Unimolecular dissociation of anions derived from succinic acid (H 2 Su) in the gas phase: HSu − and ClMgSu − . Relationship to CO 2 fixation

Abstract

Electrospray ionization of mixtures of succinic acid (here denoted H₂Su) and magnesium chloride in water/methanol give rise to ions of the type ESu⁻ (E = H or ClMg). The unimolecular dissociation of these ions was studied by collisionally induced dissociation mass spectrometry and interpreted by quantum chemical calculations (density functional theory and the composite Gaussian-4 method) of relevant parts of the potential energy surfaces. The major dissociation pathways from HSu⁻ were seen to be dehydration and decarboxylation, while ClMgSu⁻ mainly undergoes decarboxylation. The latter reaction proceeds without barrier for the reverse reaction; addition of CO₂ to a Grignard type structure ClMg(CH₂CH₂CO₂)^–. In contrast, addition of CO₂ to the analogous H(CH₂CH₂CO₂)^– ion has a substantial barrier. Dehydration of HSu⁻ gives rise to deprotonated succinic anhydride via a transition state for the key intramolecular proton transfer having an entropically favorable seven-member ring structure. The succinate systems studied here are compared to the previously reported analogous maleate systems, providing further insight to the structure–reactivity relationship.

Keywords

Tandem mass spectrometry carbon dioxide fixation unimolecular reactions gas phase ion chemistry

Introduction

Reductive fixation of CO₂ is a key step in photosynthesis and in methodologies for artificial photosynthesis.^1–5 In principle, two alternative approaches for C–C bond formation can be imagined—either first activating CO₂ reductively by making it into a carbon nucleophile capable of reacting with an electrophilic carbon of a substrate molecule RX (equation (1)) or first activating the substrate reductively and then let the activated substrate react with CO₂ (equation (2))

{CO}_{2} + {2 e}^{-} + M^{2 +} \to {MO}_{2} C

(1a)

{MO}_{2} C + RX \to {MO}_{2} CRX

(1b)

RX + {2 e}^{-} + M^{2 +} \to XMR

(2a)

XMR + {CO}_{2} \to {MO}_{2} CRX

(2b)

For M = Mg and X = halide, reaction 2 corresponds to the well-known Grignard reaction,^2,6 while more recently, the carbenoic MO₂C species formed in reaction 1a by umpolung has been spectroscopically characterized and shown to be a highly reactive carbon nucleophile.⁷

The gas phase is the ideal environment for probing reactivity in the absence of solvents, and the mass spectrometer is ideally suited for exploring elementary reaction steps with well-defined reactants and products.^6–10 Further details of the potential energy surface governing a chemical reaction can be obtained by combining mass spectrometric evidence with quantum chemical calculations. Systematic studies of this kind provide valuable information on structure/reactivity relationships of relevance also to condensed phase chemistry. A particular feature of the mass spectrometer as a chemical laboratory is that, under favorable circumstances, a reaction and its reverse reaction can be studied. For example, by activating a molecule by energy transfer (e.g. collisionally induced dissociation, CID), it may dissociate into well-defined fragments. In separate experiments, the association of the fragments may be studied. In any instance, and in particular when the details of the potential energy surfaces are known from quantum chemical calculations, insight into the reverse reaction is obtained from CID studies. Based on CID studies of carboxylate anions, Graul and Squires¹¹ realized that R⁻ fragments may add to carbon dioxide, a reaction considered to be without an energy barrier, e.g. microscopic reversibility can be established for the following equilibria

{RCO}_{2}^{-} R^{-} + {CO}_{2}

(3)

From CID studies of carboxylate-metal anions, Rodríguez-Blanco et al.¹² found that formate (HCOO⁻) and acetate (CH₃COO⁻) anions complexed with CaCl₂ dissociate by decarbonylation and loss of CH₂CO, while the analogous ZnCl₂ and MnCl₂ adduct anions decarboxylate. This was rationalized by Ca²⁺ having higher affinity for OH⁻ than for H⁻, while the opposite is true for Zn²⁺.

