Evidence for Deep Acceptor Centers in Plant Photosystem I Crystals

Article pubs.acs.org/JPCB

Evidence for Deep Acceptor Centers in Plant Photosystem I Crystals Irina Volotsenko,† Michel Molotskii,† Anna Borovikova,‡ Nathan Nelson,‡ and Yossi Rosenwaks*,† †

Department of Physical Electronics, Faculty of Engineering, and ‡Department of Biochemistry and Molecular Biology, Faculty of Life Sciences, Tel Aviv University, Tel Aviv 69978, Israel S Supporting Information *

ABSTRACT: Dry micrometer-thick crystalline photosystem I (PSI) has been shown to generate unprecedented large photovoltage under illumination. We use variable-temperature Kelvin probe force microscopy to show that deep acceptor centers are responsible for this anomalous photovoltage. We assumed that these centers are located close to the positively charged FB2+ clusters, forming a coupled center that effectively captures the photoexcited electron into a deep state. We extract the main inherent parameters of the deep centers, which are extremely important in the potential use of photosynthetic proteins in various optoelectronic devices.

■

INTRODUCTION All higher organisms on earth receive energy directly or indirectly from oxygenic photosynthesis performed by plants, green algae, and cyanobacteria. In this process light is captured and converted into chemical energy by two photosystems photosystem II (functions in water oxidation) and photosystem I (PSI). PSI generates the most negative redox potential in nature and thus largely determines the global amount of enthalpy in living systems.1 PSI emerged as a homodimeric structure containing several chlorophyll molecules over 3.5 billion years ago, and has perfected its photoelectric properties ever since. The recently determined structure of plant PSI, which is at the top of the evolutionary tree of this kind of complexes, provided the first relatively high-resolution structural model of the supercomplex containing a reaction center and peripheral antenna complexes.2 Light energy conversion to electrical energy in photosynthetic reaction centers has attracted a lot of interest in recent years.3−5 The PSI crystals are particularly interesting due to their very large optical absorption. Recently we have revealed an extremely high photovoltage (PV) in the plant PSI crystals under laser excitation.6 The induced PV gives rise to internal electric fields as large as 100 kV/cm. This electric field is among the highest ever reported in any material system. It exceeds the field (90 kV/cm) reported for such a strong ferroelectric as LiNbO3 under similar illumination intensity.7 As mentioned in ref 8, these findings may pave the way for potential use of PSI units in organic optoelectronic devices. In the general photovoltaic effect in semiconductors, the induced PV does not exceed the value of the forbidden gapEg.9 In semiconductors it is common to use the term “anomalous” when the PV is larger than the bandgap. In PSI crystals with thickness of L ∼ 4 μm the induced photovoltage is around 40 V, which exceeds by a large factor the PV corresponding to the forbidden gap of the crystals Eg ∼ 1.9 eV. Therefore, we term © 2014 American Chemical Society

this induced PV anomalous. Note that the similar anomalous photovoltaic effect is well-known in ferroelectrics and pyroelectrics.10,11 The anomalous PV in the PSI crystals was attributed by us6 to a depletion of the P700 electron donor centers following photoionization and electron transfer to a deep acceptor center. The electrons are displaced by the large donor−acceptor distance across the PSI protein inducing a huge dipole moment in each excited complex in the crystal. Alignment of these individual photoinduced dipoles in one direction generates the crystal polarization and the unusually large internal electric field (Figure 1c). The most effective location of any acceptor center regardless of its nature is in the vicinity of the positively charged iron−sulfur center FB2+ located at the opposite side of the PSI complex (Figure 1a−c). This cannot be associated with electron localization on the center FB2+ (Figure 1b), since the characteristic time of the thermal ionization of these centers does not exceed 30 ms at room temperature.12 This time is at least 3 orders of magnitude smaller than the typical PV decay time (up to a minute at room temperature) observed in our experiments. Only a complex center will possess all of the required criteria for electrical field generation under light illumination. The positively charged center FB2+ is relatively shallow but has a very large cross-section for electron capture. Therefore, the following transfer of trapped electron from the FB2+ center to the nearby located deep trap leads to the formation of a complex trap. This trap can capture the excited electrons and also localize them into relatively stable states. Such high efficiency of electron capture on positively charged deep traps in semiconductors was studied by a few groups.13−15 Hjalmason et al.14 have shown that when the deep-trap wave Received: November 10, 2014 Revised: December 13, 2014 Published: December 26, 2014 1374

