Moschoula A. Trikoupis

Self-catalysis in the gas-phase: enolization of the acetone radical cation

Abstract Because of a prohibitively large barrier, the solitary acetone radical cation, CH3C(O)CH3 + (1 +) does not rearrange, neither spontaneously nor by activation, to its more stable enol isomer, CH2C(OH)CH3 + (1a +). However, this isomerization occurs smoothly by an ion–molecule interaction with neutral acetone itself. The dimer radical cation, [ 1 + ⋯ 1 ], generated under conditions of chemical ionization dissociates to m/z 58 and collision-induced dissociation (CID) experiments show that these ions have the enol structure 1a +. Labeling experiments indicate that the reaction can be viewed as a simple 1,3-hydrogen shift within the acetone radical cation of the complex. Ab initio calculations at the CBS-Q/DZP level of theory indicate that this isomerization is best described as a proton transport catalysis rather than as a spectator model. Our calculations show that the incipient radical formed during the proton abstraction is not CH3C(O)CH2 , but rather the less stable configuration CH3C(O )CH2 stabilized by CH3C(OH)CH3+. This behaviour can be rationalized by arguments based on ion-dipole interactions. The incipient radical CH3C(O )CH2 is transformed to its more stable configuration CH3C(O)CH2 via surface crossing. However, this process does not occur via the usual “minimum to minimum crossing” but rather by the novel process of “transition state to minimum crossing”. The abstracted proton is then donated back to the oxygen atom of CH3C(O)CH2 to yield the hydrogen-bridged radical cation [ 1a + ⋯ 1 ]. The observed tautomerization of the acetone radical cation by acetone itself can be viewed as “self-catalysis”.

Benzonitrile assisted enolization of the acetone and acetamide radical cations: proton-transport catalysis versus an intermolecular H+/· transfer mechanism

Abstract The acetamide radical cation, CH 3 C(O)NH 2 ·+ , can be induced to rearrange into its more stable enol isomer, CH 2 C(OH)NH 2 ·+ , by an ion–molecule interaction with benzonitrile, C 6 H 5 CN, under conditions of chemical ionization. (This enolization does not occur unassisted because of a prohibitively high energy barrier: 26 kcal/mol, from a CBS-QB3 calculation.) The initially formed [C 6 H 5 CN ⋯ acetamide] ·+ adduct ion isomerizes to a stable hydrogen bridged radical cation [C 6 H 5 CN ⋯ HOC(NH 2 )CH 2 ] ·+ en route to its dissociation into the enol ion. Multiple collision and deuterium labeling experiments on the acetamide/benzonitrile and the previously reported acetone/benzonitrile systems, indicate that the acetone ion enolizes by way of a base-catalyzed 1,3-proton shift (“proton-transport catalysis”) but that a different mechanism must be operative in the acetamide system. Ab initio and density functional theory calculations at the PMP3//RHF/D95∗∗ and PMP3//B3LYP/D95∗∗ level of theory support a mechanism which can be described as a consecutive H + /H · transfer between the partners of the [C 6 H 5 CN ·+ ⋯ acetamide] encounter complex. The calculations provide a rationale for the observed isotope effects and lead to a tentative explanation for the differences in interaction of the title ions with benzonitrile.

Hydrogen-shift isomers of ionic and neutral hydroxypyridines: a combined experimental and computational investigation

