cooled to 243 K in a liquid nitrogen/ethanol slush bath. Then the solution of complex 3 was quickly transferred to KHMDS with a Pasteur pipette, and the obtained mixture was stirred and transferred into the IR cell at room temperature for spectrum acquisition.

• For Variable Temperature NMR Studies

Solid fac–[(dppm)Mn(CO)₃H] (3, m = 20 mg, n = 0.04 mmol) and KHMDS (m = 23 mg, n = 0.20 mmol) were placed in separate Schlenk tubes under inert atmosphere. The complex was dissolved in THF-d₈ or CD₃CN at room temperature. After that, the NMR tube under inert atmosphere and Schlenk tubes with solution and solid KHMDS were cooled to 243 K in a liquid nitrogen/ethanol slush bath. The solution of complex 3 was transferred to KHMDS with a Pasteur pipette, and the obtained mixture was stirred and quickly filtered through glass cotton directly into precooled NMR tube. The resulting NMR sample was inserted into a pre-cooled NMR probe at 243 K and then monitored with multi-nuclear NMR spectroscopy at 243 K.

3.2. General Procedure for the Interaction with Lewis Acid

• For Variable Temperature NMR Studies

The chosen amount of mer,trans–[(P(OPh)₃)₂Mn(CO)₃H] (1, m = 5.3 mg, n = 0.007 mmol) was dissolved in toluene-d₈ at room temperature and monitored in the 200–293 K temperature range. Then the solution of B(C₆F₅)₃ (7 eq., m = 25 mg, n = 0.007 mmol in 0.5 mL toluene-d₈) was added at 200 K, and the reaction mixture was monitored at 200–293 K.

Similarly, complexes 2–4, (m = 20 mg, n = 0.04 mmol) were dissolved in CD₂Cl₂, placed into an NMR tube and frozen in liquid nitrogen. Then the solution of the Lewis acid (B(C₆F₅)₃; 1 eq., n = 0.04 mmol) in CD₂Cl₂ was poured over the frozen solution in the NMR tube. Two frozen solutions were simultaneously melted in a slush nitrogen/EtOH bath at 180 K, and then the mixture obtained was monitored in the 183–293 K temperature range.

• For Variable Temperature IR Studies

The solution of mer,trans–[(P(OPh)₃)₂Mn(CO)₃H] (1, c = 0.005 M) was prepared at room temperature in toluene. Then it was placed into a cryostat and cooled to 190 K. After the spectrum of the initial complex was acquired, the solution from the cryostat was added to B(C₆F₅)₃ (10 eq., c = 0.05 M) and dissolved in a small amount of solvent kept at 190 K in a nitrogen/iPrOH slush bath. The obtained mixture was quickly returned to the cryostat and monitored in the 190–290 K temperature range.

The solutions of complexes 2–4 (c = 0.003 M) were prepared at room temperature in nBuCl (CH₂Cl₂ or toluene). Then, they were placed into a cryostat and cooled to 160 K (180 K or 190 K). After the spectrum of the initial complex was acquired, the solution from the cryostat was added to the corresponding Lewis acid B(C₆F₅)₃ (1–1.3 eq., c = 0.003–0.004 M), dissolved in a small amount of solvent and cooled to 160 K (180 K or 190 K) in a nitrogen/EtOH (nitrogen/ⁱPrOH) slush bath. The obtained mixture was quickly returned to the cryostat and monitored in the 160 (180 or 190)–290 K temperature range.

4. Conclusions

In conclusion, we have shown that tricarbonyl manganese hydride complexes exhibit dual reactivity under Lewis acid or base treatment. Even the complexes with pronounced hydride character (2, 3) can be deprotonated or, vice versa, the complex with acidic properties (1) can serve as a hydride donor. The Mn–H bond repolarization occurs at the stage of formation of non-covalent intermediates with Lewis acid or base as exemplified for complex 1. The strength of these non-covalent bonds determines whether subsequent hydrogen ion (hydride or proton) transfer will occur. The studies on the complexes bearing bidentate ligands, fac–[(L–L′)Mn(CO)₃H] (3, 4), revealed that their reactions do not follow simple mechanisms but rather involve the fac-to-mer isomerizations. The hydride abstraction from these hydrides proceeds from the mer-derivatives in which the hydride ligand located trans to electron-donating phosphine ligand is more reactive relative to fac-isomers where hydride is trans to CO ligand. Interestingly, the initial deprotonation site of complex 3 in acetonitrile was observed at the CH₂ bridge of the dppm ligand, providing the expected metal-deprotonated product via proton migration from anionic hydride intermediate. Since anionic or cationic Mn(I) complexes formed as the result of proton or hydride abstraction are potential intermediates of (de)hydrogenation reactions, quantitative values of kinetic hydricity of the hydride complexes obtained herein may correlate with their potential catalytic activity and be useful for further elucidation of reaction mechanisms.