Perylenediimide[1]
[2]
[3]
[4] (PDI), an important member of the rylene family, is a promising scaffold owing to
the prominent features of PDIs such as their excellent optical, chemical, and thermal
robustness; high fluorescence quantum yield; high electron mobility; and high molar
absorptivity. Moreover, the optical, electronic, and aggregation properties of PDIs
can be modified by judicious selection of functional groups at the N-imide, bay, ortho, or peri positions of the PDI core.[2]
[3]
[4]
[5]
[6] Therefore, PDIs have been explored for various applications in the dye and pigment
industries[5] and in optoelectronic devices.[6]
[7]
[8]
[9]
[10] Recently, bay- and N-terminal-functionalized PDIs have been explored in supramolecular chemistry.[11]
[12]
[13]
[14]
[15]
[16]
[17] In this context, PDI-based chromophores with attached donor and acceptor substituents
(D–A hybrids) might be potential candidate for use as reaction-based probes for producing
long-wavelength-absorbing dyes[18] and in the field of organic photovoltaic devices.[19]
In literature reports, basically, either two (1,7-) or four (1,6,7,12-) positions
of the bay area of PDIs (A, Scheme [1]) have been extensively used for homogeneous functionalization through nucleophilic
substitution or Suzuki/Sonogashira coupling reactions of di- or tetrahalogenated derivatives
in the presence or absence of a catalyst, to attach various aryl or alkyl groups to
the PDI core (B and C; Scheme [1]).[1]
[2]
[3]
[4]
[5]
[6] PDIs with two different functional groups in the bay positions can be prepared by multiple stepwise substitution reactions of halogenated
PDIs, but these syntheses involve cumbersome purification procedures and give the
desired product in low to moderate yield (D; Scheme [1]).[20] Few reports have appeared that discusses halogenated or nonhalogenated routes for
the synthesis of dissymmetric bay-functionalized di- or trisubstituted PDIs (E; Scheme [1]). We therefore proposed that PDI-based D–A hybrids might be synthesized by dissymmetric
bay functionalization of the PDI core.[21]
Scheme 1 General approaches for the synthesis of dissymmetric functionalized PDIs reported
in the literature
In continuation of our interest in developing PDI- and pyrimidinone-based receptors
for applications in supramolecular chemistry,[22] we report the design, synthesis, and characterization of dissymmetric bay-functionalized di- and trisubstituted PDIs through controlled nitration of monosubstituted
PDIs (Figure [1]). Density functional theory (DFT) calculations, NMR spectroscopy, electrochemical
studies, and UV-vis spectroscopy were used to elucidate the electronic and optical
behaviors of PDIs 1–5.
Figure 1 Chemical structures of dissymmetric bay-functionalized PDIs 1–5
We synthesized PDIs 6, 7, and 9 (Scheme [2]) by following the procedures reported in the literature.[12]
[23] Ceric ammonium nitrate (CAN)-mediated nitration of PDI 6 was accomplished in presence of a H2SO4–HNO3 mixture to give PDI 8. We observed that stabilization of the nitrating species in H2SO4 during nitration of PDI 6 at room temperature drastically reduced the reaction time to 10–15 min (in comparison
to the two hours reported in the literature),[23] and gave PDI 8 in >90% yield.
Scheme 2 Synthesis of dissymmetric bay-substituted perylenediimides; Reagents and conditions: (a) propargyl alcohol, K2CO3, DMF, 9 h, 90 °C; (b) HNO3 (5.64 equiv), CAN (0.47 equiv), H2SO4 (2.47 equiv), r.t., 4 h; (c) HNO3 (16.9 equiv), CAN (1.0 equiv), H2SO4 (18.9 equiv), r.t., 10 h; (d) but-1-yn-1-ol, K2CO3, DMF, 9 h, 90 °C.
With these encouraging results, we extended our nitration method to the preparation
of dissymmetric bay-functionalized di- and trisubstituted PDIs. For this purpose PDI 9 was subjected to nitration at room temperature by using a similar protocol with a
nitrating mixture consisting of CAN (1.0 equiv), HNO3 (16.9 equiv), and H2SO4 (18.9 equiv).[24] Immediately, we observed the formation of a blue-colored product that moved faster
on TLC (Rf
= 0.4) with respect to PDI 9 (Rf
= 0.3) (Scheme [2]). In the 1H NMR spectrum of this blue-colored product, we observed the presence of two singlets
at δ = 9.14 and 8.87 ppm and one doublet at δ = 8.32 ppm, corresponding to 1 H each,
clearly separated out, together with one doublet and one singlet of 1 H each at δ
= 8.55–8.57 ppm that merged with one another [see Supplementary Information (SI);
Figures S3 and S4]. In the proton–carbon heteronuclear correlation (HSQC) spectrum,
we observed five hydrogen atoms of the perylene aromatic core corresponding to five
carbon signals in the aromatic region (SI; Figure S3f). These NMR data indicate that
PDI 9 undergoes nitration at two positions to form the dissymmetric trifunctionalized PDI
2
[25] (Scheme [2]).