Abstract Apart from pyridine N-oxide ( 1a ), the C 5 H 5 NO family of stable molecules comprises, 2-, 3- and 4-hydroxypyridine ( 2a , 3a and 4a ) as well as their keto counterparts 2-, 3- and 4(1 H )-pyridone ( 2b , 3b and 4b ). This study focuses on the characterisation of their radical cations and a number of stable H-shift isomers, which are α- or β-distonic ions. This was done by using a combination of mass spectrometric experiments and computational chemistry, at the B3LYP/CBSB7 level of theory. The ionic species were identified on the basis of both their collision-induced dissociation (CID) characteristics and specific associative ion–molecule reactions with dimethyl disulfide and tert -butyl isocyanide as substrates. The distonic ions ( 1b + , 2c + , 2d + and 3c + ) were obtained by dissociative electron impact ionisation and subjected to neutralisation–reionisation mass spectrometry (NRMS). From CID spectra of the intense survivor ions, it follows that the neutral counterparts of the α-distonic ions 2c + and 3c + are viable chemical species in the rarefied gas phase. The energy-rich ylide type neutrals 1b , on the other hand, readily isomerise into pyridine N-oxide, 1a , or else dissociate. The neutral counterpart of the β-distonic ion 2d + only has a marginal stability and part of these neutrals are proposed to isomerise into energy-rich 2-pyridone molecules 2b . This is in agreement with the computational results. However, ionised 2-pyridone cannot readily be differentiated from its enol isomer 2-hydroxypyridine. In contrast, the keto isomers of ionised 3- and 4-hydroxypyridine display characteristically different CID spectra.

How do dimethyl oxalate ions CH3O-C(=O)-C(=O)-OCH3·+ break in half? Loss of CH3· + CO2 versus CH3O-C-O·

The interesting unimolecular dissociation chemistry of dimethyl oxalate (DMO) ions, CH3O-C(=O)-C(=O)-OCH3·+, has been studied by vacuum ultraviolet photoionization and tandem mass spectrometry based experiments. The measured appearance energy (AE) for the generation of CH3O-C=O+ (10. 5 eV) is not compatible with a simple bond cleavage involving the cogeneration of the radical CH3O-C=O· whose calculated AE is 11 kcal/mol higher. However, because the CH3O-C=O· radical is thermodynamically less stable than its dissociation products CH3· and CO2, by 19 kcal/mol, a two-step dissociation of ionized DMO into CH3O-C=O+ + CH3· + CO2 is energetically feasible. Collision induced dissociative ionization experiments clearly show that low energy DMO ions dissociate into CH3· + CO2 without the intermediacy of CH3O-C=O·. Experiments using a charged collision chamber further indicate that CO2 is released first, followed by loss of CH3· and not vice versa and a mechanism is proposed. The measured AE, which we assign to the two-step process, is 8 kcal/mol higher than the calculated value. This could be due to a competitive shift caused by a prominent low energy decarbonylation reaction yielding the hydrogen bridged radical cation CH2=O … H … O=C-OCH3·+. However, from metastable ion observations and AE measurements on deuterium labeled DMO ions, it follows that there is no competitive shift and that the elevated AE for the two-step process corresponds to the barrier for the first step, loss of CO2. Finally, neutralization-reionization experiments on ionized DMO and CH3O-C=O+ provide evidence for the existence of CH3O-C=O· as a kinetically stable radical.

The decarbonylation of the acetamide radical cation and the enolization of its dimer by self-catalysis

The acetamide radical cation, CH3C(=O)NH2•+, and its enol, CH2=C(OH)NH2•+, undergo several unimolecular reactions in the μs time-frame of which decarbonylation is predominant. This reaction produces the ylid ion CH2NH3•+, rather than CH3NH3•+ [J. Am. Chem. Soc. 109, 4819 (1987)]. A previously proposed mechanism via ion–dipole complexes is confirmed by the present CBS-QB3 calculations. These calculations reveal the existence of a second mechanism which proceeds via the enol ion and the distonic ion •CH2C(=O)NH3+. Both mechanisms can account for previously reported isotopic labeling experiments. Tandem mass spectrometry based experiments do not provide evidence that the non-decomposing acetamide ions rearrange to any significant extent to the more stable enol form. However, this transformation occurs smoothly by interaction with a neutral acetamide molecule (“self-catalysis”). By integration of experimental data (MS/MS/MS and labeling experiments) and ab initio calculations [CBS-Q (RHF/DZP)] three mechanisms for this assisted tautomerization have been traced. In the first mechanism the neutral acetamide component of the dimer ion accepts a C–H proton from its ionic partner and then donates the proton back to the oxygen atom. This is an example of “proton-transport catalysis”. In the second mechanism, isomerization takes place within the ionic partner via a conventional 1,3-H shift. The neutral partner serves only to lower the energy of the transition state by ion–dipole attractions. This is an example of the “Spectator” mechanism. In the third mechanism, proton transfer from the ionic partner to its neutral counterpart is followed by back-donation of a hydrogen atom. This is an example of the “Quid-pro-Quo” mechanism. The behavior of the acetamide dimer ion is compared to that of the acetone dimer ion which undergoes only proton-transport catalysis.