DOI: 10.1021/jp5112422 J. Phys. Chem. B 2015, 119, 1374−1379

Article

The Journal of Physical Chemistry B

Figure 1. PSI crystals. (a) Molecular image of PSI showing the carbon backbone (gray), chlorophylls (green), and electron-transfer chain (magenta) transferring charge induced under light illumination (black arrow), after Protein Data Bank (PDB) file 2O01. (b) Cofactors of the PSI electrontransfer chain and dielectric constant for each region. (c) Array structure of PSI complexes in the crystals. (d) Schematic of a home-built continuousflow cryostat with a heat exchanger integrated in KPFM. It includes a refrigerant transfer unit, a cryogenic unit, and a sample holder. At the bottom of the sample holder there is a heater unit which is connected to a temperature controller output.

Figure 2. Photovoltage as a function of illumination intensity. (a) Schematic structure of the layered PS I crystal consisting of active and inactive layers with thickness le and h, respectively. The large arrow represents the effective dipole induced under illumination. (b) Photovoltage (PV) as a function of exciting laser intensity. The measurements (solid triangles) and the PV model exhibit linear behavior at low light intensities (inset), and = 0.1 V and IC = 0.1 W/cm2. the fit of the data to eq 1 gives U(∞) 0

resistivity of 0.09 Ω·cm either with their oxidizing primary donor P700 (Figure 1a,b) or with their reducing (4Fe−4S)2+ iron−sulfur cluster ends known as the FB2+ center. The Kelvin probe force microscopy (KPFM) measurements were conducted using Dimension 3100 AFM (Bruker Inc.) operating in tapping mode at the cantilever resonance frequency of around 75 kHz in a glovebox with ultrapure nitrogen at relative humidity less than 0.01%. A silicon tip with platinum/iridium coating on both sides was allowed electrical contacts between tip and sample (high conductivity). A laser diode module (λ = 660 nm; maximum power, 60 mW) adapted with filters to obtain various light intensities was used in continuous wave mode for the PV measurements. For the

functions are predominantly host-like rather than impurity-like, the field of the isolated center with a small radius is not sufficient to create a bound state. Only coaction of short-range potential of impurity and long-range Coulomb potential “tail” causes the electron localization. These circumstances define a possible role of the charged centers FB2+ in activation of nearby located deep acceptors.

■

EXPERIMENTAL METHODS The preparation of the crystals consists of four stages: plant growth, isolation of PSI, crystallization of the complex, and stabilization in a solution containing polyethylene glycol (PEG).2,16 The crystals are attached to the SiC surfaces with 1375

DOI: 10.1021/jp5112422 J. Phys. Chem. B 2015, 119, 1374−1379

Article

The Journal of Physical Chemistry B

Figure 3. PV dependence on temperature. (a) PV as a function of temperature. (b) Normalized photovoltage kinetics for different temperatures: 300 K (magenta, circles), 290 K (green, squares), 280 K (blue, up triangles), 270 K (red, down triangles), and 260 K (black, diamonds). The PV saturates faster at higher temperatures. (c) Fit of the PV decay resulting in two types of traps: one with a short relaxation time τ(1) R and with negligible temperature dependence; the second with much larger relaxation time τ(2) R and a strong temperature dependence for low temperatures. (d) Fast and slow decaying components clearly observed at PV measured at low temperatures.

