Energy transfer cascade in bio-inspired chlorophyll-a/polyacrylamide hydrogel: towards a new class of biomimetic solar cells

Abstract

We explored the energy transfer dynamics of chlorophyll-aentrapped polyacrylamide hydrogel with a vision of applying this hybrid material in a bio-inspired light harvesting system. Prominent photocurrent response was observed from a simple photovoltaic assembly prepared by encapsulating the hydrogel within two electrodes. For a better understanding of the energy transport, the hybrid systems were synthesized via two different methods (in situ and swelling induced). The difference in photocurrent efficiency among the two systems could be correlated with the different dynamic behaviors of the various excitonic packets and could be explained with environment-assisted transport of photosynthesis (ENAQT). This theory predicts that energy migration in a natural photosystem relies on the balance between coherence and dephasing, owing to the respective packing geometry. Fluorescence anisotropy supports the swollen induced arrangement of chlorophyll-a packets, which are inter-connected in terms of overlapping in site energy, as evident from the broadening of the UV-Vis spectra wherein coherent spreading occurs among around 4 chlorophyll-a molecules, as revealed from the time correlated single photon count. Exciton localization within the packets is destroyed by low frequency noise to mitigate coherence trapping in a coherent classical intermediate fashion, similar to ENAQT in natural photosynthesis, resulting in greater photocurrent efficiency. Optical and photo-physical properties of the in situ sample show that charged dimers are randomly spread throughout the system. As a result, delocalization is destroyed and energy propagation becomes less effective as the exciton is more likely to recombine than trap. The experimental signature of the environment-assisted energy transfer is further supported by a simple mathematical formulation inspired from Kassal's work. The concept of using swollen induced bio-inspired soft materials for solar energy harvesting can pave the way towards a new class of biomimetic solar cell.

---

Introduction

Natural systems have inspired human innovation and technologies over the ages. As improving solar energy conversion is one of the major scientific challenges presently, mimicking the natural photosystem is of great importance. Proper understanding and mimicking of the photosynthetic process would enable us to bridge the gap between energy demand and supply, while doing this in a clean and sustainable way.¹ Photosynthetic molecular aggregates operate in a regime between the purely coherent, or noiseless, case and the completely incoherent classical case owing to the ordering of the pigment arrangement with respect to its protein environment: primarily responsible for high quantum efficiency.² Environmentally assisted quantum transport (ENAQT)-based models claim that such a regime is necessary to achieve efficient transport.^3,4 In a purely disordered arrangement, destructive interference is much more likely to arise and the excitation could then end up being trapped between pigments. In that case, the overall transport efficiency of a disordered system in the purely coherent regime is normally rather low.^5,6 Suppression of destructive interference induced by low-frequency noise, such as dephasing, can therefore have a positive effect on transport by antagonizing trapping.⁷ System–environment interactions should modulate the whole transport regime in a truly balanced fashion.^8,9 When dynamics is dominated by coherent hopping, localization can be thought of in terms of energy conservation: an excitonic state originally localized at an initial site is a superposition of energy eigenstates that exhibit only a slight overlap with an excitonic state localized at a final state with significantly different site energy. As a result, coherent hopping on its own has a low efficiency for transporting an excitation from one site to another with significantly different site energy. Coherence causes an excitation to become ‘stuck’. It might oscillate back and forth between a few sites that are strongly coupled and have similar energies, but the exciton will never venture far afield. The dephasing comes into play by destroying the destructive coherence, dephasing also destroys the localization and allows the exciton to propagate through the system. The maximum efficiency of ENAQT occurs when the de-coherence rate is comparable to the energy scales of the coherent system, as defined by the energy mismatch between states and the hopping sites.³ Natural systems with spatially intricate superstructures provide us elegant paradigms in designing new biomimetic materials with specific functionalities to perform light harvesting in a more sophisticated and efficient way. We tried to manifest our idea of exploiting the interplay between coherence and non-coherence dynamics in a bio-inspired scheme to develop a Chl-a super-structure embedded hydrogel with enhanced energy/charge transfer dynamics by suppressing excitation recombination arising from destructive interference.

The development of plant-based bio-hybrid electrodes was pioneered during the 1980s and successful immobilization of PSI onto PtO¹⁰ and pyrolytic carbon¹¹ was achieved for directed electron transport from protein to electrode. Greenbaum et al. succeeded in chemical platinization of PSI and chloroplast followed by bio-hydrogen production.^12,13 PSI-based bio-electrodes and protein films were assembled onto metallic substrates in monolayer or multilayer forms acting as a photoactive materials for direct photocurrent generation in the mA cm⁻² regime.^14–17 In 2012, both Leblanc et al.¹⁸ and Mershin et al.¹⁹ found good efficiency in PSI-loaded p-doped silicon and ZnO electrodes, respectively. Recently, pseudo dye-sensitized solar cells with PSI have also been synthesized. In 2015, Gizzie et al.²⁰ also developed a novel solar cell utilizing solid polyaniline/PSI composite material as the photosensitizing layer on a TiO₂ semiconductor anode and an evaporated Ag cathode. TiO₂ on FTO coated glass was selected as the anode. But all of these solid state materials showed current in the μA range, including complex synthesis procedures for protein isolation and conducting polymer PSI hybrid films.

