Synthesis of azulenophthalimides by phosphine-mediated annulation of 1,2-diformylazulenes with maleimides

Taku Shoji *a, Takanori Araki a, Nanami Iida a, Kota Miura a, Akira Ohta a, Ryuta Sekiguchi a, Shunji Ito b and Tetsuo Okujima c
aGraduate School of Science and Technology, Shinshu University, Matsumoto, 390-8621, Nagano, Japan. E-mail: tshoji@shinshu-u.ac.jp
bGraduate School of Science and Technology, Hirosaki University, Hirosaki 036-8561, Aomori, Japan
cDepartment of Chemistry and Biology, Graduate School of Science and Engineering, Ehime University, Matsuyama 790-8577, Ehime, Japan

Received 17th October 2018 , Accepted 27th November 2018

First published on 28th November 2018


The reaction of 1,2-diformylazulenes with several maleimides in the presence of PPh3 gave the corresponding azulenophthalimides with an ester function at the five-membered ring in moderate to good yields. The removal of the ester function from the azulenophthalimide with an N-cyclohexyl function was also achieved by the treatment with 100% H3PO4. The optical and electrochemical properties of the azulenophthalimides were investigated by UV/Vis measurements, voltammetry experiments and DFT calculations. The azulenophthalimide without the ester function exhibited remarkable emission under acidic conditions. Moreover, the azulenophthalimides displayed a significant spectral change under electrochemical reduction conditions.


Introduction

The phthalimide substructure is commonly found in many natural products and pharmaceuticals.1 In recent years, phthalimides have been explored for application to organic electronics, such as n-type semiconductors2 and fluorescent materials,3 because of their high electron mobility and excellent luminescence efficiency. Therefore, various derivatives with a phthalimide substructure have been prepared and their properties have been determined.

Recently, a variety of efficient synthetic procedures for azulene derivatives have been developed owing to their useful applications in organic electronics,4 photovoltaics,5 and electrochromic6 and stimuli-responsive materials.7 The synthesis and properties of azulene analogs of imide derivatives have also been reported by several groups (Fig. 1). Tani et al. reported the preparation of a tetraazulene-fused tetracene diimide, which shows a weak absorption band in the near-infrared region, via the Suzuki–Miyaura cross-coupling reaction of 2-azulenylboronic acid ester with tetrabromonaphthalene diimide, followed by the Scholl reaction with 2,3-dichloro-5,6-dicyano-1,4-benzoquinone (DDQ) at −60 °C.8 The tetraazulene-fused diimide is also revealed to show n-type semi-conductivity by the space charge-limited current experiment. Gao and co-workers have studied the synthesis and physical properties of diimide derivatives with a fused 2,2′-biazulene framework, as well as their polymers.9 In these research studies, they also reported that diimide derivatives behave as a high-performance field-effect transistor. These studies should definitely contribute to the applications of the azulene derivatives to organic electronic materials. However, the synthesis of much simpler derivatives, i.e., the synthesis and properties of azulene-fused phthalimides, has never been examined to date. Thus, the preparation of azulenophthalimides should greatly contribute to a better understanding of the characteristic feature of these series.


image file: c8qo01121d-f1.tif
Fig. 1 Azulene-fused diimide derivatives reported by Tani and Gao et al.

For the synthesis of azulenophthalimides, we firstly attempted the oxidative aromatization of compound 2, which was obtained by the cycloaddition of azulenosulfolene 1 with N-phenylphthalimide,10 with DDQ, but the reaction resulted in the decomposition of the compound (Scheme 1). In view of the unpromising result, we have examined another synthetic pathway to azulenophthalimides.


image file: c8qo01121d-s1.tif
Scheme 1 An attempt to the synthesis of azulenophthalimides via oxidative aromatization.

In 1979, Haddadin et al. reported an efficient synthesis of phthalimide derivatives by the reaction of o-phthalaldehydes with maleimides in the presence of P(OEt)3.11 This method provides a convenient way to give phthalimide derivatives, but the yield of the products significantly depends on the substituent on the nitrogen atom of the maleimides. More recently, Yamaguchi et al. reported the phosphine-mediated reaction of o-phthalaldehyde derivatives with maleimide in the presence of DBU to form the corresponding phthalimides in good yields.12 Since 1,2-diformylazulene derivatives can be readily prepared by the procedure reported by us, recently,13 we decided to adopt a phosphorus reagent-mediated reaction as another synthetic pathway to azulenophthalimides.

Herein, we describe the first and efficient synthesis of azulenophthalimides by the phosphine-mediated annulation reaction of 1,2-diformylazulenes with several maleimides. The electronic properties of the azulenophthalimides obtained by this research were characterized by spectroscopic measurements and theoretical calculations. The electrochemical behavior of the compounds was also investigated by cyclic voltammetry (CV) and differential pulse voltammetry (DPV) experiments.

