Theoretical study of Ni+ assisted C-C and C-H bond activations of propionaldehyde in the gas phase.

The reactions of Ni+ with propionaldehyde in the gas phase have been systematically investigated using density functional theory at the B3LYP/def2-TZVP level. The decomposition reaction mechanism has been identified. Our calculations indicated that Ni+ can assist decomposition of propionaldehyde to form Ni+CO and C2H6 through two types of reaction channel: C-C bond activation and C-H bond activation. In addition, charge decomposition analysis (CDA) was carried out to obtain a deeper understanding for orbital interaction of the initial complex. The bonding properties of the species involved were discussed by means of diverse analysis methods including electron localization function (ELF) and atoms in molecules (AIM).

Introduction

The transition metals play an important role in the catalysis, owing to their unique electronic configuration [1]. The last few decades, the reaction of transition metals with simple organic molecules in the gas phase is one of the important topics [2], [3], [4], [5], [6], [7], [8], [9], [10]. Not only because of the tremendous practical importance to the petroleum industry but also due to the fundamental importance of σ-bonds. A significant number of theoretical and experimental research have been extensively investigated at understanding the mechanisms C—C and C—H bond activations of small hydrocarbons by transition metal atoms, ions and oxide ions [11], [12], [13], [14], [15]. The experiment research has provides abundant information, such as intrinsic binding properties and thermodynamical data. The theoretical results not only rationalize the experimental findings but also provide molecular level details of catalytic processes.

Aldehydes and ketones have become promising candidates for characterizing C—C and C—H bond activations [16]. Over the last several years, the reactions of acetaldehyde and acetone with all kinds of transition metal ions have attracted significant attention, focusing mainly on the investigation of chemical bonds activation by experimental techniques and theoretical calculations. These researches indicated that the reaction product always consists of a stable, neutral organic molecule (methane, ethane, etc.) and the corresponding ionic fragment. At low energies, Cr+, Cu+ and Mn+ are unreactive with acetone, but Fe+, Co+, Ni+ react with acetone to produce C2H6 and M+CO [17], [18], [19], [20]. On the contrary, transition metal ions with high oxygen affinities react with acetone to form MO+ ions and C3H6, e.g. Sc+, V+, Ti+, Gd+ and Pr+ [21], [22]. The reactions of transition metal ions with acetaldehyde found that decarbonylation of acetaldehyde by Fe+, Co+, Cr+, Ni+ is a dominant process at low reaction energies. The reaction products are CH4 and M+CO [23], [24], [25], [26], [27], [28]. The 4Fe+, Co+, Cr+ mediated systems decarbonylation of CH3CHO only through C—C activation, but for 6Fe+, Ni+ decarbonylation reaction is take place via C—C activation and C—H activation [25].

Recently, D. J. Bellert and co-workers carried out experimental research using the single photon initiated decomposition rearrangement reaction (SPIDRR) technique and a simple theoretical research for Ni+ assisted decomposition of propionaldehyde [29]. On the basis of the experimental findings, the researchers proposed a possible reaction mechanism for the propionaldehyde decomposition. They guessed the reaction progressing along two competitive pathways. Ni+ insert into either the CH3CH2C(O)-H or CH3CH2-CHO σ-bond in the propionaldehyde, and the two different products were yielded.

However, the details of the reaction pathways and the geometrical parameters of the intermediates, transition states involved have not been fully discussed. Hence theoretical study in the reaction is still needed. In this paper, we reported a theoretical study on the reaction of C—C and C—H bond activations of propionaldehyde mediated by Ni+ using density functional theory (DFT). The molecular structures and energetics properties of all the stationary points are optimized and calculated. Charge decomposition analysis (CDA) has been carried out to understand how fragment orbitals are mixed to form complex orbital. In addition, we also studied the bonding evolution during the reaction pathways by means of electron localization function (ELF) and atoms in molecules (AIM).

Computational details

The geometries and energies of reactants, intermediates, transition states and products involved in the reaction were fully optimized by using B3LYP hybrid density functional method in conjunction with def2-TZVP basis sets for all atoms [30], [31], [32]. This selected approach has been valuated to be of good validity for the Ni+-containing system in previous papers [33], [34], [35]. Vibrational frequencies were further calculated at the same level to determine that each optimized species was a minimum or a saddle point on the potential energy surface. Local minima on the potential energy surface (PES) have no negative eigenvalue, and saddle points have only one negative eigenvalue. Intrinsic reaction coordinate (IRC) calculations [36], [37] were performed at the B3LYP/def2-TZVP level of theory to confirm the correct connection of the transition state with the corresponding reactants and products. All the calculations were carried out using the Gaussian 09 software package [38].