Scheme 3 Synthesis of dissymmetric bay-substituted perylenediimide 5; Reagents and conditions: (a) (4-formylphenyl)boronic acid, Pd(PPh3)4, toluene, 12 h, 80 °C; (b) HNO3 (5.64 equiv), CAN (0.47 equiv), H2SO4 (2.47 equiv), r.t., 4 h.
We further surmised that controlled and careful nitration of PDI 9 through portionwise addition and equivalency of the nitrating mixture might result
in the formation of a disubstituted product with nitration at only one position. We
therefore attempted another reaction with a nitrating mixture consisting of CAN (0.47
equiv), H2SO4 (2.47 equiv), and HNO3 (5.64 equiv).[26] On monitoring the progress of the reaction by TLC, we observed a new purple-colored
spot at Rf
= 0.26, compared with PDI 2 (Rf
= 0.4) (Scheme [2]). We successfully purified this compound through column chromatography [silica gel,
CHCl3–hexane (7:3)] in 90% yield; PDI 2 was also isolated in <3% yield. The 1H NMR spectrum of the product showed two singlets of 1 H each at δ = 8.51 and 9.04
ppm, and two doublets of 1 H each at δ = 8.68 and 8.78 ppm (SI; Figures S4 and S5);
two doublets of 2 H protons each merged with one another at δ = 8.64 ppm. The presence
of six correlations in the HSQC, corresponding to six ArC–H signals (SI; Figure S5)
led us to conclude that the dissymmetric disubstituted product PDI 1
[27] had been formed (Scheme [2]).
The series of dissymmetric substituted PDIs was further expanded by the synthesis
of the butyne-linked derivatives PDI 3
[28] and PDI 4
[29] (Scheme [2], left). The PDI 10 was synthesized from PDI 7 through nucleophilic substitution with but-1-yn-1-ol in the presence of a K2CO3–DMF mixture (SI; Figure S6). In the 1H NMR spectrum, the presence of two triplets at δ = 4.63 ppm (due to –OCH2) and δ = 2.22 ppm (due to –CH), and of a triplet of doublets at δ = 3.00 ppm (due
to –CH2), along with the requisite signals of the PDI core indicated that PDI 10 had been formed. Subsequently, PDI 10 was subjected to nitration, under similar conditions to those described above, to
give PDIs 3 and 4. The purity and structures of PDIs 3 and 4 were established by NMR techniques (1H and 13C NMR spectra and HSQC) (SI; Figure S7 and S8).
To widen the scope the method and to expand the series of dissymmetric substituted
PDIs, PDI 5
[30] was synthesized (Scheme [3]). To synthesize PDI 5, we first performed a Suzuki coupling of PDI 7 with (4-formylphenyl)boronic acid in the presence of Pd(PPh3)4 in toluene to give PDI 11 (SI; Figure S9). Subsequently, PDI 11 was subjected to nitration with CAN (0.47 equiv), H2SO4 (2.47 equiv), and HNO3 (5.64 equiv) to give the disubstituted PDI 5 (Scheme [3] and SI; Figure S10).
Unfortunately, even after repeated attempts, we were unable to isolate any trisubstituted
PDI derivatives. The electron-withdrawing nature of the formylbenzene group in the
bay position of the PDI inhibits the formation of trisubstituted PDI derivatives containing
two nitro groups.
The purity and structure of PDI 5 were established by 1H and 13C NMR spectroscopy. Interestingly, oxidation of the –CHO group to a –COOH group was
not observed during nitration of PDI 11 (the resonance signal of the -CHO proton appeared as singlet at δ = 10.51 in PDI
5). PDI 5 might therefore be a potential candidate for various applications after performing
desired functionalizations at one or both of the –CHO and –NO2 groups.
To obtain insight into the effects of the presence of dissymmetric substituents in
the bay positions on the electronic properties of PDIs, the molecular structures of PDIs
1–5 were optimized at the B3LYP/6-31G* level by using density functional theory (DFT)
(Figures [2] and 3 and Table [1]). In PDIs 1–4, we expected that the presence of the propargyloxy or but-1-ynyloxy groups would
enrich the electron density, whereas the presence of the –NO2 groups would induce electron deficiency in the perylene core. From molecular orbital
analysis, we observed that the HOMOs are localized on the central PDI moiety and
on the donor oxygen atom of the –OR (R = propargyl or but-1-ynyl) moiety, whereas
the LUMO is extended uniformly from the PDI moiety to the nitro moieties (Figure [2]). The LUMO and HOMO levels of PDI 1 drop to –3.86 from –3.46 eV (PDI 9) and to –6.19 from –5.84 eV (PDI 9), respectively. As expected, the presence of two nitro groups in the bay positions in PDI 2 results in a further decrease in the HOMO level to –6.47 eV, whereas drastic decrease
in the LUMO level to –4.32 eV in comparison to PDI 9 occurs (Figure [3] and Table [1]). A similar trend was observed on moving from PDI 10 to PDI 3 to PDI 4.
Figure 2 The optimized structures of PDI 1–4 (a–d; left) calculated by DFT at the B3LYP/6-31G* level. The dihedral angles and
a representation of the computed frontier molecular orbitals (HOMOs and LUMOs) of
the PDIs 1–4 are shown (a–d; right).
