Miroslav
Dudič
a,
Pavel
Lhoták
*a,
Ivan
Stibor
a,
Hana
Petříčková
b and
Kamil
Lang
*c
aDepartment of Organic Chemistry, Institute of Chemical Technology, Technická 5, 166 28, Prague 6, Czech Republic. E-mail: lhotakp@vscht.cz; Fax: +420-224 354 288; Tel: +420-224 354 280
bDepartment of Solid State Chemistry, Institute of Chemical Technology, Technická 5, 166 28, Prague 6, Czech Republic
cInstitute of Inorganic Chemistry, Academy of Sciences of the Czech Republic, 250 68, Řež, Czech Republic. E-mail: lang@iic.cas.cz; Fax: +420 220 941 502; Tel: +420 266 172 193
First published on 31st October 2003
Molecular tweezers, (thia)calix[4]arene–porphyrin conjugates, are constructed from the (thia)calix[4]arene unit serving as a scaffold and from two and/or four porphyrin units. These molecules form stable complexes with fullerenes in a toluene solution and exhibit selectivity towards C70. The observed fullerene–porphyrin contacts suggest cooperative behaviour of closely separated porphyrin units attracting C60 or C70. Measurements show efficient quenching of porphyrin fluorescence emission.
Fullerenes fit nicely to preorganized cavities and form stable complexes, in particular with deep-walled cavitands, calixarenes, homotrioxacalixarenes, cyclotriveratrylenes or cyclodextrins.4 It was also shown that in the solid state or even in solutions the curved π surface of C60 is attracted to the centre of a (metallo)porphyrin ring. Consequently, numerous elegantly designed porphyrin systems have been synthesized for studying porphyrin/fullerene interactions.5
As we have recently demonstrated, the combination of (thia)calix[4]arene and porphyrin units leads to novel conjugates with complexation abilities towards anions,6 cations7 or neutral8 molecules. In this study, we present receptor molecules capable of forming stable supramolecular complexes with fullerenes C60 and C70 in toluene with high selectivity for C70 (Scheme 1).
![]() | ||
Scheme 1 (Thia)calix[4]arene–porphyrin receptors. |
Complexation was first studied by 1H NMR titrations with fullerenes C60 and C70 in toluene-d8. Upon addition of increasing amounts of fullerene to 0.2–0.5 mM solutions of the receptors, the proton resonances of the porphyrin NH moved upfield (ca. 0.4 ppm in the 2+
C70 system)
(Fig. 1). Similarly, upfield shifts, albeit not so pronounced (approx. 0.15 ppm), were observed for the β-pyrrole protons of the porphyrin moieties. The chemically-induced shifts were shown only by porphyrin protons, while all protons of the calixarene skeleton remained unchanged (Fig. 2). These results indicate direct contact between the porphyrin moiety and fullerene.5 In addition, Job plots confirmed the formation of 1 ∶ 1 complexes of the receptors with fullerenes. Assuming the 1 ∶ 1 stoichiometry, the binding isotherms constructed from induced shifts of the NH and β-pyrrole resonances for the porphyrin and metalloporphyrin moieties, respectively, were analysed by nonlinear least-squares methods giving binding constants summarized in Table 1.9,10 In contrast, there was no 1H NMR evidence of a complex formation between reference mono-porphyrin conjugate 5 and fullerenes. This suggests that the (thia)calix[4]arene skeleton does not bind fullerenes and that the preorganization of two porphyrin units on the lower rim of calixarenes is the fundamental prerequisite for complexation of fullerenes. Hence, the calixarene skeleton serves as a molecular scaffold holding porphyrins in a suitable distance corresponding roughly to the size of fullerenes.
![]() | ||
Fig. 1 1H NMR titration of 2 (2 × 10−4 M−1) with C70 (porphyrin NH protons, 300 MHz, 298 K). The solid line is the theoretical isotherm obtained by the least-squares fit to the experimental data. Inset: Job plot for the same receptor. |
![]() | ||
Fig. 2 Partial 1H NMR spectra of the porphyrin signals: a) free receptor 1, b) receptor 1![]() ![]() |
Receptor | C60 | C70 |
---|---|---|
a Experimental error 15% unless otherwise stated.