In order to gain a deeper understanding of the chemical bonds evolution, the wavefunction files (.wfn) produced by Gaussian 09 were used as the inputs into Multiwfn to perform the electron localization function (ELF) and atoms in molecules (AIM) analysis. Charge decomposition analysis (CDA) was created, with the aim of gathering insights into the orbital interaction of the initial complex, as implemented in the Multiwfn code [39].

Results and analysis

The optimized geometries and structural parameters corresponding to all the intermediates and transition states involved along the two different reaction pathways are depicted in Fig. 1. The orbital interaction diagram of the initial complex can be directly plotted in Fig. 2. Besides, bonding analysis of all species involved in the reaction was performed. The analysis results are depicted in Fig. 3 and Table S1. The double potential energy surface and relevant energies of all species are shown in Fig. 4 and Table S2. It can be found that there are two different reaction pathways on the PES toward the reaction of Ni+ with CH3CH2CHO. A path is the C—H bond activation. The other path is activation of C—C bond. Ni+ insertion into the C—H, C—C bonds and the following rearrangement are key steps for this system. In the following sections, a detailed analysis of the reaction is described.

Fig. 1

Geometrical parameters of the intermediates and transition states involved on the double PES at the B3LYP/def2-TZVP level (bond lengths in angstrom and angles in degrees).

Fig. 2

Orbital interaction diagrams of the initial complex (molecular orbital energy in eV).

Fig. 3

ELF projection map of stationary points on the Ni+ + CH3CH2CHO reaction pathway at B3LYP/def2-TZVP level of theory.

Fig. 4

Potential energy profiles for the reactions of Ni+ + CH3CH2CHO calculated at the B3LYP/def2-TZVP level of theory.

Initial complexes

When Ni+ collides with CH3CH2CHO, the initial complex Ni+(C3H6O) with the Cs symmetry is formed. In order to gain a deeper understanding of how fragment orbitals are mixed to form complex orbital, charge decomposition analysis (CDA) was performed and the orbital interaction diagrams are plotted in Fig. 2. As shown in Fig. 2, the horizontal lines on the left, the right and the middle are corresponding to the CH3CH2CHO fragment orbitals, the Ni+ fragment orbitals and the I1a/I1b complex orbitals. The calculated results indicate that electrons have been transferred from CH3CH2CHO to Ni+ during the formation of the Ni+(C3H6O) complex. The net numbers of alpha and beta electrons of I1a are 3.9350 and 0.1254, respectively. So the net number of all electrons of I1a is 4.0604. The net numbers of alpha and beta electrons of I1b are 0.8895 and 0.2112, respectively. Thus, the net number of all electrons of I1b is 1.1001. From Fig. 2, it can be seen that the highest occupied molecular orbital (HOMO) of I1a has two orbitals of the same energy. These are both orbitals that are composed by 3dyz orbital of Ni, have a′ and a″ spatial symmetry, respectively. This orbital specifications stem from the Cs symmetry of the complex. the lowest unoccupied molecular orbital (LUMO) with a′ spatial symmetry is primarily composed by the 4s orbital of Ni. The highest occupied molecular orbital (HOMO) of I1b, which is primarily composed by Ni 3dz2 orbital, has a′ spatial symmetry. And the lowest unoccupied molecular orbital (LUMO) with a′ spatial symmetry is mainly composed of the empty 4s orbital of Ni.

Bonding analysis

In this section, the exploration of the bonding evolution of all species involved is based on the analysis of the electron localization function (ELF) and atoms in molecules (AIM). The ELF analysis shows a detailed picture of the bonding, which exhibits maxima at the most probable positions of local electron pairs and all special positions are surrounded by the basin in which there is an increased probability of finding an electron pair [40]. In the AIM analysis, the existence of the interaction positively correlated with the presence of the bond critical point (bcp) and the strength of the bond is indicated from the magnitude of the electron density at the bond critical point (bcp) [41].

The ELF projection figures of the stationary points of Ni+ + CH3CH2CHO are shown in Fig. 3. The AIM parameters [42], [43] of bond critical points (bcp) are listed in Table S1. As is shown in Table S1, Bond critical points (bcp) have been analyzed in terms of the values of the electron density ρ(r) and its Laplacian of electron density ∇(r) at the (3,−1) critical points, the Lagrangian kinetic energy G(r), the electron potential energy density V(r) and the total electron energy density E(r) or H(r), E(r) = G(r) + V(r) = H(r). The larger the electron density ρ at the bond critical points is, the stronger the interaction will be. The sign of Laplacian ∇(r) is a generally recognized criterion for discriminating bond types [44]. The positive ∇(r) implies the loose charge density at the critical point. Whereas the negative ∇(r) means the shared interaction [45]. The E(r) standard has been proven to be very applicable to characterize the nature of a chemical bond for heavy-atom systems [46], [47], [48]. Negative and positive signs suggest that the interactions are covalent character and closed-shell interactions, respectively [42], [49].

