Vibronic Origin of the Qy Absorption Tail of Bacteriochlorophyll a

Aug 18, 2017 - Institute of Physics, University of Tartu, W. Ostwald Str. 1, Tartu ... Institute of Molecular and Cell Biology, University of Tartu, R...
0 downloads 0 Views 545KB Size
Subscriber access provided by UNIVERSITY OF ADELAIDE LIBRARIES

Letter y

Vibronic Origin of the Q Absorption Tail of Bacteriochlorophyll a Verified by Fluorescence Excitation Spectroscopy and Quantum Chemical Simulations Kristjan Leiger, Juha Matti Linnanto, and Arvi Freiberg J. Phys. Chem. Lett., Just Accepted Manuscript • DOI: 10.1021/acs.jpclett.7b01704 • Publication Date (Web): 18 Aug 2017 Downloaded from http://pubs.acs.org on August 20, 2017

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

The Journal of Physical Chemistry Letters is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 17

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry Letters

Vibronic Origin of the Qy Absorption Tail of Bacteriochlorophyll a Verified by Fluorescence Excitation Spectroscopy and Quantum Chemical Simulations Kristjan Leigera, Juha Matti Linnantoa, Arvi Freiberg*a,b aInstitute of Physics, University of Tartu, W. Ostwald Str 1, Tartu 51011, Estonia bInstitute of Molecular and Cell Biology, University of Tartu, Riia 23, Tartu 51014, Estonia * E-mail: [email protected].

1

ACS Paragon Plus Environment

The Journal of Physical Chemistry Letters

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ABSTRACT: The long-wavelength tail of the lowest-energy Qy singlet absorption band of bacteriochlorophyll a in triethylamine peaking at 768.6 nm was examined by means of fluorescence excitation spectroscopy at ambient temperature of 22±1 °C. The tail, usually considered a Gaussian, is in fact weakening quasi-exponentially, being clearly evident as far as 940 nm, nearly 2400 cm-1 (~12 kBT) away from the absorption peak. Quantum chemical simulations identified vibronic transitions from the thermally populated normal modes and their overtones in the ground electronic state as the origin of this tail. Since energy transfer and relaxation processes generally depend on vibronic overlap integrals, these findings may have important implications on the interpretation of numerous photo-induced phenomena that involve bacteriochlorophyll and similar molecules, including photosynthesis.

2

ACS Paragon Plus Environment

Page 2 of 17

Page 3 of 17

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry Letters

Bacteriochlorophylls are ubiquitous in bacterial photosynthesis, as are chlorophylls in photosynthesis of plants and algae. The outstanding position of these molecules in one of the nature’s most important process relies on their unique optical and redox properties, which are still under intense experimental and theoretical scrutiny. In isolated molecules the couplings between electrons and nuclei determine the shape and structure of their optical spectra. In the solvent phase, the added guest-host interactions, basic origin of the inhomogeneous or disorder broadening, further modify the spectra. Here, we are interested in the detailed shape of the Qy singlet absorption spectrum of the bacteriochlorophyll a (BChl a) molecule at ambient (i.e., at physiological) temperatures. As the lowest-energy singlet electronic transitions in the near-infrared spectral range, the Qy vibronic transitions are instrumental in all photophysical and photochemical processes in which the BChl a molecules are involved. The shape of the Qy absorption band, a convolution of the disorder broadening and a homogenous vibronic lineshape, is usually considered to be a Gaussian, i.e., declining rapidly towards low energies. However, there is mounting literature evidence, in detail described in a topical review,1 that certain low-energy states of chlorophylllike molecules lead to functional operation of photosynthetic units, bypassing the established higher-energy mechanisms. Furthermore, a weak up-converted emission was recently observed at ambient temperature from the bacterial light-harvesting complexes containing multiple BChl a chromophores upon the 1064 nm laser excitation, very far from the usual absorption range of these samples.2 This emission, detectable even from whole bacterial cells, was linearly dependent on the excitation intensity and had the same spectral shape as the typical fluorescence accompanying the decay of the Qy excited state. Though a different explanation was favored in Ref. 2, later experiments with tunable laser (data under preparation) indicated that the observed

