Open Access Article
Steve O.
Sequeira
a,
Roberto M.
Diaz-Rodriguez
a,
Mmasinachi
Atansi
a,
James W.
Hilborn
a,
Adil
Alkaş
a,
Rosinah Liandrah
Gapare
a,
Madeleine
Dearden
a,
Em C.
Sullivan
a,
Emily B.
Brown
a,
Katherine N.
Robertson
b and
Alison
Thompson
*a
aDepartment of Chemistry, Dalhousie University, P.O. Box 15000, Halifax, Nova Scotia B3H 4R2, Canada. E-mail: alison.thompson@dal.ca
bDepartment of Chemistry, Saint Mary's University, Halifax, Nova Scotia B3H 3C3, Canada
First published on 14th August 2025
The chemistry of the azobispyrrole framework, which melds the impressive electronic capabilities of the azo (–N
N–) moiety and the pyrrole heterocycle, is significantly under-developed. Herein, the synthesis and characterisation of a series of azobispyrroles substituted with four aryl groups are presented. The steric bulk of aryl groups at the β-positions of the pyrrolic building blocks controls the degree to which azobispyrrole formation competes with aza-dipyrrin formation when using nitrobutanones as starting materials. The ability of aryl groups at the α-positions of the pyrrolic building blocks to influence conformational stability is discussed, as is the consequent control of photophysical properties.
N–R′ were first reported by Alfred Noble in 1856 and remain of considerable research interest.1 Notwithstanding significant ecotoxicity2 concerns that continue to emerge, azo-compounds are prominent as colorants in textiles3 and food,4 and their tunable light-responsive properties have led to applications across diverse areas spanning drug delivery5 to photonics.6 Recent advances include azo-based photoswitchable ligands for health research applications7 and solid-state solar thermal fuels and storage solutions.8 Although azobenzenes have dominated the field, compounds featuring at least one heterocycle attached to the azo functionality have attracted interest courtesy of the widely tunable electronic features and varied functionality that the presence of heteroatoms imparts.9 Strikingly, azobispyrroles are largely absent from this domain.
More than fifty years ago,10 the reaction of pyrrolylmagnesium bromide with tosyl azide was reported to provide the parent azobispyrrole as an orange solid. Twenty-five years passed before this material reappeared in the literature, when it was used as a monomeric starting material to enable exploration of the electrochemical properties of polyazopyrroles.11,12 Since then, these symmetrical azo compounds have featured only as a curiosity explored via ab initio studies predicting electronic properties for five-membered heteroaromatic moieties connected via double-bonds.13,14 Beyond the parent compound reported in 1971, characterised azobispyrroles remain absent aside from an elaborate and unstable Ru complex that features the azo moiety attached to two pyrroles, each of which are heavily substituted with protected alkynyl units.15 Accordingly, we initiated a project16 to explore the potential for the synthesis of azobispyrroles
Anticipating that our target azobispyrrole was likely to be highly coloured, we hypothesised that any such materials formed would be distinguishable, via their optical features, from polypyrroles or dipyrroles such as aza-dipyrrins when assessing reaction mixtures using chromatography.23,24 Consequently, without delay and yet cognisant that 2 is susceptible to multiple reaction and decomposition pathways, we exposed 2 to the classic conditions used to convert aryl amines into their corresponding diazonium salts (Scheme 1).25 Each step was conducted under an inert atmosphere, and using degassed reagents. Intent on avoiding isolation of the potentially highly reactive dry diazonium salt 3, the crude reaction mixture was added directly to a coupling solution containing 2,4-diphenylpyrrole (4). Work-up using NH3 (aq), following a known procedure for azobenzene synthesis,25 did not result in a material with properties suggestive of the desired azobispyrrole 5. Significant efforts effecting variations in work-up procedures proved equally fruitless. We further noted that after treating 2 with NaNO2/HCl, the addition of the crude reaction mixture to a coupling solution containing aniline did not generate any azo compound. However, phenylazopyrrole 7 was obtained upon reversing the roles of the coupling partners (Scheme 1, bottom). This indicates that although the formation and/or productive reaction of the pyrrolediazonium salt 3 is elusive, the reaction of the α-free pyrrole with an aryldiazonium salt provides azo compounds.26
