Influence of the substitution pattern on exciton localisation in centrosymmetric quadrupolar dyes

Abstract

Localisation of the electronic excitation via excited-state symmetry breaking (ESSB) is a characteristic property of many quadrupolar dyes in polar environments, and was shown to depend on the nature of the electron donor (D) and acceptor (A) subunits and their separation distance. Here, we compare the excited-state properties of two centrosymmetic D–π–A–π–D dyes with a dipyrrolonaphthyridinedione (DPND) acceptor and N,N-dimethylaniline donors, which only differ by the position of the –π–D arms on the DPND core. Time-resolved IR spectroscopy reveals that ESSB and hence exciton trapping on single A–π–D branch is facilitated with the donor arms in 1 and 7 positions on the DPND core compared to 3 and 9 positions, given that it occurs in a medium polar solvent for the former and only in highly polar media for the latter. This a priori unexpected difference is explained by the significantly larger dipole moment generated upon ESSB for the 1,7 isomer, as reflected by its much larger fluorescence solvatochromism compared to the other isomer. In this respect, the DPND core with its C_2h symmetry allows for a much finer tuning of the excited-state properties of quadrupolar dyes than most previously used D or A cores of higher symmetry.

---

1 Introduction

Over the past few years, excited-state symmetry breaking (ESSB) in polar environments was found to be a common phenomenon of many multipolar molecules consisting of electron donor (D)–acceptor (A) branches arranged in D–(π–A)_2,3 or A–(π–D)_2,3 motifs.^1–32 During this process, the exciton, initially distributed evenly over the whole molecule, localises, at least partially, on one D–π–A branch. As a consequence, the excited state changes from quadrupolar/octupolar to dipolar. Previous investigations revealed that the propensity of a multipolar molecule to undergo ESSB in polar solvents increases with the electron donating and withdrawing strength of the D and A constituents as well as with their separation distance. These results were rationalised with a simple excitonic model in which ESSB is only possible if the loss of interbranch coupling, V_ib, upon localisation is compensated by a gain in solvation energy.^27,33–35 The effect of substitution pattern was also investigated by comparing two D–π–A–π–D molecules with the same D and A subunits but with different symmetries, one centrosymmetric and linear and the other with the two branches forming a right angle.³⁶ Full localisation of the excitation on one branch was found to be easier, i.e. to require a less polar solvent, for the linear molecule than for the right-angled one. This was explained by the fact that the delocalised state of the right-angled molecule is already polar. Consequently the gain in solvation energy upon ESSB is smaller than for the linear one.

Here, we compare two formally centrosymmetric D–π–A–π–D molecules, which differ by the position of the π–D branches on the dipyrrolonaphthyridinedione (DPND) acceptor core, i.e., in either 1 and 7 positions, ‘Away’ from the pyrrole N atom (A1) or in 3 and 9 positions, ‘Close’ to the pyrrole N atom (C1, Fig. 1). DPNDs are recently introduced cross-conjugated dyes,³⁷ and are now well recognised for their strong and tunable emission,^38,39 large two-photon absorption cross-section,⁴⁰ and significant singlet fission yield.⁴¹ As a comparison, molecules A2 and C2 with the dimethylanilines replaced by the much weaker donor anisoles as well as the DPND core alone were also investigated. To detect occurrence of ESSB, we apply time-resolved IR absorption spectroscopy (TRIR) in the –C≡C– stretching region. If electronic excitation is delocalised evenly over the whole molecule, only the anti-symmetric –C≡C– vibration is visible in the IR spectrum.⁴² On the other hand, as soon as the centro-symmetry of the excited state is lost, the symmetric –C≡C– mode becomes IR active as well and two bands are now visible in the transient IR spectrum. We find that ESSB does not occur with anisole units as donors (A2 and C2) but is operative with both A1 and C1. However, our results reveal that the tendency to undergo ESSB depends on the substitution pattern, with A1 exhibiting a higher propensity than C1. This difference is explained by the combination of two effects: (i) a weaker conjugation, hence a smaller interbranch coupling, and (ii) a more dipolar excited state upon ESSB for A1 compared to C1.

Fig. 1 Structure of the D–π–A–π–D dyes (An, Cn) and of the DPND core model.

2 Experimental

2.1 Dyes

The dyes C1 and C2 were synthesised as described in ref. 43, while A2 was prepared as discussed in ref. 39. The synthesis of A1 was performed via Sonogashira reaction as described in the SI (Fig. S1–S3). The synthesis of core will be published separately in due course. The solvents, cyclohexane (CHX, Thermo Scientific), toluene (TOL, Thermo- Scientific), tetrahydrofuran (THF, Thermo Scientific), benzonitrile (BCN, Thermo Scientific), acetonitrile (ACN, Sigma-Aldrich), and dimethyl sulfoxide (DMSO, Roth) were of the highest commercially available purity and were used as received.

2.2 Stationary spectroscopy

Electronic absorption spectra were measured using a Cary 50 spectrometer, whereas emission spectra were recorded on a Horiba FluoroMax-4 spectrofluorometer. The fluorescence spectra were corrected using a set of secondary emissive standards.⁴⁴

2.3 Time-resolved spectroscopy

2.3.1 Time-resolved fluorescence. Fluorescence lifetime measurements were performed using the time-correlated single-photon counting (TCSPC) technique with a home-built setup described in ref. 45. Excitation was achieved with a tunable picosecond laser source (NKT photonics SuperK COMPACT with SuperK SELECT tunable multiline filter). The full width at half maximum (FWHM) of the instrument response function (IRF) was around 200 ps.