Taking these into account, in this work, we propose a scheme to increase the efficiency of natural chromophore-based simple bio-friendly light harvesting systems by mimicking the PSII antenna complex. In order to do so, we tried to stabilize Chl-a within polyacrylamide hydrogel and to correlate the light harvesting property with patterns of pigment interaction within the active material with reference to their arrangement. For the purpose of evaluating the underlying mechanism in connection with energy transfer and its dependence on Chl-a arrangements, the materials were developed both through swelling-induced and in situ methods. Hydrogels are three dimensional arrays of cross-linked biopolymers capable of encapsulating huge amounts of (∼90% of their weight) of water and biomolecules.²¹ Reaction rates and ion transport properties in “quasi-liquid” hydrogels are comparable to that in liquid media, making them a good alternative for hazardous volatile electrolytes used in conventional dye-sensitized solar cells.^22–24 Hydrogel scaffolds facilitate stabilization of degradation-prone natural fluorophores in addition to modification of absorbance spectra as a result of molecular orientation changes, inter-molecular couplings and molecule–matrix interactions, similar to pigment–protein interactions in PSII. When a soft hydrogel has swelling mismatch in different regions, an internal stress develops that can direct the orientation of flexible Chl-a molecular assemblies, arranging them through non-covalent interactions within the hydrogel matrix. Such ordered-fashion assembly can further facilitate interactions with the surrounding system of relevance by means of different modes of energy transfer.²⁵ Thorough studies of such systems may serve as a step towards biomimicking photon conversion complexes leading to fabrication of a new class of nature-inspired photovoltaic. Koo et al. achieved ∼4.7 μA cm⁻² short-circuit current and 0.37 FF using photosensitive ions embedded in the agarose hydrogel of a photovoltaic device.²⁴ The authors have already reported superfast energy hopping within Chl-a-entrapped chitosan hydrogel.²⁶ This method replaces the protein isolation procedure by simple chromatography to extract Chl-a eventually, making it cheaper. The iceberg on top is that the synthesis routes avoided involvement of any hazardous chemicals and makes it greener. The self-assembled Chl-a may have the potential to utilize the whole visible solar spectra as a result of orbital tuning similar to the antenna complexes in bacteria.²⁷ This can enhance the light harvesting property of the preliminary ensembles structure consisting of Chl-a entrapped hydrogels. Comparison between Chl-a-entrapped polyacrylamide hydrogel hybrids prepared via two different routes may lead to the realization of photocurrent enhancement due to swelling-induced dephasing and system–pigment interactions, in agreement with the ENAQT theory.²⁸ This work discusses the design of a simple, bio-inspired, soft material based photovoltaic assembly along with revealing some very interesting physical working procedures. Understanding the phenomenon can lead the path to fabrication of highly efficient, smart, biomimicking solar cells.

Materials and method

PaaChl synthesis

All the chemicals were purchased from Sigma-Aldrich and they were of analytical grade. 2 g acrylamide and 0.1 g N-N′ methylene-bis-acrylamide were mixed in 20 mL water and stirred for 30 minutes. The mixture was kept under vacuum for 30 minutes, followed by addition of 80 μL 1 M APS (ammonium persulfate) solution and 10 μL tetramethylethylenediamine (TEMED)^29,30 under continuous vacuum. pH of the solution was measured to be ∼6.5, favorable for Chl-a. The as-prepared white solution was kept under vacuum for 2 days for drying. The elastic, transparent, hydrogel was washed in water several times, wiped thoroughly, and was soaked in Chl-a overnight followed by removal of surface pigments through washing. The volume of Chl-a before and after soaking was measured and subtracted to determine the amount of Chl-a entrapped. The determined volume of Chl-a of the same concentration was used to prepare PaaChl_in situ. The concentration of Chl-a was determined from absorbance spectra using the standard formula. In the case of in situ samples, Chl-a (5 mL) was added dropwise to the liquid solution after 15 minutes of TEMED addition and was kept in desiccators for gelation. After completion of gelation, it was washed several times with water and wiped thoroughly before characterization. The whole experiment was done at room temperature (30–35 °C).

Materials characterization

UV-Vis spectra of the samples were measured using a LAMBDA 35 UV-Vis spectrometer (Perkin-Elmer). Steady state fluorescence measurements were carried out using an assembled spectrofluorometer with a 1000 W xenon source (Spectra physics, 74100). The fluorescence decay data were collected on a Hamamatsu MCP photomultiplier (R3809) and were analyzed using IBH DAS6 software. Fluorescence anisotropy measurements were carried out with a Horiba Jobin Yvon IBH, JY-IBH 5000M set up. Photocurrent responses were measured with a Keithley 238 source meter under white light illumination of 1000 W cm⁻² power. All the synthesis and characterization were performed at room temperature (30–35 °C).