Dicarboxylic acids are of interest in this respect because several are central in essential biochemical cycles involving CO₂.^13–15 In addition, two carboxylic acid functions present in the same molecule allows for a metal cation to bind simultaneously to both groups, thereby influencing the molecular conformation and thereby the reactivity.¹⁶ Systematic study of the gas phase fragmentation of deprotonated dicarboxylic acids was conducted by Grossert et al.¹⁷ showing an intimate relationship between carbon chain length and reactivity. The loss of CO₂ was the most prevalent reaction for the shortest dicarboxylic acids, while H₂O loss abundancy increased with chain length, the latter being attributed to increased conformational freedom. CID of the simplest dicarboxylic acid, oxalate, leads to concurrent loss of CO and CO₂ as reported by Soldi-Lose et al.¹⁸

Miller et al.² investigated the unimolecular dissociation of deprotonated maleic acid (HMa⁻) and ClMgMa^–. Both anions readily undergo two consecutive decarboxylations. The first decarboxylation is more facile for ClMgMa^– than for HMa^–. It was also seen that HMa^– loses water. In addition to losing a second CO₂ molecule, the fragment ion produced by decarboxylation of ClMgMa^–, ClMg(O₂C)(HCCH)⁻, was seen to lose ethyne giving rise to ClMg(η²-O₂C)⁻ without having a barrier for the reverse reaction. The latter ion corresponds to a carbenoic species of the type MCO₂ mentioned above. In other words, in the reverse reaction, ClMg(η²-O₂C)⁻ is able to form a C–C bond by addition to ethyne (Scheme 1).

Scheme 1.

Addition of ClMg(η²-O₂C)⁻ to ethyne.

In this paper, we wanted to investigate the closely related succinic acid applying the same experimental strategy. It was of interest to study the effect of replacing the central C=C double bond of maleic acid with the C–C single bond of succinic acid. This may have two consequences. Firstly, a C–C single bond allows for more conformational flexibility. Secondly, the π bond of maleic acid is able to influence the stability by electronic interaction conjugation.

Methods

Experimental details

The experiments were conducted using a modified Micromass/Waters Q-TOF 2 MS with a custom built collision gas inlet system. The instrument has a quadrupole/hexapole/time-of-flight geometry, where the quadrupole serves as the first mass filter, the hexapole as the collision cell, and finally the time-of-flight as the second mass analyzer. The sample was introduced with a 1 mL syringe using an injector flow rate of 2.0 µL min⁻¹. The concentrations of succinic acid and MgCl₂ were 1.0 mM and 2.0 mM, respectively, in all experiments. Both compounds were available as crystalline solids (succinic acid: Acros Organics, 99%) dissolved in the given solvent for the experiment at hand (H₂O/MeOH in an 80/20 v/v ratio).

The mass spectra were recorded at 3.0 kV time-of-flight (TOF) voltage for the deprotonated acid in order to allow detection of low-mass fragments (cut-off at m/z 14) and at 7.0 kV TOF voltage for the magnesium chloride adduct, as its detection was difficult at lower potentials. The electrospray capillary was operated at 3.0 kV, and the multi-channel plate (MCP) detector had an operational voltage of 2200 V throughout. The cone voltage was kept at 0–15.0 V. For deprotonated succinic acid, HSu^–, CID mass spectra were recorded in steps of 1.0 eV collision energy (lab frame) in the range 5.0–10.0 and 25–32.0 eV and in steps of 0.5 eV in the range 10.0–25.0 eV. The collision gas (argon) pressure was kept at 2.0 × 10⁻⁴ mbar during these experiments. Similarly, for ClMgSu^–, the collision energy was varied 5.0–47.0 eV (lab frame) in increments of 1.0 eV. The collision gas pressure was kept at 3.0 × 10⁻⁴ mbar during these experiments. In order to enhance the abundance of ClMg(CH₂CH₂CO₂)^– in the reaction experiments with CO₂, the cone voltage was increased to 35.0 V. In these experiments, the collision energy was set to 0.4 eV in the center-of-mass frame, while the pressure was kept at a nominal pressure of 8.65 × 10⁻⁴ mbar to ensure a sufficient frequency of collisions between the ions and the CO₂.

The succinic acid sample was also analyzed with a GC-EI-MS (Bruker Scion 436-GC) and a different ESI-MS instrument (Bruker maXis II ETD) to aid elucidation of the sample solution composition.

Appearance energies were estimated by first fitting a polynomial function to the breakdown curve of a given signal, and subsequently extrapolating the tangent line through the inflection point of the fragment ion curve to the base line. Error estimates were made on the basis of the count statistics. We realize the rather approximate nature of this method but for the present purpose, in comparing appearance energies between competing reaction channels, this semi-quantitative approach was considered appropriate. More sophisticated treatment would involve the use of the methods by the Armentrout or Chen laboratories.^19,20 However, the signal-to-noise-ratios of some of our recorded breakdown curves, in particular at very low collision gas pressure, were considered too poor to obtain accurate estimates using these methods.

Quantum chemical calculations

Geometry optimization was done with Gaussian 09 initially using the B3LYP method with aug-cc-pVTZ basis sets to improve the description of polarization in the anionic systems.^21,22 Optimized structures were recalculated using Gaussian-4 theory (G4)²³ for increased accuracy, intended to provide energy values within ca. 10 kJ mol⁻¹ of accurate experiments. Intrinsic reaction coordinate calculations were used to connect all transition states to their respective energy minima.

