Chromophores of chromophores: a bottom-up Hückel picture of the excited states of photoactive proteins

Cate S. Anstöter , Charlie R. Dean and Jan R. R. Verlet *
Department of Chemistry, Durham University, Durham, DH1 3LE, UK. E-mail:

Received 23rd August 2017 , Accepted 18th September 2017

First published on 18th September 2017

Many photoactive proteins contain chromophores based on para-substituted phenolate anions which are an essential component of their electronic structure. Here, we present a reductionist approach to gain fundamental insight into the evolution of electronic structure as the chromophore increases in complexity from phenolate to that in GFP. Using frequency- and angle-resolved photoelectron spectroscopy, in combination with electronic structure theory, the onset of excited states that are responsible for the characteristic spectroscopic features in biochromophores are determined. A comprehensive, yet intuitive picture of the effect of phenolate functionalisation is developed based on simple Hückel theory. Specifically, the first two bright excited states can be constructed from a linear combination of molecular orbitals localised on the phenolate and para-substituent groups. This essential interaction is first observed for p-vinyl-phenolate. This bottom-up approach offers a readily accessible framework for the design of photoactive chromophores.


At the core of most photoactive proteins is a small organic chromophore that acts as a light activated switch for the protein's function. Many of these natural chromophores contain the phenolate anion, Ph (Fig. 1). In the green fluorescent protein (GFP), the chromophore is often taken to be the deprotonated p-hydroxybenzylidene-2,3-dimethylimidazolinone (HBDI, Fig. 1), which is hindered from undergoing isomerisation in the protein and consequently fluoresces with a very high quantum yield.1,2 In the photoactive yellow protein (PYP), the chromophore is often taken to be deprotonated p-coumaric acid (CA, Fig. 1), which undergoes a transcis isomerisation upon near UV-excitation and serves as a mechanical switch to initiate the protein's response.3,4 The commonality of phenolates in many photoactive proteins arises from the biosynthetic pathways of chromophores that often involve tyrosine. The photophysics of photoactive protein chromophores has attracted much attention because of their underpinning role in the initial protein response to light.5–7 Gas-phase spectroscopy has been particularly valuable as it provides an unperturbed view of the excited states of the chromophores and of their dynamics.8–12 Additionally, the gas-phase environment is tractable by high-level electronic structure theory,13–16 which when combined with experiment, provides a detailed understanding of the excited state dynamics. However, the reliance on very high level computational methods can mask some of the simple physical principles that are the foundation of the overall electronic structure of the chromophores. Such principles are central to the logical development of new light-activated proteins or macromolecules, and to understanding the natural selection based on nature's basic building blocks such as the amino acids. Here, we use gas-phase photoelectron imaging of the phenolate chromophores of HBDI and CA. Based on our results, we provide a simple bottom-up picture of their bright excited states using Hückel theory together with electronic structure calculations. Specifically, we show that a simple linear combination of molecular orbitals (MO) gives rise to the characteristic spectroscopic features of these photoactive chromophores.
image file: c7cp05766k-f1.tif
Fig. 1 Schematic showing structures of chemically substituted para-phenolate chromophores with increasing complexity.

There is a wealth of experimental and computational studies on the photophysics of photoactive proteins and, specifically, their chromophores.12,17–32 These have largely focussed on the first singlet excited (S1) state of the chromophores, because this is generally the excited state that is the most relevant to the biological function of the protein. For example, the pioneering work by Andersen and co-workers showed that the absorption spectrum of HDBI has a very similar appearance in vacuum as it does in the protein environment. The S1 state of HDBI has been the focus of a large number of gas-phase studies.13,19,31,33–37 A very insightful picture of the electronic structure of the S1 state was provided by Bravaya et al., who rationalised the energetic shifts in different coloured photoactive proteins using a 3-centre allyl radical in a simple Hückel framework and a particle in the box model.38

More recently, experimental and computational work has been directed to the higher-lying excited states in the UV spectral region. These are believed to be important in the photooxidation of the photoactive proteins. For example, in GFP a second optically bright state of HBDI, S3, has been observed by action spectroscopy.19 Photoelectron spectroscopic measurements with complementary computational studies suggest that the S3 state leads to electron ejection from the protein to form hydrated electrons.13 More generally, frequency-resolved photoelectron spectroscopy has provided valuable insight into the dynamics of resonances in the isolated chromophores, including HBDI.13,39 High-level electronic structure calculations show that the S3 resonance is formed by promotion of an electron from the highest occupied molecular orbital (HOMO) to an unoccupied π* MO localised almost exclusively on the phenolate chromophore. In a similar vein to the interpretation of the S1 state, Bochenkova et al. viewed the MOs for the ground and excited electronic states using a cartoon Hückel picture to sketch the nature of transitions.19 While high-level calculations allow quantitative analysis of experimental data, the simple models based on Hückel theory allow qualitative analysis that provides deep insight into the nature of electronic states, which is critical in the rational design of new photoactive proteins and macromolecules. This is particularly so from a synthetic chemist's perspective who may not have the highly specialised skills required to perform high-level electronic structure calculations. In this study, we generalise and validate a Hückel picture using a bottom-up approach. We show that the first two bright states of photoactive proteins arise from interaction of two chromophore moieties and that the phenolate chromophore is essential to describing the UV response of photoactive chromophores.