Results & discussion

Light induced J–V characteristics of transparent PaaChl materials was investigated by sandwiching them in between two electrodes (ITO coated glass as the transparent working electrode on top, Ag coated substrate as the counter electrode, diagram given in Fig. 1) with an active area of ∼0.64 cm² and a thickness ∼1 mm. The photovoltaic response of the hydrogel composites, presented in Fig. 2A and B, represent a comparison of the responses through a bar diagram.

Fig. 1 Image of PaaChl/electrode assembly for photo-induced I–V characteristics measurements.

Fig. 2 A. Photo-induced J–V characteristics of both PaaChl_in situ and PaaChl_soak showing greater photocurrent response for the later. Inset: photo-induced J–V characteristics of bare Paam hydrogel. B. Bar diagram comparing different parameters retrieved from the J–V curves of PaaChl_soak and PaaChl_in situ.

The FF and external PCE of the PaaChl_soak/electrode assembly was found to be 0.45 and 0.59%, respectively, which are higher than that of the in situ counterpart (FF = 0.28 and PCE = 0.08%) (Table 1). The as-calculated values (though much lower than that of conventional solar cells) are higher than existing bio-inspired solar cells using natural chromophores or photosystems.^19,20,31,32 The overall power conversion efficiency of an organic solar cell depends not only on the component materials, but also on the fabrication technique, which includes choice of electrodes, thickness of active layers, absorbance efficiency, intermediate conducting layer, surface area of the cell and many other factors. As we fabricated a very simple ensemble, there is room for many-fold efficiency enhancements by fine tuning these factors. The high open circuit voltage achieved, in the case of the PaaChl_soak, is an attractive feature in the effort of harvesting solar energy found in the recently popular perovskite-based photovoltaics, which draws the most significant contribution from reduced molecular recombination.^33,34 This is in agreement with the signature of faster excitation dynamics in the soak sample favored by the environment in comparison to the in situ counterpart revealed from various photo-physical properties, discussed in later sections. Open circuit voltage measures the amount of voltage a cell can produce when it is sourcing no current and represents the maximum voltage of the cell, which is high in this case (1.162 V), as bio-inspired solar harvesting, in addition to the photovoltaic performance of the as prepared Chl-a/hydrogel based system, has other advantages in terms of cost and long-term stability of the active material. Sustainability of photo-current generation efficiency of the Chl-a/hydrogel hybrid was studied as a function of time, demonstrated in Fig. 3. PCE of the active material was found to retain ∼87% up to five months from the time of synthesis. The retention factor R was calculated as

where PCE_i is the initial power conversion efficiency and PCE_t is the power conversion efficiency taken at different time intervals.

Table 1 Short-circuit currents and open circuit voltages calculated from J–V curves of PaaChl prepared via two different routes depicting better performance for complexes prepared by swelling^a

Sample name	Jsc (μA cm^−2)	Voc (volt)	FF	PCE (%)
a PCE: power conversion efficiency, Jsc: short-circuit current density, Voc: open circuit voltage, FF: fill factor.
PaaChl_soak	109.18	1.162	0.45	0.59
PaaChl_in situ	28.68	0.9	0.28	0.08

---

Fig. 3 J–V characteristics curve of PaaChl_soak taken at different time intervals since the time of synthesis. Inset: PCE as a function of time.

For revealing the possible mechanism that bestows this higher efficiency in the swollen sample with respect to the in situ counterpart, various photo-physical studies were carried out. The main difference between the two systems is the preparation procedure, which could be correlated with Chl-a macrocycle ordering in the hydrogel matrix and further related to the difference in current output. This proposal is supported by absorbance spectra, fluorescence emission spectra, TCSPC and fluorescence anisotropy studies.

Compared to the swollen sample, the in situ counterpart shows greater sharpening and a shift in the visible absorbance spectra (shift to 684 nm for PaaChl_in situ, Fig. 4) with two more distinct bands around 430 nm and 446 nm. These indicate the probable presence of columnar face to face liquid crystalline-like structures within the hydrogel matrix with a greater degree of ordering and lesser inter-planar spacing of individual Chl-a molecules. But the degree of randomness of these ordered molecular-stacks throughout the gel is much higher, cumulatively giving rise to a heterogeneous population of stacked porphyrins, as supported by TCSPC and anisotropy (Fig. 5 and 6).^35,36 In contrast, lesser red shift to 675 nm, the large broadening of the Q band of the soaked hydrogel indicates intimate molecular interactions among porphyrin populations distributed through long range order among a variety of porphyrin core geometries. These reflect the presence of different conformational freedoms throughout the matrix interactions, leading to destructive interference with respect to coherence.³⁷

Fig. 4 Absorbance spectra of soaked and in situ samples.

Fig. 5 A. Steady state fluorescence emission spectra. B. Time correlated emission spectra of both soaked and in situ hybrids.