Results

Mass spectra, dissociation reactions and breakdown curves of HSu⁻ (m/z 117)

The results for the hydrogen succinate experiments are illustrated in the spectra of Figure 1. Starting with Figure 1(a), the mass spectrum of succinic acid electrosprayed from a solution of water and methanol, we observe the base peak at m/z 117 as the deprotonated acid, i.e. HSu⁻. This interpretation is supported by the isotope distribution pattern. The second most intense signal is for m/z 73, corresponding to decarboxylation. Water loss from HSu⁻ is consistent with the fragment ion peak at m/z 99. Both these fragment ions are attributed to in-source fragmentation, although separate GC-EI-MS experiments indicate that a small amount of succinic anhydride is present in the sample solution, which also may contribute to the m/z 99 peak.

Figure 1.

(a) Mass spectrum of succinic acid electrosprayed from an 80/20 v/v H₂O/MeOH solution. Separate experiments indicate that there is a 2% contamination of acetate (m/z 59) in the succinic acid sample. (b) CID mass spectrum of HSu⁻ (m/z 117) with collision energies averaged over 3.0–5.0 eV (E_COM) at 2.0 × 10⁻⁴ mbar Ar pressure, 3.0 kV TOF voltage. Note the different scales on the y-axes.

The CID mass spectrum is shown in Figure 1(b). The fragment ion peaks at m/z 99 and 73 appear in both spectra, supporting the assumption that decarboxylation and dehydration are giving rise to these peaks.

Additional peaks were observed in the CID-spectra with higher collision energies at m/z values 55, 45, 27 and 17 (supplementary information). Ions with m/z 55 may be formed by decarboxylation of the dehydration fragment ion at m/z 99, and/or by dehydration of the decarboxylation fragment at m/z 73. The peak at m/z 45 is either due to loss of CO or C₂H₄ from the fragment ion at m/z 73. The fragment ion at m/z 27 seems to be formed by decarbonylation of the ion with m/z 55. The fragment ion with m/z 17 is most likely the hydroxyl anion, probably formed by decarbonylation of the m/z 45 fragment ion. In addition, there is a peak at m/z 44, which we interpret to be due to carbon dioxide radical anion. Although carbon dioxide has negative electron affinity, EA = −0.60 eV,²⁴ making it metastable with respect to electron detachment, it is often observed upon CID of carboxylate ions. According to Schröder et al.,²⁵ formation of ${CO}_{2}^{-} \cdot$ by collisional activation of carboxylates becomes more facile if a secondary radical $R^{•}$ with a negative electron affinity is formed

{RCO}_{2}^{-} \to R^{•} +_{CO}^{2} -^{•}

(4)

This is consistent with propionate (m/z 73) being the direct precursor in the present experiment, since the ethyl radical (R = C₂H₅) has negative electron affinity, EA = −0.28 eV²⁶). Separate CID experiments of propionate, not shown here, support this assumption.

The above spectral assignments are summarized in equations (5) to (11) following.

{HSu}^{-} (m / z 117) \to

\begin{matrix} {\begin{matrix} C_{4} H_{3} O_{3}^{-} (m / z 99) + H_{2} O & (5) \\ C_{3} H_{5} O_{2}^{-} (m / z 73) + {CO}_{2} & (6) \\ C_{3} H_{3} O^{-} (m / z 55) + H_{2} O + {CO}_{2} & (7) \\ {HOCO}^{-} (m / z 45) + C_{2} H_{4} + {CO}_{2} & (8) \\ _{CO}^{2} -^{•} (m / z 44) + C_{2} H_{5}^{•} + {CO}_{2} & (9) \\ C_{2} H_{3}^{-} (m / z 27) + {CO}_{2} + H_{2} O + CO & (10) \\ {OH}^{-} (m / z 17) + C_{2} H_{4} + {CO}_{2} + CO \end{matrix} \end{matrix}

(11)

Breakdown graphs for the two fragment ions of lowest appearance energy are presented in Figure 2.

Figure 2.

Breakdown curves of hydrogen succinate, HSu^– (m/z 117) from a CID experiment recorded at a collision gas pressure of 2.0 × 10⁻⁴ mbar, with a TOF voltage of 3.0 kV. (a) Breakdown curves with signals corresponding to dehydration (m/z 99) and decarboxylation (m/z 73) of HSu⁻; (b) same, but the signal of m/z 99 has been multiplied by a factor of 20 in order to facilitate comparison.

The fragment ion due to decarboxylation, at m/z = 73, is the most abundant product at all energies. However, the low energy tail of the signal due to m/z 99 extends to lower energy. In fact, using the extrapolation procedure described in the Experimental section, a lower appearance energy is estimated, as evident from Table 1. The experimental data in the form of appearance energy estimates are presented in Table 1.

