Mass spectrometric studies of competitive binding of C₆₀ and C₇₀ to mesosubstituted porphyrins

Abstract

Competitive binding of $C_{60}$ and $C_{70}$ to meso-substituted porphyrins was studied by mass spectrometry (MS). Electrospray ionization MS was employed to acquire the mass spectra of 1 : 1 porphyrin-fullerene complexes formed in a mixture of mesosubstituted porphyrin and fullerite to determine the ratio of complexes between $C_{60}$ and $C_{70}$. Matrix-free laser desorption ionization MS was used to obtain the mass spectra of fullerite to measure the mole fraction of $C_{60}$ and $C_{70}$. The binding constant ratio ($K_{70}$/$K_{60}$) was determined from the mass spectral data. The difference in standard Gibbs free energy change, ${\Delta}({\Delta}G^o)_{70-60}$, for the competitive binding of $C_{60}$ and $C_{70}$ was calculated from $K_{70}$/$K_{60}$. Of the five porphyrins, tetraphenyl, tetra(4-pyridyl), tetra(4-carboxyphenyl), tetra(3,5-di-tert-butylphenyl), and tetra(pentafluorophenyl) porphyrins, the first three non-bulky porphyrins yield negative values of ${\Delta}({\Delta}G^o)_{70-60}$, whereas the other two bulky porphyrins result in positive values of ${\Delta}({\Delta}G^o)_{70-60}$. This result indicates that $C_{70}$ binding to porphyrin is thermodynamically favored over $C_{60}$ binding in non-bulky porphyrins, but disfavored in bulky ones. It also suggests that the binding mode of $C_{70}$is different between non-bulky and bulky porphyrins, which is in line with previous experimental findings of the "side-on" binding to non-bulky porphyrins and the $C_{60}$-like "end-on" binding to bulky porphyrins.

Experimental

Porphyrins 1 (97%), 2 (97%), 3 (75%), and 5, as well as fullerite and pure C60 (99.5%) were purchased from Sigma-Aldrich (St. Louis, MO). Pure C70 (99.5%) was obtained from Materials Technologies Research (Cleveland, OH). Porphyrin 4 was provided by Dr. P. D. W. Boyd (University of Auckland, New Zealand). HPLC-grade methanol and dichloromethane were purchased from J. T. Baker (Phillipsburg, NJ).

Both porphyrin and fullerite were mixed in a 3 : 1 (v/v) dichloromethane/methanol solution by 10-min sonication. The solute concentration was 50 μM for porphyrin and 168 μM for fullerite. For ESI, 0.5% formic acid was added to the sample solution. The ESI mass spectra of a porphyrin−fullerene mixture were taken with a triple quadrupole-time-of-flight (TOF) mass spectrometer (QSTAR Pulsar-i, AB Sciex, Foster City, CA), as previously described in detail.13 The LDI mass spectra of fullerite as well as pure C60 and C70 were obtained using a TOF mass spectrometer (4700 Proteomics Analyzer, AB Sciex) equipped with a Nd : YAG laser (355 nm). The known molar ratio of [C60]/[C70] = 0.2, 0.33, 1.0, 3.0, and 5.0 was prepared by dissolving 0.017/0.101, 0.029/0.101, 0.072/0.084, 0.086/ 0.034, and 0.086/0.020 mg of pure C60/C70 in 1 mL of toluene. A fullerite solution was prepared by dissolving 0.075 mg of fullerite in 1 mL of toluene.

 

Results and Discussion

The mass spectra of fullerite are shown in Figure 1(a). The molar ratio of [C60]/[C70] in fullerite was determined from a plot of the intensity ratio of I(C60+)/I(C70+) as a function of the known molar ratio of pure [C60]/[C70] (Figure 1(b)), as previously described by Cristadoro et al. 14 Since ionization energies of C60 (7.6eV15) and C70 (7.47 eV16) are different, their response factors for LDI are not identical. A least-squares fit to the data yields the slope of 0.535 as the response factor for C60 relative to C70. Thus, the measured intensity ratio of I(C60+)/I(C70+) = 1.54 ± 0.01 divided by 0.535 results in the molar ratio of [C60]0/[C70]0 = 2.88 ± 0.03, suggesting the mole fraction of 0.742 for C60 and 0.258 for C70 in fullerite, which corresponds to 125 μM of C60 and 43 μM of C70 in 168 μM of fullerite.

Figure 1.(a) LDI mass spectra of fullerite (a mixture of C60 and C70). (b) Plot of the measured intensity ratio of I(C60+)/I(C70+) as a function of the premixed molar ratio of pure C60 and C70, [C60]/[C70]. The slope is the response factor for LDI. Error bars are from standard deviations of five measurements.

The ESI mass spectra of protonated porphyrin–fullerene complexes are displayed in Figure 2(a)−(e). Porphyrin (P) forms 1 : 1 [P + H]+·Cm (m = 60, 70) complexes. Their intensity ratio, I([P + H]+·C60)/I([P + H]+·C70), significantly varies with meso-substituent, indicating that meso-substituents affect the binding equilibrium between porphyrin and fullerene. In the ESI solution containing 0.5% formic acid, all of the porphyrins are considered to be protonated, as the proton affinity of porphyrin is 233−263 kcal mol−1.13 Thus, the protonated porphyrin [P + H]+ competitively forms 1:1 [P + H]+·Cm (m = 60, 70) complexes, as expressed in eq 1.