Fig. 6 A. Random nature of the anisotropy decay curve for PaaChl_in situ. B. Bi-exponentially fitted anisotropy decay curve of the PaaChl_soak. Excitation wavelength: 405 nm; emission wavelength: 676 nm for PaaChl_soak; 682 nm for PaaChl_in situ.

The fluorescence emission spectrum of PaaChl_soak with 405 nm excitation (Fig. 5A) shows a broad red-shifted (to 676 nm) peak of very low intensity. This kind of broad emission with significant quenching compared to that of the monomeric Chl-a is signature of excitation migration from one molecular assembly to another having overlapping of excitonic energy levels.³⁸ Photoluminescence spectrum of the in situ counterpart showing a higher red shift to 682 nm indicates the formation of a charged dimer complex, which is not favorable for long range excitonic migration and a major portion of the excitation dissipates through various non-radiative pathways, vibrations or get trapped.^37,39–41 Time correlated fluorescence study (TCSPC) (Fig. 5B) shows a bi-exponential decay with short average excited state lifetime for the PaaChl_soak (1.915 ns) and PaaChl_in situ (2.107 ns), respectively. The higher quantum efficiency calculated from TCSPC (Table 2) in the swollen sample can be correlated with excitonic coupling among the porphyrin assemblies, excluding the possibility of singlet annihilation. The two fluorescence decay components of swollen sample were related to conformational heterogeneity.^42,43 The ratio of these two populations (faster 0.874 ns, 46.30 Rel. Ampl. and slower 3.26 ns, 53.70 Rel. Ampl.) correlates the fact that there is a quantitative balance between populations participating in coherence and classical dynamics. We propose that the faster component arises from coherence-like excitonic interactions among Chl-a molecules within a close vicinity. The slower component, on the other hand, may be attributed to excitation migration within supramolecular clusters. This kind of balance is absent in the case of the in situ counterpart (1.12 ns, 12.24 Rel. Ampl. and 0.198 ns, 87.76 Rel. Ampl.) where a major portion of the excited state population exhibits ultrafast decay, which can be explained by the presence of strong dipolar interactions between the molecules participating in dimer formation, as suggested by the absorbance and emission spectra. Decrease in the average excited state lifetime in the swollen sample is related to the higher intermolecular coordinative strength in 3D space, although the presence of various coupled excitonic patterns are evident from the UV-Vis data (Fig. 4). The swollen-induced arrangement gives rise to a coordinative homogeneous long range molecular pattern having different excitonic probabilities. The small disorders, as a result of site-specific inhomogeneity, can induce dephasing to assist the quantum transport and increase the efficiency, further supported via mathematical formulation of dephasing induced quantum transport (eqn (1)–(9)). Another spectroscopic observable directly related to the exciton coherence is the radiative coherence length, which is 4 porphyrin units as evident from TCSPC in swollen sample. This seems to be compatible with the estimated exciton coherence length of the photosystem in the S states.²⁷ It was calculated to be 2 for PaaChl_in situ, in agreement with the absorbance and emission data where the signature of porphyrin dimer formation was found. The major red shift, along with peak sharpening, in the absorbance spectrum in polar solvents (water in our study) is also commonly attributed to dimer formation through macrocyclic π–π interactions. But the flow of excitons is restricted as the dimers are randomly arranged within the gel matrix, as evident from the anisotropy data (Fig. 6A, discussed later in details). This may restrict the excited exciton to opt for a suitable path towards the gel–electrode interface, and loses its energy in a non-radiative way or gets trapped at various sites, rather than finding a suitable path from one arranged packet to another, as evident from the faster radiative decay rate (Table 2). The randomly oriented chromophore packets may experience such a high destructive interference that cannot be corrected via low frequency dephasing, as further proved mathematically in the later section. Thus, no conclusive evidence of coherence between pigment packets can be drawn in the case of the in situ counterpart, although adjacent molecules take part in coherence with each other. On the other hand, the swelled hydrogel experiences stress-induced ordering to make long arrangements of Chl-a molecular structures that introduce dephasing assisted quantum transport along with the increase in the number of probabilistic pathways for efficient migration.

Table 2 Parameters calculated from the time correlated fluorescence decay data for both PaaChl_soak and PaaChl_in situ^a

Sample name	Avg. excited state lifetime (ns)	Quantum yield	Radiative decay rate (ns^−1)	Non-radiative decay rate (ns^−1)	Dipole strength (D)	No. of molecules in coherence
a Excitation wavelength: 405 nm; emission wavelength: 676 nm for PaaChl_soak, 682 nm for PaaChl_in situ.
PaaChl_soak	1.915	0.34	0.178	0.34	1.03 × 10^−2	4
PaaChl_in situ	2.107	0.1	0.047	0.42	0.3 × 10^−2	2

---

For fluorescence anisotropy measurements, a polarizer was placed before the sample for anisotropy measurements. The analyzer was rotated by 90° at regular intervals and the parallel (I_∥(t)) and perpendicular (I_⊥(t)) components of the fluorescence emission were alternatively collected. Then, anisotropy r(t) was calculated using the formula:

The factor G is used for experimental correction of the polarization bias of the instrument, if any. In the present case, the G value of the setup was calculated to be 0.70.

