24666
J. Phys. Chem. B 2006, 110, 24666-24673
Structural and Electronic Characterization of the Complexes Obtained by the Interaction between Bare and Hydrated First-Row Transition-Metal Ions (Mn2+, Fe2+, Co2+, Ni2+, Cu2+, Zn2+) and Glycine Tiziana Marino,† Marirosa Toscano,† Nino Russo,*,† and Andre´ Grand‡ Dipartimento di Chimica and Centro di Calcolo ad Alte Prestazioni per Elaborazioni Parallele e Distribuite-Centro d’Eccellenza MIUR, UniVersita` della Calabria, I-87030 ArcaVacata di Rende (CS), Italy, and De´ partement de Recherche Fondamentale sur la Matie´ re Condense´ e, SerVice de Chimie Inorganique et Biologique, CEA-Grenoble, 17 Rue des Martyrs, 38054 Grenoble Cedex 9, France ReceiVed: July 20, 2006; In Final Form: September 20, 2006
The complexes formed by the simplest amino acid, glycine, with different bare and hydrated metal ions (Mn2+, Fe2+, Co2+, Ni2+, Cu2+, Zn2+) were studied in the gas phase and in solvent in order to give better insight into the field of the metal ion-biological ligand interactions. The effects of the size and charge of each cation on the organization of the surrounding water molecules were analyzed. Results in the gas phase showed that the zwitterion of glycine is the form present in the most stable complexes of all ions and that it usually gives rise to an η2O,O coordination type. After the addition of solvation sphere, a resulting octahedral arrangement was found around Ni2+, Co2+, and Fe2+ ions in their high-spin states, whereas the bipyramidaltrigonal (Mn2+ and Zn2+) or square-pyramidal (Cu2+) geometries were observed for the other metal species, according to glycine behaves as bi- or monodentate ligand. Despite the fact that the zwitterionic structure is in the ground conformation in solution, its complexes in water are less stable than those obtained from the canonical form. Binding energy values decrease in the order Cu2+ > Ni2+ > Zn2+ ≈ Co2+ > Fe2+ > Mn2+ and Cu2+ > Ni2+ > Mn2+ ≈ Zn2+ > Fe2+ > Co2+ for M2+-Gly and Gly-M2+(H2O)n complexes, respectively. The nature of the metal ion-ligand bonds was examined by using natural bond order and charge decomposition analyses.
Introduction The simplest and smallest glycine (Gly) amino acid was used extensively as a model compound for theoretical and experimental investigations devoted to the elucidation of the interactions occurring between metal cations and biological systems in many life processes.1-24 It was shown previously that these interactions are not simple to predict because different coordination modes as well as bond types are possible due to the presence of amine nitrogen and carboxylic groups in the R-amino acid backbone. Previous theoretical studies addressed the attention to the analysis of the interaction of bare alkali 2,5,6,16,18,19,21 and transition-metal cations3-10,13,15-17,21,23,24 with glycine, but until now, the metalation of Gly by hydrated cations represents a poorly explored field.15,16,25-27 To our knowledge, the few studies reported to date explored the coordination process of hydrated alkali cations by small amino acid models.25,28-32 More recently, a theoretical investigation of the gradual hydration effect on the geometry of Gly-Mn+(M ) Li, Na, K, Mg, Ca, Ni, Cu and Zn, n ) 1 or 2) complexes appeared in the literature.16 Our study focused on the structure and relative stabilities of the different chelating forms of glycine with the differently * To whom correspondence should be addressed. Fax: +39-0984492044. E-mail:
[email protected]. † Dipartimento di Chimica and Centro di Calcolo ad Alte Prestazioni per Elaborazioni Parallele e Distribuite-Centro d’Eccellenza MIUR, Universita` della Calabria. ‡ De ´ partement de Recherche Fondamentale sur la Matie´re Condense´e, Service de Chimie Inorganique et Biologique, CEA-Grenoble.
hydrated bivalent cations of some transition metals, namely, Mn2+, Fe2+, Co2+, Ni2+, Cu2+, and Zn2+ having filled or partially filled valence d orbitals. The choice of these transitionmetal ions derived mainly from the important role that they play in many biochemical processes.3,4,8,9,12,13,21,33,34 Such species, participating as multiply charged ions in biological systems, bind much more tightly to ligands because of the enhanced electrostatic potential and smaller ionic radius with respect to the nontransition-metal ions.27 A deep knowledge of these interactions can be helpful when interpreting the specific fragments formed under massspectrometry conditions. Although accurate thermodynamic information on the noncovalent interactions between alkali metal ions, amino acids, and water molecules are accessible at the experimental level,27,35-38 little material is available for the interactions of biomolecules with multiply charged metals.20,23,24 In these cases, in fact, experimental studies are restricted to the binding energy value determination for the second solvation shell39,40 because in the first one the ligands are bound too tightly to be measured by most experimental methods. Computational Methods All calculations were carried out using the Gaussian 03 package.41 The non-local hybrid three-parameter B3LYP42 density functional method together with the 6-311++G** basis set for all atoms were used to obtain the equilibrium geometries and harmonic vibrational frequencies of all of the considered structures. It is worth noting that the B3LYP functional was proven to give reliable binding energies for such a type of system. This
10.1021/jp0645972 CCC: $33.50 © 2006 American Chemical Society Published on Web 11/14/2006
Structural and Electronic Characterization is because it is able to balance reasonably the errors inherent in DFT resulting from the use of an approximate exchange functional (self-interaction error) and from the monodeterminal description of the wave function, which causes a inappropriate description of the nondynamical correlation energy. On the basis of our experience43-45 and of different studies in literature,46,47 the average error of B3LYP functional on the metal-ligand bond strengths in a variety of transition-metalcontaining systems can be quantified at about 3-5 kcal/mol. The self-consistent reaction field polarizable continuum model (SCRF/PCM) procedure48-50 was used to evaluate the solvent effect on all hydrated complexes in their fixed B3LYP/DZVP geometry. The surrounding water was modeled using the standard value of 78 for the dielectric constant. Previous studies performed at the same level of theory showed that this computational protocol was adequate to describe the metal-ligand interactions in biological systems.3,4,6,7,15,16,21,51-55 On the basis of many literature data13,14,22 suggesting that the singlet and the doublet represent the ground state, respectively, for Zn(II) and Cu (II) complexes, we performed calculations only for the most stable spin state of these cations. For iron (II), manganese (II), cobalt (II), and nickel (II), most stable naked and hydrated complexes, both high-spin and low-spin states were taken into account. Binding energies were obtained as the negative enthalpy variation (-∆H) at 298 K for the metalation processes. The nature of the metal-ligand chemical bond was established by applying the natural bond order (NBO)56 and charge decomposition (CDA) analyses.57,58
J. Phys. Chem. B, Vol. 110, No. 48, 2006 24667
Figure 1. Common geometries of Gly-M2+ complexes.