The experiment has been described in detail elsewhere.40,41 Briefly, electrospray ionisation was used to produce deprotonated anions from ∼1 mM solutions of phenol, p-methyl-phenol or p-ethyl-phenol in methanol. These phenolates (Ph, MPh and EPh, respectively) were introduced to the first of several vacuum regions by a transfer capillary before progressing along a series of ring-electrode ion guides. The anions were pulsed into a collinear Wiley–McLaren time-of-flight mass spectrometer42 by an ion trap at the end of the ion guides. Mass-selected ion packets were then irradiated by a tunable ∼6 ns laser pulse from a Nd:YAG pumped optical parametric oscillator (Continuum Surelite II-10, Horizon I), and photoelectrons produced were imaged using a velocity-map-imaging assembly.43,44 The photoelectron spectrum and photoelectron angular distributions (PADs) were extracted from raw photoelectron images using the Polar Onion Peeling algorithm.45 Photoelectron spectra were calibrated using the known atomic spectrum of I and have an experimental resolution of ∼5%. The photoelectron spectra for HBDI is reproduced from a previous study using the same experimental method.33 As detailed in previous work,46p-vinyl-phenolate, VPh, was produced from CA that underwent collisional-induced dissociation to lose CO2 in the ion guide. The anionic fragment was then mass-selected in the time-of-flight mass spectrometer before being irradiated to perform photoelectron imaging.


Hückel theory calculations were performed using the HuLiS calculator.47,48 Additional electronic structure calculations of the ground and excited states of the deprotonated anions and corresponding neutral radicals were carried out using the QChem 4.4 package.49 Initial density functional (DFT) calculations optimized the geometries of the ground states of neutral and anions. These geometries were confirmed to be energetic minima by vibrational frequency analysis. Time-dependent (TD) DFT calculations, with the Tamm–Dancoff approximation,50 confirmed the character and energetics of the excited states of the neutral accessed experimentally. The results of these calculations are presented in the ESI and used the B3LYP functional51–54 and the aug-cc-pVDZ Dunning basis set.55 Further calculations qualitatively explored changes to the character of the bright excited states of a series of para-substituted phenolates from phenolate to p-hydroxybenzylideneimidasolinone (HBI), as shown in Fig. 1. These calculations were done using the B3LYP functional and the cc-pVDZ basis set, to exclude diffuse continuum states and simplify analysis of the molecular orbitals. This deliberately low level of theory was employed to allow qualitative comparison of MOs and their relative weightings to excited states with the results of the HuLiS calculator.

To model the angle-resolved data, all anions were optimized using coupled cluster singles and doubles (CCSD) method, with the same basis set as before. Using these optimized geometries the Dyson orbitals were calculated through implementation of the equation-of-motion CCSD (EOM-CCSD) formalism.56,57 Further detail on the Dyson orbitals used to model the D0 and D1 direct detachment channels accessed experimentally can be found in previous work.46 Finally, the PADs for the two direct detachment channels were calculated using the ezDyson program (version 3.2) developed by Krylov and co-workers.58


The frequency-resolved photoelectron spectra for Ph, EPh, VPh and HBDI are summarized as 3D false-colour plots in Fig. 2(a–d), respectively. MPh is presented in the Fig. S1 (ESI) for completeness. To emphasize spectral changes as a function of photon excitation energy, , the photoelectron spectra have been normalised to a maximum intensity of one. At all , the photoelectron spectra are dominated by a feature with an electron kinetic energy (eKE) that increases linearly with increasing photon energy. This direct detachment channel corresponds to electron loss from the ground singlet state of the anion to form the ground doublet state of the corresponding neutral species, S0 + → D0 + e. A second direct detachment feature for Ph, MPh and EPh, with an onset at ∼ 3.2 eV, corresponds to the S0 + → D1 + e direct detachment channel. This channel is also seen for VPh, albeit relatively weaker and shifted to ∼ 3.5 eV. It is not observable in HBDI.
image file: c7cp05766k-f2.tif
Fig. 2 Frequency-resolved photoelectron spectra of (a) Ph, (b) EPh, (c) VPh, and (d) HBDI (taken from ref. 33) are shown. All photoelectron spectra are normalised to a unit maximum. The horizontal arrows indicate the onsets of resonances in eV. For HBDI, these are taken from ref. 19. Anisotropy parameters (β2) for (e) Ph and (f) VPh as a function of electron kinetic energy (eKE). Open circles and solid lines are experimentally and computationally determined β2 (eKE), respectively. Shaded eKE regions indicate resonance dynamics that change β2 from the calculated ones. The parameters of the resonances are taken from the indicated experimental energies at which resonances are encountered shown in (a) and (c), for (e) and (f), respectively. The Dyson orbitals used to model the anisotropy computationally are inset.

In addition to the direct detachment features, spectral broadening is observed for all the anions, where photoelectron signal is observed at lower eKE than the direct detachment peak. Such shifts have been observed in the photoelectron spectra of many anionic species,39,41,59–62 including several bio-chromophore derivatives based on phenolate.8,13,22,33,36,63 The shift towards lower eKE arises from the excitation of an electronic resonance of the anion. This resonance can undergo nuclear motion which leads to changing Franck–Condon factors with the final neutral states to which it autodetaches. The spectral features are broadly similar for all the phenolates including HBDI, suggesting that a common motif is responsible for the observed dynamics. However, the onset and width of the resonance features can be seen to change most significantly between Ph and EPh to VPh and HBDI. Additionally, the cross-section of excitation to the resonance appears to be larger for the latter two as evidenced by the larger proportion of photoelectrons lost from the autodetachment channel than from the direct channel leaving the neutral in the D1 state.