2.3.2 Electronic transient absorption spectroscopy. Ultrafast electronic transient absorption (TA) measurements were performed with a setup described in ref. 46 and based on an amplified Ti:Sapphire system (Solstice Ace, Spectra-Physics), producing 35 fs pulses centred at 800 nm with a 5 kHz repetition rate. The pump pulses were generated using a TOPAS-Prime combined with a NirUVis module (Light Conversion). The pump irradiance on the sample was between 0.15 and 0.75 mJ cm⁻². Probing was achieved from about 320 to 750 nm using white light pulses generated in a 3 mm CaF₂ plate. The polarisation of the pump pulses was at magic angle with respect to that of the probe pulses. The sample cell was 1 mm thick and the IRF had a FWHM varying between 80 and 350 fs, depending on the probe wavelength.

2.3.3 Time-resolved infra-red spectroscopy. The time-resolved infrared (TRIR) spectroscopy measurements were performed using a setup described in detail previously,⁴⁷ and based on an amplified Ti:Sapphire system (Spectra-Physics, Solstice, 100 fs, 800 nm, 1 kHz). The pump pulses were generated from the second harmonic of the output of an optical parametric amplifier (Light Conversion, TOPAS-C). Probing was achieved using the output of an optical parametric amplifier (TOPAS C, Light Conversion) combined with a non-collinear difference-frequency-mixing module (NDFG, Light Conversion). These pulses were dispersed in a Triax 190 spectrograph (Horiba, 150 lines per mm) and detected with a 2 × 64 elements MCT array (Infrared Systems Development). The samples were flowed through a cell with CaF₂ windows separated by a 500 µm spacer and had an absorbance of less than 0.3 at the excitation wavelength.

2.4 Quantum-chemical calculations

All calculations were carried out in the gas phase at the density functional theory (DFT) or time-dependent (TD) DFT level using the CAM-B3LYP functional,⁴⁸ combined with the 6-31G(d,p) basis set, as implemented in the Gaussian 16 (rev.B) package.⁴⁹

3 Results

3.1 Stationary spectroscopy and quantum-chemical calculations

The electronic absorption and emission spectra of A1 and C1 in different solvents are depicted in Fig. 2. The lowest energy absorption band of both dyes in CHX exhibits a distinct vibronic structure that becomes hardly visible in polar solvents due to broadening. This band shifts to lower energy by 1100 cm⁻¹ when going from A1 to C1. Additionally to the broadening, this band exhibits a small solvatochromism, which correlates with the solvent polarisability (Fig. S4), pointing to a non-polar ground state, as expected, and to dispersion as the dominant solute–solvent interactions.^50,51 The lower absorption band of A2 and C2 with the weak anisoleD show similar behaviour, except for a weaker broadening in polar solvents (Fig. S5).

Fig. 2 Electronic absorption (solid lines) and fluorescence spectra (dashed lines) of A1 (top) and C1 (bottom) in various solvents. CHX: cylohexane; TOL: toluene; THF: tetrahydrofuran; BCN: benzonitrile.

Much larger solvatochromism and band broadening can be observed in the fluorescence spectrum of A1 and C1 (Fig. 2), whereas only weak solvent dependence is found with A2 and C2 (Fig. S5). The red shift of the fluorescence of A1 and C1 correlates with the orientational polarisation of the solvent (Fig. S4). This points to a solvatochromism dominated by dipole–dipole interactions,⁵⁰ and suggests a dipolar excited state in polar solvents, hence a localisation of the excitation.

Interestingly, although these two dyes consist of the same D and A subunits at the same distance, the fluorescence solvatochromism of A1 is significantly larger than that of C1, with a shift of 6570 vs. 3880 cm⁻¹ when going from CHX to BCN (Fig. S4). The solvent shift is also accompanied by a marked decrease in fluorescence quantum yield, which correlates with the shortening of the fluorescence lifetime observed by going from CHX to THF. This effect is more marked with A1, namely by a factor of about 5 (2.6 and 0.5 ns) vs. a factor of less than 2 (2.8 and 1.5 ns) for C1, as shown in Table S1. The fluorescence solvatochromism of A2 and C2 is comparatively much smaller, 1400 and 1160 cm⁻¹ respectively, when going from CHX to BCN (Fig. S5). Additionally, the fluorescence lifetime is essentially the same in CHX and BCN (Table S1). These results are consistent with the much weaker electron-donating properties of anisole.

The above results can be rationalised using gas-phase quantum-chemical calculations at the (TD)-DFT level of analogues of A_n and C_n with the heptyl substituents replaced by H atoms. They point to a planar ground state for A_n and C_n. However, significant torsional disorder is predicted, given that these molecules are relative flexible for the rotation around the single bonds of the phenylethynyl arms. For A1, the barrier for torsion is slightly lower than for C1, namely, 2.0 vs. 3.1 k_BT at room temperature (Fig. S7).