Results and Discussion Experimental evidence59-63 and theoretical calculations64-71 of the molecular structure of isolated Gly revealed that this small biological molecule can exist in several stable forms. According to previous theoretical gas-phase studies on glycine, performed at the same our level of theory,63,64,69,70 the five most stable isomers of the amino acid, included in a very narrow range of energy, were chosen to be employed in the metalation process by bare and hydrated ions (Mn2+, Fe2+, Co2+, Ni2+, Cu2+, Zn2+). Because of the presence of studies concerning the interaction of Gly with Co2+ and Cu2+ in literature,3,7 the investigation concerning the complexes with naked cations was performed on the remaining four metal species. Several coordination modes to glycine isomers were considered for M2+. In particular, the starting structures involve the coordination of ions on the different electron-rich sites of ligand, in an attempt to retain the Gly intramolecular hydrogen bonds as much as possible. Among all binding possibilities that bare metal cations usually have when they interact with R-amino acids, the experimental studies71-73 indicated the η2-N,O as the preferred one because such a type of coordination gives rise to stable five-membered chelate rings. Theoretical studies, instead, showed that in several cases an alternative way for the metal binding is the η2-O,O coordination to the zwitterionic form of amino acid.3,4,6-10,50,51 It is known that the zwitterionic form in the gas phase, in the absence of other stabilizing forces, is not a stable species;59,63 however, it is stabilized in aqueous solution or in the presence of cations with which it gives rise to stable salt-bridge structures. With reference to the Gly-M2+ system, in some cases (i.e., Mn2+, Fe2+, and Ni2+) we have observed that the proton-transfer process that generates the zwitterion from the neutral form occurs spontaneously. This can be deduced from the collapse,
Figure 2. Common geometries of Gly-M2+(H2O)n (n ) 4 or 5) complexes
during the optimization, of the first into the second structure. On the contrary, previous theoretical works on Cu2+ and Co2+ ions3,7 showed that both minima are possible and that their interconversion probably requires getting through an important energy barrier. Although the metal ions under investigation can accommodate a variable number of water molecules in their coordination sphere, we considered as a starting point a hexacoordination for each of them. In most cases, this fact entailed the addition of four water molecules to the structures obtained with bare metal cations. Instead, in the systems characterized by the η1O(COOH) coordination, it was necessary to consider a pentahydrated metal ion, [M(H2O)5]2+, to reach the hexacoordination of the metal center. As far as the electronic spin states of naked and hydrated complexes are concerned, the results indicated ground states for the quintet, sextet, quartet, and triplet for Fe2+, Mn2+, Co2+, and Ni2+, respectively. This part of the investigation allowed us to establish that no variation of spin multiplicity occurs upon increasing coordination to the metal cations, except for nickel. The most stable structures obtained for Gly-M2+ and Gly2+ M (H2O)n systems were depicted in Figures 1 and 2, together with the numbering scheme. Because of some differences occurring in the cases of hydrated Mn2+, Cu2+, and Zn2+, a new figure (Figure 3) was introduced collecting the exceptions and the new labels for them. In Table 1 the main structural parameters of the optimized metallic complexes for both Gly-M2+ and Gly-M2+(H2O)n
24668 J. Phys. Chem. B, Vol. 110, No. 48, 2006
Figure 3. Common geometries of Gly-M2+(H2O)4 complexes: some particular cases.
systems are given. In Table 2 the computed absolute and relative energies for both bare and hydrated complexes are reported. Table 3 shows the charge values of metal ions into Gly-M2+, Gly-M2+(H2O)4, and M2+(H2O)4 complexes. Gly-Mn2+ Complexes. As in the case of other divalent metal cations,3,5-8,21,74 the most stable structure, Gly-Mn2+ (d) (see Figure 1), corresponds to the metal cation interacting with both oxygens of zwitterionic glycine. Mn2+-Oa and Mn2+-Ob distances are quite different being 2.038 and 2.100 Å, respectively (see Table 1). The next minimum structure, Gly-Mn2+(a), has the Mn2+ ion linked to the nitrogen and to carbonyl oxygen atoms and lies 6.8 kcal/mol above the most stable one. The Gly-Mn2+(b) and (c) relative minima lie at 12.3 and 27.2 kcal/mol above the global one, respectively. The former is characterized by the same coordination type of complex (a) although its carboxylic group is in the trans position, whereas the latter shows the η2-N,OH coordination of Mn2+ in which the less basic -OH group of the carboxyl is involved (see Figure 1). The structure (e) is not a minimum because it collapses into the most stable Gly-Mn2+ (d) during the optimization. Starting from the Gly-Mn2+ structures, four or five water molecules were added to reach the octahedral coordination of the Mn2+. Optimizations indicated that in the global minimum, which originates from the Gly-Mn2+ (d), four H2O molecules are present (see D′ of Figure 3). As can be observed from this Figure, glycine acts as a monodentate ligand. The coordination sphere of the cation assumes a bipyramidal-trigonal geometry. The further addition of a water molecule to this last minimum again proposes the glycine as a monodentate ligand but an octahedral environment of Mn2+. Through the binding energy evaluation for the reaction Gly-Mn2+(H2O)n f Gly + Mn2+(H2O)n (where n ) 4, 5), we have established that the tetrahydrated complex (D) is more stable than the corresponding pentahydrated (E) one (73.8 vs 43.8 kcal/mol, respectively). In Gly-Mn2+(H2O)4 (D) and Gly-Mn2+(H2O)5 (E), the distance between the Mn2+acceptor and the oxygen glycine atom donor is 2.068 and 2.153 Å, respectively. The differences between these two values depend on the fact that the strength with which manganese ion can bind additional ligands decreases gradually. The addition of four water molecules to the Gly-Mn2+ complexes in which the cation is η2-N,O-coordinated (see a-c