The experimental data allows key properties of the phenolate derivatives to be determined. This includes the vertical detachment energy (VDE) and adiabatic detachment energy (ADE), as well as the D0–D1 gap and the onset of the resonances (tabulated in ESI 1,). The resonance onsets were determined in two ways. They can be identified by the at which the indirect autodetachment signal can be observed (see Fig. 2(a–d)). They can also be determined from sudden changes in PADs as a function of . Fig. 2(e and f) show the experimentally determined anisotropy parameters (β2) for Ph and VPh, respectively. The experimental β2 parameter as a function of eKE were computed and their functionality followed the experimental values up to an eKE where a resonance can be accessed. Note that, although we recently showed that such an approach yields almost quantitative agreement,41,46 the PADs appear to have a constant discrepancy for Ph, although the trend is correct. The Dyson orbital approach fails to capture autodetachment from resonance leading to PADs rapid deviations from calculated behaviour.41,64 This has recently been observed in AgF and CuF, where qualitative arguments were used to account for the change in anisotropy for autodetachment from resonances.65 Examples of similar behaviour for the present systems are shown in Fig. 2(e and f) for Ph and VPh, respectively. There are two regions that show discontinuities for VPh, suggesting that there are two resonances present. The onset of the resonance derived from the photoelectron spectra and the PADs agree remarkably well with each other, providing confidence of the assignment, although we note that there is no clear indication of the second resonance for VPh, except for the much greater range over which spectral broadening occurs in this anion. The onsets of observed resonances have been included in Fig. 2(a–d) as horizontal arrows. Those for HBDI are taken from the action spectrum of the Andersen group and agree with the values determined from the frequency-resolved photoelectron spectra in Fig. 2(d).

In addition to ground state calculations, excited state calculations were performed. While the absolute energies are in poor agreement, the relative energies and patterns of states are expected to be less prone to large errors and these are used to provide a basis for the interpretation of the experimental data. The first bright resonance in Ph, MPh and EPh was calculated to be independent of substituent (see Fig. S3, ESI), in agreement with experimental observations. This resonance corresponded to the promotion of an electron from the HOMO to an unoccupied antibonding MO localised on the phenol ring, πPh*. The relevant orbital calculated by DFT for Ph is shown in Fig. 3(a) and those for MPh and EPh are in the Fig. S3 (ESI). Autodetachment from the resonance correlates with a one-electron process, in which loss of the electron from the πPh* orbital produces the D0 ground neutral state. The resonance is therefore of shape character.39 According to our calculations, for VPh there are two bright excited states. Both have mixed character with a combination of a similar πPh* orbital as well as an unoccupied MO, πR* on the vinyl unit. This πR* orbital has the appearance of the π* MO of ethene, as shown in Fig. 3(b). The weights according to TD-DFT of the two excited states in terms of these orbitals are: Ψ+ = 0.60πPh* + 0.76πR* and Ψ = 0.75πPh* − 0.58πR*, where Ψ+ is lower in energy than Ψ. The presence of two resonances in VPh accounts for the large width observed for the resonance autodetachment seen in Fig. 2(c) and the changes noted in the PADs in Fig. 2(f).

image file: c7cp05766k-f3.tif
Fig. 3 Relevant molecular orbitals (MOs) and schematic MO diagram of excited states of (a) Ph, (b) VPh, (c) BPh, and (d) HBDI. The πPh* MO (left) produces a linear combination with the πR* MO (right) to produce the excited state structure of the chromophores. The neutral state, D0, is also indicated.


A Hückel interpretation of excited states

The frequency resolved photoelectron spectra for Ph, MPh and EPh are essentially identical, which allows us to conclude that the addition of an alkyl group (R) at the para position of the phenolate does not affect the πPh* resonance energies, nor does it appear to affect the autodetachment dynamics from this resonance. This can be justified in a Hückel framework as the πPh* orbital has no electron density on the oxygen or the para-carbon atom which binds to the R group. Hence, the πPh* orbital essentially corresponds to a π* orbital of benzene with a Hückel energy of ε = αβ. As it is a localised excitation, there is also no reason to expect the dynamics to differ as R changes. However, in the case of VPh and HBDI, the R group does perturb the excited state structure. The main difference in these cases is that the R group has its own π-system that is conjugated to the phenolate.

If the R group's π electron system is conjugated with the phenolate, then the HOMO becomes delocalised over the entire molecule. For VPh the HOMO calculated using DFT is shown in Fig. 3(b), while Fig. 4(b) shows the results from a simple Hückel calculation. Both calculations show the delocalised π electron nature of the HOMO. However, inspection of the Hückel HOMO allows us to approximate it as a combination of the HOMO of the phenolate (shown in Fig. 4(a)) and the allyl radical (highlighted region in Fig. 4(b)) with the para-carbon providing a common contribution. We now consider the unoccupied MOs of VPh within the Hückel framework.

image file: c7cp05766k-f4.tif
Fig. 4 Highest occupied Hückel molecular orbital (HOMO) and lowest unoccupied molecular orbital (LUMO) and πPh* and πR* MOs of (a) Ph, (b) VPh, (c) BPh, and (d) HBI. Marked insets in (b–d) show the chromophoric MOs involved in the HOMO → LUMO transition.