According to TD-DFT calculations, the S₁ ← S₀ transition of all four dyes is dominated by a HOMO to LUMO one-electron excitation and characterised by a large oscillator strength, around 1.5–1.6 for A1 and C1 and around 1.3 for the other two (Fig. 3 and S8, Table S2). The red shift of the S₁ ← S₀ transition observed when going from A1 to C1 (Fig. 2) is also reproduced by the calculations, with the S₁ state of C1 located 0.11 eV (900 cm⁻¹) below that of A1 (Table S2). Additionally, the S₂ ← S₀ transition, mostly a HOMO−1 to LUMO excitation, is predicted to be about 0.6–0.7 eV higher and to be symmetry forbidden. This result is fully consistent with the theoretical models used to describe the lowest excited states of quadrupolar molecules, with the one-photon allowed and two-photon forbidden S₁ ← S₀ transition and the one-photon forbidden and two-photon allowed S₂ ← S₀ transition.^3,27,33,52 In the excitonic model,⁵³ these dyes can be considered as two D–π–A chromophores arranged in a collinear manner, similar to a J-dimer. The frontier molecular orbitals reveal that these first two electronic transitions involve significant charge transfer from the dimethylanilines D to the DPND core, and point to a significant increase in quadrupole moment upon excitation (Fig. 3). This change is comparatively smaller with anisole as donor, i.e. for A2 and C2, as expected (Fig. S8).

Fig. 3 Frontier molecular orbitals (CAM-B3LYP/6-31G(d,p)) of analogues of A1 and C1 with R=H (Fig. 1) involved in the S₁ ← S₀ (HOMO to LUMO) and S₂ ← S₀ (HOMO−1 to LUMO) transitions.

TD-DFT calculations also predicted a planar S₁ state, but higher barriers for torsion of the phenylethynyl arms, namely 7 and 8k_BT, for A1 and C1, respectively (Fig. S7). This difference in barrier for torsion between ground and excited states is typical of phenylethynyl-based chromophores and results in non-mirror image relationship of absorption and emission spectra, with a narrower and more structured fluorescence spectrum, as observed here in CHX.^54–57

3.2 Electronic transient absorption spectroscopy

Electronic transient absorption (TA) measurements were performed with all four dyes in the non-polar CHX and the polar BCN. The TA data were analysed globally assuming a series of successive exponential steps to obtain evolution-associated difference absorption spectra (EADS) and time constants.^46,58 These EADS and time constants do not necessarily correspond to well-defined states/species and inverse rate constants but reflect the changes in the TA spectra and their timescales.

The TA spectra recorded with A1 and C1 in CHX are qualitatively similar with a negative band above 600 nm due mostly to stimulated emission (SE) and positive bands at shorter wavelength due to S_n>1 ← S₁ excited-state absorption (ESA) (Fig. 4A, B and S9). The bleach of the S₁ ← S₀ band is hardly visible because of overlapping ESA features. The latter are also overlapping with the lower energy side of the SE of C1. Apart from the decay of all bands on the ns timescale, in agreement with the fluorescence lifetimes (Table S1), the TA spectra exhibit only little dynamics at early time, which most probably arise from vibrational and/or structural relaxation. For example, the increasing vibronic structure of the SE band of A1 is consistent with planarisation.⁵⁶

Fig. 4 Evolution-associated difference absorption spectra and time constants obtained from global analysis of the transient absorption data recorded upon photoexcitation of A1 and C1 in cyclohexane (CHX, A and B) and benzonitrile (BCN, C and D) assuming a series of successive exponential steps (A → B → ⋯ →). The negative stationary absorption and stimulated emission spectra are shown in shaded area for comparison.

By contrast, significant spectral dynamics can be observed in BCN, especially in the SE region (Fig. 4C, D and S9). They can be explained by the red shift of the SE band due to solvent relaxation and the appearance of ESA features that are hidden by the SE band at early time. The timescales of these changes are similar to those of solvent relaxation in BCN.⁵⁹ The concurrent blue shift of the ESA band around 500 nm is also consistent with the equilibration of the S₁ state and an increase of S_n>1–S₁ gap. Comparatively, much less spectral dynamics can be observed with A2 and C2 (Fig. S10 and S11), in agreement with their markedly smaller fluorescence solvatochromism.

These TA data in BCN also reveal a strong acceleration of the excited-state decay relative to CHX, with the S₁ lifetime decreasing from ns to 125 and 500 ps for A1 and C1, respectively. This decrease with increasing solvent polarity is also consistent with that observed for the fluorescence lifetime when going from CHX to THF. This effect is notably more marked for A1 than C1. Comparatively, the excited-state decays of A2 and C2 are not affected by the solvent polarity.

To find spectral evidence for a charge-transfer character of the S₁ state of A1 and C1, TA measurements were also carried with the DPND core alone and with 1 M N,N-dimethylaniline (DMA) in BCN. As illustrated in Fig. 5, in the presence of DMA, the SE band of core decays in a few ps and positive bands around 440 and 700 nm appear concurrently. Given that the radical cation of DMA absorbs relatively weakly around 460–470 nm,⁶⁰ these two bands are most probably due to the radical anion of core. Based on these results, the ESA features observed with A1 and C1 in BCN are consistent with a S₁ state characterised by significant charge transfer from the DMA donor(s) to the DPND acceptor.

Fig. 5 Evolution-associated difference absorption spectra and time constants obtained from global analysis of the transient absorption data recorded upon photoexcitation of the core (A) without and (B) with 1 M N,N-dimethylaniline (DMA) in benzonitrile assuming a series of successive exponential steps (A → B → ⋯ →). The negative stationary absorption and stimulated emission spectra are shown in shaded area for comparison.