Given that azo coupling via a pyrrolediazonium salt was unsuccessful, our attention turned to other synthetic strategies by which azobispyrroles might be isolated. Alternative methods such as oxidative coupling27,28 or reactions of diazonium salts with diarylzinc reagents29 were rendered inaccessible due to challenges with the unstable aminopyrrole 2, despite our best efforts to avoid self-coupling, aza-dipyrrin formation and decomposition. Similarly, exploration of a metal-catalysed C–N bond coupling approach involving the reaction of pyrroleboronic acid, phenylpyrroleiodonium compounds and phthalic hydrazide30,31 was stymied by our unsuccessful attempts to generate arylpyrroleiodonium triflate salts using methods adapted from work involving phenyl- and mesityl-substituted linear diaryliodonium compounds.32,33
Our ongoing work with aza-BODIPYs includes exploring the photophysical tunability that is achievable by incorporating aryl groups at the boron atom of these dipyrrolic frameworks.34 The ability of substituents at the 3,5- and the 1,7-positions of F-aza-BODIPYs to influence the efficiency of emission originates from the extent to which rotational freedom, and thus nonradiative decay, is restricted as a consequence of either steric bulk or intramolecular interactions (Fig. 1, left).35 Although we prepared several arylboryl C-aza-BODIPYs (Fig. 1, right), all suffered from negligible fluorescence emission (Φf < 0.01). This contrasts with the analogous F-aza-BODIPY frameworks, wherein B–F⋯H hydrogen bonds involving the –BF2 fragment and the ortho-H atoms of 3- and 5-phenyl groups serve to restrict molecular motion, and consequently, nonradiative decay.35
Hoping to recover aza-dipyrrin fluorescence by locking the rotation of the 1,7-substituents, we prepared the nitrobutanone 8a, bearing a bulky mesityl substituent, and reacted it with NH4OAc (Fig. 2).35,36 Reactions of this nature are extensively used for the synthesis of aza-dipyrrins, wherein a nitrogen atom occupies the meso-position of the dipyrrolic skeleton. A comprehensive study23 regarding the mechanistic route(s) towards aza-dipyrrins concluded that butanones (e.g.9, Fig. 2) likely form aza-dipyrrins (correspondingly, 10) via reaction of the corresponding α-free pyrrole D with the related nitrosopyrrole E, both formed in situ, although possibilities for other pathways were also presented. However, an unexpected red spot, appearing on a TLC plate used to monitor the progress of the reaction between nitrobutanone 8a with NH4OAc in ethanol at reflux temperature (Fig. 2), caught our attention.37 Surprisingly, despite employing established conditions, none of the anticipated aza-dipyrrin 11a was isolated. It has been noted that strategies to synthesise aza-dipyrrins bearing bulky β-substituents, from appropriately-substituted nitrobutanones, only proceed with low yields, and so the use of glacial acetic acid as an alternative solvent was suggested.35 Accordingly, reaction of 8a with 35 equiv. NH4OAc in glacial acetic acid enabled isolation of the purple aza-dipyrrin 11a in low yield. In addition, a significant amount of a free-flowing red solid was isolated, and the corresponding characterisation data identified it as the azobispyrrole 13a.16 Despite extensive variation of reaction conditions we were unable to augment the yield of 13a beyond 30%.16
Isolation of the tetra-aryl azobispyrrole 13a from the reaction mixture prompted an analysis based on mechanistic considerations, cognisant that the formation of aza-dipyrrins from 4-nitro-1,3-diarylbutanones is poorly understood. Insight regarding possible pathways concluded that various routes and rearrangements operate in parallel.23 Nevertheless, it appears reasonable that a 2,4-diarylpyrrole is formed in situ, and that nitrosation thereof ensues. Reaction of the 2,4-diarylpyrrole with its nitrosated analogue presumably follows, giving rise to aza-dipyrrins.21 It appears likely that the presence of bulky mesityl groups in 2-phenyl-4-mesitylpyrrole (D, Ar = Mes, formed from 8a) and the nitrosated analogue E (Ar = Mes) disfavours condensation between them, rendering alternative outcomes competitive. It also appears reasonable that the corresponding 2-aminopyrrole F is formed in situ,23 and that its reaction with 2-nitroso-5-phenyl-4-mesitylpyrrole (E) would, after oxidation, yield the azobispyrrole 13a. Reaction of E and F would enable the two mesityl groups to remain more distal to each other than in an aza-dipyrrin. Thus, formation of the azobispyrrole framework likely becomes more competitive as the size of the substituents at the β-position of the constituent pyrroles increases. Nevertheless, the reaction mixture is complex, and indeed both 13a and the aza-dipyrrin 11a were isolated therefrom.