The lowest-lying unoccupied MO has the appearance of the ethene antibonding orbital, πR*, as shown in Fig. 4(b). The second unoccupied MO is the LUMO of phenolate, πPh*, which is localised because of the very small coefficients (i.e. almost a node) at the para-carbon. Assuming that πR* can be approximated as the ethene π* MO, then the Hückel energies of the πPh* and πR* are degenerate with an energy of ε = αβ. We can now construct a simple MO diagram involving these πPh* and πR* orbitals, as shown in Fig. 3(b). A linear combination of πPh* and πR* MOs will lead to two new MOs: Ψ± = 0.71πPh* ± 0.71πR*. This is in very good agreement with the TD-DFT calculations that yield Ψ+ = 0.60πPh* + 0.76πR* and Ψ = 0.75πPh* − 0.58πR*. The Ψ+ MO is reduced in energy compared to Ph, MPh and EPh, in good agreement with the results shown in Fig. 2. The Ψ solution is expected to increase in energy by a similar amount relative to the πPh* MO because of the nearly equal contribution of both MOs to wavefunctions. The separation of the two states is 0.5 eV from the experiment (Fig. 2(c)), suggesting a shift of ∼0.25 eV of Ψ+ and Ψ, relative to their non-interacting MOs.

In addition to the spectral agreement, the above analysis agrees with the observed experimental changes in the resonance dynamics. The autodetachment dynamics that can be ascertained from the spectral broadening appears to be similar for the two resonances of VPh. This is in line with the fact that both have equal contributions from the πPh* MO which is of shape character and is responsible for the autodetachment as seen in Fig. 2(a and b). Additionally, the transition strength appears to have increased for both Ψ+ and Ψ in VPh compared to the excitation to the pure πPh* state in Ph. This can be rationalised by the 1ππ* character in the excitation that comes about from the bright n(allyl) → π*(ethene) transition (see inset of Fig. 4(b)).

It is also noted that energy gap between the HOMO and the πPh* MO is not drastically affected by para-substitution. This can be appreciated by inspection of Fig. 3(b), which shows that the HOMO of VPh is qualitatively the same as that of Ph because the conjugated R group is non-bonded due to a central allyl node. In fact, this argument essentially remains true for all phenolates discussed here.

We now extend this Hückel analysis to larger conjugated molecules. We begin by considering 1,3-butadienylphenolate (BPh), Fig. 3(c) and 4(c). For this molecule, R can be approximated as pentadiene bound to phenolate with the common para-carbon (see Fig. 4(c) inset). The unoccupied MOs can be considered in a similar spirit as above. The πPh* MO is the same as that of VPh and Ph. The πR* MO can be viewed as the first π* MO of butadiene (see Fig. 4(c) inset), which has a Hückel energy ε = α − 0.62β. The reduction of the orbital energy of πR* relative to πPh* means that their linear combination of πR* with πPh* results in a lower energy of Ψ+, while Ψ will remain at approximately the same energy as Ψ in VPh. The linear combination of MO will lead to Ψ+ and Ψ with non-equal coefficients for the two contributing MOs. For Ψ+, the πPh* MO is expected to have the largest contribution, while for Ψ, πR* is expected to have the largest coefficient. This is in agreement with TD-DFT calculations that shown that Ψ+ = 0.00πPh* + 0.95πR* and Ψ = 0.95πPh* − 0.00πR*. Hence, the modest lowering of the πR* orbital energy has drastically altered the nature of the excited states, with the lowest being almost exclusively of n(pentadiene) → π*(butadiene) character (see Fig. 4(c) inset). Experimentally, one would expect the photoelectron spectra to have similarities to that of Ph with the πPh* resonance shifted to ∼ 3.7 eV (as for VPh). We expect the transition energy to the πPh* resonance to incur small changes because there is no significant perturbation to either the πPh* orbital nor the HOMO from which the electron is excited with increased conjugation. In addition to the πPh* resonance, we expect a lower-lying state to be observable because of the lower orbital energy of πR* in BPh.

Using the above framework, the frequency-resolved photoelectron spectra of HBDI, which have previously been recorded33 and reproduced in Fig. 2(d), can also be interpreted. For convenience, we ignore the methyl groups of HDBI and only consider HBI. With reference to Fig. 4(d), HBI can be viewed as phenolate conjugated via an allyl bridge to the imidazole ring. As before, there are two important π* orbitals: the phenolate πPh* orbital and the πR*, where R is the imidazole-ring with ethene in the 2-position (Fig. 4(d) inset). The Hückel energy of the πR* orbital is ε = α − 0.35β compared to ε = αβ for πPh*. Hence, based on a comparison with BPh, we expect the lowest-lying exited state to be almost purely of πR* character, while at higher energy we expect to see a resonance similar to that of Ph, but again shifted to ∼ 3.7 eV. This is broadly consistent with experimental observation. The energy gap between S1 and the higher lying πPh* resonance is ∼1.2 eV, based on the action spectra from Andersen and coworkers.19 The TD-DFT calculated excited states are defined by Ψ+ = 0.00πPh* + 0.95πR* and Ψ = 0.78πPh* − 0.59πR*. Hence, as anticipated from the BPh, the S1 state is almost purely of πR* character. Note that the Ψ solution appears to have a larger than expected contribution from the πR* MO. This arises from excitation of HOMO−4 (the next lower-lying πPh MO) to the πR* MO, which is not accounted for in the current simple Hückel theory picture. Nevertheless, the dynamics of the predominantly πPh* resonance in HBDI are consistent with those observed for the other systems studied here. The results from our TD-DFT calculations are consistent with the high-level calculations by Bochenkova et al.13