3.3 Time-resolved infra-red spectroscopy

Time-resolved IR (TRIR) measurements were carried out with A1 and C1 in a series of solvents of increasing polarity (Fig. S12 and S13). A2 and C2 were only investigated in CHX and DMSO, as very little solvent dependence of the TRIR dynamics was observed (Fig. S14 and S15). Like for the TA experiments, the TRIR data were analysed globally assuming series of successive exponential steps (Fig. 6, S14 and S15).^58,61 In CHX, both A1 and C1 exhibit a single band, ESA1, in the –C≡C– region (Fig. 6). For A1, this band is significantly narrower and is shifted by +20 cm⁻¹ relative to C1. For both molecules, this bands rises and narrows during the first ∼20 ps before decaying on the ns timescale, in agreement with the S₁ state lifetime determined from TA and time-resolved fluorescence. This early narrowing can be attributed to vibrational and structural relaxation,⁶² similarly to the early dynamics observed in the TA spectra. Additionally to ESA1, another broad positive feature is present on the high-frequency side of the weak negative feature around 2200 cm⁻¹ due to the ground-state bleach of the –C≡C– stretch band (Fig. S6). This high-frequency band, ESAe, which extends above 2600 cm⁻¹, is assigned to a mid-IR electronic transition from the delocalised S₁ state, as supported by the data discussed below.^24,42 ESAe is significantly more intense with C1 and probably overlaps with ESA1. The mixing of electronic and vibrational transitions with C1 could possibly be at the origin of the much larger bandwidth of ESA1. As illustrated in Fig. S14 and S15, the TRIR spectra recorded with A2 and C2 exhibit similar behaviour.

Fig. 6 Evolution-associated difference absorption spectra and time constants obtained from global analysis of the time-resolved IR absorption data recorded upon photoexcitation of A1 and C1 in various solvents assuming a series of successive exponential steps (A → B → ⋯ →). CHX: cyclohexane; THF: tetrahydrofuran; BCN: benzonitrile; DMSO: dimethyl sulfoxide. Given the poor solubility of A1 in DMSO, a 80 : 20 (v/v) DMSO/BCN mixture was used.

The presence of a single –C≡C– stretching band in the TRIR spectrum points to a symmetric distribution of the excitation on the two D–π–A branches with only the antisymmetric stretching mode IR allowed.^10,42 In principle, the large bandwidth recorded with C1 could also be due to the presence of two weakly split bands and could, thus, indicate an uneven distribution of the excitation. This can however be excluded by the results obtained in polar solvents, as described below.

The TRIR spectra measured in the weakly polar TOL are very similar to those in CHX (Fig. 6). However, significant changes are observed when increasing solvent polarity to THF as illustrated in Fig. 6. The early spectra recorded with A1 are dominated by ESA1 at the same frequency as in CHX but with a larger width and by the broad ESAe, which is now as intense as ESA1. Additionally, a weak and relatively narrow vibrational band, ESA2, can be observed on the high-frequency side of ESA1. Over the first 10 ps after excitation, ESA2 increases and becomes dominant, while ESA1 and ESAe decay concurrently. Afterwards, all three bands decay in parallel in 600–700 ps.

The early TRIR spectra measured with C1 in THF also exhibit intense ESA1 and ESAe and a weak ESA2 (Fig. 6). However, the intensity increase of ESA2 and parallel decrease of ESA1 and ESAe, which occur during the first few ps, is much smaller than for A1. Consequently, after these changes, ESA1 and ESAe remain dominant. After these early dynamics, all three bands decay concurrently on the ns timescale.

The presence of two –C≡C– stretching bands in THF can be interpreted as evidence of ESSB and a lopsided distribution of the excitation over the molecule.^10,33 Consequently, the symmetric –C≡C– stretching mode is no longer IR forbidden. The early dynamics with the increase of ESA2 and partial decrease of the other two bands reflect the ESSB dynamics. They occur on a timescale similar to those of solvent motion in agreement with previous findings that ESSB is driven by a gain in solvation energy.^27,35,63 The difference in the ESA1/ESA2 intensity ratio between A1 and C1 suggests that the extent of asymmetry is larger for A1. This conclusion is consistent with the stronger intensity decrease of ESAe, attributed to an electronic transition of the delocalised excited state, observed with A1 compared to C1.

Similar early dynamics are observed upon further increasing the solvent polarity, i.e., in BCN and DMSO (Fig. 6). Due to the limited solubility of A1 in DMSO, measurements with this dye were carried out in a 80 : 20 (v/v) DMSO/BCN mixture. For A1, only ESA2 is present in the late spectra, pointing to a full localisation of the excitation on one A–π–D side of the molecule. In this case again, the timescale of the early dynamics are similar to those of solvent motion.⁵⁹ By contrast, for C1 in BCN, both ESA1 and ESAe remain very intense after equilibration, although ESA2 is more visible than in THF. This band becomes dominant only in the highly polar DMSO, but the other two bands are still visible contrary to A1.

These results reveal that full ESSB, i.e., exciton trapping on a single branch, is operative with A1 in BCN and more polar solvents, whereas excitation in C1 remains partially delocalised.

In agreement with the TA measurements in CHX and BCN, these TRIR data point to a continuous decrease of the relaxed S₁ state lifetime with increasing solvent polarity, namely, from a few ns in CHX to 600, 130 and 42 ps for A1 in THF, BCN and DMSO/BCN, respectively. This effect is not as marked for C1, whose excited-state lifetime changes from a few ns in CHX and THF to 540 and 82 ps in BCN and DMSO.

Comparatively, the TRIR spectra of the molecules with anisole donors, A2 and C2, exhibit little dynamics with a single vibrational band in the –C≡C– stretching region and a possible broad electronic background in CHX as well as in DMSO (Fig. S14 and S15). Moreover, the excited-state lifetime is on the ns timescale in both solvents. These results suggest that electronic excitation remains evenly delocalised independently of the solvent polarity and the position of the π–D arms.