Results and discussion

Synthesis

As an initial investigation, the annulation reaction of 1,2-diformylazulene derivative 3 with N-methylmaleimide was selected as a model reaction for the optimization of the reaction conditions by varying phosphorus reagents and solvents. The effect of phosphine reagents was examined by using 1,4-dioxane as a solvent under an Ar atmosphere. The yield of the product significantly depended on the phosphorus reagents employed as shown in Table 1.
Table 1 Optimization of the reaction conditions

image file: c8qo01121d-u1.tif

Entry PR3 Solvent Temperature 4a [%] 5 [%]
a Isolated yield. b 1.2 equiv. of PPh3 were employed. c 3.0 equiv. of PPh3 were employed.
1 P(OEt)3[thin space (1/6-em)]b 1,4-Dioxane Reflux 0 0
2 P(nBu)3[thin space (1/6-em)]b 1,4-Dioxane Reflux 27 19
3 PPh3[thin space (1/6-em)]b 1,4-Dioxane Reflux 50 37
4 PPh3[thin space (1/6-em)]c 1,4-Dioxane Reflux 58 0
5 PPh3[thin space (1/6-em)]b DMF 70 24 0
6 PPh3[thin space (1/6-em)]b DMF 100 81 0
7 PPh3[thin space (1/6-em)]b Toluene 100 45 27
8 PPh3[thin space (1/6-em)]b DMSO 100 44 0


Even though the successful conditions for phthalimide synthesis reported by Haddadin et al. were adopted for the reaction of 1,2-diformylazulene derivative 3 with N-methylmaleimide, the desired azulenophthalimide 4a was not obtained by the reaction in 1,4-dioxane in the presence of P(OEt)3 (entry 1). Since the phosphite ester was unsuitable for the reaction, the annulation reaction was investigated by using phosphine reagents for further investigation. When the alkylphosphine P(nBu)3 was employed in the reaction, the targeted product 4a was obtained in 27% yield, along with 5 in 19% yield (entry 2). The structure of by-product 5 was clarified by single crystal X-ray analysis as shown in Fig. 2. The formation of the geometrical isomer of 5 in this reaction was not observed by the NMR experiments of the products. The reaction of 3 in the presence of PPh3 improved the yield of 4a (50%), along with the formation of 5 in 37% yield (entry 3). When an excess PPh3 (3 equiv.) was used in the reaction, 4a was obtained as a sole product in 58% yield, but the yield was not much improved (entry 4). Among the phosphine reagents tested, PPh3 was found to be the best with respect to the product yield. Thus, the solvent effects in the PPh3-mediated condensation reaction of 3 with N-methylmaleimide were examined for further improvement of the product yield (entries 5–8). As a result, we found that DMF afforded the best result (81% yield) among the solvents examined. Changing the reaction temperature from 100 °C to 70 °C resulted in a remarkable decrease in the product yield (entry 5). Therefore, the reaction conditions using PPh3 as a phosphine reagent in DMF at 100 °C were selected for further investigation for the synthesis of azulenophthalimides.


image file: c8qo01121d-f2.tif
Fig. 2 ORTEP drawing of 5 (CCDC 1529064); ellipsoids are drawn at the 50% probability level.14

Utilizing the optimized reaction conditions, we examined the scope of the annulation reaction of 3 with several maleimides having a functional group on the nitrogen atom. The yield and the structure of the products are summarized in Table 2. Generally, the product could be obtained in good yields, independent of the substituents on the nitrogen atom of the maleimides.

Table 2 Synthesis of azulenophthalimides 4b–4g

image file: c8qo01121d-u2.tif

Entry R Product Yielda [%]
a Isolated yield.
1 Et 4b 79
2 image file: c8qo01121d-u3.tif 4c 83
3 H 4d 64
4 image file: c8qo01121d-u4.tif 4e 77
5 image file: c8qo01121d-u5.tif 4f 82
6 image file: c8qo01121d-u6.tif 4g 81


N-Alkylmaleimides reacted with 3 under the optimized conditions to give the desired azulenophthalimides 4b and 4c in 79% and 83% yields, respectively. Likewise, the reaction of N-unsubstituted maleimide afforded the corresponding product 4d in 64% yield. Similar to the results described above, maleimides with an aryl substituent on the nitrogen atom were also reacted under the conditions to afford azulenophthalimides 4e–4g (4e: 77%, 4f: 82%, 4g: 81%). From these results, the formation of azulenophthalimides was not affected by the functional group on the nitrogen atom of the maleimides, since the products were obtained in almost high yields, similarly.

Two reaction pathways are considered for the formation of azulenophthalimides by this reaction. One is starting from the Wittig reaction to give 5, followed by an intramolecular condensation reaction of 5 to form the azulenophthalimide skeleton (Scheme 2). In contrast, the other starts with an intermolecular condensation reaction, followed by the intramolecular Wittig reaction (Scheme 3).


image file: c8qo01121d-s2.tif
Scheme 2 Plausible reaction mechanism for the formation of azulenephthalimide 4a from 3via the Wittig reaction product 5.

image file: c8qo01121d-s3.tif
Scheme 3 Plausible reaction mechanism for the formation of azulenephthalimide 4a from 3 starting from the intermolecular condensation reaction.

To elucidate the reaction mechanism, the Wittig reaction of 3 with the ylide15 prepared from N-methylmaleimide was examined. Thus, the reaction of 3 with the phosphorus ylide gave compound 5 in 54% yield, but further reaction to form 4a did not proceed under the reaction conditions. Although the reaction of 5 with PPh3 in DMF afforded 4a in moderate yield (51%), the reaction was not completed even though the reaction time was prolonged for 48 h. As illustrated in Scheme 2, the formation of 4a might be proceeded by aldol-type condensation, in which PPh3 behaves as a base. These results indicate that the mechanism should not be the main route to the product 4a from 3, from the viewpoint of the product yield and the prolonged reaction period from 5 to 4a.