Table 1.

Appearance energies with standard deviations for hydrogen succinate (HSu⁻) fragment ions.

	eV	kJ mol⁻¹
m/z 99	1.7₄ ± 0.1₄	168 ± 13
m/z 73	2.2₅ ± 0.2₀	217 ± 19
m/z 55	3.1₉ ± 0.2₀	308 ± 19
m/z 45	5.4₇ ± 0.3₂	528 ± 31
m/z 44	5.5₅ ± 0.2₆	536 ± 25
m/z 27	4.7₂ ± 0.3₁	455 ± 30
m/z 17	6.6₇ ± 0.1₀	644 ± 10

Mechanisms for dehydration and decarboxylation of HSu⁻

Next, a survey of the potential energy landscape was conducted. All results from this endeavor can be found in the supplementary information. Here, we will discuss the most prominent features. The potential energy diagram of relevance to decarboxylation and dehydration is shown in Figure 3.

Figure 3.

Potential energy diagram (G4 (0K)) for dehydration (right hand side) and decarboxylation (left hand side) of HSu⁻ (H-1), energies in kJ mol⁻¹.

Water loss is identified to occur via the isomeric intermediates H-2B and H-4B. The pathway H-1 → H-TS3B → H-4B → H-TS4B → H-3B + H₂O is seen to be the energetically most economical with a barrier of 188 kJ mol⁻¹. Both mechanisms lead to the formation of H-3B, deprotonated succinic anhydride, analogous to dehydration of neutral succinic acid as reported by both Chen et al.²⁷ and Li et al.²⁸

Decarboxylation of HSu⁻ yielding H-2A, propionate, follows a simple mechanism in which the heterolytic C–C bond cleavage is coupled with a “hidden” proton transfer from the carboxylic acid group to the emerging carbanionic site. Two conformeric transition states were located, of which the one of lowest energy, H-TS1A, is at 194 kJ mol⁻¹.

Both decarboxylation reactions occur in one single reaction step from the most stable conformer of HSu^–, indicating that they are entropically more favorable than the dehydration pathways, the latter requiring two reaction steps. This simple fact may explain why decarboxylation is kinetically dominating, although dehydration has lower critical energy. The difference in critical energy is computed to be 6 kJ mol⁻¹.

It should also be mentioned that in the absence of the hidden proton transfer, heterolytic C–C bond dissociation leads directly to the simultaneous formation of three fragments—ethene, carbon dioxide and hydroxycarbonyl anion, HOCO⁻ at 320 kJ mol⁻¹ (not shown in Figure 3, see supplementary information).

Mass spectra, dissociation reactions and breakdown curves of ClMgSu⁻ (m/z 175)

Results from the experiments with the mixture of succinic acid and magnesium chloride are displayed in Figure 4. The relative intensity distribution found for the group of signals in the range m/z 175–179 in the normal mass spectrum of Figure 4(a) is consistent with the theoretical isotopic distribution of ClMgSu⁻. The group of peaks m/z 129–135 is largely due to MgCl₃⁻ anion, but with a contribution (ca. 6 %) from [(ClMgSu) − CO₂] ⁻. The CID mass spectrum of ClMgSu⁻ (m/z 175), displayed in Figure 4(b), suggests that the only major fragmentation reaction is decarboxylation. At higher collision energies (supplementary information), a minor peak at m/z 129 indicates loss of H₂ from the decarboxylation fragment ion. In addition, a fragment ion with m/z 103 was observed at higher collision energies, most likely due to ClMg(η²-O₂C)⁻,^7,16 in analogy with the aforementioned CID experiments with ClMgMa⁻ mentioned in the Introduction.² Also, at high collision energies, fragment ion peaks are found for m/z 87, 71 and 59. In short, the assignments are as follows ((12) to (17)).

{ClMgSu}^{-} (m / z 175) \to

\begin{matrix} {\begin{matrix} {ClMgC}_{3} H_{4} O_{2}^{-} (m / z 131) + {CO}_{2} & (12) \\ {ClMgC}_{3} H_{2} O_{2}^{-} (m / z 129) + {CO}_{2} + H_{2} & (13) \\ {ClMgCO}_{2}^{-} (m / z 103) + {CO}_{2} + C_{2} H_{4} & (14) \\ {ClMgC}_{2} H_{4}^{-} (m / z 87) + {2 CO}_{2} & (15) \\ C_{3} H_{3} O_{2}^{-} (m / z 71) + {CO}_{2} + HMgCl & (16) \\ {ClMg}^{-} (m / z 59) + {2 CO}_{2} + C_{2} H_{4} \end{matrix} \end{matrix}

(17)

Figure 4.

