Formation of zwitterionic fullerodendron using a new DBN-focal dendron.

A new poly(amidoamine) dendron having 1,5-diazabicyclo[4.3.0]non-5-ene (DBN) at the focal point was synthesized. Interestingly, formation of zwitterionic fullerodendrons (λ(max) = 930 nm for C(60) and 795 nm for C(70)) were observed by Vis-NIR spectroscopy upon the reaction of C(60) or C(70) with the DBN-focal dendron. In particular, the C(70) anion was effectively stabilized by the site isolation effect of the dendritic wedge. The half-life of fullerodendron 12b having C(70) anion at the focal point reaches 7,345 min, which is 20 times longer than that of complex between C(60) and pristine DBN. Furthermore, in order to confirm the structure of the zwitterionic complex, fullerodendron 12a was reprecipitated from benzonitrile/1,2,4-trimethylbenzene, and was observed using IR spectroscopy and APPI-MS.

Introduction

The sensing of fullerenes (C60, C70, and higher n class="Chemical">fullerenes) is gaining considerable interest because of their versatile applications in material science and nanotechnology, such as solar cells and field effect transistors (FETs). From this point of view, reversible formation of fullerodendrons is important [1-4], since the fullerodendron is known to be very soluble fullerene derivative. Meanwhile, Hirsch and co-workers reported that 1,8-diazabicyclo[5.4.0]undec-7-ene (DBU) reacts with C60 to give a zwitterionic complex via single electron transfer and radical recombination [5]. Recently, Nagata and coworkers have reported that this reaction is useful for large scale separation of C60 from a fullerene mixture because of selective complexation of higher fullerenes with DBU [6]. This zwitterion formation is known to be reversible reaction. However, the resulting complexes are susceptible to air-oxidation; they must therefore be handled in an inert atmosphere. Although many studies on the fullerodendron having neutral C60 or C70 moiety at the focal point have been described [7-29], there are few examples of incorporation of unstable fullerene species into dendritic architectures. Meanwhile, several groups have reported that covalent incorporation of highly unstable subunits into the dendritic architecture is effective to obtain active-site mimics for enzymes [30] and/or highly reactive species as an isolable compound [31] because of site isolation effect of the dendrimer [32-34]. In this context, site-isolation of dendritic substituent expected to be effective to stabilize the unstable fullerene species, such as fullerene anion. These backgrounds prompted us to investigate stabilization of the zwitterionic complex between fullerene and bicyclic amidines using the site isolation effect of the dendron. Herein we report the synthesis of a new DBN-focal dendron, poly (amidoamine) dendron having 1,5-diazabicyclo [4.3.0] non-5-ene (DBN) at the focal point together with the formation of zwitterionic fullerodendrons via complexation of fullerene, either C60 or C70, and DBN moiety of the dendron.

Results and Discussion

DBN-n class="Chemical">focal poly(amidoamine) dendrons 1a and 1b were synthesized by the use of the divergent method shown in Scheme 1. A focal point of the dendron, 9-benzyloxycarbonyl-1,5-diazabicyclo [4.3.0] non-5-ene (2), was prepared as described by Kumagai et al. [35]. Compound 2 was allowed to react with ethylenediamine to afford DBN derivative 3. Then, treatment of 3 with methyl acrylate produced dendron 4. Subsequent reaction of 4 with methyl acrylate in the presence of triethylamine (TEA) produced dendron 1a in 40% yield. This three-step process can be repeated to prepare dendron 1b in 26% yield. The structures of dendrons 1a and 1b were confirmed by 1H- and 13C-NMR spectroscopies and MALDI-TOF-MS. In the 1H-NMR spectra of dendron 1a, multiplet peaks around 4.9 ppm are attributable to the methine proton of DBN’s 9-position of diastereomeric mixture of dendron 1a. Furthermore, a broad peak around 3.2 ppm, which represents the methine proton of DBN’s 7-position, was observed. In the 13C-NMR spectra of dendron 1a, the chemical shift at δ 166.7 is the imine carbon signal of the focal point. The MALDI-TOF-MS spectrum of 1a showed a molecular ion peak at m/z 469.27 (1a, C22H37N4O7 requires m/z 469.26) using positive-ion mode. Figure 1 shows that the MALDI-TOF-MS spectrum of 1b exhibits a molecular ion peak at m/z 870.04 (1b, C40H69N8O13 requires m/z 869.49) using positive-ion mode.

Scheme 1.

Syntheses of dendrons 1a and 1b.

Figure 1.

MALDI-TOF MS spectrum of dendron 1b.

