Introduction

π-Conjugated oligomers and polymers attracted considerable attention from the very start as a promising class of semiconductors, chemosensors, and various electronic devices . Although silicon and inorganic materials still play a major role in the development of modern electronics, the prospects for using organic electronic materials as an alternative are becoming increasingly clear. One of the advantages of those materials is a possibility to fine-tune useful properties by simply varying of the π-conjugated backbone and side-chain substituents .

π-Conjugated oligomers consisting of alternating C≡C bonds and aromatic nuclei, commonly, have a rigid, rod-like structure and exhibit high charge carriers’ mobility . 1,4-Diaryl-1,3-butadiynes is a particular class of such compounds. Both theoretical and experimental studies revealed that the side groups of 1,4-diaryl-1,3-butadiynes have a significant impact on their useful characteristics . For example, single-molecule conductivity, nonlinear optical properties, and the ability to serve as photosensitizers of singlet oxygen production have been identified in porphyrin-based butadiynes , 1,3-butadiyne-linked oligoporphycenes , and 1,3-butadiyne-linked amines . A wide variety of applications was proposed for graphdiynes (2D allotropes of graphene), including electrocatalysts and energy devices, which exploit the carbon-rich nature, porous framework, and expanded π-electron system of these compounds . And this is not a complete list.

Recently, we reported on the synthesis of 1,4-diaryl-1,3-butadiynes 1–4 based on the “proton sponge” [1,8-bis(dimethylamino)naphthalene, DMAN] (Figure 1) . In the present work we describe the synthesis of a new family of proton sponge-based butadiynes 5 bearing arylethynyl substituents of different electronic nature. Oligomers 5 having electron-withdrawing groups on the aryl termini are interesting as push–pull A–π–D–π–D–π–A systems, whereas the counterpart with an electron-donating methoxy group can be converted into a D–π–A–π–A–π–D system by protonation of the proton sponge fragments (Figure 2). Moreover, oligomers 5 are cross-conjugated π-systems. “A cross-conjugated compound may be defined as a compound possessing three unsaturated groups, two of which although conjugated to a third unsaturated center are not conjugated to each other” . It is easy to see that there are two π-conjugation paths in molecules 5: a donor–acceptor conjugation path (Figure 2, highlighted in blue) and the π-conjugation of naphthalene rings through a butadiyne linker (highlighted in green). In comparison to linearly conjugated materials, oligomeric and polymeric compounds with a fully cross-conjugated carbon backbone are relatively unexplored . Molecules of this type serve not only as objects of fundamental research into the phenomena of cross-conjugation, electron transfer, and quantum interference , but are also considered as promising molecular switches and transistors , NLO materials , and suitable starting compounds for syntheses involving multiple Diels–Alder additions . All these facts motivated us to undertake the current study. X-ray crystallography, UV–vis spectroscopy and cyclic voltammetry were applied to analyze the extent of π-electron conjugation and the efficiency of the particular donor–acceptor conjugation path in chromophores 5.

Figure 1: Proton sponge-based 1,4-diaryl-1,3-butadiynes synthesized previously and in this study.

Figure 2: Target oligomers as push–pull and cross-conjugated π-systems.

Results and Discussion Synthesis

The target oligomers 5 can be synthesized by a Glaser oxidative dimerization of monomers 6 (Scheme 1). The obvious route for the synthesis of the latter is the sequential alkynylation of 2,7-diiodonaphthalene 8.

Scheme 1: Synthetic strategy for target oligomers 5.

In accordance with this strategy, diiodide 8 was cross-coupled with copper(I) arylacetylides (Castro–Stephens reaction, method A) and arylacetylenes (Sonogashira reaction, method B). In all cases, even when using a small excess of 8, in addition to the desired monoalkynyl derivative 7, a double alkynylation product 9 was formed (Table 1). The Sonogashira coupling was somewhat more efficient, yielding compounds 7a–e in 42–62% yields, but also gave higher amounts of products 9a–e (10–30%). Thus, the Pd- and phosphine-free Castro–Stephens coupling was a good enough alternative to synthesize alkynes 7. The structure of the double alkynylation product 9e was confirmed by X-ray diffraction data (see Supporting Information, File 1, Figure S60).

Table 1: Synthesis of 7-(arylethynyl)-2-iodo-DMAN 7.