I1a and I1b species. The ELF analysis of the complex reveals that the absence of a disynaptic valence basin between the Ni atom and O atom of CH3CH2CHO, uses it as proof that no covalent bond is formed, which confirms that the interaction can be considered as an electrostatic interaction. According to the AIM analysis, there is a (3,−1) critical point between the Ni and O atoms, but the corresponding charge density is really low (ρ(bcp) = 0.093 au for I1a. ρ(bcp) = 0.086 au for I1b). The ∇(bcp) value is small and positive (∇(bcp) = 0.591 au for I1a. ∇(bcp) = 0.514 au for I1b). The energy density E(r) is negative but |E(r)| is quite small (E(r) = −0.018 au for I1a. E(r) = −0.017 au for I1b). All available evidence points to the fact that no covalent bond is formed between the Ni atom and O atom of CH3CH2CHO.

I2a species. There is an increasing trend in the ρ(bcp) of the Ni—C1 and Ni—H1 bonds. The∇(bcp) is decreased. E(r) is still negative (E(r) = −0.087 au for Ni—H1 bond. E(r) = −0.062 au for Ni—C1 bond), but |E(r)| is increased. The ELF analysis indicates that the formation of Ni—C1 and Ni—H1 bonds at this step, as evidenced by the strengthening of the V(Ni,C1) and (Ni,H1) basins, giving place to the disappeared of a disynaptic V(C1—H1) basin.

I2b species. Our ELF analysis reveals that the C1—C2 bond is completely broken and Ni—C1, Ni—C2 bond formed at this process. This fact is evidenced by the lack of V(C1,C2) disynaptic valence basin and its replacement by a disynaptic V(Ni,C1) and V(Ni,C2) basins. This description is consistent with the results of AIM analysis, which shows the inexistence of the bond critical point between the C1 and C2 atoms. The ρ(bcp) and |E(r)| of the Ni—C1 and Ni—C2 bonds have a stable increase. The ∇(bcp) also has a marked decrease, although it is still positive.

I3 species. There are two ways to form species I3. One is completely broken of the C1—C2 covalent bond of T2a. This is evidenced by the lack of a disynaptic V(C1,C2) basin in the ELF analysis. The fact is also confirmed by AIM analysis, which shows a bcp (3,−1) vanish between the C1 and C2 atoms. The other is H1-migration of T2b from C1 atom to Ni atom to form Ni—H1 bond, as evidenced that a bcp (3,−1) exists between the Ni and H1 atoms. The ρ(bcp) and ∇(bcp) are 0.135 au and −0.016 au, respectively.

I4 species. At this stage, the precursor (I4) is generated with the transferring of H1 from the Ni atom to C2 atom. The I4 intermediate can directly dissociate into the Ni+CO and C2H6. The appearance of the V(C2,H1) basin as shown in the ELF analysis. According to the AIM analysis, the ρ(bcp) and E(r) of the C2—H1 bond is 0.254 au and −0.246 au, respectively, indicating the covalent nature of the C2—H1 bond.

Reaction mechanism

All geometrical parameters of the intermediates and transition states involved on the doublet state are shown in Fig. 1. The relevant energies of all species are listed in Table S2. As shown in Fig. 4. There are two different pathways. Six minima (noted as I1a, I1b, I2a, I2b, I3, I4, respectively) and five first-order saddle points (noted as T1a, T1b, T2a, T2b, and T3, respectively) have been located along the reaction coordinate. ‘intermediates’ are specified by an ‘I’, while ‘T’ is the abbreviation for ‘transition states’. Moreover, ‘a’ denotes the C—H bond activation path, while ‘b’ stands for the Ni+ insertion into the C—C bond.

An encounter complex Ni+(C3H6O) is initially formed by the linkage of Ni+ with the O atom of propionaldehyde. We found that the interaction of Ni+ and CH3CH2CHO leads to two different encounter complexes, suggesting that the metal ion interacts with the two different lone pair orbitals of the oxygen atom. Once the complex is formed. The C1—H1 and C1—C2 bonds can be activated by the metal ion. Structurally, the two encounter complexes have overall Cs symmetry, and the symmetry plane is defined as Ni+—O—C1—C2. The Ni+—O—C1 angle of I1a and I1b is 137.8 and 119.4, respectively. The equilibrium Ni+—O distance severally is 1.888 Å and 1.935 Å. Upon binding to Ni+, the most significant change in propionaldehyde is the lengthening of C1—O bond. Because of oxygen polarizing charge toward Ni+, the C1—O bond is weakened. Energetically, the two complexes lie at below the ground state reactants. The stability of two different structures is nearly the same (226.1 kJ/mol for I1a and 243.1 kJ/mol for I1b), indicating coexistence of them in the gas phase.