Marino et al. of Figure 1) gives rise to three systems all involving the glycine ligand still bidentate and an octahedral disposition around the cation. Metal-ligand distances are reported in Table 1. From these values, it emerges that Mn2+ is always more strongly bound to the oxygen atom. Relative energy values (see Table 2) show that hydrated complexes follow the same stability order like that of naked species although the gaps between them are enhanced by the presence of water molecules. Gly-Fe2+ Complexes. Previous theoretical studies, performed on glycine-Fe2+ complexes with different multiplicities, established that for these systems the quintet represents the ground state.9,10 Contrary to the study of Ai et al.9 in which only two glycine conformers were selected for the cation binding, we have considered five low-energy isomers. It is worth noting that form e (see Figure 1) does not exist as a stable minimum because during the optimization process a spontaneous proton shift between the -OH group and amine nitrogen occurs, giving rise to the most stable isomer (d) (see Figure 1). The relative energy values propose the following order of stability for the Gly-Fe2+ adducts: d > a > b > c. As in the case of other divalent cations studied here and in other works,3,5-8,21,74 the structure with which the ligand prefers to coordinate the metal cation is the zwitterionic one. In the complex (d), Fe2+-Oa and Fe2+-Ob distances are 1.988 and 2.058 Å, respectively (see Table 1). The addition of water molecules causes noticeable changes neither in the coordination modes nor in the stability order of the complexes (see Table 2). It is interesting that the metal center in all five hydrated complexes appears to be hexacoordinated, showing a preference for the octahedral geometrical environment (see Figure 2). In the hydrated complex (D), Fe2+-Oa and Fe2+-Ob distances become longer than those in the corresponding naked species to balance the effects of incoming water molecules and the consequent steric hindrance. Form A, in which the η2-N,O coordination is present, lies 4.1 kcal/mol above the global minimum and proposes again a major affinity of the cation toward oxygen rather than the nitrogen atom (see Table 1 for bond lengths). The B and C omplexes lie at high energy, although the gaps (8.1 and 19.3 kcal/mol) that separate them from the most stable structures (D and A) appear reduced by the presence of water molecules. The hydrated complex (E) represents the only case in which Gly behaves as monodentate ligand favoring the entrance of a fifth water molecule. Gly-Co2+ Complexes. As mentioned before, because of the presence in the literature of a B3LYP investigation on the interaction of glycine with a Co2+ bare ion,7 computations in this case were performed only for the hydrated form of the cation. For purpose of comparison with the other bare metal ions examined here, we have reported, in Table 2, the results of this previous theoretical work7 together with our new data. The Gly-Co2+(H2O)n (n ) 4 or 5) systems containing the divalent ion in the doublet state are on average less stable by 15.0 kcal/mol with respect to the corresponding adducts with the cation in the quartet spin state (4F); thus, the discussion will concern only these last complexes. The Gly-Co2+(H2O)n complexes follow the same stability order found by Sodupe et al.7 for the analogous systems without water: D > A > B > C (see Table 2). As observed for most of the cations examined, the energy gaps between hydrated complexes become smaller with respect to those of naked systems.
Structural and Electronic Characterization
J. Phys. Chem. B, Vol. 110, No. 48, 2006 24669
TABLE 1: B3LYP/6-311++G** Distances (Å) between Metal Ions and Coordinating Atoms of the Glycine Molecule for Both Gly-M2+(H2O)n (n ) 4 or 5, M ) Mn2+, Fe2+, Co2+, Ni2+, Cu2+, and Zn2+) and Gly-M2+ (in parentheses) Systems species
M2+‚‚OadC
A (a) B (b) C (c) D′ (d) E (e)
2.168 (1.986) 2.170 (1.966)
M2+‚‚ObdC
M2+‚‚NH2
M2+‚‚OH
2.332 (2.175) 2.350 (2.193) 2.276 (2.241)
2.248 (2.012)
2.230 (2.121) 2.284 (2.137) 2.192 (2.081)
2.239 (1.943)
2.185 2.180 2.147
2.155
2.128 (2.004) 2.134 (2.009) 2.098 (1.998)
2.098 (1.893)
2.044 2.053 2.016
2.127
2.032(2.157) 2.181(2.040) 2.118(2.004)
2.204(1.925)
Mn2+
2.068 (2.038) 2.153
-(2.100) Fe2+
A (a) B (b) C (c) D (d) E (e)
2.147 (1.918) 2.088 (1.900)
A (a) B (b) C (c) D (d) E (e)
2.108 2.096
A (a) B (b) C (c) D (d) E (e)
2.056 (1.897) 2.035 (1.883)
A′ (a) B′ (b) C (c) D′′ (d) E (e)
1.997 1.977
A (a) B (b) C (c) D′ (d) E (e)
2.143 (1.933) 2.105 (1.917)
2.069 (1.988) 2.131
2.458 (2.058) Co2+
2.084 2.086
2.293 Ni2+
2.081 (1.932) 2.055 (-)
2.207 (1.971) Cu2+
1.959 Zn2+
1.993 (1.992) 2.080 (1.912)
-(2.050)
In the Gly-Co2+(H2O)4 (D) global minimum, the cobalt ion lies on the plane containing the two oxygens of the zwitterionic form of the Gly (Co2+-Oa and Co2+-Ob distances are 2.084 and 2.293 Å, respectively) and the oxygens of the two water molecules (at 2.123 and 2.099 Å). The other water molecules occupy the axial positions. The resulting geometry is almost octahedral (see Figure 2). Both the A and B species exhibit the η2-N,O coordination mode, whereas the C complex has η2-N,OH coordination. System E appears to be monocoordinated. In this last structure, the cobalt ion interacts with the carbonyl oxygen (η1-O,COOH coordination) of the neutral glycine conformer that shows an intramolecular hydrogen bond having a length of 2.086 Å. This hydrogen bond becomes longer (2.251 Å) after the interaction of Gly with the bare cation7 and shorter (1.678 Å) when the hydrated cobalt is involved. This can be explained by the fact that the bare ion, contrarily to the hydrated species, subtracts negative charge from the oxygen to which it is directly coordinated and hence, because of an inductive effect, from the hydroxyl one, weakening its donor nature. The remaining species Gly-Co2+(H2O)4 (A-C) lying at 3.4, 11.6, and 17.6 kcal/mol above the global minimum exhibited distorted octahedral geometry. Gly-Ni2+ Complexes. The only information concerning the interaction of glycine with the nickel cation derives from previous theoretical works that examine the Ni+-glycine complexes.4,6 For both Gly-Ni2+ and Gly-Ni2+(H2O)n systems, our B3LYP/
6-311++G** calculations indicated the triplet electronic configuration as the preferred one. In agreement with the experimental evidence,13,14 the complexes containing the nickel cation with different multiplicity show a different coordination geometry. In particular, in the triplets ligands assume an octahedral disposition, whereas in the singlets a square-planar disposition is adopted. For the nickel, contrary to the other cations, structure e does not exist because it collapses into the most stable one (d), during the optimization procedure. The global minimum (d) shows the usual η2-O,O coordination of the cation with the two Ni2+-O distances slightly different (1.932 Å with Oa and 1.971 Å with Ob). It is worth noting that the most stable complex deriving from the interaction of glycine with Ni+ exhibits an η2-N,O coordination of metal4 as other mono-charged species. All Gly-Ni2+ minima are enclosed in a range of energy of 20.7 kcal/mol (see Table 2). In particular, the η2-N,O complex (a) arising from the canonical form of Gly lies at only 1.3 kcal/ mol. Starting from the same canonical form with a trans orientation of carboxylic group, a complex (b) less stable by 7.1 kcal/mol with respect to the global minimum (see Table 2 and Figure 1) is obtained. Again, the distances of metal cationcarbonyl oxygen are smaller than those with nitrogen (1.897 and 2.004 Å in a and 1.883 Å and 2.009 in b, respectively). In the less-stable structure (c), the metal ion is linked to the hydroxyl oxygen and amino nitrogen atoms with distances of 1.893 and 1.998 Å, respectively. The Gly-Ni2+(H2O)4 complexes, compared to the correspond-
24670 J. Phys. Chem. B, Vol. 110, No. 48, 2006
Marino et al.