The anisotropy decay curve shows (Fig. 6) a more complex nature for the in situ synthesized hybrid rather than the swollen one. The decay curve is random and could not fit into any particular function . A dispersion of data points during first 5 ns is visible. This was followed by steady decay paths after longer times which can arise due to the presence of multiple excited states each following different depolarization processes arranged in a complete inhomogeneous manner throughout the matrix. The nature of anisotropy becomes super complex in such cases.^44,45 The initial rapid decay of different emitters for attaining quasi-equilibrium between excited states corresponds to the dispersion.⁴⁶ The relatively slower and steady decay at long time intervals arises due to molecular motion. We found bi-exponentially fitted decay parameters for the later counterpart, indicating the presence of two major porphyrin conformations.⁴⁷ The faster decay component of 55 ps, along with the high initial anisotropy value (r₀ = 0.5), may be attributed to fast resonance energy migration within strongly coupled Chl-a molecular assemblies.^48,49 Depending on the depolarization factor of the instrument, depolarization due to orientation of the absorption and emission dipoles, r₀, can vary from −0.5 to 1.⁵⁰ Strong dipolar interactions among porphyrin molecules in close vicinity gives rise to the initial anisotropy of 0.5. The longer decay component (0.433 ns) may be attributed to relaxation through dephasing of excited states, which further corroborates the presence of coherence-classical interplay among the long range ordering of Chl-a assemblies. Distance and orientation angle between adjacent dipoles taking part in resonance were calculated to be 58 A° and 33°, respectively (Table 3). The concept of swelling-induced ordering of porphyrin molecular assemblies is presented in Scheme 1.

Table 3 Parameters calculated from bi-exponential fitting of the fluorescence anisotropy decay curve of PaaChl_soak. The PaaChl_in situ could not be fitted^a

R	Exciton migration distance	τET	τRot
a R: effective radius of interaction; τET: faster decay component of anisotropy associated with energy transport; τRot: slower decay component of anisotropy associated with rotational relaxation.
61 Å	58 Å	55 ps	0.44 ns

---

Scheme 1 Schematic representation of the probable arrangement of Chl-a molecules within the Chl-a-embedded polyacrylamide hydrogel synthesized via two different methods.

Theoretical study

For further credence to the experimental results, we also carried out a simple mathematical formulation, inspired by Kassal's work.⁵¹ It reveals the effects of dephasing on the dynamics of density matrix which is directly related to charge and energy transfer. Let us consider a linear system comprising of N no. of identical molecules coupled to their respective nearest neighbors. The Hamiltonian of this system can be described as:here V is the coupling strength (considered to be 1). This Hamiltonian is equivalent to a system of coupled two level energy states with initial state m. The excitation may be irreversibly lost from each site at an equal rate ‘μ’ due to different processes, such as exciton recombination, or it can be trapped at a particular site |τ> at a rate ‘κ’. The effect of these attenuations is incorporated in the Hamiltonian as a non-Hermitian part:

Now we consider the environmental effects to be purely dephasing, acting independently on all sites with an equal rate ‘γ’. To introduce the interaction with surroundings we use the relaxation that depends on the random fluctuation of the surroundings, not the oscillatory quantum mechanical phenomenon. Therefore, the stochastic Liouville equation becomes:

or,

This general equation can be used to understand the dynamics of both the diagonal and off-diagonal elements of a reduced density matrix where R is the relaxation or dephasing operator described as:

(Rρ)_jk = −2γ(1 − δ_jk)ρ_jk (4)

Eqn (4) depicts that this dephasing factor is zero for all the off-diagonal elements of the density matrix. Let us consider the coherence between two states j & k, whose energy varies from system to system in an ensemble but are essentially constant for any given system. So, the average of the density matrix for the ensemble can be written as

where is the mean value of ρ_jk at t = 0,

E_jk = H_jj − H_kk

for a particular system.

G(E_jk) is the normalized distribution of E_jk for the ensemble expressed as:

where σ is the standard deviation and E₀ is the ground state of the system. Upon solving eqn (5), the time correlated density matrix can be expressed as:

For homogeneous systems, σ = 0; and for inhomogeneous systems, σ = 0 to 1.

Inserting this result in eqn (3) we have,

The total contribution of the dephasing operator was calculated by integrating the dephasing factor over a long time, resulting in:

∼0 for inhomogeneous systems whereσ ≠ 0 and for homogeneous systems where σ = 0.

This is in agreement with the experimental data wherein a better photocurrent generation efficiency for the PaaChl_soak was observed and the stress-induced homogeneous arrangement of Chl-a assemblies was assumed to obtain the greater contribution of dephasing operator, which in turn corrects the destructive interference towards better quantum transport, in comparison to PaaChl_in situ, which was completely random, as predicted from anisotropy and TCSPC data.