Ar   Yield, % method А method B 7 9 7 9 Ph a 52 18 62 30 4-MeO-C6H4 b 41 2.5 51 21 4-CF3-C6H4 c 45 11 53 22 4-CN-C6H4 d 24 9 49 26 4-NO2-C6H4 e 30 5 42 10

The further alkynylation of compounds 7a–e was carried out using trimethylsilylacetylene and the Pd(PPh3)2Cl2/CuI/Et3N/DMSO catalytic system giving rise to dialkynyl derivatives 10a–e in high yields (Scheme 2). Column chromatography of trimethylsilyl derivatives 10a–e on Al2O3 resulted in their quantitative desilylation with the formation of the target monomers 6a–e, thus eliminating the need to remove the trimethylsilyl protection. Pure samples of compounds 10a–e could be obtained by extraction of the reaction mixture with hexane followed by recrystallization of the crude product from ethanol.

Scheme 2: Synthesis of 7-(arylethynyl)-2-ethynyl-DMAN 6.

Next, the oxidative dimerization of terminal alkynes 6a–e was carried out in an aerobic medium in the CuI/TMEDA/iPr2NH system at room temperature, which proved to be effective in the synthesis of butadiynes 1–4 (Scheme 3). The desired diarylbutadiynes 5a–e were obtained in good yields regardless of the substituent R in the benzene ring. Treatment of the latter with fluoroboric acid in dichloromethane gave double salts 11a–e.

Scheme 3: Synthesis of 1,4-diaryl-1,3-butadiynes 5 and their salts 11.

X-ray structures

Slow evaporation of solutions of butadiynes 5 in the CHCl3/EtOAc system made it possible to grow single crystals of samples 5b, 5d, and 5e suitable for X-ray diffraction studies (Figure 3 and Figure 4). Crystals of compound 5c were grown up using CHCl3/EtOH solvent, and it was unexpectedly found that keeping this compound in the above system for a month leads to its partial heterocyclization to benzo[g]indole 12 (Scheme 4 and Figure S59 in Supporting Information File 1). The structure of compound 12 was unambiguously established by X-ray diffraction analysis (see Supporting Information File 1, Figure S61). We assumed that this transformation is facilitated by hydrogen chloride, which is formed during the oxidation of chloroform with atmospheric oxygen. We also succeeded in growing crystals of the salt 11c in the MeCN/EtOH system (Figure 5). Unfortunately, good crystals of other salts have not been obtained.

Figure 3: Molecular structures of compounds 5b (top), 5d (middle), and 5e (bottom).

Figure 4: Views on the molecular backbone of compounds 5b (top), 5d (middle), and 5e (bottom) along the naphthalene rings plane (hydrogen atoms omitted).

Scheme 4: Transformation of butadiyne 5c into benzo[g]indole 12.

Figure 5: Molecular structure of compound 11c: frontal (top; BF4− omitted) and side views (bottom; hydrogen atoms omitted).

From Figure 3 and Figure 4 it is easy to see that all molecules 5b, 5d, and 5e are rather distorted, including naphthalene cores, butadiyne and acetylene linkers. The main structural parameters of diynes 5 that characterize the degree of this distortion are presented in Table 2, where ϕ1 is the angle between the planes of the benzene ring and the neighboring naphthalene system, ϕ2 is the angle between the averaged planes of the naphthalene rings, ∠Cx‒Cy–Cz is the bond angle of the carbon–carbon bonds in the butadiyne linker, Θ is the C2(2′)–C3(3′)–C6(6′)–C7(7′) torsion, N···N is the internitrogen distance in the DMAN fragments, Σ∠N is the sum of the C–N–C angles of the NMe2 groups, and φ is the N1(1′)–C1(1′)–C8(8′)–N8(8′) torsion. For salt 11c, two additional parameters characterizing the hydrogen N–H···N bond are given, e.g., the N–H bond lengths and the angle between them (∠N–H···N).

Table 2: Some structural parameters of oligomers 5 and salt 11c (X-ray data).