TABLE 2: Gas-Phase B3LYP/6-311++G** Absolute (au) and Relative (kcal/mol) Energies at 0 K for Both Gly-M2+ and Gly-M(H2O)n2+ (n ) 4 or 5, M ) Mn2+, Fe2+, Co2+, Ni2+, Cu2+, and Zn2+) Complexesa Gly-Mn 2+(H2O)n Gly-Mn2+ A (a) B (b) C (c) D′(d) E (e) A (a) B (b) C (c) D (d) E (e) A (a) B (b) C (c) D (d) E (e) A (a) B (b) C (c) D (d) E (e)
-1434.814230 -1434.805467 -1434.781669 -1434.825063
6.8 12.3 27.2 0.0
-1740.785454 -1740.778062 -1740.761532 -1740.801225 -1817.252264
Gly-Fe2+ -1547.489279 -1547.480817 -1547.453855 -1547.496927
4.8(4.3,b 2.3c) 10.1 27.0 0.0(0.0,b0.0c)
-1853.465285 -1853.458958 -1853.441010 -1853.471796 -1930.135361
2.3e
-1972.504118 -1972.490988 -1972.481462 -1972.509507 -2048.968757
Gly-Co2+ 6.3d 11.1d 30.7d 0.0 c (3.0) 35.4d (36.5)
0.0e
Gly-Ni2+ -1792.049272 -1792.040057 -1792.018357 -1792.051381
A′ (a) B′ (b) C (c) D′′(d) E (e)
Gly-Cu2+ f 6.1 11.8 27.2 0.0 14.6
A (a) B (b) C (c) D′(d) E (e)
Gly-Zn2+ -2063.101481 -2063.091188 -2063.067933 -2063.102318 -2063.050952
-2098.039880 -2098.032822 -2098.016542 -2098.041136 -2174.501942
1.3 7.1 20.7 0.0
0.5(5.0g) 6.9(-) 21.6(24.9g) 0.0(0.0g) 31.7(41.8g)
9.9 14.5 24.9 0.0 Gly-Fe2+(H2O)n 4.1 8.0 19.3 0.0
µ
PCM
BE
4.232 4.196 8.051 3.634 2.152
0.0 3.6 5.8 1.5
170.1 (67.7)
4.345 3.916 7.989 4.469 1.986
0.0 4.5 5.3 8.5
187.8 (62.1)
0.0 4.7 5.0 4.4
203.3 (61.3)
0.0 2.5 5.8 5.8
208.9 (77.6)
Gly-Co2+(H2O)n 3.4 3.597 11.6 4.610 17.6 7.317 0.0 5.128 1.477 Gly-Ni2+(H2O)n 0.8 5.2 15.4 0.0
4.126 4.348 7.068 5.375 0.875
-2230.197702 -2230.189547 -2230.163891 -2230.203282 -2306.656349
Gly-Cu2+(H2O)n 3.5 2.682 8.6 4.956 24.7 6.891 0.0 5.844 0.690
0.0 3.3 8.3 1.9
243.0 (83.5)
-2369.077590 -2369.070326 -2369.053675 -2369.093294 -2445.540363
Gly-Zn2+(H2O)n 9.8 3.515 14.4 3.848 24.8 7.458 0.0 4.435 1.029
0.0 3.2 5.5 2.2
203.2 (67.4)
a Solvent relative energies for hydrated systems are in kcal/mol. Binding energy values for both naked and (hydrated) systems are given at 298 K. Dipole moments (Μ) are in Debye (D). b B3LYP, from ref 9. c BHLYP, from ref 9. d B3LYP, from ref 7. e BHLYP, from ref 7. f B3LYP/ D95++(d,p), from ref 3. g MP2/basis1, from ref 8.
TABLE 3: Charge on the Metal Ion (Q) Is in |e| species Mn2 Fe2+ Co2+ Ni2+ Cu2+ Zn2+
Gly-M
2+
1.688 1.608 1.472 1.637
2+
2+
M (H2O)4
Gly-M (H2O)4
1.716 1.632 1.631 1.527 1.582 1.760
1.665 1.608 1.566 1.528 1.491 1.738
ing species without water molecules, are enclosed in a smaller range of energy (15.4 kcal/mol vs 20.7 kcal/mol) (see Table 2). The global minimum, as for the Gly-Ni2+ (d) system, corresponds to the structure in which the metal cation interacts with both oxygen atoms of the zwitterionic form of glycine (D). Here, glycine still acts as a bidentate ligand and the Ni2+-Oa and Ni2+-Ob distances are 2.081 and 2.207 Å (see Table 1), respectively. The two donor atoms of the ligand lie on the equatorial plane of the octahedral environment around the metal center (see D of Figure 2). The A and D complexes, already very similar in energy in the case of the interaction of Gly with the bare cation, become practically degenerate (0.8 kcal/mol) in the case upon hydration; thus, both structures have the same probability of existing. In all hydrated complexes, the Ni2+ metal ion adopts an octahedral geometry.