b UV–vis titration, toluene, 294 K, K/M−1: (7.6![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
||
1 | 4920 | 21![]() |
2 | 2340 | 15![]() |
1Zn e | 8600![]() ![]() |
27![]() ![]() ![]() |
2Zn e | 2710![]() ![]() |
37![]() ![]() ![]() |
3 | 3510 | 3330 |
4 | 3420 | 6350 |
3Zn | c | c |
4Zn e | 2300![]() ![]() |
d |
5 | No interaction | No interaction |
6 | 3500 | 7920 |
7 | 1460 | 14![]() |
In order to assess the influence of the preorganization of the calixarene unit we designed and synthesized receptors with the functionalized upper rim.6 Namely, calix[4]arene derivatives 6 and 7 were immobilised in the cone conformation with two porphyrin units being connected to the upper rim via ureido functions (Scheme 2). These calixarene–porphyrins possess similar complexation ability towards fullerenes as the receptors 1–4. The 1H NMR titration experiments revealed that compounds 6 and 7 exhibit higher complexation ability towards C70. This phenomenon is especially pronounced in the case of tetraacetate derivative 7
(cf. 1.5×
103 M−1 for 7-C60 and 14.5
×
103 M−1 for 7-C70).
![]() | ||
Scheme 2 Calix[4]arene–porphyrin receptors. |
Further indications of complexation were obtained by UV–vis titrations. Electronic absorption spectra of 1–4 and 1Zn–4Zn have typical porphyrin features, however, the Soret bands are considerably broadened and split into at least two components when compared to a single porphyrin unit. It indicates intramolecular exciton coupling between closely separated porphyrin units due to the spatial flexibility of the amide spacer connecting them with the calixarene skeleton.6–8 After addition of C60 or C70 the original split Soret band underwent significant hypochromicity and well-defined isosbestic points appeared (Fig. 3). Evidently, the high spectral sensitivity of the receptors 1–4 and 1Zn–4Zn is due to exciton coupling since the features of the Soret band are strongly affected by interaction with fullerenes and by a porphyrin–porphyrin relative orientation. Exciton coupling does not influence the complexation because the receptors 6 and 7 with much less extent of coupling6 bind fullerenes similarly as 1–4 (Table 1, 1H NMR results). However, in this case the spectral changes are too small for quantitative interpretation. No interaction occurred between the model compounds 5 or 5,10,15,20-tetraphenylporphyrin (TPP) and fullerenes in solution since no spectral changes were observed up to 50 equiv. of fullerenes. It is noteworthy that the values of K are comparable with those obtained by 1H NMR experiments (Table 1). It demonstrates compatibility of both methods and no concentration dependent effects on the function of the receptors.
![]() | ||
Fig. 3 Difference UV–vis spectra showing the Soret band of 1 (1.1 μM) after addition of C70 in toluene. The isosbestic point is at 429 nm. The arrows follow changes due to increasing concentrations of C70, varied from 0 to 64 μM. Inset: Benesi–Hildebrand plot to the experimental data. |
The complexation was also evidenced by fluorescence spectroscopy. While the lifetime of 1Zn (1.92 ns) did not show any changes upon addition of fullerenes, steady-state fluorescence was strongly quenched (Fig. 4). Evidently, quenching is a consequence of photoinduced electron transfer between 1S of the porphyrin moiety and C70.11 The fluorescence decay–time profiles indicate that the lifetime of porphyrin in the complexes is less than 100 ps, i.e. below the time resolution of our instrument.
![]() | ||
Fig. 4 Steady-state fluorescence spectra of 1.5 μM 1 (a) in the presence of 35.8 μM (b) and 71.6 μM C70 (c) in toluene. |
Comparison of 1, 2, 6, 7 and model compounds (5, TPP) clearly indicates that the cooperative effect of two porphyrin units is crucial for the fullerene complexation. The UV–vis and fluorescence results reveal that the complex enables porphyrins and fullerenes to undergo electronic coupling. The preorganization of the lower or upper rim of (thia)calixarene with two cofacially oriented porphyrin moieties creates a cavity where fullerene can be inserted to form the 1 ∶ 1 complex. A substantial increase of selectivity for C70 is observed for the tetraacetate (7) over tetrapropoxy (6) calixarene although no effect was anticipated due to their similar size. The respective substituents were also reported to affect hydrogen bonding between anions and ureido functions at the upper rims of 6 and 7.6 Evidently this subtle change to the calixarene structure effectively influences the binding affinity at the opposite upper rim. It appears to be an important finding because it renders an efficient way to modulate the binding selectivity by simple functionalization of the lower rim. Introduction of four porphyrins on the lower rim of the receptors (3, 4, 4Zn) does not improve the fullerene complexation (cf. 4.9×
103 M−1 for 1–C60 and 3.5
×
103 M−1 for 3–C60) and leads to the loss of the fullerene recognition (cf.KC70/KC60 is 4.3 and ∼1 for 1 and 3, respectively). We suppose that fullerenes are not well complemented by the four-armed calixarenes and that the stacking of the porphyrin units within the molecule constrains the four-point binding motif.