Here α is the absorption coefficient, le and h are the thicknesses of the active and inactive layers, respectively, It is the light intensity at the illuminated crystal surface, IC is a characteristic light intensity above which the PV becomes nonlinear and then saturates, and the factor

temperature dependence measurements the standard KPFM was modified by a home-built continuous-flow cryostat as schematically shown in Figure 1d. It includes a designed chuck, a cryogenic unit, a refrigerant transfer unit, and a sample holder. At the bottom of the sample holder there is a heater unit which is connected to a temperature controller output. It provides precise control and temperature stability using a proportional− integral−derivative (PID-mode) controller, Lake Shore 321. Refrigerant input and output is carried out through thin and flexible stainless steel siphons, which allow a free chuck movement and, as a result, prevent additional noises in the system.

U0(∞) =

RESULTS AND DISCUSSION In this work we use variable-temperature KPFM to study the mechanism responsible for the anomalous PV in the PSI crystals. We have used purified plant PSI crystals (Figure 1c) placed on the cleaned C-face doped SiC substrates (Xiamen Powerway Advanced Materials Co., Ltd.). Figure 2a shows the schematics of the photoexcited PSI crystal in which the electrons are generated only in the layers containing the photosynthetic reaction centers; these active layers are separated by inactive layers, in which no photoelectron transfer takes place, as shown in Figure 2a. The various deep-trap parameters are extracted from steady-state and time-dependent PV measurements conducted as a function of temperature and of laser intensity. Figure 2b shows a typical PV measured at room temperature as a function of the exciting laser intensity. For thick PSI crystal with a thickness larger than the light penetration depth (∼1/α), the total PV is given by6 ⎛ I ⎞ 1 ln⎜1 + t ⎟ α(le + h) ⎝ IC ⎠

(2)

defines the maximal contribution of the surface active layer to the total PV generated under strong illumination It ≫ IC. Here e is the elementary charge, ε is the static dielectric constant, and NS(0) is the density of the neutral acceptor centers in the surface active layer. Fitting eq 1 to the KPFM measurements (Figure 2b) with le = 5 μm, h = 5.5 μm, and α = 104 cm−1 gives the prefactor U(∞) and IC = 0.1 W/cm2. 0 The surface trap density is determined using eq 2. The potential difference between the donor and acceptor sides of the active layer depends on the constant ε within this layer. In the PSI crystal the dielectric constant changes from ∼3 for the P700−A0 protein region up to ∼6.6 for the Fx−FB area as shown in Figure 1b.17 Therefore, the effective dielectric constant of the PSI complex was estimated in proportion to the relative contribution of each region,18 and was found to be 5.2. Using eq 2 with ε = 5.2 gives a surface trap density of NS(0) = 6 × 1010 cm−2. Our crystal contains one PSI complex in the asymmetric unit, with unit cell parameters a = 120.65 Å, b = 189.09 Å, and c = 129.39 Å, and nonorthogonal angle 91.24°.18 Assuming that each PSI complex includes one primary electron donor P700, the surface density of centers then becomes Nmax = 1/(ab sin φ) = 4.4 × 1011 cm−2. S Comparing this maximum possible density with the experimentally obtained NS(0) shows that only about 14% of the centers contribute to the PV in our samples.

■

U (It) = U0(∞)

4πeNS(0)le ε

(1) 1376

DOI: 10.1021/jp5112422 J. Phys. Chem. B 2015, 119, 1374−1379

Article

The Journal of Physical Chemistry B

Figure 4. Trapping. (a) Electron-transfer pathway upon illumination in PSI crystals. The exited electron moves away from the excited state P700* across the electron transport chain toward the positive charged FB2+ center having the large cross-section for electron capture. The trapped electron has low binding energy, and it is ionized thermally within 10−30 ms at room temperature. This state is stabilized when the electron is trapped in the stable deep acceptor center located near the FB2+ cluster. The recombination pathway is depicted as direct to P700 for the sake of clarity; the actual back-pathway is through the electron transport states. (b) Schematic of electron capture in the deep trap in the PSI crystal. The electron capture from the FB+ center to the deep trap (black solid arrow) is a result of the crossing of the shallow and deep centers’ energy terms. The local vibration states of deep trap are excited as a result of transfer and then relax to the ground phonon state by the cascade of single-phonon transitions (blue dotted arrows). The thermal ionization of deep traps occurs in two steps as in the case of electron capture. First an electron moves from the stable deep state to the shallow trap and only then to the recombination pathway states. The total thermal ionization energy is 0.56 eV (red dashed arrows).