Gly-Cu2+ Complexes. In a previous theoretical work, Sodupe et al.3 investigated the interaction of naked copper dication with glycine. Thus, as in the case of Co2+,7 we studied only the possible adducts obtainable with the hydrated form of ion. The results concerning our investigation were collected in Table 2 together with those present in literature.3 The ground state for all of the examined Gly-Cu2+ (H2O)4 complexes is that of the doublet. From a first glance at the Table, it emerges that the stability trend of the complexes is not affected by the addition of the water molecules, even if these latter still reduce the gap separating the different conformers. As before, the preferred structure is obtained on the interaction of hydrated metal ion with the zwitterionic form of glycine. However, from a geometrical point of view, this complex shows some differences with respect to the analogue without water. In fact, although in the Gly-Cu2+ complex the cation is involved in a quite η2-O,O(CO2-) symmetrical bicoordination,3 in the hydrated species, Gly-Cu2+(H2O)4, the Cu2+ appears to be monocoordinated (D′′ of Figure 3) with a Cu2+-Oa distance of 1.959 Å. Instead, the other oxygen (Ob) is involved in a bond, whose length is 1.697 Å, with the hydrogen atom of one of water molecules. The original octahedral geometry around the ion undergoes rearrangements during the optimization procedure, generating a square-planar pyramidal structure with the oxygen
Structural and Electronic Characterization atom of Gly occupying the basal position and a water molecule at the vertex that is far from the copper ion more than the equatorial water molecules (D′ of Figure 3). Other interesting results emerge from the analysis of the A′ and B′ structures (see Figure 3) characterized by a η2-N,O coordination. In fact, in both of these complexes, lying at 3.5 (A′) and 8.6 kcal/mol (B′) above the global minimum, the Gly acts as a bidentate ligand through the carbonyl oxygen and amino nitrogen, but a water molecule moves away from the metal center remaining in the second shell of hydration. The resulting geometries are square-pyramidal structures (see Figure 3). From these results, it seems clear that the copper cation rejects the hexacoordination and tends to assume the penta one even if the fifth ligand is always very far from it. This is in agreement with other investigations on similar systems11-14 in which authors suggested that the copper ion (II) generally has a preference for the square-planar geometry. It was also noteworthy that the copper (II) complexes are subjected to the JahnTeller distortion that causes the lability and plasticity of the ligand in axial position.11 However, a further confirmation of these observations arises from the low stability of the remaining adduct (C) (see Figure 2) in which the ion appears hexacoordinated. The monocoordinated complex Gly-Cu2+(H2O)5 (E), outwardly octahedral, shows that two axial water molecules are linked to the copper ion with distances that are about 0.3 Å longer than the other bonds with the equatorial ligands. Gly-Zn2+ Complexes. The B3LYP/6-311++G** protocol was used to study the interaction of Gly with the naked zinc ion because it was studied previously at the MP2 level.8 Both B3LYP and MP2 calculations8 on the Gly-Zn2+ systems agree about the stability order of obtained complexes and suggest the same lowest-energy structure (d) (see Table 2 and Figure 1). However, the η2-N,O coordination of Zn2+ to the canonical form of glycine yields an adduct (A) that lies at only 0.5 kcal/ mol above the global minimum. A similar situation was encountered in the case of the Gly-Ni2+species. Adducts c and e represent the highest-energy minima lying at 21.6 and 31.7 kcal/mol, respectively (24.9 and 41.8 kcal/ mol at MP2 level).8 No comparison is possible for the lowlying structure (b) that shows the same coordination pattern of the glycine in complex a but presents a different orientation of the hydroxyl group of the carboxyl (see Figure 2). Except for complex D′, the coordination geometry of the hydrated cation in the lowest-energy minima Gly-Zn2+(H2O)n (n ) 4 or 5) remains very similar to that found in the Gly-Zn2+ systems (see Figure 1 and Table 2). In all cases, the Zn2+OdC and Zn2+-N bond lengths appear to be slightly longer compared to those in the Gly-Zn2+ systems. The larger differences can be observed in the case of the d and e structures (see Table 1). Contrary to what occurs for all other metal ions examined before, except for Mn2+, hydrated complexes of zinc are separated by larger energetic gaps with respect to those without water (see Table 2). For instance, the energy difference that separates complexes A and D′, almost degenerate in the absence of water, becomes 9.8 kcal/mol with a clear predominance of D′. This behavior is also in contrast to that observed for the complexes of Gly and other amino acids with alkali metal ions, in which the addition of explicit water molecules to the cation favors the species involving the canonical form of glycine rather than the zwitterionic one.17,25 This is because the presence of
J. Phys. Chem. B, Vol. 110, No. 48, 2006 24671 solvent around the metal ion decreases its effective charge, reducing its ability to stabilize charge separation in the zwitterion. The discrepancies found for the hydrated complexes of zinc and manganese ions are probably due to major repulsive contributions, induced by the presence of water molecules, between the valence shell electrons (d5 and d10 for Mn2+ and Zn2+, respectively) and those of the donor atoms present in the ligand molecules. The Gly-Zn2+(H2O)4 global minimum (D′) is a monocoordinated system in which the zwitterionic glycine engages the bond with zinc through one of the oxygen atoms (Zn2+-Oa 1.993 Å, see Table 1 and Figure 2). In the corresponding complex without water, the cation is, instead, coordinated to both oxygen atoms of the carboxylate group of glycine (Zn2+-Oa is 1.992 Å and Zn2+-Ob is 2.050 Å, see Table 1). The addition of the four water molecules produces the same effect found in the cases of Mn2+ and Cu2+ in which the switch from the bi- to monocoordination, in going from naked to hydrated cation complexes, was the consequence of a bond formation of 1.749 Å between the hydrogen of one of water molecules and one of the oxygens of the carboxylate group. Thus, the final geometry around the metal ion in the Gly-Zn2+(H2O)4 (D′) system is bipyramidal-trigonal (see Figure 2). Solvation Effect. To add further biological relevance to our study, we have considered it interesting to include solvent effects. PCM computations in aqueous solution on hydrated complexes provide some remarkable news. The most common one concerns the stability inversion occurring between the complex arising from the zwitterionic form of Gly and those obtained from the canonical isomer (see Table 2). Iron and nickel ion complexes are an exception because the most stable species after the absolute minimum are the B complexes. As expected, the new trend in water solvent depends on the dipole moment (µ) of the various species (see Table 2). The largest hydration energy is obtained for the C complexes, but the energetic gain is not sufficient to compensate their intrinsic gas-phase instability. In agreement with our findings, a previous study27 on the interaction of biological systems with metal ions underlines the minor stability of the complexes of Li+, Na+, and K+ deriving from the zwitterion forms with respect to those obtained by the canonical ones. However, a deeper look to the relative energy values reported in Table 2 shows that energetic gaps between the various complexes of each metal ion are sensibly smaller with respect to those in the gas phase. This means that all species are probable in solution. NBO and CDA Analysis. NBO and CDA analyses were performed only for the lowest-energy structure of all Gly-M2+ and Gly-M2+(H2O)n complexes in order to establish the metal ion-ligand bond nature. For the first class of compounds, our calculations suggest a charge transfer occurring from ligand to cation. The degree of these transfers can be deduced by the net charge values on the metal ions reported in Table 3. As it is known, a charge transfer implies the presence of almost an interaction with a covalent contribution. Actually, because in all Gly-M2+ examined complexes glycine acts as a bidentate ligand, we found that both M2+-Oa and M2+-Ob bonds exhibit a partially covalent character.