Dissociation chemistry of the hydrogen-bridged radical cation [CH2O ⋯ H ⋯ OC–OCH3]·+: proton transport catalysis and charge transfer

Abstract Tandem mass spectrometry based experiments on the decarbonylation products of ionized methyl-β-hydroxypyruvate (MHP) and dimethyloxalate (DMO) show that the hydrogen-bridged radical cation (HBRC) CH2O ⋯ H ⋯ OC–OCH3·+ is a stable species in the gas phase. Its low energy dissociation products are protonated methylformate, HOC(H)OCH3+, and the formyl radical, HCO·. The HBRC isomers HOCH2C(O)OCH3·+ (ionized methylglycolate) and (CH3O)2CO·+ (ionized dimethylcarbonate) show the same dissociation characteristics. Deuterium labeling experiments dictate that loss of HCO· from the title HBRC cannot be formulated as a simple H shift from the formaldehyde moiety to the C atom of the OC·–OCH3 group. Ab initio molecular orbital (MO) calculations support the proposal that this dissociation proceeds via sequential transfers of a proton, electron, and another proton within ion–dipole complexes. The first step in this rearrangement process is a 1,2-proton shift catalyzed by a formaldehyde dipole. This yields an ion/dipole complex, CH2O ⋯ H–C(O)OCH3·+, that is in the correct configuration for electron transfer to occur at the energetic threshold dictated by experiment. The resulting intermediate triggers the transfer of yet another proton from the formaldehyde unit, thereby generating another stable H-bridged radical cation viz. HCO ⋯ H ⋯ OC(H)OCH3·+. This final intermediate dissociates with little or no activation energy into HOC(H)OCH3+ and HCO·. It is further predicted by the calculations that ionized methylglycolate isomerizes into the title HBRC by a fairly high barrier that makes the communication between ionized methylglycolate and dimethylcarbonate via the title ion quite unlikely; instead an alternative route for this communication is proposed.

The heat of formation of sulfine, CH2SO, revisited: a CBS-QB3 study

Abstract The heat of formation ( ΔH f ) of sulfine, CH 2 SO, has been determined by the CBS-QB3 quantum chemical method, using 10 reactions, including the isodesmic reaction CH 2 SO+SO 2 →CH 2 S+SO 3 . The derived ΔH f of sulfine, −30±6 kJ / mol at 298 K, lies midway between two previously calculated values: −9±14 and −52±10 kJ/mol. The CBS-QB3 derived ΔH f (0 K) was very recently validated against the very accurate Weizmann-1 ′ (W1 ′ ) method [L.N. Heydorn, et al., Z. Phys. Chem. 215 (2001) 141] and there is excellent agreement between the two methods, within 3 kJ/mol. Our recommended value is evaluated against experimental observables, such as the measured proton affinity of CH 2 SO and the appearance energy of CH 2 S + OH from dimethyl sulfoxide ions, CH 3 S(O)CH 3 + .

Generation and characterization of ionic and neutral (CH3OBH)+/· and (CH3BOH)+/· in the gas phase by tandem mass spectrometry

The isomeric dicoordinated borinium ions CH3O–B–H+ and CH3–B–OH+ are generated upon electron ionization of trimethylborate and methyl boronic acid, respectively. The connectivity of the ions is confirmed by collision-induced dissociation experiments on magnetic deflection type tandem mass spectrometers. Neutralization–reionization experiments on these structurally characterized ions indicate that the neutral radicals CH3O–B–H· and CH3–B–OH· are viable species in the gas phase. Calculations at the G2 level of theory were performed to obtain thermochemical data on the title isomers and their main dissociation products. The calculations also provide a rationale for the moderate yield of the neutrals generated in the experiments: the vertical electron transfer processes for both systems are associated with particularly unfavourable Franck-Condon factors.