3

ACS Paragon Plus Environment

The Journal of Physical Chemistry Letters

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

emission could be due to common fluorescence of excitonically coupled BChl a chromophores that were directly excited via their weak absorption tails. The above findings strongly suggest that absorption bands of photosynthetic complexes may stretch significantly more towards the red side of the optical spectrum than commonly believed. Because energy transfer and relaxation generally depend on the overlap between absorption and fluorescence spectra, the extended lineshapes may have important implications on virtually all aspects of photosynthetic primary processes. It thus appears very timely to pay more thorough attention to the spectral lineshapes of photosynthetic complexes. As a first step toward this goal, we looked at the absorption lineshape of the pure solvated BChl a. The simpler system allows more involved computational modeling, hence more complete analysis of the observed spectra, also avoiding the typical complications of the photosynthetic pigment-protein complexes such as the presence of different chromophores and the strong inter-chromophore interactions. The common optical absorption spectroscopy, being usually limited to 3-4 orders of magnitude of relative absorbance, is not well suited for the very low level absorption measurements we are presently seeking. Yet for the lowest-energy transition and emitting molecule like in the current case, one can instead apply the zero-background fluorescence excitation spectroscopy, which, compared with the high background level absorption technique, excels with much higher sensitivity and dynamic range. Superb selectivity is yet another advantage of the fluorescence excitation method because specifically emitting molecules can be easily distinguished from the bulk of different emitting or non-emitting molecules. Figure 1 shows an overview of absorptance (1 – transmission) and fluorescence excitation spectra of BChl a dissolved in dried triethylamine (TEA). The two representations of Figure 1 with ordinary linear (A) and logarithmic (B) intensity scales were used to highlight strong (A) and weak (B) signal regions of the spectra. In all cases the excitation spectrum was fitted to the 4

ACS Paragon Plus Environment

Page 4 of 17

Page 5 of 17

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry Letters

normalized absorptance spectrum determined for the same sample as described in the Methods section. The long-wavelength edge of the experimental fluorescence excitation range of 720940 nm was determined by the strength of the fluorescence signal for recording with reasonable signal-to-noise ratio.

Figure 1. The normalized absorptance (black solid line) and fluorescence excitation (red symbols) spectra of BChl a in dried TEA measured at ambient temperature of 22±1 ºC. The signal intensity scale is linear in part A and logarithmic in part B to emphasize the strong peak and weak tail regions of the spectra, respectively. Note also the reciprocal (linear in energy) wavelength scale. The insert shows the conventional absorption (blue) and fluorescence (red, corrected for spectral sensitivity of the set-up) spectra of BChl a in the same solvent. Color bars at the bottom part of the insert schematically demarcate the monitored fluorescence ranges in the

5

ACS Paragon Plus Environment

The Journal of Physical Chemistry Letters

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

blue, red and intermediate (shown as green) regions of excitation spectrum measurements, as described in Methods part. Figure 1A demonstrates a reasonable overlap of the absorptance and fluorescence excitation spectra around the spectral peak. Beyond about 820 nm both spectra are practically indistinguishable from the background noise. The peak wavelength of 768.6 nm for the dried sample is nearly 3 nm blue shifted compared to the earlier published value,3 see discussion below. Figure 1B clearly visualizes the weak quasi-exponentially falling red edge tail of the fluorescence excitation spectrum. It is absolutely clear from this figure that the BChl a absorption does not stop at 940 nm where the signal is over 6 orders of magnitude weaker than it is at 768.6 nm, the maximum of the Qy transition, but continues well beyond this mark. This agrees with the fact of observing the up-converted fluorescence of BChl a upon the 1064 nm laser excitation in Ref. 2 (see also Figure 2). In dry TEA at ambient temperature the individual BChl a molecules are penta-coordinated, i.e., their central Mg atoms are bound to one solvent molecule as an axial ligand.3–5 The BChl a molecules as co-factors in almost all bacterial light-harvesting complexes are similarly pentacoordinated. Traces of water in TEA are known to initiate the growth of larger oligomers by binding the BChl a-s together via hydrogen bonding.6,7 Such oligomeric complexes generally show modified spectral shapes compared to isolated molecules. In our experience, both the absorptance and fluorescence excitation spectra of BChl a dissolved in previously un-dried TEA, as supplied by the producer, demonstrated a 2-3 nm red shifted maximum while the shape and falling rate of the red edge tail generally did not change (spectra not shown). We thus explain the above maximum position difference with the earlier published data3 with different amount of trace water molecules in the respective samples.