In view of the results, the reaction mechanism illustrated in Scheme 3 seems to be more plausible for the one-pot formation of 4a from 3. Initially, the enolate formed in situ by the reaction of PPh3 with maleimide is added to the formyl group at the 2-position of the azulene ring to give intermediate A. Then, the intermediate A is transformed into phosphorus ylide B by the proton transfer. The intramolecular Wittig reaction of B proceeds to afford the intermediate C, which was aromatized to give azulenophthalimide 4a by the elimination of water. Although the Wittig reaction to give 5 should compete with the nucleophilic addition in some solvents, such as 1,4-dioxane, the nucleophilic addition of the enolate to the formyl group on 3 should be faster under the optimized conditions from the results on the reaction outcome described above.

The synthesis of the derivatives having two azulenophthalimide moieties in a molecule was investigated in a similar manner as described above. The alkoxy moiety of the ester group on 1,2-diformylazulene 3 was converted from the methoxy to n-butoxy group in order to avoid the solubility problem caused by the introduction of multiple azulenophthalimide units in a molecule (Scheme 4). The transesterification reaction of 2-methylazulene derivative 6 with a methoxycarbonyl function was performed by the treatment with metallic sodium in n-butanol to afford 7 in 84% yield.16 The 2-methylazulene derivative 7 having an n-butoxycarbonyl group was transformed into the corresponding 3-formylazulene derivative 8 in 97% yield by the Vilsmeier formylation reaction. Eventually, the preparation of 1,2-diformylazulene 9 was achieved by the conversion of the methyl moiety at the 2-position of 8 to the enamine by the reaction with N,N-dimethylformamide dimethyl acetal (DMFDMA), followed by the oxidative cleavage with NaIO4.9


image file: c8qo01121d-s4.tif
Scheme 4 Preparation of 1,2-diformylazulene 9 from 2-methylazulene derivative 6.

The reaction of 9 with 4,4′-bis(maleimidodiphenyl)methane and N,N’-1,3-phenylenedimaleimide in the presence of PPh3 afforded the corresponding bis(azulenophthalimides) 10 (23%) and 11 (20%), but in low yields (Scheme 5). Despite the ester moiety having a long alkyl chain in the alkoxy moiety, compounds 10 and 11 showed considerable insolubility in common organic solvents. The poor solubility of these compounds is probably due to the π–π stacking of the molecules based on the extension of the π-conjugated system.


image file: c8qo01121d-s5.tif
Scheme 5 Synthesis of bis(azulenophthalimides) 10 and 11.

In order to obtain the parent azulenophthalimide without the ester function at the five-membered ring, the deesterification reaction of 4c was investigated by the treatment with 100% H3PO4 due to the high solubility of the compound. As a result, azulenophthalimide 4c was transformed into the parent derivative 12 in quantitative yield (99%) by the treatment with 100% H3PO4 at 100 °C (Scheme 6). Since the 1,3-positions of the azulene ring are highly reactive to the electrophilic substitution reaction, the product 12 could be applied for further transformation by the introduction of functional groups at the five-membered ring.


image file: c8qo01121d-s6.tif
Scheme 6 Deesterification of 4c by the treatment with 100% H3PO4.

Spectroscopic properties

Azulenophthalimides 4a–4g and 10–12 were fully characterized using their spectral data as shown in the Experimental section. The signal assignment of 1H NMR was accomplished by COSY experiments. High-resolution mass spectra ionized by the FAB method displayed the presumed molecular ion peaks of the products. These results prove the correctness of the structure of the novel compounds.

In the 1H NMR spectrum, the ring proton signals in the aromatic region of compound 12 showed a pronounced high-field shift, compared with those of 4c, due to the removal of the electron-withdrawing ester function from 4c (Fig. 3). It is known that the ring-fused azulene derivatives at the five-membered ring induce a remarkable bond-alternation at the seven-membered ring, which appeared in the coupling constant on 1H NMR at the azulene ring.17 Azulenophthalimides 4a–4g and 10–12 exhibited the characteristic feature of the bond-alternation in their 1H NMR spectra. For instance, the coupling constant 3J(H4,H5) of 4c and 12 at the seven-membered ring showed that both have J = 8.6 Hz, which implies the single-bond nature in this part. However, the larger coupling constant was observed in the vicinal protons H5 and H6 of 4c (J = 11.2 Hz) and 12 (J = 11.5 Hz), which indicates the double-bond character in this part. Therefore, these results suggest that the aromaticity of the azulene moiety of 4c and 12 is reduced by the ring-fusion of the phthalimide unit at the five-membered ring.


image file: c8qo01121d-f3.tif
Fig. 3 1H NMR spectra of (a) 4c and (b) 12 in CDCl3 (500 MHz).

In order to elucidate the aromaticity for each ring of azulene, 4a′ and its deesterified form 4a′′, in which the isopropyl group was replaced by a hydrogen atom, nucleus-independent chemical shift (NICS) calculations were performed at the GIAO/HF/6-31G* level. The NICS(0) and NICS(1) values of azulene, 4a′ and 4a′′ are shown in Table 3. Indeed, the NICS(0) and NICS(1) values for each ring of 4a′ and 4a′′ showed a low field shift, compared with those of the parent azulene (Table 3). The outcomes imply the decrease in the aromaticity both at the five- and seven-membered rings of 4a′ and 4a′′ as predicted from the bond-alternation observed from the 1H NMR spectra. The lower aromaticity of 4a′ and 4a′′ in the azulene part should be attributed to the high aromaticity of the benzene ring in the phthalimide moiety to avoid the contribution of the unstable ortho-quinoidal structure in their resonance structures (Scheme 7).