The analysis for HBI can be compared with the Hückel interpretation of the S1 state provided by the Krylov group.38 They took an even more reductionist view of the S1 state and argued that it can be viewed as a transition from the n(allyl) → π*(ethene). Although this view is elegant in its simplicity, a slight extension of this depiction provides a much more detailed and far-reaching picture. We suggest that HBDI can be viewed as a phenolate unit and the crotonaldehyde radical (see Fig. 4(d) inset). The S1 state corresponds to a transition from the singly-occupied orbital of the crotonaldehyde radical (which has the appearance of the n(allyl) MO) to the lowest unoccupied πR* orbital, where R = acrolein, as shown in Fig. 4(d). Hence, the S1 is best described by n(crotonaldehyde) → π*(acrolein). The bright shape resonance identified at = 3.8 eV, S3, corresponds a πPh → πPh* transition localised on the phenolate and is remarkably close to the 3.7 eV that was anticipated from the simple bottom-up picture. The slight discrepancy can be attributed to the additional mixture of the HOMO−1 → πR* transition in the Ψ excited state for HBDI. The dynamics of the S3 state have been the subject of some controversy.13,33,36 In a previous study, we argued that the frequency resolved photoelectron spectra were most consistent with autodetachment from S3 and internal conversion to the lower-lying dark S2 state which subsequently autodetached.33 However, Bochenkova et al. showed that the spectral broadening observed in Fig. 2(d) could be reproduced as an autodetachment from the S3 state, without invoking any internal conversion.13 Based on the above Hückel arguments, the fact that similar spectral broadening is seen in Ph and HBDI suggest that the dominant decay is simple autodetachment from the πPh* orbital.

In principle, the above analysis is also valid for the PYP chromophore, CA. In fact, the Hückel structure of the relevant MOs in CA is essentially identical to those of HBI! Specifically, the same n(crotonaldehyde) → π*(acrolein) is expected to describe the S1 state while the higher lying electronic resonances will be predominantly of πPh* character. Hence, one may expect similar excited state structures and dynamics for the two. A study by Andersen and coworkers measured the lower lying (S1) state of CA to be at an excitation energy of 2.9 eV, which is very close to the experimental electron affinity of 2.91 ± 0.05 eV,21 as in HBDI.14 Based on our analysis, we expect the bright resonance to lie at ∼ 3.7 eV.

Relevance of Hückel picture to excited state dynamics

Although the Hückel picture is rudimentary, it provides useful insight into the basic electronic structure of bio-chromophores involving phenolates, which have been the subject of much recent interest. As the conjugation is extended further, the πR* MO is lowered in energy relative to the πPh* MO, consistent with the particle in a box picture. This trivially explains why red fluorescent proteins absorb at longer wavelengths.38 Overall, the lower the energy of the πR* MO is relative to that of the πPh* MO, the more dominant its contribution to the lower-lying excited state and the lower excitation energy to the Ψ+ excited state.

A key result here is that the VPh chromophore is integral to building a bottom-up understanding of biochromophores, as it represents the species in which the characteristic spectroscopic properties of photoactive proteins emerge. The mixing of the two chromophore units produces the two bright states seen in several larger bio-chromophores. In particular the excited state around = 3.7 eV has been the subject of much recent interest. The autodetachment from this resonance has been discussed in terms of photo-oxidation of GFP.13,66 Our results show that this autodetachment process can be viewed quite simply as the loss of an electron from the lowest energy πPh* of the phenolate (or benzene), in agreement with high-level electronic structure calculations.13 Hence, to probe the dynamics and details of this autodetachment process, it appears not necessary to study the complex HBDI molecule as the same dynamics can be observed in Ph. This is a pleasing conclusion and shows how simple chromophores of chromophores can provide exquisite insight into the dynamics of complex bio-molecules and how a gas-phase bottom-up approach can yield genuine insight into complex molecules.

One must of course recognise that the Hückel approach has major limitations and does not provide quantitative insight. Specifically, Hückel theory only predicts the nature and energies of MOs and not excited states. For excited states with character involving more than one MO, the linear combination of MOs method used here is insightful, but can only be applied to a certain extent. When excited states are combinations of many MOs, the Hückel approach fails and high-level computational methods are required. This is likely the case for higher-lying excited states. Additionally, the Hückel approach provides little information regarding the shape of excited state potential energy surfaces, although we note that bond-orders are readily calculated in Hückel theory and yield some indication of the likely initial motion away from the Franck–Condon region. Nevertheless, there is no provision for predicting dynamics, which is often one of the most important features of a chromophore's function. Finally, the protein environment also plays a deterministic role. For example, the similarity in electronic structure of the PYP and GFP chromophores does not map onto their photophysical properties or biological function; in GFP, the protein structure inhibits isomerisation which leads to the fluorescent properties of the protein, while in PYP, it does not and isomerisation with internal conversion is the main decay mechanism. Nevertheless, we hope that the simple electronic structure and reductionist picture of photoactive protein chromophores will be useful in the development of new chromophores, particularly in synthetic laboratories that are often guided by qualitative electronic structure arguments rather than high-level ab initio calculations.


We have presented a combined frequency- and angle-resolved photoelectron and computational study of the phenolate chromophores that make up the biochromophores of photoactive proteins. A Hückel theory approach is employed to provide an understanding of the evolution of electronic structure as different para-substituents are incorporated into phenolate. For non-conjugated substituents, the electronic structure simply resembles that of bare phenolate. For conjugated substituents, a linear combination of molecular orbitals localised on the phenolate and substituent lead to the observed electronic excited states. For p-vinyl-phenolate, it is this effect that leads to the observation of a second bright excitation characteristic of chromophores in photoactive proteins. As the conjugation of the substituent increases, the lowering of its orbital energy defines the character of the first bright state, while the second bright state has predominantly phenolate π antibonding character. In these cases, the S1 excited state can be described as a transition from the non-bonding orbital on the substituent including the para-carbon of phenolate to the first π antibonding orbital of the substituent excluding the para-carbon. Our results provide an intuitive and accessible framework for the logical design of photoactive chromophores.