Proposed energy transfer mechanism

The whole process of energy transfer begins with photon absorbance by Chl-a molecules close to the upper surface (Fig. 1) followed by exciton generation. As suggested from the absorbance and TCSPC spectra, the excitons are delocalized over pigment molecules participating in aggregates, comparable to charge transfer excitons found in organic heterojunction solar cells. Fluorescence anisotropy studies of the PaaChl_soak shows the signature of resonance energy transfer within the molecular assemblies homogeneously arranged in the matrix. No specific indication about the exciton migration method could be achieved for PaaChl_in situ as the fluorescence anisotropy was completely random and does not follow any particular pattern. Taking into account all these results, the authors suggest the following mechanism of charge separation and transfer:

1. Excitons are produced after the absorbance of photons and delocalized among approximately 4 Chl-a molecules within packets. Adjacent pigment assemblies are arranged in such a fashion, through swollen-induced ordering, that their overlapping wave functions create probabilistic pathways for the excitons to propagate. The combined spatial and energetic landscape of the complex enables the excitation to move to the lowest energy state, no matter where the excitation starts.

2. Owing to the strong inter-molecular couplings and overlapping site-energies, the exciton migrates down the energy ladder through various Chl-a assemblies towards the target electrode.⁴⁴ Difference among electrode work functions and the built-in potential results in exciton dissociation. The free charges are collected to the respective electrodes to generate current flow.

3. The efficiency of photocurrent generation in PaaChl_in situ is much lower than that in the PaaChl_soak. In the case of the in situ sample, a major portion of the energy fails to reach hydrogel-electrode interface owing to a lack of coordination among randomly distributed excited states. On the other hand, environment-assisted homogeneous ordering of the emitting states provided a suitable energy migration pathway that resulted in higher current flow.

In summary, we propose that the swollen-induced arrangement of Chl-a packets are inter-connected in terms of overlapping in site energy, as evident from the broadening of the UV-Vis spectra. Exciton delocalization length within the packet was found to be 4 by TCSPC where coherent spreading occurs, and then it is localized within the packets and becomes resistive to further ventures. Noise destroys this localization effect and mitigates coherence trapping. The exciton then propagates in more a classical way from one packet to another, which already has overlapping in the site energy state (Scheme 2). In the in situ sample, due to the random orientation of energy states, the delocalization is destroyed and interactions between various pigment species are hampered, resulting in less effective energy propagation. The exciton is more likely to recombine than trap in this scenario.^52,53 The in situ sample shows low energy transfer efficiency. Destructive interference is high in such inhomogeneous arrays of excited states. Energy transport in the light harvesting system basically works in the regime of competition between coherent excitonic dynamics and localized trapping states optimizing the efficient energy transfer. Coherence can help avoid trapping of excitation in between two pigments (oscillatory in nature) by suppressing destructive interference and enhancing delocalization of the exciton. However, if the system is fully coherent, the charge or energy would simply oscillate and no population transfer will occur. Therefore, a natural photosystem works in a regime somewhere midway of the two, achieving the maximum transfer probability. This happens in the case of the PaaChl_soak, giving ultrafast energy transfer.⁵³

Scheme 2 Representation of probable energy and charge transfer dynamics within Chl-a embedded polyacrylamide hydrogel. Excitons are produced after absorbance of photons and delocalized among around 4 Chl-a molecules within packets. Adjacent pigment assemblies are arranged in such a fashion through swollen induced ordering that their overlapping wave functions create probabilistic pathways for the excitons to propagate. Difference among electrode work functions and built in potential result in exciton dissociation. The free charges are collected to respected electrodes, generating current flow.

Conclusions

We successfully evaluated a bio-friendly material involving natural pigments for efficient solar energy harvesting, mimicking the photosystem antenna complex. In an effort to do that, Chl-a molecules were entrapped within a polyacrylamide hydrogel and this resulted in conversion of solar energy with higher power efficiency compared to existing biomimetic photosystem-based solar cells. Light induced J–V measurements of the swollen induced hybrid yields an average photocurrent density of ∼109 μA cm⁻² with high open circuit voltage of ∼1.162 V and power conversion efficiency ∼0.59%. The feasibility of solar energy conversion is established in an approach that can bypass the involvement of any complex procedure and enhance the pigment absorption property. Within the hydrogel scaffold, the neighboring Chl-a molecules interact and form supramolecular assemblies. Excitons created upon the absorbance of light energy gets delocalized over such assemblies. Transmission of excitonic energy from one pigment packet to another occurs via resonance energy transfer. The potential difference at the hydrogel/electrode interface results in charge separation and flow of carriers to the respective electrodes. The higher power conversion efficiency compared with that of currently existing photosystem-induced solar cells can be correlated with the excitonic dynamic behavior in the hybrid material prepared via two different routes. In the two different synthesis methods, ordering of molecules throughout the hydrogel matrix varies in terms of excitonic coupling of Chl-a molecules. This in turn affects the dephasing influenced correction of destructive interference, as in a natural photosystem, to optimize the interplay between coherence and non-coherence, which gives rise to higher power conversion efficiency in the swollen counterpart when kept in between two electrodes. This phenomenon is further supported by a simple mathematical formulation inspired by Kassal's work revealing the effects of dephasing on the dynamics of the density matrix related to charge and energy transfer. This work opens up a new paradigm of fabricating biomimetic, cheap and easily synthesizable artificial light harvesting systems by coordinating the underlying mechanism with an optimized fabrication procedure.