for the photocarrier generation and their capture on the deep trap, and ℏω is the incident photon energy. IC (eq 4) is a function of the cross-section for the donor photoionization σS and the quantum yield η. The product ησS can be found from the value of IC and from the PV τR obtained when the light is switched off. Using IC = 0.1 W/cm2 and τR = 60 s at room temperature in eq 4, the product ησS is around 4 × 10−20 cm2 for ℏω = 1.88 eV. σS cannot be determined since η is not known for the PSI complexes. The efficiency for electron generation by absorbed photon is close to 100%;19,20 however, only a little part of the excited electrons η are involved in a formation of the internal field. Assuming a reasonable value of photoionization cross-sections for the deep donor center, σS ∼ 10−18−10−17 cm2, which is similar to that in semiconductors,21 the quantum yield of the electron transfer from the donor to stable acceptor state is η ∼ 10−3−10−2. Figure 4a shows the proposed electron-transfer pathway in the illuminated PSI crystals based on the measured deep traps. The electron is first trapped by the positively charged FB2+ cluster,22 having a large cross-section for capture. This process can be described by the Lax model23 (see also its elaboration by Abakumov et al.24). Following these models, the electron is first trapped in highly excited states of the charged center and then losses its energy via a cascade of single-phonon radiationless transitions. However, even the ground state of the FB+ center is shallow with a lifetime of tens of milliseconds at room temperature.12 The electron binding energy and its lifetime are largely increased when the electron is trapped in the significantly deeper acceptor center located near FB+ cluster as schematically shown in Figure 4b. When the distance between the FB+ center and the deep trap is small and their wave functions strongly overlap, we refer to it as a “coupled center”. Following Henry and Lang25 vibrations of a lattice modulate the depth of the potential wells of the traps. In this model the electron capture results from lattice vibrations causing the crossing of shallow and deep centers’ energy terms. Transitions between terms can occur at not only their

The PV saturation value as a function of temperature is shown in Figure 3a. The PV decreases with increasing temperature due to the increase in the rate of the thermal ionization of the deep traps and consequently the drop in their charge. The slow PV decay (Figure 3b,d) when the light is turned off indicates the existence of deep acceptor centers. The PV characteristic decay time can be fitted in the simplest form with one decay time τR as6 ⎛ t ⎞ Uoff (t ) = Uoff (0) exp⎜ − ⎟ ⎝ τR ⎠

(3)

where Uoff(0) is the PV when the excitation is switched off. Figure 3b shows the normalized PV as a function of time for different temperatures; for the highest temperature (magenta circles, 300 K) the PV rise and decay times are much smaller than those measured at low temperatures. The PV decay curves definitely exhibit two different time constants which may be an indication of deep traps at two different energies. Figure 3d shows that the first type of traps has a short relaxation time τ(1) R of around 10 s at room temperature and does not exceed 100 s at low temperatures. The second one has a much larger relaxation time τ(2) R and a very strong temperature dependence at low temperatures: τ(2) R is 60 s at room temperature and around 1400 s at 260 K. The first type of the traps is relatively shallow acceptor centers that easily ionizes at room temperature. They cannot give significant contribution to the large PV. Because the relaxation time strongly depends on the thermal ionization energy, we argue that the second type of deep traps (Figure 3c) is responsible for the long PV decay. The characteristic light intensity (Figure 2b) is6

IC =

ℏω ησSτR

(4)

where τR is the charge relaxation time governed by thermal ionization of the deep acceptor, σS is the photoionization crosssection of the P700 donor centers, η is the quantum efficiency 1377

