162
L. Mammino
individually and also simultaneous binding to two or three geometrically suitable
sites. Particular attention was given to including all the predictably stable complexes.
The 58 complexes thus calculated can be considered a sufficiently representative
(though not exhaustive) ensemble.
The relative energies and the molecule-ion affinity (MIA) of the complexes provide indications about the Cu
2+ binding preferences for the different sites, showing
greater preference for Cu
2+ to bind simultaneously to O25 and O30, followed by
simultaneous binding to O25, O30 and O12, simultaneous binding to O8 and O18,
and binding to O30 alone (the preferred site when the ion binds to only one site).
Complexes in which the ion binds to these four sites account for relative energies up
to 22.65 kcal/mol and for the best MIA values. In all the calculated complexes, the
charge on the copper ion is reduced to less than +1, proving the ability of FNGB to
reduce the ion by transferring an electron to it and becoming a molecular ion with a
positive charge and an unpaired electron. Spin density maps show that this electron
is distributed in regions of the molecule far away from the ion.
Single point calculations in solution were added to evaluate the changes in the relative energies and other properties of the complexes, whose changes can realistically
be identified using a continuum solvent model with a single-point approach.
Tables reporting the values of relevant properties for all the calculated complexes
(relative energy, MIA, natural and Mulliken charge and Mulliken spin density on
the ion, distances of the ion from its binding site/s, comparisons of IHB parameters
in the uncomplexed FNGB molecule and in the complexes, etc.) and figures showing the geometries of the calculated uncomplexed conformers and of the calculated
complexes and the spin density maps in the complexes, as well as graphs illustrating
the trends of relevant quantities, are included in the Electronic Supporting Material
(ESM). Their numbering is independent of the numbering of the tables and figures
included in the text; their numbers are preceded by the letter S to clearly distinguish
them when they are referred to in the text.
2 Computational Details
Calculations were performed at the same level of theory utilized in the study of the
complexes of other antioxidant ACPLs with a Cu
2+ ion [6–8], i.e., Density Functional
Theory (DFT) with the B3LYP functional [35–37]. The conformational study of the
uncomplexed molecule utilised the 6-31+G(d,p) basis set, as the inclusion of diffuse
and polarisation functions had proved particularly relevant for DFT descriptions
of IHBs in ACPLs [26–28, 32, 33]). The calculation of complexes utilised the 631+G(d,p) basis set for the C, O and H atoms and the LANL2DZ pseudopotential [38]
for Cu
2+ , as this combination is known to enable better estimation of the moleculeion interaction for transition metals complexes [39, 40]. The reasons for selecting the
B3LYP functional have been explained in detail in [6] and are not repeated here; it
suffices to recall that the comparability of values (such as relative stabilities or binding
energies) remains good and, therefore, B3LYP is suitable and widely utilized [15]
L. Mammino
individually and also simultaneous binding to two or three geometrically suitable
sites. Particular attention was given to including all the predictably stable complexes.
The 58 complexes thus calculated can be considered a sufficiently representative
(though not exhaustive) ensemble.
The relative energies and the molecule-ion affinity (MIA) of the complexes provide indications about the Cu
2+ binding preferences for the different sites, showing
greater preference for Cu
2+ to bind simultaneously to O25 and O30, followed by
simultaneous binding to O25, O30 and O12, simultaneous binding to O8 and O18,
and binding to O30 alone (the preferred site when the ion binds to only one site).
Complexes in which the ion binds to these four sites account for relative energies up
to 22.65 kcal/mol and for the best MIA values. In all the calculated complexes, the
charge on the copper ion is reduced to less than +1, proving the ability of FNGB to
reduce the ion by transferring an electron to it and becoming a molecular ion with a
positive charge and an unpaired electron. Spin density maps show that this electron
is distributed in regions of the molecule far away from the ion.
Single point calculations in solution were added to evaluate the changes in the relative energies and other properties of the complexes, whose changes can realistically
be identified using a continuum solvent model with a single-point approach.
Tables reporting the values of relevant properties for all the calculated complexes
(relative energy, MIA, natural and Mulliken charge and Mulliken spin density on
the ion, distances of the ion from its binding site/s, comparisons of IHB parameters
in the uncomplexed FNGB molecule and in the complexes, etc.) and figures showing the geometries of the calculated uncomplexed conformers and of the calculated
complexes and the spin density maps in the complexes, as well as graphs illustrating
the trends of relevant quantities, are included in the Electronic Supporting Material
(ESM). Their numbering is independent of the numbering of the tables and figures
included in the text; their numbers are preceded by the letter S to clearly distinguish
them when they are referred to in the text.
2 Computational Details
Calculations were performed at the same level of theory utilized in the study of the
complexes of other antioxidant ACPLs with a Cu
2+ ion [6–8], i.e., Density Functional
Theory (DFT) with the B3LYP functional [35–37]. The conformational study of the
uncomplexed molecule utilised the 6-31+G(d,p) basis set, as the inclusion of diffuse
and polarisation functions had proved particularly relevant for DFT descriptions
of IHBs in ACPLs [26–28, 32, 33]). The calculation of complexes utilised the 631+G(d,p) basis set for the C, O and H atoms and the LANL2DZ pseudopotential [38]
for Cu
2+ , as this combination is known to enable better estimation of the moleculeion interaction for transition metals complexes [39, 40]. The reasons for selecting the
B3LYP functional have been explained in detail in [6] and are not repeated here; it
suffices to recall that the comparability of values (such as relative stabilities or binding
energies) remains good and, therefore, B3LYP is suitable and widely utilized [15]