Parameter   5b (R = OMe) 5d (R = CN) 5e (R = NO2) 11ca     A B A B A = B A = B ϕ2 torsion, °   17.4 34.4 0 0 0 ϕ1 torsion, °   1.4 12.7 28.2 29.0 17.9 32.4 12.8 ∠C2‒Cα‒Сβ, °
∠Cα‒Cβ‒Сγ, °
∠Cβ‒Cγ‒Сδ, °
∠Cγ‒Cδ‒С2′, °   176.9
172.9
176.0
170.2 177.5
177.7
177.9
176.7 172.2
177.6 173.5
178.2 176.1
177.7 ∠C7(7')‒Ca(a')‒Сb(b'), °
∠Ca(a')‒Cb(b')‒Сc(c'), °   173.2
178.3 173.3
179.1 174.1
178.5 174.5
176.9 174.5
172.2 168.5
174.2 175.0
176.2 C2(2′)–C3(3′)–C6(6′)–C7(7′) torsion Θ, °   19.6 15.3 12.8 9.5 18.6 0.8 0.0 N···N distance, Å   2.859 2.808 2.779 2.772 2.848 2.553 2.570 sum of the C–N–C angles Σ∠N, ° N1(1′) 357.3 357.6 354.4 354.9 357.6 340.0 340.5 N8(8′) 359.3 357.0 357.3 353.5 358.6 341.0 340.9 N1(1′)–C1(1′)–C8(8′)–N8(8′) torsion φ, °   29.0 21.7 20.9 8.4 31.6 2.7 2.5 N–H···N bond lengths, Å   – – – – – 1.03
1.54 1.01
1.59 ∠N–H···N, °   – – – – – 166 162

aStructural parameters of two independent molecules are given.

In all cases naphthalene fragments linked by a 1,3-butadiyne axis take a trans position relative to each other. Despite formally symmetrical structure of diynes 5, naphthalene rings A and B of molecules 5b and 5d differ in their structural parameters. At the same time, the monomer fragments of the nitro derivative 5e are identical. In the case of 5e, the naphthalene rings lie in parallel planes, while in crystals of 5b and 5d the angle between the average planes of naphthalene nuclei A and B reaches 17° and 34°, respectively. Molecule 5d demonstrates the largest rotation angles of the aryl termini with respect to the naphthalene rings (ϕ1 = 28–29°). The dimethylamino groups of 5 are strongly flattened (the Σ∠N value varies from 353.5° to 359.3°), which is characteristic of ortho-substituted proton sponges . The observed N···N distances are slightly larger than those typical for ortho-disubstituted DMAN derivatives .

Molecule 5b is the most distorted, as evidenced by the significant twisting of naphthalene rings (ΘA = 19.6° and ΘB = 15.3°), the largest internitrogen distance (2.859 and 2.808 Å for rings A and B, respectively), the largest deviation of the dimethylamino groups from the naphthalene ring plane (φA = 29.0°, φB = 21.7°), as well as deviation of the bond angles of the butadiyne linker from the standard value of 180° by 3–10°. Deviations of bond angles in acetylene bridges are ≈1–7°. The methoxy derivative 5b has the most complex crystal packing with a large number of different nonvalent interactions (see Supporting Information File 1, Figures S62 and S63).

The molecule of cyano derivative 5d is characterized by the least distortion of the DMAN fragments in the series (twisting Θ = 9.45 and 12.83°, torsions φA = 20.9° and φB = 8.4°, bond angle deviations in both butadiyne and acetylene linkers do not exceed 6°). In the crystal packing of 5d (see Supporting Information File 1, Figures S64 and S65), the DMAN fragments do not participate in nonvalent interactions and do not form short contacts. The recurring motif in the crystals is the coordination of the benzene meta proton by the nitrogen atom of the C≡N group.

The structural parameters of both monomer fragments of nitro derivative 5e are identical. This fact, together with the parallelism of the naphthalene ring planes, indicates the existence of an inversion center in the molecule. Molecule 5e is characterized by the largest N1(1′)–C1(1′)–C8(8′)–N8(8′) torsion angle φ (31.6°) in the series. Other structural parameters are close to those of the molecule 5b. In the crystal packing (Figure S66 in Supporting Information File 1), molecules 5e tend to approach π-donor DMAN and π-acceptor p-nitrophenyl fragments, and the shortest distance between the two molecules is 2.810 Å (Figure S67 in Supporting Information File 1).

The alternation of the C–C bond lengths in the aryl rings of molecules 5d and 5e may indirectly indicate the conjugation of the π-donor fragment with the π-acceptor p-nitrophenyl or p-cyanophenyl fragments. The qr parameter, calculated according to equation (Figure 6) and characterizing the quinoid character of the aryl ring, was proposed for D–π–A systems. This parameter is a good indication for intramolecular charge transfer from the donor to the acceptor moiety in the ground state. In benzene, the qr value is equal to 0. In a fully quinoid ring, the qr was found to be equal to 0.10–0.12.

Figure 6: Calculation of the qr parameter.