image file: c8qo01121d-s7.tif
Scheme 7 Presumed resonance structures of 4a′ and 4a′′.
Table 3 The NICS(0) and NICS(1) of azulene, 4a′ and 4a′′

image file: c8qo01121d-u7.tif

Sample Ring NICS(0) [ppm] NICS(1) [ppm]
Azulene A −8.25 −9.90
B −21.55 −21.03
 
4a′ C −7.15 −9.38
D −12.51 −13.61
E −11.60 −12.71
 
4a′′ F −4.82 −7.09
G −13.51 −14.95
H −12.10 −13.06


The NICS values of the seven-membered C-ring of 4a′ denoted the higher-field shift relative to those of the F-ring of 4a′′, whereas the opposite phenomenon was observed between those of their five-membered rings. Since the NICS value of the C-ring of 4a′ is similar to that of the tropylium ion itself (−7.6 ppm),18 the contribution of the tropylium ion structure formed by the resonance with the ester function increased the aromaticity of the C-ring of 4a′′, compared with that of the F-ring of 4a′′ (Scheme 7).

The absorption maxima and molar extinction coefficient (log ε) of azulenophthalimides 4a–4g and 10–12 in CH2Cl2 are summarized in the experimental section. The UV/Vis spectra of 4c and 12 in CH2Cl2 and that of 12 in 30% CF3CO2H/CH2Cl2 and the fluorescence spectrum of 12 in 30% CF3CO2H/CH2Cl2 are shown in Fig. 4 and 5, respectively.


image file: c8qo01121d-f4.tif
Fig. 4 UV/Vis spectrum of 4c in CH2Cl2 (blue line), 12 in CH2Cl2 (pink line) and 12 in 30% CF3CO2H/CH2Cl2 (light-blue line); the dotted lines represent the magnification (×50) of the spectrum of 4c in CH2Cl2 and 12 in CH2Cl2 in the visible region.

image file: c8qo01121d-f5.tif
Fig. 5 UV/Vis spectrum (blue line) and fluorescence spectrum (pink dot-line) of 12 in 30% CF3CO2H/CH2Cl2.

The UV/Vis spectra of 4a–4g in CH2Cl2 showed a relatively strong absorption band at around 400 nm and a weak absorption band at around 570 nm. The absorption bands of 4a–4g were little affected by the substituent on the nitrogen atom of the phthalimide moiety, which showed almost the same maximum absorption wavelengths with each other. These results imply the small contribution of the substituent on the nitrogen atom of the phthalimide moiety toward the electronic nature of the molecules.

It is well known that azulene derivatives display a color change depending on pH, so called halochromism, under acidic conditions.19 The halochromic behavior was observed in 12 from the UV/Vis spectra. The solvent dependence of the color change is shown in Fig. 4. The longest wavelength absorption band of 12 (λmax = 618 nm) showed a bathochromic shift, compared with that of the corresponding ester derivative 4c (λmax = 573 nm) and compound 12 showed a green color in CH2Cl2 solution [Fig. 6(a)]. When the UV/Vis spectrum of 12 was obtained in 30% CF3CO2H/CH2Cl2, the absorption band in the visible region disappeared, along with the generation of a new absorption band at λmax = 391 nm, owing to the generation of cationic species [12 + H]+ by the protonation of the azulene moiety [Fig. 4 and 6(a) and Scheme 6]. The reverse neutralization of [12 + H]+ with Et3N as a base recovered the original green color of 12.


image file: c8qo01121d-f6.tif
Fig. 6 (a) The solution of 12 in CH2Cl2 (left) and in 30% CF3CO2H/CH2Cl2 (right) under ambient conditions; (b) the solution of 12 in CH2Cl2 (left) and in 30% CF3CO2H/CH2Cl2 (right) under the photo-irradiation (λEX = 365 nm).

To elucidate the origin of the absorption bands in the visible region of 4c, 4e and 12, theoretical calculations were performed using time-dependent density functional theory (TD-DFT) at the B3LYP/6-31G** level.20 The frontier Kohn–Sham orbitals and their energy levels are summarized in the ESI.

The HOMOs and LUMOs of 4c, 4e and 12 exhibited very similar orbital coefficients, independent of the substituents on azulene and phthalimide moieties. The results show the small electronic interaction of the substituents on the nitrogen atom of the azulenophthalimides. The calculations disclosed that the longest wavelength absorption band at around 570–620 nm of 4c, 4e and 12 is caused by the transition from the HOMO to the LUMO, i.e., the π–π* transition in the azulene moiety. The calculated HOMO–LUMO gap of 12 (2.77 eV) was lower than those of 4c (2.86 eV) and 4e (2.86 eV), which reflects the bathochromic shift in the longest wavelength absorption band of 12 in the visible region, compared with those of 4c and 4e. On the other hand, the energy levels of the HOMO (−5.59 eV) and LUMO (−2.73 eV) of 4c were almost the same as those of 4e (HOMO = −5.61 eV, LUMO = −2.75 eV). These consequences mean the less contribution of the phenyl group on 4e to the π-extension, due to the orthogonal arrangement of the phenyl group to the phthalimide moiety, as also expected by the UV/Vis measurements (Table 4).