Conflict of interest

The manuscript was written with equal contributions of all the authors. All authors have given approval to the final version of the manuscript. The authors declare no competing financial interests.

Abbreviations

Chl-a	Chlorophyll-a
Paam	Polyacrylamide hydrogel
PaaChl	Chl-a entrapped polyacrylamide hydrogel hybrid
PaaChl_soak	Hybrid prepared by soaking polyacrylamide hydrogel in Chl-a
PaaChl_in situ	Hybrid prepared by in situ synthesis
PSI/II	Photosystem I/II
TCSPC	Time correlated single photon count
Anisotropy	Fluorescence anisotropy
ENAQT	Environmentally assisted quantum transfer
PCE	Power conversion efficiency
FF	Fill factor

Acknowledgements

The authors are thankful to Prof. S. Baitalik of Department of Chemistry, Jadavpur University for his help in time correlated fluorescence measurements. Help rendered by Dr Pushan Banerjee of School of Energy Sciences, Jadavpur University in photocurrent measurements is also deeply acknowledged. Financial assistance offered to authors Pubali Mandal and Debmallya Das by UGC-UPE 2, and CSIR, Govt. of India respectively, are highly appreciated. We are also thankful to anonymous referees who have helped towards improvement of this article by giving valuable comments.

References

1. P. L. Marek and T. Silviu, Energy Environ. Sci., 2011, 2366–2378.
2. A. Ishizaki and G. R. Fleming, Annu. Rev. Condens. Matter Phys., 2012, 3, 333–361.
3. P. Rebentrost, M. Mohseni, I. Kassal, S. Lloyd and A. Aspuru-Guzik, New J. Phys., 2009, 11, 033003.
4. K. M. Gaab and C. J. Bardeen, J. Chem. Phys., 2004, 121, 7813–7820.
5. J. Cao and R. J. Silbey, J. Phys. Chem. A, 2009, 113, 13825–13838.
6. A. W. Chin, A. Datta, F. Caruso, S. F. Huelga and M. B. Plenio, New J. Phys., 2010, 12, 1–5.
7. S. Mostarda, F. Levi, D. Prada-Gracia, F. Mintert and F. Rao, Nat. Commun., 2013, 4, 2296.
8. F. Levi, S. Mostarda, F. Rao and F. Mintert, Rep. Prog. Phys., 2015, 78, 082001.
9. C. Smyth, G. Oblinsky and G. D. Scholes, Phys. Chem. Chem. Phys., 2015, 17, 30805–30816.
10. E. Y. Katz, A. Y. Shkuropatov, O. I. Vagabova and V. a. Shuvalov, Biochim. Biophys. Acta, Bioenerg., 1989, 976, 121–128.
11. E. Katz, J. Electroanal. Chem., 1994, 365, 157–164.
12. E. Greenbaum, Science, 1985, 230, 1373–1375.
13. J. W. Lee, I. Lee, P. D. Laible, T. G. Owens and E. Greenbaum, Biophys. J., 1995, 69, 652–659.
14. P. N. Ciesielski, C. J. Faulkner, M. T. Irwin, J. M. Gregory, N. H. Tolk, D. E. Cliffel and G. K. Jennings, Adv. Funct. Mater., 2010, 20, 4048–4054.
15. D. Mukherjee, M. May, M. Vaughn, B. D. Bruce and B. Khomami, Langmuir, 2010, 26, 16048–16054.
16. A. K. Manocchi, D. R. Baker, S. S. Pendley, K. Nguyen, M. M. Hurley, B. D. Bruce, J. J. Sumner and C. a. Lundgren, Langmuir, 2013, 29, 2412–2419.
17. B. Samuel Ko, B. Babcock, G. K. Jennings, S. G. Tilden, R. R. Peterson, D. Cliffel and E. Greenbaum, Langmuir, 2004, 20, 4033–4038.
18. G. Leblanc, G. Chen, E. a. Gizzie, G. K. Jennings and D. E. Cliffel, Adv. Mater., 2012, 24, 5959–5962.
19. A. Mershin, K. Matsumoto, L. Kaiser, D. Yu, M. Vaughn, M. K. Nazeeruddin, B. D. Bruce, M. Graetzel and S. Zhang, Sci. Rep., 2012, 2, 1–7.
20. E. A. Gizzie, J. Scott Niezgoda, M. T. Robinson, A. G. Harris, G. Kane Jennings, S. J. Rosenthal and D. E. Cliffel, Energy Environ. Sci., 2015, 8, 3572–3576.