4 Discussion

The trends in the extent of ESSB revealed by the TRIR measurements are fully consistent with the observed fluorescence solvatochromism (Fig. 2 and S5). On the one hand, the negligible solvent dependence of the emission spectra of A2 and C2 agrees with a quadrupolar excited state. On the other hand, the substantial emission solvatochromism of A1 and C1 points to a dipolar character of the excited state. Furthermore, the larger solvent dependence measured with A1 compared to C1 suggests that, in polar media, the excited state of A1 is more dipolar than that of C1.

The absence of ESSB with anisole as donor can be easily explained by its weak electron donating ability, reflected by its very positive oxidation potential of 1.76 V vs. SCE relative to 0.78 V vs. SCE for DMA.^64,65 These values should be compared with the reduction potential of the DPND of −1.12 V vs. SCE,⁴³ and the energy of its S₁ state of 2.35 eV. According to the Weller equation (eqn (S1)),⁶⁶ photoinduced electron transfer between DPND and anisole is highly endergonic, i.e. ΔG_ET = 0.53 eV, whereas it is energetically favourable with DMA with ΔG_ET = −0.45 eV. Therefore, only weak CT from the core to the anisole ends should be expected upon photoexcitation of A2 and C2. Consequently, their S₁ state should be only weakly quadrupolar. By contrast, much larger CT and quadrupolar moment can be expected for the S₁ state of A1 and C1.

More surprising is the finding that ESSB is easier with A1 than with C1, despite them having the same D and A strength and C_2h symmetry. To explain this, we consider the simple excitonic model mentioned in the introduction, where both D–π–A–π–D dyes can be approximated to a J dimer of D–π–A with the Davydov splitting of the two lowest delocalised electronic excited states, S₁ and S₂, corresponding to twice the interbranch coupling, 2V_ib (Fig. 7, left).^27,35 Exciton localisation on a single branch leads to a loss of the interbranch coupling and is, thus, endergonic by V_ib. However, if the gain of solvation energy upon going from the symmetric quadrupolar excited state to the dipolar symmetry-broken state, ΔE_s, exceeds V_ib, then ESSB is energetically favourable (Fig. 7, right). As a consequence, the higher tendency of A1 to undergo ESSB compared to C1, must originate from a smaller V_ib and/or a larger ΔE_s.

Fig. 7 Energetics of excited-state symmetry breaking. (left) Exciton localisation in non-polar media is not operative because the excitonic stabilisation energy of the delocalised state, V_ib, is lost. (right) In polar solvents, the quadrupolar excited state, S_1,Q, is stabilised by solvation energy, E_s,Q, which is smaller than E_s,D, the solvation energy of the dipolar excited state, S_1,D. If the gain in solvation energy, ΔE_s = E_s,D–E_s,Q, is larger than V_ib, excited-state symmetry breaking is energetically favourable, ΔE_SB < 0.

The red shift of the S₁–S₀ absorption and emission spectra of C1 in CHX relative to those of A1 (Fig. 2) can be explained by a larger conjugation in the S₁ state of the former and should be associated with a larger V_ib. In principle, the splitting of the symmetric and antisymmetric vibrations of two coupled vibrators increases with the magnitude of the coupling.^67,68 Therefore, the larger frequency splitting of ESA2 and ESA1 measured with C1 relative to A1 in THF and BCN, namely 85 vs. 75 cm⁻¹ (Fig. 6), could also reflect a larger V_ib value. Finally, calculations of the nucleus-independent chemical shift (NICS) values for A2 and C2 reported in ref. 39 predict a slightly higher aromaticity of the pyrrole rings upon 3,9- (i.e. Cn) than 1,7-substitution (An), supporting a higher conjugation for 3,9-substitution and possibly a larger interbranch coupling for C1. Nevertheless, all these differences between An and Cn are small and might not be sufficient to account for the higher tendency of A1 to undergo ESSB.

However, a major difference between A1 and C1 can be observed in the fluorescence solvatochromism, which is more than 50% larger for A1 (Fig. 2 and S4). This implies that the relaxed S₁ state of A1 is much more stabilised by polar solvation than that of C1, pointing, thus, to a significantly larger gain in solvation energy, ΔE_s, than for C1. This higher stabilisation of the relaxed S₁ state of A1 in polar media is also confirmed by the significantly shorter excited-state lifetime compared to C1, i.e., by a faster non-radiative decay due to the reduced S₁–S₀ gap.

The gain in solvation energy, ΔE_s, was shown to be well approximated by taking half of the fluorescence band shift measured by going from a non-polar to a given polar solvent.²⁷ Based on this, ΔE_s in BCN can be estimated to be larger for A1 than C1 by about 0.15 eV. This is a substantial difference that can easily account for the higher propensity of A1 toward exciton trapping.

The larger ΔE_s of A1 indicates that the symmetry-broken state of A1 is more polar than that of C1. To try understanding the origin of this effect, we performed quantum-chemical calculations of analogues of A1 and C1 but with only a single –π–D branch, namely A1sb and C1sb. Calculations of the permanent dipole moment of these single-branched molecules predicted unrealistically large values, between 4.5 and 5 D, in the ground state and only a minor increase, around 0.7 D, in the S₁ state. A large ground-state dipole moment is also predicted for A2sb, although this molecule exhibits negligible absorption solvatochromism (Fig. S5C), pointing to weak ground state dipole moment. Similar results were obtained with other functionals and basis sets, such as ωB97XD/def2-TZVPP. Reliable calculation of permanent electric dipole moments using standard DFT based methods is known to be problematic for such systems.^69,70 Therefore, the results of these dipole moment calculations were no longer considered.