With the first tetra-aryl azobispyrrole 13a
16 in hand, complementing the few reported phenylazopyrroles,9,26,38 focus moved to assessing the scope for the isolation of systems bearing alternative aryl groups. We were particularly interested in exploring the requirements for steric bulk of the two aryl substituents in analogues of the nitroso- and amino-intermediates E and F, respectively. A series of chalcones (14) were thus prepared via condensation of the appropriate acetophenone and benzaldehyde, as shown in Table 1, and providing a number of novel compounds. These chalcones were then reacted with nitromethane under Michael addition conditions to provide the corresponding nitrobutanones 8b–n. Reaction of 8b–n with NH4OAc was then explored. Given the complexity of the product mixtures, we were gratified to isolate the first series of azobispyrroles, comprising seven additional members (13b–i, Fig. 3). Intriguingly, aza-dipyrrin 10 was the only product isolated from the reaction of the diphenylated nitrobutanone 8n, with no trace of the corresponding tetraphenylazobispyrrole apparent on TLC or during work-up. As such, inclusion of a mesityl group or other similarly bulky group at the β-position of the pyrrolic units seems key to promoting azobispyrrole formation in reactions of intermediates of type E with those of type F (Fig. 2). Indeed, reaction of 8i, to locate a 4-bromo-2-methylphenyl substituent at the β-positions of an azobispyrrole, provided only trace product.
| Entry | R1 | R2 | Yield 14 (%) | Yield 8 (%) |
|---|---|---|---|---|
| a | Mes | Ph | 90 | 80 |
| b | (2,6-diCl)C6H4 | Ph | 87 | 80 |
| c | Mes | (2,5-Dimethyl)C6H4 | 78 | 68 |
| d | Mes | (4-Chloro)C6H5 | 94 | 63 |
| e | (2,6-Dichloro)C6H4 | (4-Chloro)C6H5 | 66 | 89 |
| f | Mes | 1-Naphthyl | 98 | 77 |
| g | Mes | 2-Naphthyl | 71 | 80 |
| h | Mes | 2-Anthracenyl | 67 | 56 |
| i | (4-Bromo-2-methyl)C6H4 | (4-Bromo)C6H5 | 86 | 75 |
| j | Mes | 2-Thienyl | 67 | 89 |
| k | (2,6-Dichloro)C6H4 | 2-Thienyl | 70 | Quan. |
| l | Mes | (4-Methoxy)C6H4 | 41 | 90 |
| m | (2,6-Dichloro)C6H4 | (4-Methoxy)C6H4 | Quan. | Quan. |
| n | Ph | Ph | — | 79 |
In contrast, the substituent at the α-position of the pyrrolic units could easily be varied using this synthetic sequence. Phenyl rings bearing alkyl (13c) and chloro substituents (13d–e) could be incorporated at the α-position, as could fused aryl groups such as 1- and 2-naphthyl and 2-anthracenyl groups (13f–h). In contrast, reactions of nitrobutanones (8j-m) bearing electron-rich substituents failed to provide azobispyrroles, perhaps as a consequence of the anticipated enhanced reactivity of the corresponding amino-pyrroles (F) bearing α-thienyl and α-anisyl substituents.
Photochemical characterisation of azobispyrroles 13a–h (Table 2) was then undertaken. In each case, solvent-dependent absorption properties were apparent. With the exception of 13h, the azobispyrroles were soluble in DMF, MeCN, chloroform, CH2Cl2 and acetone, and less soluble in MeOH. Although emission was in all cases negligibly weak, significant Stokes shifts were observed, likely attributable to aggregation or other such structural changes between ground and excited states.16,39–41 Indeed, more concentrated solutions in MeCN were seen to produce a fine precipitate upon standing.16 We observed no evidence of E–Z switching upon irradiation of these materials, which is unsurprising given the steric encumbrances that would be experienced between mesityl groups in a Z configuration.