Calculations based on the bond lengths in the aryl fragments determined by X-ray diffraction analysis gave the following average values of the qr parameter: 0.012 for 5d and 0.014 for 5e. For comparison, the same parameter calculated for N,N-dimethyl-4-nitroaniline is 0.038 (X-ray data from reference ). Therefore, the π-charge transfer from the donor DMAN to the acceptor aryl ring of 5d and 5e is extremely modest in the ground state. It should be also noted that the CNaph–N bonds of 5e are the shortest in the series (1.379–1.380 Å), which may also indirectly indicate a more pronounced conjugation of the dimethylamino groups with the nitro group.

There are two types of independent non-equivalent dications, marked in blue and green, and two types of BF4− anions, marked in red and yellow, in the crystal structure of salt 11c (Figure S68 in Supporting Information File 1). Monomer fragments in both are identical (Table 2, Figure 5). The trifluoromethyl group of one independent molecule is disordered with an occupancy of fluorine atoms of 0.54/0.46, which makes the molecule asymmetric. The second independent molecule has an inversion center. Compared to the free bases 5 discussed above, the protonated form 11с demonstrates almost complete planarization of the naphthalene backbone due to the disappearance of steric and electrostatic stress between the NMe2 groups. The nitrogen atoms in the DMAN fragments strongly approach during protonation and practically do not deviate from the average naphthalene ring plane. The dimethylamino groups naturally become more pyramidal and the CNaph–N bonds lengthen. All these changes are typical for protonated DMAN derivatives . Interestingly, the NH protons are localized on the 1-NMe2 and 1′-NMe2 groups adjacent to the butadiyne linker and do not move away from each other at the maximum distance closer to the aryl substituents. Noteworthy is the bending of the acetylene linker (in one of the independent molecules, the bond angle is only 168.5°) and the greater linearity of the butadiyne fragment, which may indicate a more pronounced conjugation between two DMAN fragments than between DMAN and p-trifluoromethylphenyl rings. By the way, the qr parameters calculated for two independent molecules of 11c were 0.008 and 0.009, which is slightly less than in the case of compounds 5. As for the crystal packing of 11c, BF4− anions of two types (“red” and “yellow”) interact with cations in different ways. The “red” anion hangs over the cationic centers of both independent molecules, which are almost perpendicular to each other, while the “yellow” one participates mainly in the coordination with the hydrogen atoms of the NMe2 groups. Such a distribution of counterions apparently ensures mutually perpendicular packing of almost linear molecules (Figure S69, Supporting Information File 1).

UV–vis spectra and redox properties

As stated above, oligomers 5 are cross-conjugated π-systems. For cross-conjugated structures, the main question is about the preferential conjugation path. For oligomers 5, two different directions of electron density transfer are possible (Figure 7): between two DMAN fragments through the butadiyne linker (highlighted in green) and between the DMAN and aryl rings through the acetylene bridge (highlighted in blue). Obviously, the “butadiyne path” includes a longer conjugation chain. Noncovalent interactions of molecules in crystals and packing effects do not allow one to strictly judge the charge transfer in the oligomers 5. Therefore, we analyzed their UV–vis spectra (Table 3, Figure 8). The functional groups R are located at the far ends of the oligomeric chain and, from the steric point of view, cannot have a significant effect on conformational transformations of 5. All differences in optical properties must be of an electronic nature. It is obvious that the same “butadiyne conjugation pathway” (marked in green) is realized in compound 1, while the conjugation chain in monomers 6 is identical to the “blue” one in oligomers 5 (Figure 7). Thus, the UV–vis spectra of compounds 1 and 6 were used for comparison (Table 3, Figure 8).

Figure 7: Two π-conjugation ways in oligomers 5.

Table 3: Summary of the UV–vis spectraa of monomers 6, oligomers 5 (in CHCl3), and salts 11 (in MeCN).

R   Monomer 6 Oligomer 5 Salt 11 λmax, nm (lg ε) λmax, nm (lg ε) λonset, nm Egopt, eVb λmax, nm (lg ε) λonset, nm Egopt, eVb H a 393 (3.98) 432 (4.51) 519 2.39 382 sh (4.28) 397 3.12 OMe b 405 sh (3.92) 423 (4.54) 518 2.39 387 sh (4.31) 415 2.99 CF3 c 400 (4.08) 437 (4.64) 524 2.37 382 sh (4.30) 390 3.18 CN d 441 sh (4.08) 449 (4.54) 547 2.27 383 sh (4.24) 392 3.16 NO2 e 456 sh (3.92) 453 (4.58) 594 2.09 385 sh (4.19) 396 3.13

aAbsorption maxima measured in the corresponding solutions at c = 10−5 M. bThe optical gap estimated from the onset point of the absorption spectra: Egopt = 1240/λonset.