The 1H NMR study of the lower-rim substituted calixarene 1 and thiacalixarene 2 did not reveal differences in the conformational behaviour. However, a high downfield shift of the NH amidic signals (11.11 ppm for 1, 11.13 ppm for 2; chemical shifts were concentration independent) might indicate that the preorganization of 1–4 is strengthened via intramolecular hydrogen bonding. Our attempts to grow suitable monocrystals for X-ray studies of structural motifs have failed. Hence, we synthesized model compounds, 4-methylphenyl diamides 8–11 (Scheme 3), with the similar aromatic amide structural fragments. We succeeded in growing suitable crystals of 8 and 11 using slow evaporation of an ethyl acetate/ethanol solution.
![]() | ||
Scheme 3 Model compounds for crystallographic study. |
The classical calix[4]arene 8 adopts the cone conformation where both amidic hydrogens are engaged in hydrogen bonding with neighbour oxygen atoms of the –O–CH2– and OH groups. The corresponding NH⋯O distances are in the range of 2.06–2.20 Å (Fig. 5a). The resulting hydrogen bonding array stabilizes the calixarene core with approx. C2 symmetry and contributes to the preorganization of the lower rim with two coplanar aromatic units separated by 3.54 Å (Fig. 5b).
![]() | ||
Fig. 5 Solid-state structures of derivatives 8 (a,b) and 11 (c,d). |
The hydrogen bonding arrangement within thiacalixarene 11 is completely different. The NH group of an amidic unit is connected to the carbonyl CO group (NH⋯O distance
=
2.11 Å) of the diametrical amide arm (Fig. 5c). This intramolecular bonding leads to the overall unsymmetrical structure with one amide NH function facing towards the C
O group of a neighbour molecule. The distance of this intermolecular hydrogen bonding (NH⋯O) is of 1.92 Å. The unsymmetrical organisation of the hydrogen bonds results in structural changes to the lower rim. Consequently, the mutual position of the two 4-methylphenyl units is no longer coplanar as within 8, but rather almost perpendicular (Fig. 5d).
Although we cannot confirm the same hydrogen bonding patterns in the solution, these results are indicative of the specific behaviour of both molecular systems. The solid-state conformational preferences could be used as fundamentals for the explanation of calixarene vs. thiacalixarene complexation ability towards fullerenes.
Metallation of 1 and 2 (1Zn, 2Zn) does not reduce the binding constants opposed to the Boyd's experimental observation that free-base porphyrins bind C60 more strongly.5f It is clear that a number of effects can influence binding constants and the ordering of free-base and metalloporphyrins can be ascribed to subtle interplay of charge transfer, electrostatics and solvation energy effects. The solvation effects can be very important as documented by our observation that the receptors do not bind fullerenes in more polar 1,2-dichlorobenzene as follows from 1H NMR and UV–vis experiments. This can be ascribed to the fact that attractive electrostatic interactions contribute approximately 50–60% to the total attractive interactions.12
The most interesting feature is the preference of C70 over C60.13,14 For example, the receptor 2 with K of 1.6×
104 M−1 and 2.3
×
103 M−1 for C70 and C60, respectively, gives the C70/C60 selectivity of about 7. Despite relatively flexible nature of the porphyrin tweezers, the presented receptors can efficiently differentiate between C70 and C60. The higher affinity towards C70 was suggested to be a result of the ovoid shape allowing maximization of C70–porphyrins interaction.13a
In conclusion, we have introduced novel fullerene receptors, molecular tweezers, constructed from (thia)calix[4]arene and porphyrin moieties. The complexation occurs in a toluene solution and can be quantitatively investigated using common spectroscopic methods (UV–vis, NMR, fluorescence). The selectivity towards C70 opens the way towards self-assembling systems and new separation and purification methods for fullerenes.
8 Yield: 56%, white crystals, mp: 304–306°C (ethyl acetate). 1H NMR (300 MHz, CDCl3)
δ: 2.30 (s, 6H, CH3Ar), 3.57 (d, J
=
13.50 Hz, 4H, eq. ArCH2Ar), 4.23 (d, J
=
13.20 Hz, 4H, ax. ArCH2Ar), 4.62 (s, 4H, OCH2CO), 6.78 (t, J
=
7.70 Hz, 2H, H-arom.), 6.90 (t, J
=
6.90 Hz, 2H, H-arom.), 7.02 (t, J
=
8.80 Hz, 8H, H-arom.), 7.15 (d, J
=
7.40 Hz, 4H, H-arom.) 7.25 (d, J
=
8.30 Hz, 4H, H-arom.), 8.30 (s, 2H, ArOH), 10.13 (s, 2H, –CONH). FAB MS m/z
(rel. int.) 719 [M
+
1]+
(100).