DOI: 10.1021/jp5112422 J. Phys. Chem. B 2015, 119, 1374−1379

Article

The Journal of Physical Chemistry B intersection but also a little below due to quantum tunneling.24 The electron-transfer process from the shallow to the deep trap is shown in Figure 4b (black solid arrow); this electron transfer is a thermally activated process with an activation energy WFD. As a result, the deep trap is excited to a high local vibration state; as the energy difference between the intersection energy and the ground-state energy of the deep trap is significantly larger than the phonon energy, the electron cascades from the excited local vibration state of the deep trap to its ground state via a multiphonon process24,25 as shown in Figure 4b (blue dotted arrows). These vibrations are then dispersed in the PSI crystal, and the acceptor center relaxes into its ground electron and phonon states. The back-electron-transfer from the ground state of the deep trap to the shallow FB center is a thermally activated process with an activation energy WA (Figure 4b, red dashed arrows). The thermal ionization of the deep center as well as electron capture includes two steps. The electron is first excited from the deep trap to the nearby FB2+ center. The probability for backtransition from FB+ to the deep trap is small due to an activated nature of electron capture in the deep trap. An electron that does not move from the shallow FB+ center to the deep trap is transferred to the states of the recombination pathway (Figure 4b) within 10−30 ms,12 moves across the complex through this pathway, and recombines with the P700+ donor center that was ionized upon illumination. WA is extracted from the PV decay curve of Figure 3c. Using an Arrhenius’s dependence of the electron lifetime (of the second type traps) τ(2) R ∝ exp(WA/kBT) on temperature, we obtain an activation energy of 0.56 ± 0.03 eV. To the best of our knowledge, the deep traps inside PSI crystals have not been studied; nevertheless this value is considered reasonable since it is similar to the thermal activation energy of about 0.7 eV reported by Krause et al.26 for the electron transfer from secondary acceptor ferredoxin to the primary acceptor FB cluster19 in the PSI complex of living plants (Chlorella algae). Tikhonov27 has noted that the energy required for the same electron transfer in plant cells is 0.3−0.4 eV. In summary, we have shown that the anomalously large PV in PSI crystals is due to electrons trapped in deep centers located in the vicinity of the shallow positive charged FB2+ clusters. The thermal ionization energy, concentration, and probability of the excited electron transfer from the donor to the deep acceptor traps were determined. In spite of the fact that the molecular nature of these centers is still unclear, their existence may help to understand a wide range of phenomena related to photoexcited photosynthetic proteins.