Table 4 Electronic transitions for 4c, 4e, 12 and [12 + H]+ derived from the computed values based on the TD-DFT calculations at the B3LYP/6-31G** level and those of the experimental results
Compound Experimental Computed values
λ max [nm] (log[thin space (1/6-em)]ε) λ max [nm] (strength) Composition of banda (amplitude) H–L gap [eV]
a H = HOMO; L = LUMO.
4c 573 (3.02) 567 (0.0152) H → L (0.9709) 2.86
406 (3.99) 400 (0.0369) H → L+1 (0.7575)
432 (4.12)
 
4e 568 (3.02) 566 (0.0163) H → L (0.9708) 2.86
409 (3.90) 397 (0.0454) H → L+1 (0.7719)
435 (4.02)
 
12 618 (3.06) 593 (0.0197) H → L (0.9744) 2.77
412 (3.99) 438 (4.04) 406 (0.0566) H → L+1 (0.7905)


The azulene derivatives do not exhibit the emission from the S1 → S0 transition because of the fast internal conversion in the S1 state, but the derivatives show uncommon emission from the S2 → S0 transition in contrast to the Kasha rule.21 Similar to the usual azulene derivatives, compound 12 did not show fluorescence in CH2Cl2, whereas the remarkable luminescence was observed in 30% CF3CO2H/CH2Cl2 at λFL = 444 nm (λEX = 390 nm) with the Stokes shift of 53 nm [Fig. 5 and 6(b)]. This phenomenon should be ascribed to the formation of the tropylium ion structure [12 + H]+ by the protonation under acidic conditions, which eliminates the contribution of the azulene form that takes part in the quenching, so the light-emission phenomenon of the phthalimide moiety is restored.

Electrochemical properties

To clarify the electrochemical behavior of azulenophthalimides 4a–4g and 12, redox potentials of these products were measured by CV and DPV.22 The measurements were carried out by using a standard three-electrode configuration. Tetraethylammonium perchlorate (0.1 M) in benzonitrile was used as a supporting electrolyte with platinum wire auxiliary and disk working electrodes. All measurements were performed under an argon atmosphere, and potentials were related to the reference electrode formed from Ag/AgNO3, by using Fc/Fc+ as an internal reference, which discharges at +0.15 V under these conditions. The redox potentials are summarized in Table 5. The cyclic voltammograms of 4a obtained at different scan rates are shown in Fig. 7.
image file: c8qo01121d-f7.tif
Fig. 7 Cyclic voltammograms for the reduction of 4a in benzonitrile (1 mM) containing Et4NClO4 (0.1 M) as the supporting electrolyte at different scan rates.
Table 5 Redox potentialsa of azulenophthalimides 4a–4g and 12
Sample Method E 1 ox [V] E 2 ox [V] E 1 red [V] E 2 red [V]
a V versus Ag/AgNO3, 1 mM in benzonitrile containing Et4NClO4 (0.1 M), Pt electrode (internal diameter: 1.6 mm), scan rate = 100 mV s−1 and internal reference (Fc/Fc+ = +0.15 V). b Anodic potential (Epa). c Cathodic potential (Epc).
4a CV +1.02b −1.32c
−0.40b
(DPV) (+0.96) (−1.15) (−1.27)
4b CV +1.07b −1.37c
−0.32b
(DPV) (+0.94) (−1.28) (−1.90)
4c CV +1.02b −1.39c
−0.33b
(DPV) (+0.95) (−1.27) (−1.93)
4d CV +1.01b −1.31c
−0.48b
(DPV) (+0.96) (−1.25)
4e CV +1.05b −1.30c
−0.32b
(DPV) (+0.97) (+1.84) (−1.24)
4f CV +1.04b −1.30c
−0.30b
(DPV) (+0.99) (−1.24) (−1.93)
4g CV +1.08b −1.32c
−0.31b
(DPV) (+0.99) (+1.38) (−1.25) (−1.92)
12 CV +0.70b −1.49c
−0.71b
(DPV) (+0.61) (+0.87) (−1.46)


Both the electrochemical oxidation and reduction of azulenophthalimides 4a–4g and 12 exhibited irreversible waves in CV. Supposing that the delays in electron transfer and/or structural changes in the molecules under the redox reaction are responsible for the irreversibility of the voltammograms, the reversibility should be improved by decreasing the scan rate. On the other hand, when the irreversibility is derived from the instability of the generated radical ion, increasing the scan rate improves the reversibility. Although the reduction wave of 4a was measured by CV at different scan rates, the reversibility of the redox wave was not improved as shown in Fig. 7. This result suggests the considerable instability of the radical anionic states, so the original neutral species should not be regenerated owing to the decomposition or subsequent reaction.23 Moreover, since the redox potentials of 4a–4g were almost the same, the substituent on the nitrogen atom of the phthalimide moiety does not affect the reactivity toward the electrochemical reaction as anticipated by the UV/Vis measurements and DFT calculations.

Among the azulenophthalimides examined by voltammetric analysis, compound 12 without the ester function on the azulene ring showed the lowest oxidation (E1ox = +0.61 V) and the highest reduction potentials (E1red = −1.46) in DPV. In consequence, the HOMO and LUMO levels of these compounds are clearly influenced by the ester substituent on the azulene ring, as predicted by the theoretical calculations, although the substituent on the nitrogen atom of the phthalimide moiety does not affect the levels.

We have reported the synthesis of various azulene-substituted redox active chromophores for the purpose of constructing stable organic electrochromic materials under redox conditions.24 Although the electrochromic behavior of the derivatives, in which the azulene rings are connected to the aromatic ring directly or through the π-conjugation, has been examined in these studies, that of the ring-fused azulene derivatives has been scarcely investigated so far. The ring-fused azulene derivatives may become a new partial structure for practical organic electrochromic materials. Thus, to examine the electrochromic behavior of the azulene derivatives, the spectral changes of the azulenophthalimide derivatives 4a–4g and 12 were monitored by visible spectroscopy under the constant voltage redox conditions.