24672 J. Phys. Chem. B, Vol. 110, No. 48, 2006 The overlap that gives rise to these bonds always occurs between a 2p orbital of oxygen atoms and a hybrid sd orbital of the metal ion. The only expected exception concerns the GlyZn2+ complex in which the d10 cation uses its 4s orbital. In the NBO analysis, the asymmetric coordination of the metal ion, reflected in the different length of M2+-Oa and M2+-Ob bonds, is described by a different participation of Oa and Ob to the molecular orbital composition. The lowest and highest covalent contribution is found for the Mn2+ and Ni2+ ions, respectively. NBO and CDA analyses are substantially in agreement between them. The NBO analysis applied to the Gly-M2+(H2O)n complexes indicated that no charge transfer occurs between the ligand and the hydrated ions. This can be argued not only by the absence of any orbital overlap but also by the comparison between the values of the metal charge into the free and coordinated aquocomplexes (see Table 3). The presence of water molecules makes the coordination of an additional ligand more difficult not only for the reduced orbital availability of the metal but also for the evident steric hindrance. Binding Energies. The experimental2,75 study of the gasphase interaction of alkali metal cations (Li+, Na+, K+) with amino acids allowed the determination of absolute metal ion affinity values. Instead, this information is lacking for the most part of multiply charged metal ions for which only a few relative values were reported.20,38,76 In this work, we have computed the absolute binding energies at 298 K for the most stable systems of both Gly-M2+ and GlyM2+ (H2O)n and the values are reported in Table 2 where we have also included some available theoretical values.3,7,8 Metal ion affinities for the naked and hydrated systems follow the trends Cu2+ > Ni2+ > Zn2+ ≈ Co2+ > Fe2+ > Mn2+ and Cu2+ > Ni2+ > Mn2+ ≈ Zn2+ > Fe2+ > Co2+, respectively. Usually, the selectivity of a ligand toward Lewis acids is analyzed taking into account the hardness of cations. In the cases examined here, all of the metal ions have shown a preference for the oxygen atom of the glycine rather than the nitrogen one (D′), suggesting a hard Lewis-acid nature. However, the Fe2+, Co2+, Ni2+, Cu2+, and Zn2+ that display intermediate behavior according to the hard-soft acid-base (HSAB) theory interact with glycine more strongly than the Mn2+ ion that is the only really hard species. This leaves us to think that the covalent contribution to the bond is an important factor in determining the trend of BE values. In fact, as can be argued from the charge values on the metal ions into the complexes (see Table 2), this contribution is small in the case of Mn2+ because of its scarce polarizability, whereas it is quite large for Cu2+ and Ni2+ that occupy the first positions in the sequence of metal-ion affinities. The rationalization of the remaining positions in the trend is difficult to do because glycine complexes with transition-metal ions represent a very special category in which factors such as hardness, ionic radii size, and charge do not always correlate well with the BE’s values. From Table 2, it is evident that coordination of the water molecules to the metallic centers causes a considerable reduction (about 100.0-160.0 kcal/mol) of the binding energies. This can be explained not only by the fact that the solvated cation has a minor orbital availability but also by the different nature of the bond and by the different metal coordination type. In the presence of water molecules, the interaction established by metal ions with the carboxylate group of the zwitterionic
Marino et al. glycine assumes a pronounced electrostatic character everywhere. The disappearance of the covalent contribution to the bond entails the absence of a distinctive character in the interaction of glycine with the various metal ions and in these conditions, the Mn2+ for which we should expect the most favorable interaction with the hard site of ligand, closes the gap showing the largest affinity after the Cu2+ and Ni2+ cations. This still does not explain the reason for the major affinity of copper and nickel ions for glycine, although, as far as the Cu2+ is concerned, we can hypothesize a particular effect on the internal polarizability of the whole complex due its nature as an open-shell system. The large affinity of Ni2+ for the amino acids, although not easy to elucidate, was also proposed many times by other previous works.16 Conclusions In this work we have investigated the possible complexes that glycine forms with different naked and hydrated transitionmetal dications in their ground and excited states. Results showed that stable systems can be obtained with all metal species. The lowest electronic spin state for the complexes was determined to be the sextet, the triplet, the quintet, the quartet, the doublet, and the singlet for Mn2+, Ni2+, Fe2+, Co2+, Cu2+, and Zn2+, respectively. The complexes involving the naked cations are characterized by an η2O,O coordination type that proposes the glycine zwitterion as the most favored form of the ligand. Except for Mn2+ and Zn2+, the presence of explicit water molecules on cations reduces the gap between the complexes obtained by the zwitterion and canonical forms of glycine so that to approach the alkali metal complexes behavior. The further addition of solvent effects leads to the definitive slight predominance of the canonical glycine in the most stable complexes. The bond involving the naked ions has a mixed ioniccovalent nature with the covalent contribution that decreases in going from Cu2+ to Ni2+ to Co2+ to Fe2+ to Zn2+ to Mn2+. The absolute metal ion affinity trend suggests that the interaction is influenced by the covalent contribution to the metal ion-ligand bond and that the largest values are obtained for copper and nickel cations. Addition of the water molecules in the number necessary to realize an octahedral hydration shell changes the coordination type of the metals that usually appear monocoordinated to an oxygen atom of the glycine zwitterion. The most stable complexes exhibit different geometries depending on the involved metal ion. In particular, we obtain an octahedral disposition of ligands in the case of Fe2+, Ni2+, and Co2+. For Zn2+ and Mn2+ adducts, the bipyramidal-trigonal geometry is the favored one. Finally, around Cu2+ a distorted square-pyramidal arrangement is present. The bond between the ligand and the hydrated metal ions is purely electrostatic. The most relevant effect due to the presence of water molecules is a substantial decrease of the binding energies because of the different orbital availability of the metal centers and the different coordination and nature of the bond. The trend of the hydrated metal ion affinities undergoes some modification with respect to that obtained for the naked species, but the preference of the ligand for copper and nickel cations is also confirmed in this case.