(a) Mass spectrum of a mixture of succinic acid and MgCl₂ in an 80/20 v/v H₂O/MeOH solution. (b) CID mass spectrum of ClMgSu⁻ (m/z 175) with collision energies averaged over 3.0–5.0 eV (E_COM) at 3.0 × 10⁻⁴ mbar Ar pressure, 7.0 kV TOF voltage.

The appearance energies of the fragment ions were estimated from the breakdown curves in Figure 5, summarized in Table 2.

Figure 5.

Breakdown curves of ClMgSu^– (m/z 175) from a CID experiment recorded at a collision gas pressure of 3.0 × 10⁻⁴ mbar, with a TOF voltage of 7.0 kV. The ion count is a function of the centre-of-mass collision energy. Note the different scales: (a) Breakdown curves with measured intensities for all fragment ion signals. (b) Breakdown curves where the decarboxylation signal intensity has been multiplied by 0.03 in order to facilitate comparison with the rest of the signals.

Table 2.

Appearance energies with standard deviations for magnesium chloride succinate (ClMgSu⁻) fragment ions.

	eV	kJ mol⁻¹
m/z 131	2.4₈ ± 0.0₇	239 ± 70
m/z 129	3.8₁ ± 0.2₉	371 ± 28
m/z 103	4.4₈ ± 0.1₈	432 ± 17
m/z 87	4.1₈ ± 0.4₂	403 ± 41
m/z 71	4.4₃ ± 0.9₇	427 ± 94
m/z 59	3.7₄ ± 0.2₆	361 ± 25

The appearance energy for decarboxylation is measured to 239 kJ mol⁻¹, which within experimental uncertainty is not significantly different from that of the analogous reaction for HSu⁻ at 217 kJ mol⁻¹. No fragment ion peak at m/z 99 was observed that would correspond to the loss of ClMgOH—analogous to water loss from HSu^–.

Mechanisms for decarboxylation of ClMgSu⁻

The computed potential energy diagram describing CO₂ loss from ClMgSu⁻ is shown in Figure 6. Two isomeric forms of ClMgSu^–, M-1A and M-1B, serve as the direct precursors for the two dissociation channels. These isomers differ only in how of one of the carboxylate groups interact with the magnesium atom and are of similar energy. Decarboxylation from the most stable conformer has the highest barrier, yet interconversion between the two conformers has a modest barrier at 20 kJ mol⁻¹, which is of little significance to the overall reaction kinetics. This indicates that the reaction: M-1B → M-TS1B → M-2B → M-3B + CO₂ is the more favorable pathway. The product ion formed in this reaction, M-3B, is seen to be a Grignard type reagent by having an Mg–C bond. The alternative pathway: M-1A → M-TS1A → M-2A → M-3A + CO₂ may be relevant at higher collision energies. The spiro compound formed in this reaction, M-3A, will be discussed in more detail below. An interesting point is that both decarboxylation reactions have no barriers to the reverse reactions, meaning that addition of CO₂ to M-3B (and M-3A) should be favorable. This was investigated in separate experiments on reactions between m/z 131 ions and CO₂ at low collision energy and relatively high CO₂ collision gas pressure which showed the formation of ions with m/z 175. Although not conclusive, this finding supports ion our interpretation. In stark contrast, the energy diagram for dissociation of HSu^– of Figure 3 shows a significant barrier for CO₂ addition to propionate.

Figure 6.

Potential energy diagram (G4 (0 K)) showing the decarboxylation of ClMgSu⁻, energies in kJ mol⁻¹.

Mechanism for formation of ClMg(η²-O₂C)⁻

Starting from the spiro compound M-3A, the product ion of the high energy pathway for CO₂ loss from ClMgSu^–, subsequent loss of ethene leads to the formation of the bidentate species ClMg(η²-O₂C)⁻(M-5): M-3A → M-TS2 → M-4 → M-5 + C₂H₄ (Figure 7). This species, corresponding to an activated form of CO₂, has previously been described as a carbon nucleophile having all the characteristics of a carbene.^7,16 The observation that the reaction M-5 + C₂H₄ → M-3A is a typical carbene addition to a carbon–carbon double bond strengthens this description.

Figure 7.

Potential energy diagram (G4 (0 K)) for the formation of ClMg(η²-O₂C)⁻(M-5), energies in kJ mol⁻¹.