The formation of zwitterionic complex of dendrons 1a or 1b with C60 was observed by Vis/NIR spectrum, as reported by Hirsch et al. (Scheme 2) [5]. In a typical experiment, C60 (0.200 mg, 0.278 × 10−3 mmol) was dissolved in n class="Chemical">benzonitrile (3 mL), to which dendron 1b (9.49 mg, 0.0109 mmol) was added under an Ar atmosphere. Subsequently, the Vis/NIR spectrum of the solution was recorded.

Scheme 2.

(a) Structure of zwitterionic complex 7. (b) Formation of zwitterionic fullerodendrons 9a or 9b.

Figure 2 shows the Vis/NIR spectrum of the reaction of dendron 1b with C60. The decay of the absorbance of n class="Chemical">C60 anion radical (λmax = 1083 nm) is accompanied by the evolution of the zwitterion (λmax = 930 nm) [5]. These assignments were confirmed by the reaction of C60 with pristine DBN, in which radical ion pair (λmax = 1083 nm) and zwitterionic complex 7 (λmax = 930 nm) were also observed. Although the zwitterionic fullerodendrons 9a and 9b were stable under an Ar atmosphere for 24 h at room temperature, they slowly decomposed on exposure to air, as reported by Nagata et al. [6]. We examined the time course of the absorbance of zwitterionic complexes 7, 9a and 9b (λmax = 930 nm) in the presence of atmospheric oxygen (Figure 3). The half-lives of the zwitterion complexes 7, 9a and 9b were estimated as 377, 397 and 463 min, respectively. Although a clear difference of half-lives between zwitterions 7 and 9a could not be found, the highest generation 9b, which might be stabilized by the site isolation effect of dendritic wedge, showed the longest half-life.

Figure 2.

Vis-NIR spectra of the zwitterion 9b and radical ion pair 8b in benzonitrile. Inset: Vis-NIR spectrum of zwitterion 7.

Figure 3.

Time profile of the concentration of the zwitterions 7 (▪), 9a (•), and 9b (▴).

In order to obtain a more stable zwitterionic fullerodendron, we examined the respective reactions of n class="Chemical">dendrons 1a and 1b with C70, which has higher electron affinity than C60 (Scheme 3). In a typical experiment, C70 (0.220 mg, 0.262 × 10−3 mmol) was dissolved in benzonitrile (3 mL), to which dendron 1b (9.49 mg, 0.0109 mmol) was added under an Ar atmosphere. Subsequently, the Vis/NIR spectrum of the solution was observed. Figure 4 shows the Vis/NIR spectrum of the reaction of dendron 1b with C70. The decay of the absorbance at 1,380 nm is accompanied by the evolution of the absorbance at 795 nm. We interpreted that these absorbances were derived from a radical ion pair (λmax = 1,380 nm) and zwitterionic complex (λmax = 795 nm), as reported by Fukuzumi et al. [36]. These assignments were confirmed by the reaction of C70 with pristine DBN, in which a radical ion pair (λmax = 1,380 nm) and zwitterionic complex 10 (λmax = 795 nm) were also observed. Although the zwitterionic fullerodendrons 12a and 12b were stable under an Ar atmosphere for 5 days at room temperature, they decomposed slowly on exposure to air as well as in the case of C60. We examined the time course of the absorbance of zwitterionic complexes 10, 12a and 12b (λmax = 795 nm) in the presence of atmospheric oxygen (Figure 5). The half-lives of the zwitterion complexes 10, 12a, and 12b were estimated as 1445, 4800 and 7345 min, respectively. The half-lives of zwitterions having dendritic wedge, compounds 12a and 12b, were longer than that of 10. Comparing half-lives of fullerodendrons 12a and 12b, it is obvious that the stability of zwitterions depends on the generation of the dendron unit. This result indicated that zwitterionic complexes 12a and 12b might be stabilized by the site isolation effect of the dendritic wedge.

Scheme 3.

(a) Structure of zwitterionic complex 10. (b) Formation of zwitterionic fullerodendrons 12a or 12b.

Figure 4.

Vis-NIR spectra of the zwitterions 12b and radical ion pair 11b in benzonitrile. Inset: Expanded Vis-NIR spectrum of radical ion pair 11b.

Figure 5.

Time profile of the concentration of the zwitterions 10 (▪), 12a (•), and 12b (▴).