Structural and Electronic Characterization Acknowledgment. Financial support from the Universita` degli Studi della Calabria and Regione Calabria (POR Calabria 2000/2006, misura 3.16, progetto PROSICA) is gratefully acknowledged. References and Notes (1) Wyttenbach, T.; Witt, M.; Bowers, M. T. J. Am. Chem. Soc. 2000, 122, 3458-3464. (2) Hoyau, S.; Norrman, K.; McMahon, T. B.; Ohanessian, G. J. Am. Chem. Soc. 1999, 121, 8864-8875. (3) Bertran, J.; Rodriguez-Santiago, L.; Sodupe, M. J. Phys. Chem. B 1999, 103, 2310-2317. (4) Rodriguez-Santiago, L.; Sodupe, M.; Tortajada, J. J. Phys. Chem. A 2001, 105, 5340-5347. (5) Hoyau, S.; Pelicier, J.-P.; Rogalewicz, F.; Hoppilliard, Y.; Ohanessian, G. Eur. J. Mass Spectrom. 2001, 7, 303-311. (6) Pulkkinen, S.; Noguera, M.; Rodriguez-Santiago, L.; Sodupe, M.; Bertran, J. Chem.sEur. J. 2000, 6, 4393-4399. (7) Constantino, E.; Rodriguez-Santiago, L.; Sodupe, M.; Tortajada, J. J. Phys. Chem. A 2005, 109, 224-230. (8) Rogalewicz, F.; Ohanessian, G.; Gresh, N. J. Comput. Chem. 2000, 21, 963-973. (9) Ai, H.; Bu, Y.; Li, P.; Li, Z.; Hu, X.; Chen, Z. J. Phys. Org. Chem. 2005, 18, 26-34. (10) Xu, J. J. Mol. Struct.: THEOCHEM 2005, 757, 171-174. (11) Hattori, T.; Toraishi, T.; Tsuneda, T.; Nagasaki, S.; Tanaka, S. J. Phys. Chem. A 2005, 109, 10403-10409 (12) Rulı´sˇek, L.; Havlas, Z. J. Phys. Chem. B 2003, 107, 2376-2385. (13) Rulı´sˇek, L.; Havlas, Z. J. Am. Chem. Soc. 2000, 122, 10428-10439. (14) Rulı´sˇek, L.; Vondras´ek, J. J. Inorg. Biochem. 1998, 71, 115-127. (15) Remko, M.; Rode, B. M. Struct. Chem. 2004, 15, 223-232. (16) Remko, M.; Rode, B. M. J. Phys. Chem. A 2006, 110, 1960-1967. (17) Cerda, B. A.; Wesdemiotis, C. J. Am. Chem. Soc. 1995, 117, 97349739. (18) Cerda, B. A.; Hoyau, S.; Ohanessian, G.; Wesdemiotis, C. J. Am. Chem. Soc. 1998, 120, 2437-2448. (19) Talley, J. M.; Cerda, B. A.; Ohanessian, G.; Wesdemiotis, C. Chemistry 2002, 8, 1377-88. (20) Corral, I.; Mo´, O.; Ya´n˜ez, M.; Moran, D.; Radom, L.; Salpin, J.Y.; Tortajada, J. Chem. A. Eur. J. 2006 12, 6787-6796. (21) Marino, T.; Russo, N.; Toscano, M. J. Inorg. Biochem. 2000, 79, 179-185. (22) Holm, R. H.; Kennepohl, P.; Solomon, E. I. Chem. ReV. 1996, 96, 2239-2314. (23) Rogalewicz, F.; Hoppilliard, Y.; Ohanessian, G. Int. J. Mass Spectrom. 2000, 204, 267-280. (24) Rogalewicz, F.; Hoppilliard, Y.; Ohanessian, G. Int. J. Mass Spectrom. 2001, 206, 45-52. (25) Jockusch, R. A.; Lemoff, A. S.; Williams, E. R. J. Am. Chem. Soc. 2001, 123, 12255-12265. (26) Ye, S. J.; Moision, R. M.; Armentrout, P. B. Int. J. Mass Spectrom. 2005, 240, 233-233. (27) Rodgers, M. T.; Armentrout, P. B. Acc. Chem. Res. 2004, 37, 989998. (28) Robertson, E. G.; Hockridge, M. R.; Jelfs, P. D.; Simons, J. P. Phys. Chem. Chem. Phys. 2001, 3, 786-795. (29) Robertson, E. G.; Hockridge, M. R.; Jelfs, P. D.; Simons, J. P. J. Phys. Chem. A 2000, 104, 11714-11724. (30) Mons, M.; Dimicoli, I.; Tardivel, B.; Piuzzi, F.; Robertson, E. G.; Simons, J. P. J. Phys. Chem. A 2001, 105, 969-973. (31) Carles, S.; Lecomte, F.; Schermann, J. P.; Desfrancois, C. J. Phys. Chem. A 2000, 104, 10662-10668. (32) Carles, S.; Desfrancois, C.; Schermann, J. P.; Smith, D. M. A.; Adamowitz, L. J. Chem. Phys. 2000, 112, 3726-3734. (33) Hoyau, S.; Ohanessian, G. J. Am. Chem. Soc. 1997, 119, 20162024. (34) Shoeib, T.; Rodriquez, C. F.; Siu, M.; Hopkinson, A. C. Phys. Chem. Chem. Phys. 2001, 3, 853. (35) Armentrout, P. B. J. Mass Spectrom. 2000, 200, 219-241. (36) Moision, R. M.; Armentrout, P. B. Phys. Chem. Chem. Phys. 2004, 6, 2588-2599. (37) Moision, R. M.; Armentrout, P. B. J. Phys. Chem. A 2002, 106, 10350-10362. (38) Koizuma, H.; Muntean, F.; Armentrout, P. B. J. Chem. Phys. 2004, 120, 756-766. (39) Rodriguez-Cruz, S. E.; Jockush, R. A.; Williams, E. R. J. Am. Chem. Soc. 1999, 121, 8898-8906. (40) Peschke, M.; Blades, A. T.; Kebarle, P. J. Phys. Chem. A 1998, 102, 9978-9985. (41) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb, M. A.; Cheeseman, J. R.; Montgomery, J. A., Jr.; Vreven, T.; Kudin, K.