The electrochromic behavior of 4a–4g resembled each other, reflecting their similar electronic properties (see the ESI). The electrochemical reduction of 4a–4c developed a new absorption band at around 450–570 nm, along with an isosbestic point at λ = 570 nm, due to the formation of anionic species. The reverse oxidation of the reduced species of 4a–4c showed a decrease in the newly appeared absorption band, but the recovery of the original spectra of 4a–4c was observed incompletely. Azulenophthalimide 4d without the substituent on the nitrogen atom displayed reversible spectral changes under the redox conditions, although spectral changes were not so significant.

Azulenophthalimides 4e–4g with an aryl substituent on the nitrogen atom also showed spectral changes under electrochemical redox conditions. For example, the electrochemical reduction of 4e exhibited a gradual increase in the absorption band at around 500 nm, along with a slight diminishment of the original absorption band at around 570 nm. The electrochemical oxidation of the reduced species of 4e returned to almost the original spectrum (Fig. 8). A similar electrochromic behavior of 4e–4g suggests that the redox reaction should proceed at the azulene ring rather than at the phthalimide moiety, as also predicted by the voltammetry experiments.


image file: c8qo01121d-f8.tif
Fig. 8 Continuous change in the visible spectrum of 4e: constant-voltage electrochemical reduction at −1.75 V (top) and reverse oxidation of the reduced species at ±0 V (bottom) in benzonitrile containing Et4NClO4 (0.1 M) at 20 s intervals.

Similar to the results described above, azulenophthalimide 12 without the ester function on the azulene ring also displayed the spectral changes under the electrochemical reduction conditions. The electrochemical reduction of 12 at −1.75 V showed the development of an absorption band at around 550 nm, along with an isosbestic point at λ = 635 nm. The newly appeared absorption band of 12 exhibited a bathochromic shift, compared with those of the anionic species of 4a–4g. In the reverse oxidation of the reduced species of 12, absorption bands of the anionic species of 12 gradually decreased and coincidently recovered the original absorption bands, but incompletely (Fig. 9).


image file: c8qo01121d-f9.tif
Fig. 9 Continuous change in the visible spectrum of 12: constant-voltage electrochemical reduction at −1.75 V (top) and the reverse oxidation of the reduced species at ±0 V (bottom) in benzonitrile containing Et4NClO4 (0.1 M) at 20 s intervals.

Conclusions

We have described herein a novel procedure for the synthesis of azulenophthalimides with several substituents on the nitrogen atom from 1,2-diformylazulene derivatives, which are readily available by the procedure reported by us, recently. This method should open up a new synthetic pathway to azulenophthalimides that are difficult to synthesize by the previous methods.

By utilizing the procedure, azulenophthalimide derivatives 4a–4g, 10 and 11 were prepared starting from 1,2-diformylazulene derivatives 3 and 9 in moderate to good yields. The UV/Vis measurements, voltammetry experiments and DFT calculations revealed the optical and electrochemical properties of 4a–4g, which are not significantly affected by the substituent on the nitrogen atom of the phthalimide moiety. The deesterified compound 12, which was obtained by the treatment of 4c with 100% H3PO4, did not show the emission in common organic solvents, but a remarkable fluorescence was observed under the acidic conditions because of the formation of the azulenium ion structure by the protonation at the five-membered ring.

The azulenophthalimides displayed significant spectral changes attributing to the formation of anionic species under the electrochemical reduction conditions. However, the reverse oxidation of the reduced species of 4a–4g and 12 did not recover their original spectra, which should be attributed to the instability of the generated anionic species.

As mentioned in the introduction section, phthalimides should be applied to semiconductors and fluorophores. Further functionalization of the azulenophthalimides prepared in this study may induce practical optical and electrochemical properties required for the development of electronic materials. The construction of novel π-electron systems based on azulenophthalimides is currently in progress in our laboratory.

Conflicts of interest

There are no conflicts to declare.

Acknowledgements

This work was supported by the JSPS KAKENHI Grant Number 17K05780. The authors are grateful to Professor Dr Mitsunori Oda (Shinshu University) for valuable comments on this study. We thank Dr Shigeki Mori (Advanced Research Support Center, Ehime University) for helping us to analyze the X-ray diffraction results of compound 5.