The absorption maxima and half-lives of zwitterions 7, 9, 10 and 12 were summarized in Table 1. Comparing the half-lives of zwitterionic complexes 7, 9, 10 and 12, we can conclude that anionic fullerene moieties of n class="Chemical">zwitterionic fullerodendrons 9b and 12b are stabilized by the site isolation effect of the dendritic wedge. In particular, zwitterionic fullerodendron 12b, which has a C70 moiety at the focal point, showed remarkable stability compared with 10, which does not have a dendritic wedge, and fullerodendron 9b, which contains C60 moiety at the focal point. These observed results might be due to higher electron affinity of C70 than C60, and the difference of anion delocalization between C60 and C70. In marked contrast with the complete anion delocalization of C60, the localized anion of C70 are known to be the reason of regioselective addition reactions [37]. Furthermore, the structure of zwitterionic fullerodendron 12a, which could be isolated by reprecipitation from benzonitrile/1,2,4-trimethylbenzene, was confirmed by IR spectroscopy and APPI-MS. In the IR spectrum of fullerodendron 12a, the –C = N absorbance (1a, 1,680 cm−1) of the DBN+ moiety is split into two bands at 1,663 and 1,674 cm−1. This splitting occurs because the two nitrogens in the DBN+ moiety are not identical; therefore two –C = N vibrations appeared as reported by Hirsch et al. [5]. The APPI-MS showed a molecular ion peak at m/z 1,309.52 (C92H37N4O7 requires m/z 1,309.27 [MH+]) and fragment peaks at 1,223.49 ([MH+]-CH2CH2COOCH3) and 1,077.52 ([MH+]-CH2CH2COOCH3-(CH2CH2COO)2) as shown in Figure 6.

Table 1.

Absorption maxima and half-lives of the zwitterionic complexes 7–12 in air.

Compound	λmax [nm]	half-life [min] a
7	930	377
9a	930	397
9b	930	463
10	795	1445
12a	795	4800
12b	795	7345

Half-lives were estimated using pseudo-first order decays of the absorption spectra.

Figure 6.

APPI MS spectrum of zwitterionic fullerodendron 12a.

Experimental Section

NMR spectra were measured using a spectrometer (AL 300; JEOL). Matrix-assisted laser desorption ionization time-of-flight mass spectroscopy (MALDI-TOF-MS) was performed on a mass spectrometer (Autoflex; Bruker Daltonics Inc.) using n class="Chemical">dithranol (1,8-dihydroxy-9-anthrone) as a matrix. Atmospheric pressure photo ionization mass spectroscopy (APPI-MS) was performed on a BRUKER micrOTOF focus-Kci mass spectrometer equipped with an APCI ionization unit. The GPC experiments (LC-918V; Japan Analytical Industry Co.) were performed using JAIGEL 1H, 2H (eluent: chloroform). The UV/vis spectra (λmax in nm (ε)) were measured using a spectrophotometer (UV-3150, Shimadzu Corp.). Infrared (IR) spectra were measured using a spectrophotometer (Avatar 360T2: Thermo Nicolet). The reagents were obtained from Wako Pure Chemical Industries Ltd., Tokyo Kasei Kogyo Co. Ltd. Co., Aldrich Chemical Co. Inc., and Frontier Carbon Co. The reagents used as reaction solvents were further purified using general methods.

Preparation of dendron 1a

A suspension of 9-benzyl-1, 5-diazabicyclo [4.3.0] non-5-ene 2 (690 mg, 2.67 mmol) in n class="Chemical">methanol (18 mL) was added dropwise to a stirred solution of ethylenediamine (30.1 g, 534 mmol) in methanol (18 mL) at room temperature. The mixture was stirred continuously for 1 day. After removal of the solvent, the residue was washed with excess diethyl ether to obtain compound 3, which was used for a following reaction without further purification. A mixture of 3 (530 mg, 1.88 mmol), methyl acrylate (3.24 g, 37.6 mmol), and methanol (50 mL) was stirred at 45 °C for 3 days. After removal of the solvent, the residue was purified using silica-gel column chromatography (eluent, chloroform/methanol = 15/1) to obtain compound 4. Compound 4 (900 mg, 2.34 mmol), methyl acrylate (4.02 g, 46.8 mmol), triethylamine (0.240 g, 2.34 mmol), and methanol (63 mL) were stirred at 45 °C for 1 day. After removal of the solvent, the residue was purified by silica-gel column chromatography (eluent, chloroform/methanol = 15/1) to afford dendron 1a (720 mg, 1.54 mmol) as a yellow oil in 40% yield: 1H-NMR (300 MHz, CDCl3) δ 2.18–2.24 (m, 3H), 2.43–2.47 (t, J = 6.6 Hz, 4H), 2.61–2.65 (dt, J = 4.8, 6.6 Hz, 2H), 2.75–2.79 (t, J = 6.6 Hz, 4H), 2.82–2.86 (m, 2H), 3.15–3.24 (m, 1H), 3.36–3.52 (m, 5H), 3.63–3.69 (m, 8H), 3.73 (s, 3H), 3.80–3.84 (t, J = 6.6 Hz, 2H), 4.87–4.92 (m, 1H), 8.10 (t, J = 4.8 Hz, 1H); 13C-NMR (CDCl3) δ 18.9, 24.0, 30.7, 31.6, 32.6, 37.4, 41.8, 44.7, 49.0, 51.9, 52.3, 52.5, 68.2, 77.2, 166.7, 168.7, 171.0, 173.1; IR (neat) νmax = 1,664, 1,731 cm−1; MALDI-TOF Mass Found: m/z 469.27. Calcd. for C22H37N4O7: [MH+], 469.26.