J. Phys. Chem. B, Vol. 110, No. 48, 2006 24673 N.; Burant, J. C.; Millam, J. M.; Iyengar, S. S.; Tomasi, J.; Barone, V.; Mennucci, B.; Cossi, M.; Scalmani, G.; Rega, N.; Petersson, G. A.; Nakatsuji, H.; Hada, M.; Ehara, M.; Toyota, K.; Fukuda, R.; Hasegawa, J.; Ishida, M.; Nakajima, T.; Honda, Y.; Kitao, O.; Nakai, H.; Klene, M.; Li, X.; Knox, J. E.; Hratchian, H. P.; Cross, J. B.; Bakken, V.; Adamo, C.; Jaramillo, J.; Gomperts, R.; Stratmann, R. E.; Yazyev, O.; Austin, A. J.; Cammi, R.; Pomelli, C.; Ochterski, J. W.; Ayala, P. Y.; Morokuma, K.; Voth, G. A.; Salvador, P.; Dannenberg, J. J.; Zakrzewski, V. G.; Dapprich, S.; Daniels, A. D.; Strain, M. C.; Farkas, O.; Malick, D. K.; Rabuck, A. D.; Raghavachari, K.; Foresman, J. B.; Ortiz, J. V.; Cui, Q.; Baboul, A. G.; Clifford, S.; Cioslowski, J.; Stefanov, B. B.; Liu, G.; Liashenko, A.; Piskorz, P.; Komaromi, I.; Martin, R. L.; Fox, D. J.; Keith, T.; Al-Laham, M. A.; Peng, C. Y.; Nanayakkara, A.; Challacombe, M.; Gill, P. M. W.; Johnson, B.; Chen, W.; Wong, M. W.; Gonzalez, C.; Pople, J. A. Gaussian 03, revision A.1; Gaussian, Inc.: Wallingford, CT, 2004. (42) Stevens, P. J.; Devlin, F. J.; Chablowski, C. F.; Frisch, M. J. J. Phys. Chem. 1994, 98, 11623-11631. (43) Marino, T.; Russo, N.; Toscano, M. Inorg. Chem. 2001, 40, 64396443. (44) Russo, N.; Toscano, M.; Grand, A. J. Phys. Chem. B 2001, 105, 4735-4741. (45) Russo, N.; Toscano, M.; Grand, A. J. Am. Chem. Soc. 2001, 123, 10272-10279. (46) Siegbhan, P. E. M. J. Biol. Inorg. Chem. 2006, 11, 695-701. (47) Furche, F.; Perdew, J. P. J. Chem. Phys. 2006, 124, 044103-(127). (48) Miertos, S.; Scrocco, E.; Tomasi, J. Chem. Phys. 1981, 55, 117129. (49) Miertos, S.; Tomasi, J. Chem. Phys. 1982, 65, 239-245. (50) Cossi, M.; Barone, V.; Commi, R.; Tomasi, J. J. Chem. Phys. 1996, 255, 327-335. (51) Marino, T.; Russo, N.; Toscano, M. J. Mass Spectrom. 2002, 37, 786-791. (52) Belcastro, M.; Marino, T.; Russo N.; Toscano M. J. Mass Spectrom. 2005, 40, 300-306. (53) Marino, T.; Russo, N.; Toscano, M. J. Phys. Chem. B 2003, 107, 2588-2594. (54) Belcastro, M.; Marino, T.; Russo N.; Sicilia, E. J. Phys. Chem. A 2004, 108, 8407-8410. (55) Mazzuca, D.; Russo, N.; Toscano, M.; Grand, A. J. Phys. Chem. B 2006, 110, 8815-8824l, (56) Glendening, E. D.; Reed, A. E.; Carpenter, J. E.; Weinhold, F. NBO, version 3.1. (57) Gorelsky, S. I. AOMIx: Program for Molecular Orbital Analysis; York University: Toronto, 1997, http://www.sg-chem.net/. (58) Gorelsky, S. I.; Lever, A. B. P. J. Organomet. Chem. 2001, 635, 187-196. (59) Suenram, R. D.; Lovas, F. J. J. Mol. Spectrosc. 1978, 72, 372382. (60) Iijima, K.; Tanaka, K.; Onuma, S. J. Mol. Struct. 1991, 246, 257266. (61) Godfrey, P. D.; Brown, R. D. J. Am. Chem. Soc. 1995, 117, 20192023. (62) Chapo, C. J.; Paul, J. B.; Provencal, R. A.; Roth, K.; Saykally, R. J. J. Am. Chem. Soc. 1888, 10, 12956-12957. (63) Stepanian, S. G.; Reva, I. D.; Radchenko, E. D.; Rosado, M. T. S.; Duarte, M. L. T. S.; Fausto, R.; Adamowicz, L. J. Phys. Chem. A 1998, 102, 1041-1054. (64) Barone, V.; Adamo, C.; Lelj, F. J. Chem. Phys. 1995, 102, 364370. (65) Kaschner, R.; Hohl, D. J. Phys. Chem. 1998, A102, 5111-5116. (66) Remko, M.; Rode, B. M. Chem. Phys. Lett. 2000, 316, 489-494. (67) Remko, M.; Rode, B. M. Phys. Chem. Chem. Phys. 2001, 3, 46674673. (68) Wang, W.; Pu, X.; Zheng, W.; Wong, N. B.; Tian, A. Chem. Phys. Lett. 2003, 370, 147-153. (69) Csaszar, A. G. J. Am. Chem. Soc. 1992, 114, 9568-9575. (70) Csaszar, A. G.; Perczel, A. Prog. Biophys. Mol. Biol. 1999, 71, 243-309. (71) Yamada, S.; Terashima, S.; Wagatsuma, M. Tetrahedron Lett. 1970, 1501-1504. (72) Siegel, H.; Martin, R. B. Chem. ReV. 1982, 82, 385-426. (73) Freeman, H. C. AdV. Protein Chem. 1967, 22, 257-424. (74) Strittmatter, E. F.; Lemoff, A. S.; Williams, E. R. J. Phys. Chem. A 2000, 104, 9793-9796. (75) Kish, M. M.; Ohanessian, G.; Wesdemiotis, C. Int. J. Mass Spectrom. 2003, 227, 509-524. (76) Rodgers, M. T.; Armentrout, P. B. Mass Spectrom. ReV. 2000, 19, 215-247.