Discussion

Hydrogen succinate (HSu⁻) is here seen to display an analogous dissociation pattern to hydrogen maleate (HMa⁻) previously reported by Miller et al.,² with losses of carbon dioxide and water being the dominating reactions for both ions. Furthermore, both ClMgSu^– and ClMgMa^– exclusively lose CO₂ at low collision energies. While the experimental estimates of the appearance energies provide indispensable guidelines for assessment of the computational models, the computed energies are considered more accurate. In this respect, the compound quantum chemical method G4 used here has previously been shown to perform well in reproducing energy barriers for a selection of 76 reactions with a mean unsigned deviation value of 4 kJ mol⁻¹.²⁹ The computed critical energies of the most abundant maleate and succinate reactions are compared in Table 3.

Table 3.

G4 (0 K) critical energies in kJ mol⁻¹ for the main pathways of dissociation for succinate and maleate adducts with H⁺ and MgCl⁺.

	HSu⁻	ClMgSu⁻	HMa⁻	ClMgMa⁻
−CO₂	194	185	180	153
−H₂O	188	–	234	–

In comparing CO₂ loss from HSu⁻ and HMa⁻, we see that the latter has a slightly lower barrier. The reaction mechanisms are similar in producing the propionate and acrylate anions, respectively, by intramolecular transfer of a proton to the carbanionic site gradually formed as the CO₂ molecule is departing.

However, the mechanisms for dehydration are different, and the barrier for dehydration of HMa^– is 46 kJ mol⁻¹ higher for than for HSu^–. While the products deprotonated succinic anhydride + H₂O are 160 kJ mol⁻¹ relative to HSu^–, deprotonated maleic anhydride + H₂O are 230 kJ mol⁻¹ relative to HMa⁻. This considerable stability difference is linked to major structural differences between the geometries and electronic structures of deprotonated succinic anhydride and deprotonated maleic anhydride molecules. In the former, the charge can be effectively delocalized to the nearby oxygen atoms, while in the latter, the restrictions imposed by the C=C double bond hampers this. As a result, the energy landscapes for the dehydration reactions are quite different.

Decarboxylation is the major fragmentation reaction for both ClMgSu^– and ClMgMa^–. Both have two isomers that differ by Mg coordination to two and three oxygens, respectively, of which the 3-coordinated isomers are lowest in energy. For ClMgSu^–, the 2-coordinated is 13 kJ mol⁻¹ higher in energy, while for ClMgMa^–, it is only 1 kJ mol⁻¹ higher, which may account for some of the difference in that the most facile decarboxylation reaction mechanism for ClMgMa⁻ has a critical energy of 153 kJ mol⁻¹, but for ClMgSu⁻ at 185 kJ mol⁻¹. The other contributing factor lies in the relevant product stabilities. The Mg–C bond distance is 0.02 Å shorter in the Grignard species for Ma than for Su, indicating a stronger interaction. Separate G4 calculations show that the following isodesmic reaction is exothermic by 20 kJ mol⁻¹

C_{2} H_{4} + {ClMgC}_{2} H_{5} \to C_{2} H_{6} + {ClMgC}_{2} H_{3}

(18)

thereby supporting this conclusion.

The formation of ClMg(η²-O₂C)⁻ from both ClMgMa⁻ and ClMgSu⁻ uncovered an interesting opportunity to describe the differences in the reactivity of this carbenoid species with ethene and ethyne. Addition of ClMg(η²-O₂C)⁻ (M-5) to ethyne and ethene differs as illustrated in Scheme 2. ClMg(η²-O₂C)⁻ adds to a terminal carbon of ethyne in a single reaction step giving rise to the carbanion M-3 C. On the other hand, addition to ethene occurs in two steps. In the first step, ClMg(η²-O₂C)⁻ adds to a terminal carbon of ethene giving rise to the carbanion M-4, which in a second step cyclizes giving the significantly more stable spiro isomer M-3A. Both addition to ethyne and ethene proceed without an energy barrier. The reaction energies are 97 kJ mol⁻¹ and 82 kJ mol⁻¹, respectively.

Scheme 2.

Nucleophilic addition of ClMg(η²-O₂C)⁻ to ethyne (top) and ethene (bottom).

Polino et al.³⁰ performed quantum chemical calculations (RQCISD(T)/CBS// B3LYP/6-311++G(d,p)) of reactions of singlet methylene (¹CH₂) with ethyne and ethene, that form cyclopropene and cyclopropane, respectively, with reaction energies of 364 and 416 kJ mol⁻¹. The ethene and ethyne affinity of ClMg(η²-O₂C)⁻, which is an electronic singlet in its lowest energy state, is therefore seen to be significantly lower than that of singlet methylene. In addition, the ethyne and ethene affinities are inverted. However, it should be taken into account that the adduct between ClMg(η²-O₂C)⁻ and ethyne was found to be non-cyclic. Separate calculations show that the binding energy of M-5 with ethyne in a cyclic product is 47 kJ mol⁻¹, and it is thus less stable than M-3A (the product of the lower reaction in Scheme 2) by 35 kJ mol⁻¹.