Conflicts of interest

There are no conflicts to declare.


This work was supported by the European Research Council (Starting Grant 306536). This work was conducted using the resources of the iOpenShell Center for Computational Studies of Electronic Structure and Spectroscopy of Open-Shell and Electronically Excited Species (

Notes and references

  1. G. H. Patterson, S. M. Knobel, W. D. Sharif, S. R. Kain and D. W. Piston, Biophys. J., 1997, 73, 2782–2790 CrossRef CAS PubMed .
  2. R. Y. Tsien, Annu. Rev. Biochem., 1998, 67, 509–544 CrossRef CAS PubMed .
  3. M. A. van der Horst and K. J. Hellingwerf, Acc. Chem. Res., 2004, 37, 13–20 CrossRef CAS PubMed .
  4. U. K. Genick, S. M. Soltis, P. Kuhn, I. L. Canestrelli and E. D. Getzoff, Nature, 1998, 392, 206–209 CrossRef CAS PubMed .
  5. S. R. Meech, Chem. Soc. Rev., 2009, 38, 2922–2934 RSC .
  6. M. Chattoraj, B. A. King, G. U. Bublitz and S. G. Boxer, Proc. Natl. Acad. Sci. U. S. A., 1996, 93, 8362–8367 CrossRef CAS .
  7. M. Zimmer, Chem. Rev., 2002, 102, 759–782 CrossRef CAS PubMed .
  8. R. F. Gunion, M. K. Gilles, M. L. Polak and W. C. Lineberger, Int. J. Mass Spectrom. Ion Processes, 1992, 117, 601–620 CrossRef CAS .
  9. C. R. S. Mooney, D. A. Horke, A. S. Chatterley, A. Simperler, H. H. Fielding and J. R. R. Verlet, Chem. Sci., 2013, 4, 921–927 RSC .
  10. J. Tay, M. A. Parkes, K. Addison, Y. Chan, L. Zhang, H. C. Hailes, P. C. Bulman Page, S. R. Meech, L. Blancafort and H. H. Fielding, J. Phys. Chem. Lett., 2017, 8, 765–771 CrossRef CAS PubMed .
  11. M. Almasian, J. Grzetic, J. Van Maurik, J. D. Steill, G. Berden, S. Ingemann, W. J. Buma and J. Oomens, J. Phys. Chem. Lett., 2012, 3, 2259–2263 CrossRef CAS PubMed .
  12. L. Lammich, J. Rajput and L. H. Andersen, Phys. Rev. E: Stat., Nonlinear, Soft Matter Phys., 2008, 78, 51916 CrossRef PubMed .
  13. A. V. Bochenkova, C. R. S. Mooney, M. A. Parkes, J. L. Woodhouse, L. Zhang, R. Lewin, J. M. Ward, H. C. Hailes, L. H. Andersen and H. H. Fielding, Chem. Sci., 2017, 8, 3154–3163 RSC .
  14. T. Rocha-Rinza, O. C. J. Rajput, A. Gopalan, D. B. Rahbek, L. H. Andersen, A. V. Bochenkova, A. A. Granovsky, K. B. Bravaya, A. V. Nemukhin, K. L. Christiansen and M. B. Nielsen, J. Phys. Chem. A, 2009, 113, 9442–9449 CrossRef CAS PubMed .
  15. L. H. Andersen and A. V. Bochenkova, Eur. Phys. J. D, 2009, 51, 5–14 CrossRef CAS .
  16. K. B. Bravaya and A. I. Krylov, J. Phys. Chem. A, 2013, 117, 11815–11822 CrossRef CAS PubMed .
  17. S. Bonsma, R. Purchase, S. Jezowski, J. Gallus, F. Könz and S. Völker, ChemPhysChem, 2005, 6, 838–849 CrossRef CAS PubMed .
  18. E. V. Gromov, I. Burghardt, H. Köppel and L. S. Cederbaum, J. Am. Chem. Soc., 2007, 129, 6798–6806 CrossRef CAS PubMed .
  19. A. V. Bochenkova, B. Klaerke, D. B. Rahbek, J. Rajput, Y. Toker and L. H. Andersen, Angew. Chem., Int. Ed., 2014, 53, 9797–9801 CrossRef CAS PubMed .
  20. J. Rajput, D. B. Rahbek, L. H. Andersen, T. Rocha-Rinza, O. Christiansen, K. B. Bravaya, A. V. Erokhin, A. V. Bochenkova, K. M. Solntsev, J. Dong, J. Kowalik, L. M. Tolbert, M. Åxman Petersen and M. Brøndsted Nielsen, Phys. Chem. Chem. Phys., 2009, 11, 9996 RSC .
  21. M. A. Parkes, C. Phillips, M. J. Porter and H. H. Fielding, Phys. Chem. Chem. Phys., 2016, 18, 10329–10336 RSC .
  22. C. R. S. Mooney, M. A. Parkes, A. Iskra and H. H. Fielding, Angew. Chem., Int. Ed., 2015, 54, 5646–5649 CrossRef CAS PubMed .
  23. I.-R. Lee, W. Lee and A. H. Zewail, Proc. Natl. Acad. Sci. U. S. A., 2006, 103, 258–262 CrossRef CAS PubMed .
  24. L. H. Andersen, A. V. Bochenkova, J. Houmøller, H. V. Kiefer, E. Lattouf and M. H. Stockett, Phys. Chem. Chem. Phys., 2016, 18, 9909–9913 RSC .