However, the frontier MOs involved in the S₁ ← S₀ transition as well as the associated charge density difference (CDD) surfaces suggest a possible origin of the higher polarity of the localised S₁ state of A1. As illustrated in Fig. 8, the increase of electronic density upon S₁ ← S₀ excitation is not the same for the two carbonyl oxygen atoms. For both A1sb and C1sb, it is the largest for the carbonyl oxygen located close to the pyrrole ring bearing the –π–D arm. For A1sb, this atom is much farther from the DMA donor than for C1sb, with a O–N distance of 12.7 Å vs. 7.5 Å. Given that dipolar solvation energy scales with the square of the dipole moment, such difference should be sufficient to account for a larger gain in solvation energy upon ESSB in A1.

Fig. 8 Charge density difference surfaces calculated for the S₁ ← S₀ transition of single branched analogues of A1 and C1. Yellow and blue colours stand for decrease and increase in electronic density, respectively. The red arrows designate the carbonyl oxygen atom undergoing the largest increase in electronic density upon excitation. For A1sb, this atom is located further apart from the electron-donating group, suggesting a more polar excited state than for C1sb.

5 Conclusions

Previous investigations of D–(π–A)₂ and A–(π–D)₂ molecules evidenced how excited-state symmetry breaking and exciton localisation depend on the push–pull character of the constituents, their separation distance and the overall symmetry of the molecule. The present study with two centrosymmetric D–π–A–π–D dyes, which only differ by the position of the –π–D arms on the DPND acceptor core reveals that the substitution pattern has also an effect on the tendency of the electronic excitation to localise on one A–π–D side in polar media. This difference can be explained by the C_2h symmetry of the DPND core. Because of it, substitution of an electron donor in position 3, close to a carbonyl group, or in position 1 away from a carbonyl, leads to different charge-transfer characters in the excited state. Consequently, the dipole moment of the symmetry broken S₁ state is larger when the donors are in positions 1 and 7 (A1) than in positions 3 and 9 (C1). In this respect, using DPND as a core allows for a finer tuning of the nature of the excited state than most of the donor or acceptors cores used so far, which have a D_2h symmetry, and for which substitution on either side is equivalent. These findings illustrate how apparently subtle structural differences can have strong impact on the nature of the excited state of such multipolar dyes. A deeper understanding of these structural effects would certainly facilitate further developments of quadrupolar dyes and fluorophores for specific applications.

Conflicts of interest

There are no conflicts to declare.

Data availability

Supplementary information: synthesis of A1, solvatochromism, time resolved fluorescence, transient electronic and IR absorption spectra and global analysis, and quantum-chemical calculations. See DOI: https://doi.org/10.1039/d6cp00121a.

All data can be downloaded from https://doi.org/10.26037/yareta:4qg2qkc5sfdelb65tzyoxzfa6a.

Acknowledgements

This work was financially supported by the the University of Geneva, and by the Polish National Science Centre, Poland (UMO-2018/30/M/STS/00460). The authors thanks Dr. Łukacz Kielesiński for providing a sample of core. The computations were performed at the University of Geneva using the Baobab high-performance computing (HPC) service.