Conclusion

We have investigated the dissociative patterns of hydrogen succinate and magnesium chloride succinate anions and compared them with the analogous maleate systems. The main fragmentation reactions of hydrogen succinate and maleate were decarboxylation and dehydration. The demands for decarboxylation of the two were found to be quite similar; however, dehydration was more facile for the succinate system. This is attributed to the increase in conformational freedom going from double to single bonds in the carbon backbone, emphasizing the connection between structure and dissociative mechanism as has been reported in the past.^2,17,18 The dissociation of magnesium chloride succinate and maleate adduct anions occurs mainly through decarboxylation, which is most demanding for the former. The increased conformational freedom of hydrogen succinate compared to hydrogen maleate facilitates water loss, yet makes decarboxylation more difficult as the stabilization by the intramolecular bond between the carboxylate oxygen atoms and the H⁺ and MgCl⁺ ligands is stronger than in the latter, structurally more rigid anion. The nucleophilic metal-CO₂ adduct anion is observed as a fragmentation product in both systems, the survey of which has provided interesting details in relation to its reactivity with ethyne and ethene. The ClMg(η²-O₂C)⁻ adduct anion adds to these hydrocarbons in one and two steps, respectively. Addition to the former yields a structure with a single bond between ethyne and ClMg(η²-O₂C)⁻, while the equivalent reaction with the latter forms a structure with a double bidentate bonding moiety.

Footnotes

Acknowledgements

The authors are grateful to Glenn B.S. Miller for valuable suggestions.

Authors’ note

This article is dedicated to the memory of Peter Derrick, a great friend and colleague.

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by the Norwegian Research Council by the Grant No. 179568/V30, to the Centre of Theoretical and Computational Chemistry through their Centre of Excellence program, by the grant No. 249788 to the project “The chemistry of CO₂ activation and fixation”, and the Norwegian Supercomputing Program (NOTUR) through a grant of computer time (Grant No. NN4654K).

References

Drees

Cokoja

Kühn

. Recycling CO₂? Computational considerations of the activation of CO₂ with homogeneous transition metal catalysts. Chem Cat Chem 2012; 4: 1703–1712.

Miller

Fäseke

Uggerud

. Unimolecular dissociation of anions derived from maleic acid (MaH₂) in the gas phase: MaH^– and MaMgCl^– – relationship to Grignard chemistry and reductive CO₂ fixation. Eur J Mass Spectrom 2015; 21: 545.

Freund

H-J

Messmer

. On the bonding and reactivity of CO₂ on metal surfaces. Surf Sci Lett 1986; 172: A333.

Song

. Global challenges and strategies for control, conversion and utilization of CO₂ for sustainable development involving energy, catalysis, adsorption and chemical processing. Catal Today 2006; 115: 2–32.

Aresta

Dibenedetto

Dutta

. Energy issues in the utilization of CO₂ in the synthesis of chemicals: the case of the direct carboxylation of alcohols to dialkyl-carbonates. Catal Today 2017; 281: 345–351.

O’Hair

RAJ

Vrkic

James

. Gas-phase synthesis and reactivity of the organomagnesates CH₃MgL₂.⁻ (L = Cl and O₂CCH₃): from ligand effects to catalysis. J Am Chem Soc 2004; 126: 12173–12183.

Dossmann Soldi-Lose

Afonso

Lesage

et al.

Formation and characterization of gaseous adducts of carbon dioxide to magnesium, (CO₂)MgX⁻ (X=OH, Cl, Br). Angew Chem Int Ed 2012; 51: 6938–6941.

Guo

Qian

Liao

et al.

ESI-MS studies on the mechanism of Pd(0)-catalyzed three-component tandem double addition-cyclization reaction. J Am Chem Soc 2005; 127: 13060–13064.

O’Hair

RAJ

Khairallah

. gas phase ion chemistry of transition metal clusters: production, reactivity, and catalysis. J Clust Sci 2004; 15: 331–363.

10.

Falvo

Fiebig

Dreiocker

et al.

Fragmentation reactions of thiourea- and urea-compounds examined by tandem MS, energy-resolved CID experiments, and theory. Int J Mass Spectrom 2012; 330–332: 124–133.

11.

Graul

Squires

. Gas-phase acidities derived from threshold energies for activated reactions. J Am Chem Soc 1990; 112: 2517–2529.

12.

Rodríguez-Blanco

Jobst

Luider

et al.

Generation and dissociation of RCOOCaCl₂⁻ and other carboxylate-substituted superhalogens: CO₂ capture and implications for structure analysis. Chem Plus Chem 2013; 78: 1184–1189.

13.