  25. P. B. Coto, D. Roca-Sanjuán, L. Serrano-Andrés, A. Martín-Pendás, S. Martí and J. Andrés, J. Chem. Theory Comput., 2009, 5, 3032–3038 CrossRef CAS PubMed .
  26. W. L. Ryan, D. J. Gordon and D. H. Levy, J. Am. Chem. Soc., 2002, 124, 6194–6201 CrossRef CAS PubMed .
  27. C. Ko, B. Levine, A. Toniolo, L. Manohar, S. Olsen, H. J. Werner and T. J. Martínez, J. Am. Chem. Soc., 2003, 125, 12710–12711 CrossRef CAS PubMed .
  28. S. Smolarek, A. Vdovin, D. L. Perrier, J. P. Smit, M. Drabbels and W. J. Buma, J. Am. Chem. Soc., 2010, 132, 6315–6317 CrossRef CAS PubMed .
  29. H. Kuramochi, S. Takeuchi and T. Tahara, J. Phys. Chem. Lett., 2012, 3, 2025–2029 CrossRef CAS .
  30. A. Toniolo, S. Olsen, L. Manohar and T. J. Martínez, Faraday Discuss., 2004, 127, 149–163 RSC .
  31. S. B. Nielsen, A. Lapierre, J. U. Andersen, U. V. Pedersen, S. Tomita and L. H. Andersen, Phys. Rev. Lett., 2001, 87, 228102 CrossRef CAS PubMed .
  32. E. Kamarchik and A. I. Krylov, J. Phys. Chem. Lett., 2011, 2, 488–492 CrossRef CAS .
  33. C. W. West, J. N. Bull, A. S. Hudson, S. L. Cobb and J. R. R. Verlet, J. Phys. Chem. B, 2015, 119, 3982–3987 CrossRef CAS PubMed .
  34. J. Lazzari-Dean, A. I. Krylov and K. B. Bravaya, Int. J. Quantum Chem., 2015, 115, 1258–1264 CrossRef CAS .
  35. A. V. Bochenkova and L. H. Andersen, Faraday Discuss., 2013, 163, 297 RSC .
  36. C. R. S. Mooney, M. A. Parkes, L. Zhang, H. C. Hailes, A. Simperler, M. J. Bearpark and H. H. Fielding, J. Chem. Phys., 2014, 140, 205103 CrossRef PubMed .
  37. A. Svendsen, H. V. Kiefer, H. B. Pedersen, A. V. Bochenkova and L. H. Andersen, J. Am. Chem. Soc., 2017, 139, 8766–8771 CrossRef CAS PubMed .
  38. K. B. Bravaya, B. L. Grigorenko, A. V. Nemukhin and A. I. Krylov, Acc. Chem. Res., 2012, 45, 265–275 CrossRef CAS PubMed .
  39. C. S. Anstöter, J. N. Bull and J. R. R. Verlet, Int. Rev. Phys. Chem., 2016, 35, 509–538 CrossRef .
  40. J. Lecointre, G. M. Roberts, D. A. Horke and J. R. R. Verlet, J. Phys. Chem. A, 2010, 114, 11216–11224 CrossRef CAS PubMed .
  41. L. H. Stanley, C. S. Anstöter and J. R. R. Verlet, Chem. Sci., 2017, 8, 3054–3061 RSC .
  42. W. C. Wiley and I. H. McLaren, Rev. Sci. Instrum., 1955, 26, 1150–1157 CrossRef CAS .
  43. A. T. J. B. Eppink and D. H. Parker, Rev. Sci. Instrum., 1997, 68, 3477–3484 CrossRef CAS .
  44. D. A. Horke, G. M. Roberts, J. Lecointre and J. R. R. Verlet, Rev. Sci. Instrum., 2012, 83, 63101 CrossRef PubMed .
  45. G. M. Roberts, J. L. Nixon, J. Lecointre, E. Wrede and J. R. R. Verlet, Rev. Sci. Instrum., 2009, 80, 53104 CrossRef CAS PubMed .
  46. C. S. Anstöter, C. R. Dean and J. R. R. Verlet, J. Phys. Chem. Lett., 2017, 8, 2268–2273 CrossRef PubMed .
  47. Y. Carissan, D. Hagebaum-Reignier, N. Goudard and S. Humbel, HuLiS Code: Lewis Embedded in Hückel Theory, 2008,
  48. Y. Carissan, D. Hagebaum-Reignier, N. Goudard and S. Humbel, J. Phys. Chem. A, 2008, 112, 13256–13262 CrossRef CAS PubMed .
  49. Y. Shao, Z. Gan, E. Epifanovsky, A. T. B. Gilbert, M. Wormit, J. Kussmann, A. W. Lange, A. Behn, J. Deng, X. Feng, D. Ghosh, M. Goldey, P. R. Horn, L. D. Jacobson, I. Kaliman, R. Z. Khaliullin, T. Kuś, A. Landau, J. Liu, E. I. Proynov, Y. M. Rhee, R. M. Richard, M. A. Rohrdanz, R. P. Steele, E. J. Sundstrom, H. L. Woodcock, P. M. Zimmerman, D. Zuev, B. Albrecht, E. Alguire, B. Austin, G. J. O. Beran, Y. A. Bernard, E. Berquist, K. Brandhorst, K. B. Bravaya, S. T. Brown, D. Casanova, C.-M. Chang, Y. Chen, S. H. Chien, K. D. Closser, D. L. Crittenden, M. Diedenhofen, R. A. DiStasio, H. Do, A. D. Dutoi, R. G. Edgar, S. Fatehi, L. Fusti-Molnar, A. Ghysels, A. Golubeva-Zadorozhnaya, J. Gomes, M. W. D. Hanson-Heine, P. H. P. Harbach, A. W. Hauser, E. G. Hohenstein, Z. C. Holden, T.