The species H3CCH2CO—Ni+—H (I2a) may be formed by the metal ion insertion into the C1—H1 bond of propionaldehyde. This complex I2a has C1 symmetry and lies at −96.02 kJ/mol below the ground state reactants. The bond lengths of Ni+—C1 and Ni+—H1 are calculated as 1.888 Å and 1.476 Å, respectively. And a C1—Ni+—H1 angle is 89.11. The intermediate I2a and the encounter complex I1a are connected by T1a, as confirmed by IRC calculations. The system needs to overcome the activation energy barrier of 132.6 kJ/mol. The imaginary frequency is 410.4i cm−1 corresponding primarily to the H1 atom migrates from the C1 atom to the Ni atom to form the Ni—H1 bond. The Ni—H1 distance is 1.495 Å. It can be found that the geometry of T1a takes a large similar to I2a. Thus it is indeed the late transition state on the PES.

The complex I2a can be rearranged to form species I3 after the system surpasses the second transition state T2a, which is labeled by imaginary frequency 221.3i cm−1. The energy of T2a amounts to −16.30 kJ/mol. And corresponds to the highest barrier along the path, i.e., it is a rate-determining transition state in path a. The Ni+—H1, Ni+—C1, Ni+—C2 distances of species I3 are 1.492 Å, 1.834 Å, 1.997 Å, respectively, indicating that the Ni+ center interacts with all of these atoms.

Along the dissociation coordinates, the complex I3 can be rearranged to form a direct precursor (I4) of the dissociation products via transition state T3, which has an imaginary frequency of 498.4i cm−1. The species T3 lies at −66.13 kJ/mol below the reactants’ asymptote. Nonreactive dissociation of the Ni+(CO)-C2H6 would give rise to the decomposition products: Ni+CO and C2H6. The overall decomposition process of Ni+ + CH3CH2CHO is calculated to be exothermic by 149.5 kJ/mol. Our results indicated that the energy of the whole reaction pathway on this PES is lower than the reactant’s asymptote.

Next, we turn to the alternative C—C activation path for the formation of species I3. This branch involves the initial complex I1b. Subsequent Ni+ inserts into the C1—C2 bond to form species H3CH2C-Ni+-CHO (I2b). In H3CH2C-Ni+-CHO (I2b) structure, the Ni+—C1, Ni+—C2 bond distances are 1.832 Å, 1.899 Å, and a C1—Ni+—C2 angle is 107.3. Energetically, species I2b is calculated to be −74.47 kJ/mol below the separated reactants, or 21.55 kJ/mol higher than the corresponding C—H insertion minimum I2a. The encounter complex I1b and species I2b are connected by transition state T1b. The energy of the transition state is −51.83 kJ/mol. Note that T1b is rate-determining transition states in path b.

Intermediate I2b can also be converted to the species I3 through transition state T2b. The hydrogen atom is shifted from C1 atom to the transition metal atom to form species I3a. Energetically, the transition state T2b is calculated to be −55.04 kJ/mol lower than the corresponding transition state along the C—H bond activation pathway (T2a).

Conclusions

In this work, a specific theoretical calculation has been performed to further understand the gas phase reaction mechanism details of Ni+ with propionaldehyde using density functional theory. All geometries along the potential energy surfaces were fully optimized at B3lYP/def2-TZVP level. The orbital interaction of the initial complex has been performed by CDA. Bonding evolution was analyzed on the basis of ELF and AIM calculations. Those conclusions can be summarized as follows:

We found that the decomposition reaction of Ni+ + CH3CH2CHO has two types of reaction channels: C—C activation and C—H activation. This reaction contains four elementary steps of encounter complexation, C—C and C—H bond activations, H-shift and nonreactive dissociation. The second step in the C—H bond activation process has to surmount a highest barrier, which constitutes the rate-determining for this path. The rate-determining step of the C—C bond activation path is the C—C bond break, which has an energy barrier of 44.18 kJ/mol. The formation of stable precursor (H6C2-Ni+(CO)) is the reaction with the higher exothermicity, −281.9 kJ/mol, and the C-C activation path is the most favorable pathway.

The ELF and AIM analysis give important insights into the chemical bonds evolution of the reactions. The results of the analysis indicated that the initial complex, Ni+(C3H6O), is formed by electrostatic interaction, because no bond is formed between Ni atom and O atom of CH3CH2CHO.