Figure 2.ESI mass spectra of 1:1 porphyrin−fullerene complexes in a porphyrin−fullerite mixture: (a) tetraphenyl porphyrin (1); (b) tetra(4−pyridyl) porphyrin (2); (c) tetra(4-carboxyphenyl) porphyrin (3); (d) tetra(3,5-di-tert-butylphenyl) porphyrin (4); (e) tetra(pentafluorophenyl) porphyrin (5).

where [[P + H]+·Cm] and [Cm] are the equilibrium concentrations of complex and uncomplexed fullerene present in solution, respectively, and Km is the 1 : 1 binding constant. By assuming identical ESI response factors for the two complex ions, we can replace the concentration ratio [[P + H]+·C70]/[[P + H]+·C60] with the intensity ratio I([P + H]+·C70)/I([P + H]+·C60). Moreover, the ratio of [C60]/[C70] is set to be equal to the initial concentration ratio of [C60]0/[C70]0 = 2.88 ± 0.03 in fullerite, because the concentrations of complexed fullerenes are negligible as compared to the initial concentrations of C60 (125 μM) and C70 (43 μM). The complexed fullerene concentrations (0.5−3.3 μM for C60 and 0.3−0.9 μM for C70) are estimated by multiplying the initial concentration of porphyrin and the intensity ratio of the complexed porphyrin ions to the sum of all porphyrin-containing ions.

Values of I([P+H]+·C70)/I([P + H]+·C60) for porphyrins 1−5 are obtained from the ESI mass spectra (Figure 2(a)−(e)). Calculated K70/K60 values are listed in Table 1. K70/K60 values are greater than one; 1.58 ± 0.02, 1.62 ± 0.02, and 1.26 ± 0.03 for 1, 2, and 3, respectively, indicating that equilibrium is shifted toward the binding of C70 for the three non-bulky porphyrins. On the other hand, K70/K60 values are less than one; 0.54 ± 0.01 and 0.81 ± 0.01 for 4 and 5, respectively, suggesting that the binding of C60 is favored over the C70 binding for the two bulky porphyrins.

In comparison, fluorescence studies yielded K70/K60 =8.7 in toluene for tetra(octadecyloxyphenyl) porphyrin,11 and UV/visible and fluorescence studies resulted in K70/K60 =1.9 and 9.0 in toluene for calix[4]arene-linked bisporphyrins with pentafluorophenyl and 3,5-di-tert-butylphenyl meso-substituents, respectively.6 Although other studies show that porphyrins favor the binding of C70 over C60 in toluene, we find that non-bulky porphyrins prefer the binding of C70, whereas bulky ones favor the binding of C60.

Table 1.aErrors are from the standard deviation of the signal-to-noise ratio and curve fitting. bΔ(ΔGo)70−60 = −RT ln(K70/K60) cΔ(ΔE)70−60 is the difference in binding energy. 13

The change in Gibbs free energy for the reaction 1, Δ(ΔGo)70−60 = −RT ln(K70/K60), is calculated from K70/K60. R is the gas constant and T is the temperature (298 K). Values of Δ(ΔGo)70−60 are listed in Table 1 along with the change in binding energy at 0 K, Δ(ΔEo)70–60 = ΔEo70 −ΔEo60, obtained from density functional theory (DFT) calculations.13 DFT predicted the two different modes of C70 binding to porphyrins, “side-on” and C60-like “end-on” binding. The equatorial belt of C70 is used for the side-on binding, whereas the pole of C70 is used for the end-on binding. The side-on binding offers more contact area to porphyrin than the C60-like end-on binding. Porphyrins 1−3 showing negative values of Δ(ΔGo)70–60 have Δ(ΔEo)70–60 values of −2.6 to −2.1 kcal mol−1 for the side-on binding and of −0.8 to −0.4 kcal mol−1 for the end-on binding. On the other hand, porphyrins 4 and 5 presenting positive values of Δ(ΔGo)70−60 have Δ(ΔEo)70−60 values of −0.9 to 2.2 kcal mol−1 for the side-on binding and of −0.3 to −0.5 kcal mol−1 for the end-on binding. The binding energy also suggests a preference for the side-on binding of C70 to non-bulky porphyrins, but no preference toward either side-on or end-on binding of C70 to bulky porphyrins.

Meanwhile, the entropy change involved in the competitive binding of C60 and C70 to porphyrin can be estimated by considering molecular symmetry.17 The symmetry number (σ) is 60 for spherical C60 (Ih),18 10 for ovoid C70 (D5h),18 and 4 for square planar porphyrin. Porphyrin−fullerene complexes have the symmetry number of 4 for the C60 binding or the end-on binding of C70, and 2 for the side-on binding of C70. The change in symmetry number results in the entropy change by −R ln(3) for the side-on and by −R ln(6) for the end-on binding of C70 relative to the C60 binding, respectively.

Thus, the contribution of entropy, −TΔ(ΔS)70−60, to the change in Gibbs free energy is 0.65 and 1.06 kcal mol−1 for the side-on and end-on binding of C70, respectively. The entropy factor based on molecular symmetry favors the side-on binding over the end-on binding and the C60 binding over the C70 binding.

 

Conclusion

Meso-substituents in porphyrin affect porphyrin−fullerene interactions. Porphyrins with non-bulky substituents thermodynamically favor the binding of C70 that offers more contact area than the binding of C60, whereas those with bulky substituents show thermodynamic preference for the C60 binding that exerts less steric hindrance to meso-subsituents. The molecular symmetry yields the negative entropy change for the binding of C70 over C60, but less negative entropy change for the side-on binding of C70 than the end-on binding.