Preparation of dendron 1b

A suspension of 4 (630 mg, 1.64 mmol) in methanol (11 mL) was added dropwise to a stirred solution of n class="Chemical">ethylenediamine (19.7 g, 328 mmol) in methanol (11 mL) at room temperature. The mixture was stirred continuously for 1 day. After removal of the solvent, the residue was washed with excess diethyl ether to obtain compound 5, which was used for the following reaction without further purification. A mixture of 5 (560 mg, 1.28 mmol), methyl acrylate (4.40 g, 51.2 mmol), and methanol (69 mL) was stirred at 45 °C for 3 days. After removal of the solvent, the residue was purified using silica-gel column chromatography (eluent, chloroform/methanol = 10/1) to obtain compound 6. A methanol solution (26 mL) of compound 6 (740 mg, 0.946 mmol), methyl acrylate (1.63 g, 19.0 mmol), and triethylamine (0.10 g, 0.989 mmol) was stirred at 45 °C for 1 day. After removal of the solvent, the residue was purified by silica-gel column chromatography (eluent, chloroform/methanol = 10/1) and GPC to afford the dendron 1b (377 mg, 0.434 mmol) as a yellow oil in 26% yield: 1H-NMR (CDCl3) δ̣ 2.00–2.20 (m, 3H), 2.35 (t, J = 6.0 Hz, 8H), 2.44 (t, J = 6.3 Hz, 4H), 2.43–2.44 (m, 1H), 2.49–2.54 (dt, J = 5.4, 6.0 Hz, 4H), 2.60–2.65 (dt, J = 4.5, 6.0 Hz, 2H), 2.69 (t, J = 6.0 Hz, 8H), 2.76 (t, J = 6.3 Hz, 4H), 3.07–3.15 (m, 2H), 3.26–3.45 (m, 8H), 3.53–3.59 (m, 3H), 3.60 (s, 12H), 3.73 (s, 3H), 3.84 (t, J = 6.0 Hz, 2H), 4.87–4.93 (m, 1H), 7.13 (t, J = 5.4 Hz, 2H), 8.49 (t, J = 4.5 Hz, 1H); 13C-NMR (CDCl3) δ 18.5, 24.0, 25.6, 30.5, 32.7, 37.1, 37.6, 40.2, 41.9, 49.0, 49.2, 49.7, 50.8, 51.6, 52.3, 52.8, 67.6, 71.4, 166.5, 168.3, 171.0, 172.3, 173.1; IR (neat) νmax = 1681, 1733 cm−1; MALDI-TOF Mass Found: m/z 870.04. Calcd. for C40H69N8O13: [MH+], 869.49.

Formation of zwitterionic fullerodendrons 9 or 12

A solution of dendron 1a or 1b (5.16 mg/9.56 mg, 0.0110 mmol) in n class="Chemical">benzonitrile (1.5 mL) was added to a solution of C60 or C70 (0.18 mM) in benzonitrile (1.5 mL). The solution was investigated using Vis/NIR.

Preparation of zwitterionic fullerodendron 12a

A solution of dendron 1a (11.2 mg, 0.0239 mmol) in n class="Chemical">benzonitrile (2 mL) was added to a solution of C70 (1 mg) in 1,2,4-trimethylbenzene (5 mL). The mixture was stirred for 1 h at 25 °C under an Ar atmosphere. Centrifugation (4000 g) of the suspension for 30 min gave a brown precipitation: IR (neat) νmax = 1,663, 1,674, 1,731 cm−1; APPI MS Found: m/z 1309.52. Calcd. for C92H37N4O7: [MH+], 1309.27.

Conclusions

The results described herein show the first example of a DBN-n class="Chemical">focal dendron and the formation of a zwitterionic fullerodendron, observed by UV-vis-NIR spectra, having an anionic fullerene moiety at the focal point. In particular, the C70 anion was effectively stabilized by the site isolation effect of the dendritic wedge. The lifetime of zwitterionic fullerodendron 12b formed by the reaction of C70 with the DBN-focal dendron 1b is approximately 20 times longer than that of zwitterionic complex between C60 and DBN. It is notable that the reversible formation of zwitterionic fullerodendrons potentially applicable to sensing fullerenes, because the absorption maximum of an anionic fullerene moiety should depend on the number of the carbon atoms in a fullerene. Further work is in progress to explore the selective sensing of the fullerene family using zwitterionic fullerodendrons.