■

(2) Amunts, A.; Toporik, H.; Borovikova, A.; Nelson, N. Structure Determination and Improved Model of Plant Photosystem I. J. Biol. Chem. 2010, 285, 3478−3486. (3) Gur, I.; Fromer, N. A.; Geier, M. L.; Alivisatos, A. P. Air-Stable All Inorganic Nanocrystal Solar Cells Processed from Solution. Science 2005, 310, 462−465. (4) Blankenship, R. E.; et al. Comparing Photosynthetic and Photovoltaic Efficiencies and Recognizing the Potential for Improvement. Science 2011, 332, 805−809 (full reference is given in the Supporting Information). (5) Lee, J.-H.; Lee, J. H.; Lee, Y. J.; Nam, K. T. Protein/Peptide Based Nanomaterials for Energy Application. Curr. Opin. Biotechnol. 2013, 24, 599−605. (6) Toporik, H.; Carmeli, I.; Volotsenko, I.; Molotskii, M.; Rosenwaks, Y.; Carmeli, C.; Nelson, N. Large Photovoltages Generated by Plant Photosystem I Crystals. Adv. Mater. 2012, 24, 2988−2991. (7) Glass, A. M.; von der Linde, D.; Negran, T. J. High-Voltage Bulk Photovoltaic Effect and the Photorefractive Process in LiNbO3. Appl. Phys. Lett. 1974, 25, 233−235. (8) Toporik, H.; Carmeli, I.; Volotsenko, I.; Molotskii, M.; Rosenwaks, Y.; Carmeli, C.; Nelson, N. Biomaterials: High-Voltage Plant Proteins. Nature 2012, 485, 282−283. (9) Pankove, J. I. The Anomalous Photovoltaic Effect. Phys. Status Solidi A 1980, 61, 127−132. (10) von Baltz, R. Theory of the Anomalous Bulk Photovoltaic Effect in Ferroelectrics. Phys. Status Solidi B 1978, 89, 419−429. (11) Fridkin, V. M.; Popov, B. N. Anomalous Photovoltaic Effect in Ferroelectrics. Phys.-Usp. 1978, 21, 981−991. (12) Baba, K.; Itoh, S.; Hastings, G.; Hoshina, S. Photoinhibition of Photosystem I Electron Transfer Activity in Isolated Photosystem I Preparations with Different Chlorophyll Contents. Photosynth. Res. 1996, 47, 121−130. (13) Pantelides, S. T.; Sah, C. T. Theory of Localized States in Semiconductors. I. New Results Using an Old Method. Phys. Rev. B 1974, 10, 621−637. (14) Hjalmarson, H. P.; Vogl, P.; Wolford, D. J.; Dow, J. D. Theory of Substitutional Deep Traps in Covalent Semiconductors. Phys. Rev. Lett. 1980, 44, 810−813. (15) Hui, H. T. An Improved Effective Mass Theory Equation for Phosphorus Doped in Silicon. Solid State Commun. 2013, 154, 19−24. (16) Amunts, A.; Ben-Shem, A.; Nelson, N. Solving the Structure of Plant Photosystem I Biochemistry is Vital. Photochem. Photobiol. Sci. 2005, 4, 1011−1015. (17) Semenov, A. Y.; Mamedov, M. D.; Chamorovsky, S. K. Photoelectric Studies of the Transmembrane Charge Transfer Reactions in Photosystem I Pigment-Protein Complexes. FEBS Lett. 2003, 553, 223−228. (18) Ben-Shem, A.; Nelson, N.; Frolow, F. Crystallization and Initial X-ray Diffraction Studies of Higher Plant Photosystem I. Acta Crystallogr., Sect. D: Biol. Crystallogr. 2003, 59, 1824−1827. (19) Chitnis, P. R. Photosystem I. Plant Physiol. 1996, 111, 661−669. (20) Chitnis, P. R. PHOTOSYSTEM I: Function and Physiology. Annu. Rev. Plant Physiol. Plant Mol. Biol. 2001, 52, 593−626. (21) Ridley, K. Quantum processes in semiconductors; Oxford University Press: Oxford, U.K., 1999. (22) Vassiliev, I. R.; Antonkine, M. L.; Golbeck, J. H. Iron-Sulfur Clusters in Type I Reaction Centers. Biochim. Biophys. Acta 2001, 1507, 139−160. (23) Lax, M. Cascade Capture of Electrons in Solids. Phys. Rev. 1960, 119, 1502−1523. (24) Abakumov, V. N.; Perel, V. I.; Yassievich, I. N. Nonradiative Recombination in Semicondactors; North-Holland: Amsterdam, 1991. (25) Henry, C. H.; Lang, D. V. Nonradiative Capture and Recombination by Multiphonon Emission in GaAs and GaP. Phys. Rev. B 1977, 15, 989−1016. (26) Krause, H.; Kretsch, G.; Gerhardt, V. Differential Equations for Delayed Fluorescence Kinetics in Living Plants. J. Lumin. 1984, 31− 32, 885−887.

ASSOCIATED CONTENT

S Supporting Information *

Complete ref 4. This material is available free of charge via the Internet at http://pubs.acs.org.

■

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. Notes

The authors declare no competing financial interest.

■

REFERENCES

(1) Nelson, N. Photosystems and Global Effects of Oxygenic Photosynthesis. Biochim. Biophys. Acta 2011, 1807, 856−863. 1378

DOI: 10.1021/jp5112422 J. Phys. Chem. B 2015, 119, 1374−1379

Article

The Journal of Physical Chemistry B (27) Tikhonov, N. The Molecular Energy Converters in Living Cells. Soros Educ. J. 1997, 7, 10.

1379

DOI: 10.1021/jp5112422 J. Phys. Chem. B 2015, 119, 1374−1379