21. E. M. Ahmed, J. Adv. Res., 2015, 6, 105–121.
22. S. Lin, H. Yuk, T. Zhang, G. A. Parada, H. Koo, C. Yu and X. Zhao, Adv. Mater., 2015, 1–9.
23. S. Nohara, H. Wada, N. Furukawa, H. Inoue, M. Morita and C. Iwakura, Electrochim. Acta, 2003, 48, 749–753.
24. H. J. Koo, S. T. Chang, J. M. Slocik, R. R. Naik and O. D. Velev, J. Mater. Chem., 2011, 21, 72.
25. M. Arifuzzaman, Z. L. Wu, T. Kurokawa, A. Kakugo and J. P. Gong, Soft Matter, 2012, 8, 8060.
26. P. Mandal, J. S. Manna, D. Das and M. K. Mitra, Photochem. Photobiol. Sci., 2015, 14, 786–791.
27. B. R. J. P. Grimm, W. Rudiger and H. Sheer, Chlorophylls and Bacteriochlorophylls, Springer, 2006.
28. L. Chen, P. Shenai, F. Zheng, A. Somoza and Y. Zhao, Molecules, 2015, 20, 15224–15272.
29. R. E. Rivero, M. Molina and C. Barbero, Sens. Actuators, B, 2014, 190, 270–278.
30. J. Chen and K. Park, J. Controlled Release, 2000, 65, 73–82.
31. S. A. Trammell, A. Spano, R. Price and N. Lebedev, Biosens. Bioelectron., 2006, 21, 1023–1028.
32. I. Carmeli, L. Frolov, C. Carmeli and S. Richter, J. Am. Chem. Soc., 2007, 129, 12352–12353.
33. E. Edri, S. Kirmayer, D. Cahen and G. Hodes, J. Phys. Chem. Lett., 2013, 4, 897–902.
34. W. Yang, Y. Yao and C. Q. Wu, J. Appl. Phys., 2015, 117(9), 095502.
35. X. F. Wang, O. Kitao, E. Hosono, H. Zhou, S. Sasaki and H. Tamiaki, J. Photochem. Photobiol., A, 2010, 210, 145–152.
36. L. Li, S. Kang, J. Harden, Q. Sun, X. Zhou, L. Dai, A. Jakli, S. Kumar and Q. Li, Liq. Cryst., 2008, 35, 233–239.
37. B. A. Gregg, M. A. Fox and A. J. Bard, J. Phys. Chem., 1989, 93, 4227–4234.
38. S. Vasil’ev, S. Wiebe and D. Bruce, Biochim. Biophys. Acta, Bioenerg., 1998, 1363, 147–156.
39. A. Pascal, Z. Liu, K. Broess, B. Van Oort, H. Van Amerongen, C. Wang, P. Horton, B. Robert, W. Chang and A. Ruban, Nature, 2005, 436, 134–137.
40. K. Wu, H. Zhu and T. Lian, Acc. Chem. Res., 2015, 48, 851–859.
41. K. Wu, G. Liang, Q. Shang, Y. Ren, D. Kong and T. Lian, J. Am. Chem. Soc., 2015, 137, 5–8.
42. T. V. Dracheva, V. I. Novoderezhkin and A. P. Razjivin, Photosynth. Res., 1996, 49, 269–276.
43. M. Fujitsuka, A. Okada, S. Tojo, F. Takei, K. Onitsuka, S. Takahashi and T. Majima, J. Phys. Chem. B, 2004, 108, 11935–11941.
44. A. J. Cross and G. R. Fleming, Biophys. J., 1986, 50, 507–512.
45. Z. Bajzer, M. C. Moncrieffe, I. Penzar and F. G. Prendergast, Biophys. J., 2001, 81, 1765–1775.
46. A. J. Cross and G. R. Fleming, Biophys. J., 1984, 46, 45–56.
47. D. Kim and A. Osuka, Acc. Chem. Res., 2004, 37, 735–745.
48. V. Barzda, C. J. de Grauw, J. Vroom, F. J. Kleima, R. van Grondelle, H. van Amerongen and H. C. Gerritsen, Biophys. J., 2001, 81, 538–546.
49. S. Lloyd and M. Mohseni, New J. Phys., 2010, 12, 1–6.
50. S. S. Vogel, C. Thaler, P. S. Blank and S. V. Koushik, Time-Resolved Fluorescence Anisotropy, CRC Press, 2009.
51. I. Kassal and A. Aspuru-Guzik, New J. Phys., 2012, 14, 053041.
52. M. Mohseni, P. Rebentrost, S. Lloyd and A. Aspuru-Guzik, J. Chem. Phys., 2008, 129, 174106.
53. Y. C. Cheng and G. R. Fleming, Annu. Rev. Phys. Chem., 2009, 60, 241–262.

---

Footnote

† Electronic supplementary information (ESI) available: Describing various equations used to calculate physical parameters from TCSPC and anisotropy data. See DOI: 10.1039/c6ra16780b

---