Figure 3 General representation of the energy levels (HOMOs and LUMOs) for PDIs 1–5 calculated by DFT with the B3LYP/6-31G* basis set. Blue and red bars represent the
HOMO and LUMO levels (in eV), respectively.
Table 1Calculated and Experimental Parameters for PDIs 1–5
|
Compound
|
Calculated by DFT (B3LYP/6-31G*)
|
Calculated by cyclic voltammetry
|
Angles of twist (°)
|
λmax (nm)
|
|
HOMO (eV)
|
LUMO (eV)
|
E
g (eV)
|
HOMO (eV)
|
LUMO (eV)
|
E
g (eV)
|
|
|
|
PDI 9
|
–5.84
|
–3.46
|
2.30
|
–5.94
|
–3.55
|
2.39
|
1.0, 1.1
|
531
|
|
PDI 1
|
–6.19
|
–3.86
|
2.33
|
–6.16
|
–3.88
|
2.28
|
9.6, 17.0
|
541
|
|
PDI 2
|
–6.47
|
–4.32
|
2.14
|
–6.18
|
–3.96
|
2.22
|
14.5, 25.7
|
554
|
|
PDI 10
|
–5.79
|
–3.42
|
2.37
|
–5.90
|
–3.54
|
2.36
|
7.1, 4.4
|
536
|
|
PDI 3
|
–6.14
|
–3.82
|
2.32
|
–6.25
|
–3.83
|
2.42
|
11.3, 17.1
|
538
|
|
PDI 4
|
–6.42
|
–4.28
|
2.13
|
–6.03
|
–3.90
|
2.13
|
15.5, 25.8
|
556
|
|
PDI 5
|
–6.40
|
–4.09
|
2.31
|
–6.11
|
–3.99
|
2.12
|
16.0, 23.0
|
525
|
When the propargyloxy group was replaced with a formylbenzene group in PDI 5, the LUMO level increased to –4.09 eV in comparison to PDI 4, indicating minor destabilization. The energy gap (E
g) between the HOMO and LUMO levels of PDIs 2 and 4 are ~2.14 eV, which is lower than those of PDIs 9/1 and 10/3. Structure optimization of the ground-state molecular structures of the dissymmetric
bay-substituted PDIs 1–5 also revealed that the two naphthalene rings of the PDI moiety have different twist
angles to compensate for the strain caused by the presence of di-or trisubstitution
at the bay positions of the PDI moiety. In the trisubstituted derivatives 2 and 4, the approximate dihedral angles between the two naphthalene rings are 25.7°, whereas
for the disubstituted derivatives 1 and 3, the dihedral angles between the two naphthalene rings are about 17° (Figure [3]). Similarly, the dihedral angle for PDI 5 is 18.0° (SI; Figure S11). These results suggest that PDIs 1–5 have a nonplanar structure in comparison to PDIs 9 and 10 (SI; Figures S12–16).
The electrochemical response of dissymmetric functionalized PDI 1–5 were also investigated by cyclic voltammetry (CV) in dichloromethane (vs. Ag/AgCl)
(SI; Figure S17). The LUMO and HOMO energy levels of PDIs 1–5 were calculated from the cyclic voltammograms of the respective compounds by using
the equation E
LUMO= –[E
red(onset) + 4.4] eV and E
HOMO = –[E
ox(onset) + 4.4] eV;[31] the results are given in Table [1]. The LUMO and HOMO energy levels calculated by DFT agreed with those determined
experimentally by CV.
The absorption spectra of PDIs 1–5, 9, and 10 were recorded in MeCN (Figure [4]). Significant differences were observed in the absorption spectra of PDIs 1–5 compared with those of PDIs 9 and 10. Broadening, along with loss of vibronic structures of the absorption spectra, were
observed in cases of PDIs 1–5, and the absorption maxima of PDIs 2 and 4 appeared at 554 and 556 nm, showing bathochromic shift of 23 and 20 nm in comparison
to PDIs 9 and 10, respectively.
Figure 4 The UV–vis spectra of PDIs 1–5, 9, and 10 recorded in MeCN
We assumed that the broadening of the absorption spectra of the dissymmetric substituted
PDIs was due to either a large twist angle of the naphthalene rings of the PDI moiety,
induced by the presence of substituents, or to an increase in conjugation between
the donor substituent and the PDI moiety. However, on moving from PDI 9 to PDIs 1 and 2, a minor red shift was observed (Table [1]), so it is probable that twisting of the naphthalene rings in the PDI moiety is
responsible for the spectral broadening and the decrease in conjugation. Despite the
greater angles between the two naphthalene units of the perylenediimide moieties in
1–4, the bathochromic shift in the absorption spectrum might be attributable to internal
charge transfer.
In conclusion, we have synthesized and characterized the novel dissymmetric bay-substituted PDIs 1 – 5, which might find potential applications in several fields, for example, as reaction-based
near-IR colorimetric probes or in the development of long-wavelength-absorbing dyes.