Notes and references

1. S. Amthor, C. Lambert, S. Dümmler, I. Fischer and J. Schelter, J. Phys. Chem. A, 2006, 110, 5204–5214.
2. R. Stahl, C. Lambert, C. Kaiser, R. Wortmann and R. Jakober, Chem. – Eur. J., 2006, 12, 2358–2370.
3. F. Terenziani, A. Painelli, C. Katan, M. Charlot and M. Blanchard-Desce, J. Am. Chem. Soc., 2006, 128, 15742–15755.
4. F. Terenziani, C. Sissa and A. Painelli, J. Phys. Chem. B, 2008, 112, 5079–5087.
5. S. Easwaramoorthi, J.-Y. Shin, S. Cho, P. Kim, Y. Inokuma, E. Tsurumaki, A. Osuka and D. Kim, Chem. – Eur. J., 2009, 15, 12005–12017 , S12005/1–S12005/11.
6. C. Sissa, A. Painelli, M. Blanchard-Desce and F. Terenziani, J. Phys. Chem. B, 2011, 115, 7009–7020.
7. B. Carlotti, E. Benassi, A. Spalletti, C. G. Fortuna, F. Elisei and V. Barone, Phys. Chem. Chem. Phys., 2014, 16, 13984–13994.
8. A. Rebane, M. Drobizhev, N. S. Makarov, G. Wicks, P. Wnuk, Y. Stepanenko, J. E. Haley, D. M. Krein, J. L. Fore, A. R. Burke, J. E. Slagle, D. G. McLean and T. M. Cooper, J. Phys. Chem. A, 2014, 118, 3749–3759.
9. B. Carlotti, E. Benassi, C. G. Fortuna, V. Barone, A. Spalletti and F. Elisei, ChemPhysChem, 2016, 17, 136–146.
10. B. Dereka, A. Rosspeintner, Z. Li, R. Liska and E. Vauthey, J. Am. Chem. Soc., 2016, 138, 4643–4649.
11. B. Dereka, A. Rosspeintner, M. Krzeszewski, D. T. Gryko and E. Vauthey, Angew. Chem., Int. Ed., 2016, 55, 15624–15628.
12. N. Dozova, L. Ventelon, G. Clermont, M. Blanchard-Desce and P. Plaza, Chem. Phys. Lett., 2016, 664, 56–62.
13. S. Lee and D. Kim, J. Phys. Chem. A, 2016, 120, 9217–9223.
14. J. S. Beckwith, A. Rosspeintner, G. Licari, M. Lunzer, B. Holzer, J. Fröhlich and E. Vauthey, J. Phys. Chem. Lett., 2017, 8, 5878–5883.
15. T. M. Cooper, J. E. Haley, D. M. Krein, A. R. Burke, J. E. Slagle, A. Mikhailov and A. Rebane, J. Phys. Chem. A, 2017, 121, 5442–5449.
16. B. Dereka, A. Rosspeintner, R. Stężycki, C. Ruckebusch, D. T. Gryko and E. Vauthey, J. Phys. Chem. Lett., 2017, 8, 6029–6034.
17. L. G. Łukasiewicz, H. G. Ryu, A. Mikhaylov, C. Azarias, M. Banasiewicz, B. Kozankiewicz, K. H. Ahn, D. Jacquemin, A. Rebane and D. T. Gryko, Chem. – Asian J., 2017, 12, 1736–1748.
18. Y. Sonoda, J. Lumin., 2017, 187, 352–359.
19. S. A. Kurhuzenkau, M. Y. Colon Gomez, K. D. Belfield, Y. O. Shaydyuk, D. J. Hagan, E. W. Van Stryland, C. Sissa, M. V. Bondar and A. Painelli, J. Phys. Chem. C, 2018, 122, 5664–5672.
20. B. Bardi, M. Krzeszewski, D. T. Gryko, A. Painelli and F. Terenziani, Chem. – Eur. J., 2019, 25, 13930–13938.
21. H. Y. Chung, J. Oh, J.-H. Park, I. Cho, W. S. Yoon, J. E. Kwon, D. Kim and S. Y. Park, J. Phys. Chem. C, 2020, 124, 18502–18512.
22. L. G. Łukasiewicz, M. Rammo, C. Stark, M. Krzeszewski, D. Jacquemin, A. Rebane and D. T. Gryko, ChemPhotoChem, 2020, 4, 508–519.
23. X. Niu, Z. Kuang, M. Planells, Y. Guo, N. Robertson and A. Xia, Phys. Chem. Chem. Phys., 2020, 22, 15743–15750.
24. B. Dereka, D. Svechkarev, A. Rosspeintner, A. Aster, M. Lunzer, R. Liska, A. M. Mohs and E. Vauthey, Nat. Commun., 2020, 11, 1925.
25. S. L. Bondarev, T. F. Raichenok, S. A. Tikhomirov, N. G. Kozlov, T. V. Mikhailova and A. I. Ivanov, J. Phys. Chem. B, 2021, 125, 8117–8124.
26. A. Cesaretti, A. Spalletti, F. Elisei, P. Foggi, R. Germani, C. G. Fortuna and B. Carlotti, Phys. Chem. Chem. Phys., 2021, 23, 16739–16753.
27. Z. Szakács, F. Glöcklhofer, F. Plasser and E. Vauthey, Phys. Chem. Chem. Phys., 2021, 23, 15150–15158.
28. H. Usta, B. Cosut and F. Alkan, J. Phys. Chem. B, 2021, 125, 11717–11731.
29. Z. Wei, S. Sharma, A. M. Philip, S. Sengupta and F. C. Grozema, Phys. Chem. Chem. Phys., 2021, 23, 8900–8907.
30. M. Fakis, V. Petropoulos, P. Hrobárik, J. Nociarová, P. Osuský, M. Maiuri and G. Cerullo, J. Phys. Chem. B, 2022, 126, 8532–8543.
31. S. Guo, W. Liu, Y. Wu, J. Sun, J. Li, H. Jiang, M. Zhang, S. Wang, Z. Liu, L. Wang, H. Wang, H. Fu and J. Yao, J. Phys. Chem. Lett., 2022, 13, 7547–7552.
32. B. Dereka, N. Maroli, Y. M. Poronik, D. T. Gryko and A. A. Kananenka, Chem. Sci., 2024, 15, 15565–15576.
33. A. I. Ivanov, B. Dereka and E. Vauthey, J. Chem. Phys., 2017, 146, 164306.
34. A. I. Ivanov and V. G. Tkachev, J. Chem. Phys., 2019, 151, 124309.
35. P. Verma, M. Tasior, P. Roy, S. R. Meech, D. T. Gryko and E. Vauthey, Phys. Chem. Chem. Phys., 2023, 25, 22689–22699.