| Entry | Solvent | UV-Vis Absorption Maxima (λmax abs nm) | Molar Absorptivity ε (mol L−1 cm−1) × 104 | Fluorescence emission (λmax, nm) | Stokes Shifta (nm) |
|---|---|---|---|---|---|
| a Negligible; too small for quantum yield (Φf) to be accurately calculated. | |||||
| 13a | MeCN | 504 | 6.3 | 602 | 98 |
| 522 | 6.3 | ||||
| DMF | 504 | 5.1 | 583 | 79 | |
| 532 | 5.4 | 606 | |||
| 13b | MeCN | 500 | 2.5 | 601 | 101 |
| 520 | 2.3 | ||||
| DMF | 504 | 2.1 | 604 | 100 | |
| 530 | 2.2 | ||||
| 13c | MeCN | — | — | — | — |
| DMF | 496 | 6.2 | 588 | 92 | |
| 13d | MeCN | 506 | 7.1 | 607 | 101 |
| 528 | 7.0 | ||||
| DMF | 508 | 5.5 | 610 | 102 | |
| 538 | 5.6 | ||||
| 13e | MeCN | 500 | 6.5 | 606 | 106 |
| 520 | 6.4 | ||||
| DMF | 506 | 7.1 | 607 | 101 | |
| 534 | 7.2 | ||||
| 13f | MeCN | 496 | 2.5 | 604 | 108 |
| DMF | 498 | 2.1 | 603 | 105 | |
| 13g | MeCN | 520 | 2.0 | 604 | 84 |
| 544 | 1.9 | 629 | |||
| DMF | 522 | 8.3 | 605 | 83 | |
| 554 | 8.2 | 634 | |||
| 13h | MeCN | 546 | 1.5 | 628 | 82 |
| 576 | 1.5 | 662 | 86 | ||
| DMF | 550 | 0.7 | 622 | 72 | |
| 584 | 0.7 | 668 | 84 | ||
Intriguingly, the absorption spectra of most of the azobispyrroles in this series featured two distinct maxima (Fig. 4). A solution of 13a in MeCN was incubated for 1 hour at each of 0 °C and 40 °C and analysed at each temperature, and the dual maxima persisted. However, compounds 13c and 13f, whereby the flanking α-aryl substituent imparts greater proximal steric bulk by virtue of the ortho methyl substituent (13c) and the position at which the aryl is connected to the pyrrole (13f), exhibited a single absorbance maximum (Fig. 4). This behaviour is mirrored in the emission profile, notwithstanding the likely contributions of aggregation.16
As shown in Fig. 5, azobispyrroles 13a, 13g and 13h constitute a series with increasingly extended conjugation, courtesy of the arene attached to the α-position of each pyrrole unit via (formal) sequential fusion to additional benzene units. As the size of the π-system increases, a significant red-shift of absorbance is observed (13aλmax,abs 504 nm, 13gλmax,abs 520 nm, 13hλmax,abs 546 nm; all in MeCN). However, compound 13f, whereby the α-position of the pyrroles feature 1-naphthyl groups instead of the 2-naphthyl groups of 13g, retains the absorption characteristics of the phenyl-substituted 13a. The solid-state structures of 13a, 13b, 13f and 13g are reported herein (see SI for full details). Via single-crystal X-ray diffraction, we observe that the arene at the α-position of the pyrroles in 13g (and 13a) makes an angle of ∼20° with the azobispyrrole core (Fig. 6). Meanwhile, the 1-naphthyl substituents of 13f (Fig. 7) deviate distinctly from coplanarity with the pyrrole–N
N–pyrrole core, with an angle of ∼40° between the planes (Fig. 7). Presumably then, the red-shifted absorption characteristics in solution arise from conformations where the terminal aryl group lies closer to coplanarity with the azobispyrrole core, maximising conjugation across the entire chromophore. Furthermore, two polymorphs of 13a were obtained: both exhibit comparable characteristics as regards the extent of coplanarity of the flanking phenyl groups and the central core, despite their different crystalline arrangements and solvation states (see SI for full details). Coplanarity in these systems is evidently compromised by steric factors and enhanced by hydrogen bonding (see SI for full details). The inclusion of solvent (whether water, methanol or chloroform) and the hydrogen bonding it introduces (both as donor and acceptor) must offer more stabilization in the solid-state, promoting non-coplanarity of the terminal aryl groups with the central pyrrole–N
N–pyrrole core, than does the conjugation afforded by the peripheral aromatic groups (whether phenyl or naphthyl) lying in the same plane as the central core. Nevertheless, the different skeletal attachment points of the naphthyl groups in 13f and 13g to the pyrrole units clearly affects both the photophysical properties in solution, and the solid-state structural features.
Synthesis and characterisation data. See DOI: https://doi.org/10.1039/d5ob01104c
CCDC 2449926 (13b), 2449927 (13a), 2442944 (13g) and 2442945 (13f) contain the supplementary crystallographic data for this paper.42a–d
| This journal is © The Royal Society of Chemistry 2025 |