Cheng

Thomas

Rouvière

. Biological conversion of cyclic alkanes and cyclic alcohols into dicarboxylic acids: biochemical and molecular basis. Appl Microbiol Biotechnol 2002; 58: 704–711.

14.

White

Williams

Landau

. Metabolism of pyruvate by rat adipose tissue in vitro. Arch Biochem Biophys 1968; 126: 552–557.

15.

Krebs

Johnson

. Metabolism of ketonic acids in animal tissues. Biochem J 1937; 31: 645.

16.

Miller

GBS

Esser

Knorke

et al.

Spectroscopic identification of a bidentate binding motif in the anionic magnesium-CO₂ complex (ClMgCO₂.⁻). Angew Chem 2014; 126: 14635–14638.

17.

Grossert

Fancy

White

. Fragmentation pathways of negative ions produced by electrospray ionization of acyclic dicarboxylic acids and derivatives. Can J Chem 2005; 83: 1878–1890.

18.

Soldi-Lose

Schröder

Schwarz

. Computational and experimental exploration of the gas-phase chemistry of alkyloxalate ions ROCOCOO⁻ (R=H, CH₃, C₂H₅, i-C₃H₇, and t-C₄H₉). Int J Mass Spectrom 2008; 269: 62–70.

19.

Rodgers

Ervin

Armentrout

. Statistical modeling of collision-induced dissociation thresholds. J Chem Phys 1997; 106: 4499–4508.

20.

Narancic

Bach

Chen

. Simple fitting of energy-resolved reactive cross sections in threshold collision-induced dissociation (T-CID) experiments. J Phys Chem A 2007; 111: 7006–7013.

21.

Dunning

. Gaussian basis sets for use in correlated molecular calculations. I. The atoms boron through neon and hydrogen. J Chem Phys 1989; 90: 1007.

22.

Prascher

Woon

Peterson

et al.

Gaussian basis sets for use in correlated molecular calculations. VII. Valence, core-valence, and scalar relativistic basis sets for Li, Be, Na, and Mg. Theor Chem Acc 2011; 128: 69–82.

23.

Curtiss

Redfern

Raghavachari

. Gaussian-4 theory. J Chem Phys 2007; 126: 084108.

24.

Compton

Reinhardt

Cooper

. Collisional ionization of Na, K, and Cs by CO₂, COS, and CS₂: molecular electron affinities. J Chem Phys 1975; 63: 3821.

25.

Schröder

Schalley

Harvey

et al.

On the formation of the carbon dioxide anion radical CO₂⁻• in the gas phase. Int J Mass Spectrom 1999; 185: 25–35.

26.

DePuy

Gronert

Barlow

et al.

The gas-phase acidities of the alkanes. J Am Chem Soc 1989; 111: 1968–1973.

27.

Chen

H-T

Chang

J-G

Musaev

et al.

Computational study on kinetics and mechanisms of unimolecular decomposition of succinic acid and its anhydride. J Phys Chem A 2008; 112: 6621–6629.

28.

L-C

Zha

Tian

A-M

. Theoretical study on the microcosmic mechanism of succinic acid dehydration to produce succinic anhydride. Acta Chim Sin 2005; 63: 1545–1550.

29.

Curtiss

Redfern

Raghavachari

. Assessment of Gaussian-4 theory for energy barriers. Chem Phys Lett 2010; 499: 168–172.

30.

Polino

Klippenstein

Harding

et al.

Predictive theory for the addition and insertion kinetics of ¹CH₂ reacting with unsaturated hydrocarbons. J Phys Chem A 2013; 117: 12677–12692.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.31 MB

Unimolecular dissociation of anions derived from succinic acid (H 2 Su) in the gas phase: HSu − and ClMgSu − . Relationship to CO 2 fixation

Abstract

Keywords

Introduction

Methods

Experimental details

Quantum chemical calculations

Results

Mass spectra, dissociation reactions and breakdown curves of HSu− (m/z 117)

Mechanisms for dehydration and decarboxylation of HSu−

Mass spectra, dissociation reactions and breakdown curves of ClMgSu− (m/z 175)

Mechanisms for decarboxylation of ClMgSu−

Mechanism for formation of ClMg(η2-O2C)−

Discussion

Conclusion

Footnotes

Acknowledgements

Authors’ note

Declaration of conflicting interests

Funding

References

Supplementary Material

Mass spectra, dissociation reactions and breakdown curves of HSu⁻ (m/z 117)

Mechanisms for dehydration and decarboxylation of HSu⁻

Mass spectra, dissociation reactions and breakdown curves of ClMgSu⁻ (m/z 175)

Mechanisms for decarboxylation of ClMgSu⁻

Mechanism for formation of ClMg(η²-O₂C)⁻