9 Yield: 76%, white crystals, mp: 160–162°C (ethyl acetate–ethanol 4 ∶ 1). 1H NMR (300 MHz, CDCl3)
δ: 1.08 (s, 18H, But), 1.28 (s, 18H, But), 2.29 (s, 6H, CH3Ar), 3.51 (d, J
=
13.70 Hz, 4H, eq. ArCH2Ar), 4.22 (d, J
=
13.20 Hz, 4H, ax. ArCH2Ar), 4.60 (s, 4H, OCH2CO), 6.98 (d, J
=
6.80 Hz, 4H, H-arom.), 6.99 (s, 4H, H-arom.), 7.12 (s, 4H, H-arom.), 7.27 (d, J
=
6.80 Hz, 4H, H-arom.), 7.71 (s, 4H, H-arom.), 8.09 (s, 2H, ArOH), 10.15 (s, 2H, –CONH). FAB MS m/z
(rel. int.) 943 [M
+
1]+
(100).
10. Yield: 65%, colourless crystals, mp: decomp. 280–282°C (CHCl3–MeOH). 1H NMR (300 MHz, CDCl3)
δ: 2.31 (s, 6H, CH3Ar), 4.75 (s, 4H, OCH2CO), 6.85 (t, J
=
8.00 Hz, 2H, H-arom.), 6.94 (t, J
=
7.70 Hz, 2H, H-arom.), 7.05 (d, J
=
8.25 Hz, 4H, H-arom.), 7.33 (d, J
=
8.00 Hz, 4H, H-arom.) 7.46 (d, J
=
7.70 Hz, 4H, H-arom.), 7.72 (d, J
=
7.70 Hz, 4H, H-arom.), 8.49 (s, 2H, ArOH), 10.05 (s, 2H, –CONH). FAB MS m/z
(rel. int.) 791 [M]+
(100).
11 Yield: 72%, colourless crystals, mp: 251–252°C (CHCl3–ethanol 5 ∶ 1). 1H NMR (300 MHz, CDCl3)
δ: 1.09 (s, 18H, But), 1.29 (s, 18H, But), 2.30 (s, 6H, CH3Ar), 4.70 (s, 4H, OCH2CO), 7.05 (d, J
=
8.30 Hz, 4H, H-arom.), 7.36 (d, J
=
8.50, 4H, H-arom.), 7.50 (s, 4H, H-arom.), 7.71 (s, 4H, H-arom.), 8.54 (s, 2H, ArOH), 10.18 (s, 2H, –CONH). FAB MS m/z
(rel. int.) 1015 [M
+
1]+
(100).
X-Ray data for 11: C58H66N2O6S4·C2H6O: Mr=
1057.47, monoclinic system, space group P 21/c, a
=
18.037(3)
Å, b
=
17.973(2)Å, c
=
19.697(2)Å, β
=
112.25(1)°, V
=
5910.0(13)Å3, Z
=
4, Dc
=
1.19 g·cm−3, μ(CuKα)
=
1.881 cm−1, crystal size 0.05
×
0.4
×
0.8 mm. Data were measured at 293 K on an Enraf-Nonius CAD4 diffractometer with graphite monochromated CuKα radiation (λ
=
1.5418Å). The structure was solved by direct methods,16 sulfur and nitrogen atoms were anisotropically and oxygen and carbon atoms refined by full matrix least-squares on F values17 to final R
=
0.113, Rw
=
0.101 and S
=
1.170 with 280 parameters using 2500 independent reflections (θmax
=
69.95°). Hydrogen atoms were located from expected geometry and were not refined. Hydrogen atoms linked to the oxygen and nitrogen atoms were not found. Three of four t-Bu groups were disorder so that the model of disorder was used. CCDC reference number 220091. See http://www.rsc.org/suppdata/nj/b3/b307988k/ for crystallographic data in .cif or other electronic format.
The fluorescence spectra were recorded on a Perkin-Elmer LS 50B luminescence spectrophotometer. All emission spectra were corrected for the characteristics of the detection monochromator and photomultiplier. The absorbances of 1 and TPP were adjusted to the same value at the excitation wavelength of 515 nm. Because TPP does not interact with the receptors the inner filter effect due to added fullerene can be eliminated by comparison of intensity of the receptor with that of TPP.
The binding constants were assessed from the 1H NMR titration experiments using initial concentrations of the receptors ranging from 0.2 to 0.5 mM. The concentration of fullerene C60 or C70 was gradually increasing to cover the range of saturation up to 90%. The induced chemical shifts of NH signals were recorded for 1–7. Due to the absence of the NH signals in Zn-derivatives 1Zn–4Zn, the shifts of the β-pyrrole protons were plotted against the concentration of fullerene to construct titration curves. Titration data were analysed using the original non-linear regression curve-fitting computer program OPIUM.10
This journal is © The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2004 |