Notes and references

  1. J. Vamecq, P. Bac, C. Herrenknecht, P. Maurois, P. Delcourt and J. P. Stables, J. Med. Chem., 2000, 43, 1311–1319 CrossRef CAS; W. Gao, C. E. Bohl and J. T. Dalton, Chem. Rev., 2005, 105, 3352–3370 CrossRef; E. C. Lawson, D. K. Luci, S. Ghosh, W. A. Kinney, C. H. Reynolds, J. Qi, C. E. Smith, Y. Wang, L. K. Minor, B. J. Haertlein, T. J. Parry, B. P. Damiano and B. E. Maryanoff, J. Med. Chem., 2009, 52, 7432–7445 CrossRef; K.-E. Kil, P. Poutiainen, Z. Zhang, A. Zhu, J.-K. Choi, K. Jokivarsi and A.-L. Brownell, J. Med. Chem., 2014, 57, 9130–9138 CrossRef; P. Wu and T. E. Nielsen, Chem. Rev., 2017, 117, 7811–7856 CrossRef.
  2. X. Guo, A. Facchetti and T. J. Marks, Chem. Rev., 2014, 114, 8943–9021 CrossRef CAS; R. R. San Juan, A.-J. Payne, G. C. Welch and A. F. Eftaiha, Dyes Pigm., 2016, 132, 369–377 CrossRef; H. T. Turan, İ. Yavuz and V. Aviyente, J. Phys. Chem. C, 2017, 121, 25682–25690 CrossRef; X. Song, J. Zhao, W. Zhang and L. Chen, Chin. Chem. Lett., 2018, 29, 331–335 CrossRef; T. A. Welsh, A. Laventure, T. Baumgartner and G. C. Welch, J. Mater. Chem. C, 2018, 6, 2148–2154 RSC.
  3. C. Gatanaga, B. Yang, Y. Inadomi, K. Usui, C. Ota, T. Katayama, H. Suemune and M. Aso, ACS Chem. Biol., 2016, 11, 2216–2221 CrossRef CAS; A. G. Griesbeck, B. Öngel and M. Atar, J. Phys. Org. Chem., 2017, 30, e3741 CrossRef; B. Li, Z. He, H. Zhou, H. Zhang, W. Li, T. Cheng and G. Liu, Dyes Pigm., 2017, 146, 300–304 CrossRef; H. Yan, X. Meng, B. Li, S. Gea and Y. Lu, J. Mater. Chem. C, 2017, 5, 10589–10599 RSC; R. Orita, M. Franckevičius, A. Vyšniauskas, V. Gulbinas, H. Sugiyama, H. Uekusa, K. Kanosue, R. Ishigea and S. Ando, Phys. Chem. Chem. Phys., 2018, 20, 16033–16044 RSC.
  4. Y. Yamaguchi, Y. Maruya, H. Katagiri, K.-I. Nakayama and Y. Ohba, Org. Lett., 2012, 14, 2316–2319 CrossRef CAS; Y. Yamaguchi, K. Ogawa, K.-I. Nakayama, Y. Ohba and H. Katagiri, J. Am. Chem. Soc., 2013, 135, 19095–19098 CrossRef; J. Yao, Z. Cai, Z. Liu, C. Yu, H. Luo, Y. Yang, S. Yang, G. Zhang and D. Zhang, Macromolecules, 2015, 48, 2039–2047 CrossRef; Y. Yamaguchi, M. Takubo, K. Ogawa, K.-I. Nakayama, T. Koganezawa and H. Katagiri, J. Am. Chem. Soc., 2016, 138, 11335–11343 CrossRef.
  5. E. Puodziukynaite, H.-W. Wang, J. Lawrence, A. J. Wise, T. P. Russell, M. D. Barnes and T. Emrick, J. Am. Chem. Soc., 2014, 136, 11043–11049 CrossRef CAS; H. Nishimura, N. Ishida, A. Shimazaki, A. Wakamiya, A. Saeki, L. T. Scott and Y. Murata, J. Am. Chem. Soc., 2015, 137, 15656–15659 CrossRef; Y. Chen, Y. Zhu, D. Yang, S. Zhao, L. Zhang, L. Yang, J. Wu, Y. Huang, Z. Xu and Z. Lu, Chem. – Eur. J., 2016, 22, 14527–14530 CrossRef; P. Cowper, A. Pockett, G. Kociok-Köhn, P. J. Cameron and S. E. Lewis, Tetrahedron, 2018, 74, 2775–2786 CrossRef.
  6. T. Shoji, S. Ito, K. Toyota, M. Yasunami and N. Morita, Chem. – Eur. J., 2008, 14, 8398–8408 CrossRef CAS; T. Shoji, S. Ito, T. Okujima and N. Morita, Chem. – Eur. J., 2013, 19, 5721–5730 CrossRef; T. Shoji, M. Maruyama, A. Maruyama, S. Ito, T. Okujima and K. Toyota, Chem. – Eur. J., 2014, 20, 11903–11912 CrossRef; T. Shoji, T. Araki, S. Sugiyama, A. Ohta, R. Sekiguchi, S. Ito, T. Okujima and K. Toyota, J. Org. Chem., 2017, 82, 1657–1665 CrossRef.
  7. M. Murai, S. Iba, H. Ota and K. Takai, Org. Lett., 2017, 19, 5585–5588 CrossRef CAS; Y. M. Poronik, L. M. Mazur, M. Samoć, D. Jacquemin and D. T. Gryko, J. Mater. Chem. C, 2017, 5, 2620–2628 RSC; C. M. López-Alled, A. Sanchez-Fernandez, K. J. Edler, A. C. Sedgwick, S. D. Bull, C. L. McMullin, G. Kociok-Köhn, T. D. James, J. Wenk and S. E. Lewis, Chem. Commun., 2017, 53, 12580–12583 RSC.
  8. T. Koide, M. Takesue, T. Murafuji, K. Satomi, Y. Suzuki, J. Kawamata, K. Terai, M. Suzuki, H. Yamada, Y. Shiota, K. Yoshizawa and F. Tani, ChemPlusChem, 2017, 82, 1010–1014 CrossRef CAS.