The influence of the size and structure of a spectator alkyl group on the relative rates of alkyl radical elimination from ionised tertiary amines

The relative rates of alkyl radical elimination by α-cleavage of ionised amines of general structure, CH3CH2CH2(CH3CH2)CHN(CH3)R+• or CH3CH2CH2CH2(CH3CH2CH2)CHN(CH3)R+•, where R = n-C n H2n + 1 (n = 1–10, 12 or 14), iso-C5H11, CH2CH(CH3)C2H5, neo-C5H11 or CH2CH2C(CH3)3, are reported. The size of the spectator alkyl group, R, affects the ratio of ethyl and propyl radical loss from metastable ionised amines containing a 3-hexyl group; the slight preference for expelling the smaller ethyl radical increases initially before gradually falling as n increases from 1 to 14. In contrast, the degree of branching in isomeric pentyl or hexyl spectator alkyl groups has very little effect on the ratio of ethyl and propyl radical loss. For faster reactions occurring in the ion source, there is at most a marginal variation in the ratio of ethyl to propyl radical loss as n increases. In general, the kinetic energy release which accompanies loss of either radical increases slowly with n, but remains small, as would be expected if the reaction occurred without appreciable reverse critical energy. Similar trends are found for the relative rates of propyl and butyl radical elimination from ionised amines containing a 4-octyl group.

Dissociation reactions of low-energy pentenyl methyl ether radical cations C 5 H 9 OCH 3

The dissociation chemistry of the low-energy C5H9OCH3•+ ions generated from the 13 isomeric pentenyl methyl ethers derived from stable alkenols has been studied. This was done by examining their metastable ion characteristics, in conjunction with 2H and 13C-labelling as well as collision-induced dissociation and neutralisation–reionisation experiments. The influence of the position and substitution pattern of the double bond on the chemistry of these C6H12O•+ species is considered. The closely similar reactions of C2H5CH=CHCH2OCH3•+, 3•+, CH2=CH–CH(C2H5)OCH3•+, 4•+, and CH2=C(C2H5)CH2OCH3•+, 13•+, point to a common chemistry, which is rationalised in terms of facile 1,2-H and 1,2-C2H5 shifts via distonic ions. Each of the other isomers displays a distinct, though often related, chemistry. The eight allylic ionised ethers easily lose CH3• to produce C5H9O+ oxonium ions, whose structure was established by CID experiments; ions 3•+ / 4•+ / 13•+ also readily expel C2H5• to give C4H7O+ ions of structure CH2=CH–C+(H)OCH3. Elimination of CH3OH is also significant for 3•+ / 4•+ / 13•+ and for (CH3)2C=CHCH2OCH3•+, 8•+, and CH3CH=C(CH3)CH2OCH3•+, 11•+. Besides expelling CH3• and/or C2H5• and CH3OH, the three homoallylic isomers undergo dissociations which are (almost) absent for their allylic counterparts: thus, both CH3CH=CH(CH2)2OCH3•+, 2•+, and CH2=CH–CH(CH3)CH2OCH3•+, 10•+, lose H• and H2O, whereas CH2=C(CH3)CH2CH2OCH3•+, 7•+, is unique in predominantly losing CH2O. For the losses of CH2O and H2O mechanisms are proposed in which ion–neutral complexes of the type [C5H10•+ / CH2O] and [C6H10•+ / H2O] are key intermediates. The behaviour of the non-(homo)allylic isomer, CH2=CH(CH2)3OCH3•+, 1•+, is similar to that of 2•+ but the reactions occur in different proportions. A mechanism for the facile loss of an alkyl radical from 1+ is proposed in which 1,4-H shifts and distonic ions as well as communication with ionised cyclopentyl methyl ether, 14•+, play an important role.