-C. Jagau, H. Ji, B. Kaduk, K. Khistyaev, J. Kim, J. Kim, R. A. King, P. Klunzinger, D. Kosenkov, T. Kowalczyk, C. M. Krauter, K. U. Lao, A. D. Laurent, K. V. Lawler, S. V. Levchenko, C. Y. Lin, F. Liu, E. Livshits, R. C. Lochan, A. Luenser, P. Manohar, S. F. Manzer, S.-P. Mao, N. Mardirossian, A. V. Marenich, S. A. Maurer, N. J. Mayhall, E. Neuscamman, C. M. Oana, R. Olivares-Amaya, D. P. O’Neill, J. A. Parkhill, T. M. Perrine, R. Peverati, A. Prociuk, D. R. Rehn, E. Rosta, N. J. Russ, S. M. Sharada, S. Sharma, D. W. Small, A. Sodt, T. Stein, D. Stück, Y.-C. Su, A. J. W. Thom, T. Tsuchimochi, V. Vanovschi, L. Vogt, O. Vydrov, T. Wang, M. A. Watson, J. Wenzel, A. White, C. F. Williams, J. Yang, S. Yeganeh, S. R. Yost, Z.-Q. You, I. Y. Zhang, X. Zhang, Y. Zhao, B. R. Brooks, G. K. L. Chan, D. M. Chipman, C. J. Cramer, W. A. Goddard, M. S. Gordon, W. J. Hehre, A. Klamt, H. F. Schaefer, M. W. Schmidt, C. D. Sherrill, D. G. Truhlar, A. Warshel, X. Xu, A. Aspuru-Guzik, R. Baer, A. T. Bell, N. A. Besley, J.-D. Chai, A. Dreuw, B. D. Dunietz, T. R. Furlani, S. R. Gwaltney, C.-P. Hsu, Y. Jung, J. Kong, D. S. Lambrecht, W. Liang, C. Ochsenfeld, V. A. Rassolov, L. V. Slipchenko, J. E. Subotnik, T. Van Voorhis, J. M. Herbert, A. I. Krylov, P. M. W. Gill and M. Head-Gordon, Mol. Phys., 2015, 113, 184–215 CrossRef CAS .
  50. S. Hirata and M. Head-Gordon, Chem. Phys. Lett., 1999, 314, 291–299 CrossRef CAS .
  51. P. J. Stephens, F. J. Devlin, C. F. Chabalowski and M. J. Frisch, J. Phys. Chem., 1994, 98, 11623–11627 CrossRef CAS .
  52. A. D. Becke, J. Chem. Phys., 1993, 98, 5648–5652 CrossRef CAS .
  53. C. Lee, W. Yang and R. G. Parr, Phys. Rev. B: Condens. Matter Mater. Phys., 1988, 37, 785–789 CrossRef CAS .
  54. S. H. Vosko, L. Wilk and M. Nusair, Can. J. Phys., 1980, 58, 1200–1211 CrossRef CAS .
  55. T. H. Dunning, J. Chem. Phys., 1989, 90, 1007–1023 CrossRef CAS .
  56. C. M. Oana and A. I. Krylov, J. Chem. Phys., 2009, 131, 124114 CrossRef PubMed .
  57. C. Melania Oana and A. I. Krylov, J. Chem. Phys., 2007, 127, 234106 CrossRef PubMed .
  58. S. Gozem and A. I. Krylov, ezDyson,
  59. J. Schiedt and R. Weinkauf, J. Chem. Phys., 1999, 110, 304–314 CrossRef CAS .
  60. C. W. West, J. N. Bull, E. Antonkov and J. R. R. Verlet, J. Phys. Chem. A, 2014, 118, 11346–11354 CrossRef CAS PubMed .
  61. D. A. Horke, Q. Li, L. Blancafort and J. R. R. Verlet, Nat. Chem., 2013, 5, 711–717 CrossRef CAS PubMed .
  62. J. N. Bull, C. W. West and J. R. R. Verlet, Chem. Sci., 2015, 6, 1578–1589 RSC .
  63. D. J. Nelson, W. K. Gichuhi, E. M. Miller, J. H. Lehman and W. C. Lineberger, J. Chem. Phys., 2017, 146, 74302 CrossRef PubMed .
  64. C. W. West, J. N. Bull, E. Antonkov and J. R. R. Verlet, J. Phys. Chem. A, 2014, 118, 11346–11354 CrossRef CAS PubMed .
  65. T.-C. Jagau, D. B. Dao, N. S. Holtgrewe, A. I. Krylov and R. Mabbs, J. Phys. Chem. Lett., 2015, 6, 2786–2793 CrossRef CAS PubMed .
  66. A. Acharya, A. M. Bogdanov, B. L. Grigorenko, K. B. Bravaya, A. V. Nemukhin, K. A. Lukyanov and A. I. Krylov, Chem. Rev., 2017, 117, 758–795 CrossRef CAS PubMed .


Electronic supplementary information (ESI) available: Experimental and computational energetics; frequency-resolved photoelectron spectra of MPh; anisotropy across two direct detachment channels; excited states of MPh and EPh. See DOI: 10.1039/c7cp05766k
The small contribution of the second unoccupied π* MO of phenolate is noted but this is same in all molecules and so may be expected to have the same contribution in all the anions discussed here.

This journal is © the Owner Societies 2017