  9. H. Xin, C. Ge, X. Yang, H. Gao, X. Yang and X. Gao, Chem. Sci., 2016, 7, 6701–6705 RSC; H. Xin, C. Ge, X. Jiao, X. Yang, K. Rundel, C. R. McNeill and X. Gao, Angew. Chem., Int. Ed., 2018, 57, 1322–1326 CrossRef CAS; H. Xin, J. Li, C. Ge, X. Yang, T. Xue and X. Gao, Mater. Chem. Front., 2018, 2, 975–985 RSC; T. Shoji and S. Ito, Sci. China: Chem., 2018, 61, 973–974 CrossRef.
  10. K. Fujimori, K. Yamane, M. Yasunami and K. Takase, Heterocycles, 1997, 45, 2093–2095 CrossRef CAS.
  11. M. J. Haddadin, B. J. Agha and R. F. Tabri, J. Org. Chem., 1979, 44, 494–497 CrossRef CAS.
  12. C. Yuan, S. Saito, C. Camacho, S. Irle, I. Hisaki and S. Yamaguchi, J. Am. Chem. Soc., 2013, 135, 8842–8845 CrossRef CAS PubMed.
  13. T. Shoji, T. Araki, N. Iida, Y. Kobayashi, A. Ohta, R. Sekiguchi, S. Ito, S. Mori, T. Okujima and M. Yasunami, Eur. J. Org. Chem., 2018, 1145–1157 CrossRef CAS.
  14. CCDC 1529064 (for 5) contain the supplementary crystallographic data for this paper.
  15. M. Oda, T. Nakamura, M. Neha, D. Miyawaki, A. Ohta, S. Kuroda and R. Miyatake, Eur. J. Org. Chem., 2014, 5976–5985 CrossRef CAS.
  16. W. Treibs, C. Vollrad and M. Reimann, Justus Liebigs Ann. Chem., 1961, 648, 164–184 CrossRef CAS.
  17. D. J. Bertelli and P. Crews, Tetrahedron, 1970, 26, 4717–4728 CrossRef CAS; T. Morita, T. Nakadate and K. Takase, Heterocycles, 1981, 15, 835–838 CrossRef; K. Fujimori, T. Fujita, K. Yamane, M. Yasunami and K. Takase, Chem. Lett., 1983, 1721–1724 CrossRef; Y. Lu, D. M. Lemal and J. P. Jasinski, J. Am. Chem. Soc., 2000, 122, 2440–2445 CrossRef; S. Ito, T. Kubo, M. Kondo, C. Kabuto, N. Morita, T. Asao, K. Fujimori, M. Watanabe, N. Harada and M. Yasunami, Org. Biomol. Chem., 2003, 1, 2572–2580 RSC.
  18. P. V. R. Schleyer, C. Maerker, A. Dransfeld, H. J. Jiao and J. R. V. Hommes, J. Am. Chem. Soc., 1996, 118, 6317–6318 CrossRef CAS.
  19. D. H. Reid, W. H. Stafford, W. L. Stafford, G. McLennan and A. Voigt, J. Chem. Soc., 1958, 1110–1117 RSC; E. C. Kirby and D. H. Reid, J. Chem. Soc., 1960, 494–501 RSC; T. Shoji, S. Ito, K. Toyota and N. Morita, Eur. J. Org. Chem., 2010, 1059–1069 CrossRef CAS; T. Shoji, S. Ito, J. Higashi and N. Morita, Eur. J. Org. Chem., 2011, 5311–5322 CrossRef; E. Amir, R. J. Amir, L. M. Campos and C. J. Hawker, J. Am. Chem. Soc., 2011, 133, 10046–10049 CrossRef; M. Murai, S.-Y. Ku, N. D. Treat, M. J. Robb, M. L. Chabinyc and C. J. Hawker, Chem. Sci., 2014, 5, 3753–3760 RSC.
  20. B3LYP/6-31G** time-dependence density functional calculations were performed with the following program: Spartan'10, Wavefunction, Irvine, CA, 2010.
  21. A. J. Wurzer, T. Wilhelm, J. Piel and E. Riedle, Chem. Phys. Lett., 1999, 299, 296–302 CrossRef CAS; S. Y. Kim, G. Y. Lee, S. Y. Han and M. Lee, Chem. Phys. Lett., 2000, 318, 63–68 CrossRef.
  22. Low solubility of 11 and 12 toward the solvent hampered the voltammetry experiments.
  23. D. W. Leedy and D. L. Muck, J. Am. Chem. Soc., 1971, 93, 4264–4270 CrossRef CAS; C. Sánchez-Sánchez, E. Pérez-Inestrosa, R. García-Segura and R. Suau, Tetrahedron, 2002, 58, 7267–7274 CrossRef; A. Houmam and E. M. Hamed, Chem. Commun., 2012, 48, 11328–11330 RSC; G. Oksdath-Mansilla, A. A. Heredia, J. E. Argüello and A. B. Peñéñory, Photochem. Photobiol. Sci., 2015, 14, 726–736 RSC.
  24. S. Ito and N. Morita, Eur. J. Org. Chem., 2009, 4567–4579 CrossRef CAS; S. Ito, T. Shoji and N. Morita, Synlett, 2011, 2279–2298 CrossRef; T. Shoji and S. Ito, Chem. – Eur. J., 2017, 23, 16696–16709 CrossRef; T. Shoji, K. Miura, T. Araki, A. Maruyama, A. Ohta, R. Sekiguchi, S. Ito and T. Okujima, J. Org. Chem., 2018, 83, 6690–6705 CrossRef; T. Shoji and S. Ito, Adv. Heterocycl. Chem., 2018, 126, 1–54 CrossRef.

Footnote

Electronic supplementary information (ESI) available: Experimental details, 1H, 13C NMR, HRMS, UV/Vis spectra, cyclic voltammograms, and frontier Kohn–Sham orbitals of reported compounds. CCDC 1529064. For ESI and crystallographic data in CIF or other electronic format see DOI: 10.1039/c8qo01121d

This journal is © the Partner Organisations 2019