36. B. Dereka, E. Balanikas, A. Rosspeintner, Z. Li, R. Liska and E. Vauthey, J. Phys. Chem. Lett., 2024, 15, 8280–8286.
37. B. Sadowski and D. T. Gryko, Chem. Sci., 2023, 14, 14020–14038.
38. B. Sadowski, M. F. Rode and D. T. Gryko, Chem. – Eur. J., 2018, 24, 855–864.
39. K. Skonieczny, L. Kielesinski, M. Grzybowski and D. T. Gryko, Chem. Commun., 2025, 61, 5602–5606.
40. B. Sadowski, H. Kita, M. Grzybowski, K. Kamada and D. T. Gryko, J. Org. Chem., 2017, 82, 7254–7264.
41. L. Wang, L. Lin, J. Yang, Y. Wu, H. Wang, J. Zhu, J. Yao and H. Fu, J. Am. Chem. Soc., 2020, 142, 10235–10239.
42. E. Vauthey, J. Phys. Chem. Lett., 2022, 13, 2064–2071.
43. M. Grzybowski, I. Deperasińska, M. Chotkowski, M. Banasiewicz, A. Makarewicz, B. Kozankiewicz and D. T. Gryko, Chem. Commun., 2016, 52, 5108–5111.
44. J. A. Gardecki and M. Maroncelli, Appl. Spectrosc., 1998, 52, 1179–1189.
45. J. Sissaoui, D. S. Budkina and E. Vauthey, J. Phys. Chem. Lett., 2023, 14, 5602–5606.
46. J. S. Beckwith, A. Aster and E. Vauthey, Phys. Chem. Chem. Phys., 2022, 24, 568–577.
47. M. Koch, R. Letrun and E. Vauthey, J. Am. Chem. Soc., 2014, 136, 4066–4074.
48. T. Yanai, D. P. Tew and N. C. Handy, Chem. Phys. Lett., 2004, 393, 51–57.
49. M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuseria, M. A. Robb, J. R. Cheeseman, G. Scalmani, V. Barone, G. A. Petersson, H. Nakatsuji, X. Li, M. Caricato, A. V. Marenich, J. Bloino, B. G. Janesko, R. Gomperts, B. Mennucci, H. P. Hratchian, J. V. Ortiz, A. F. Izmaylov, J. L. Sonnenberg, D. Williams-Young, F. Ding, F. Lipparini, F. Egidi, J. Goings, B. Peng, A. Petrone, T. Henderson, D. Ranasinghe, V. G. Zakrzewski, J. Gao, N. Rega, G. Zheng, W. Liang, M. Hada, M. Ehara, K. Toyota, R. Fukuda, J. Hasegawa, M. Ishida, T. Nakajima, Y. Honda, O. Kitao, H. Nakai, T. Vreven, K. Throssell, J. A. Montgomery Jr., J. E. Peralta, F. Ogliaro, M. J. Bearpark, J. J. Heyd, E. N. Brothers, K. N. Kudin, V. N. Staroverov, T. A. Keith, R. Kobayashi, J. Normand, K. Raghavachari, A. P. Rendell, J. C. Burant, S. S. Iyengar, J. Tomasi, M. Cossi, J. M. Millam, M. Klene, C. Adamo, R. Cammi, J. W. Ochterski, R. L. Martin, K. Morokuma, O. Farkas, J. B. Foresman and D. J. Fox, Gaussian 16 Rev. B.01, 2016.
50. P. Suppan, J. Photochem. Photobiol., A, 1990, 50, 293–330.
51. T. Renger, B. Grundkötter, M. E.-A. Madjet and F. Müh, Proc. Natl. Acad. Sci. U. S. A., 2008, 105, 13235–13240.
52. C. Sissa, F. Terenziani and A. Painelli, J. Phys. Chem. A, 2008, 112, 8697–8705.
53. M. Kasha, Radiat. Res., 1963, 20, 55–70.
54. M. I. Sluch, A. Godt, U. H. F. Bunz and M. A. Berg, J. Am. Chem. Soc., 2001, 123, 6447–6448.
55. S. J. Greaves, E. L. Flynn, E. L. Futcher, E. Wrede, D. P. Lydon, P. J. Low, S. R. Rutter and A. Beeby, J. Phys. Chem. A, 2006, 110, 2114–2121.
56. I. Fureraj, D. S. Budkina and E. Vauthey, Phys. Chem. Chem. Phys., 2022, 24, 25979–25989.
57. R. Ringström, F. Edhborg, Z. W. Schroeder, L. Chen, M. J. Ferguson, R. R. Tykwinski and B. Albinsson, Chem. Sci., 2022, 13, 4944–4954.
58. I. H. M. van Stokkum, D. S. Larsen and R. van Grondelle, Biochim. Biophys. Acta, Bioenerg., 2004, 1657, 82–104.
59. M. L. Horng, J. A. Gardecki, A. Papazyan and M. Maroncelli, J. Phys. Chem., 1995, 99, 17311–17337.
60. T. Shida, Electronic Absorption Spectra of Radical Ions, Elsevier, Amsterdam, 1988, vol. 34.
61. J. S. Beckwith, C. A. Rumble and E. Vauthey, Int. Rev. Phys. Chem., 2020, 39, 135–216.
62. P. Hamm, S. M. Ohline and W. Zinth, J. Chem. Phys., 1997, 106, 519–529.
63. A. E. Nazarov and A. I. Ivanov, J. Phys. Chem. B, 2020, 124, 10787–10801.
64. J. Eriksen and C. S. Foote, J. Phys. Chem., 1978, 82, 2659–2662.
65. N. Mataga, T. Asahi, Y. Kanda, T. Okada and T. Kakitani, Chem. Phys., 1988, 127, 249–261.
66. D. Rehm and A. Weller, Isr. J. Chem., 1970, 8, 259–271.
67. P. Hamm, M. Lim and R. M. Hochstrasser, J. Phys. Chem. B, 1998, 102, 6123–6138.
68. H. Kwon, K. Osawa, J. G. Seol, S. Sung, D. Kim and Y. S. Kim, J. Chem. Phys., 2023, 158, 214303.
69. A. O. Lykhin, D. G. Truhlar and L. Gagliardi, J. Chem. Theory Comput., 2021, 17, 7586–7601.
70. H. S. Clifford, M. R. Hennefarth, D. G. Truhlar and L. Gagliardi, J. Phys. Chem. Lett., 2025, 16, 10400–10409.

---