6

ACS Paragon Plus Environment

Page 6 of 17

Page 7 of 17

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry Letters

In order to theoretically simulate the experimental lineshape, one needs to understand its physical origin. We propose that the extended long-wavelength absorbance of BChl a is due to vibronic transitions from the thermally populated (hot) vibrational states of the ground electronic state to the vibrational states of the Qy singlet excited electronic state. This qualitative notion is experimentally well justified by the familiar from the literature (see e.g., Ref. 3) narrowing of the Qy absorption and fluorescence spectra upon cooling of the sample. We specifically made sure that this narrowing also concerned the red edge tail of the fluorescence excitation spectrum. The major effect upon cooling down to 200 K was an increasing negative slope of the red edge tail when presented in the half-logarithmic scale of Figs. 1B and 2 (data not shown). Such behavior is expected in case of cooling out of the Boltzmann population of different normal vibrations as well as their overtones, see below.

Figure 2. Comparison of the measured BChl a excitation spectrum (red symbols, the same as in Fig. 1) with the calculated absorption spectrum (black solid line). Linear extrapolation of the experimental dependence is shown by dashed red line. The calculations took into account 315 fundamental modes together with their 20 overtones. See text for further details.

7

ACS Paragon Plus Environment

The Journal of Physical Chemistry Letters

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

An absorption spectrum at 22 ºC calculated on the above assumptions is shown by the solid black line in Figure 2, as fitted to the experimental absorption spectrum. The latter spectrum is recalculated from the actually measured fluorescence excitation, i.e. absorptance spectrum. A qualitatively good correlation between the experimental and calculated spectra is observed, although the experimental signal tends to be systematically weaker in the far-red tail region past ~870 nm. To understand this latter discrepancy, we will briefly comment on some aspects of our theoretical modeling, see Methods for further details. A methylbacteriochlorophyllide a (MeBChlide a) molecule that lacks the long phytyl tail was used to model vibronic transitions of BChl a in TEA. This truncation is justified by former quantum chemical calculations8,9 which demonstrated that the lowest singlet excited electronic state of BChl a originates from the transitions between the molecular orbitals contributed just by the atoms of bacteriochlorin skeleton. Also, experimental studies have shown that (B)Chls and the corresponding methyl(bacterio)chlorophyllide derivatives have similar electronic absorption and vibrational spectra.10–12 It was further confirmed that in both the optimized electronic ground state and Qy excited state geometries a TEA molecule is coordinated to the central Mg atom of the MeBChlide a molecule. From the two possible penta-coordinated MeBChlide a : TEA 1:1 complexes the so called β-TEA coordination complex (according to IUPAC nomenclature of tetrapyrroles13) has the lowest energy, i.e. the highest stability. This energy is furthermore about 6 kcal/mol lower than the total energy per TEA in the hexa-coordinated 1:2 complex.14 Thus the β-TEA coordination complex structure was chosen for vibronic calculations in the present work. Vibronic intensities as determined by the Franck-Condon (FC) factors were calculated for 315 different modes and their overtones. No combination modes have been included. The sum of the FC factors is normalized to keep the total magnitude of an electronic transition moment 8

ACS Paragon Plus Environment

Page 8 of 17

Page 9 of 17

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry Letters

unchanged. This, however, defines that the values of the FC coefficients depend on the size of the vibrational space (number of the modes and their overtones) used in the calculations. It was checked that at least 15 overtones of each mode were required to simulate the red side of the Qy absorption band at 296 K. Increasing the number of overtones up to 20 did not have any significant effect on the shape of the calculated spectrum. Figure 3 shows a bar-code energy distribution of the applied vibrational modes in the ground electronic state. It can be seen that while the fundamental modes have significant gaps in energy distribution, inclusion of overtones creates a virtual continuum of the states, enabling the smooth red absorption tail observed. A more detailed analysis of the FC factors revealed that overtones of only 70 lowest-energy normal modes (out of total 315 modes considered) primarily contribute to the long-wavelength tail of the absorption spectrum, while a small “hunch” apparent around 840-850 nm takes its origin from overtones of the lowest 25 normal modes. The analysis also established that the vibrational overtones forming the long tail are mainly localized on the 1-pyrrole ring or on adjacent pyrrole rings of the bacteriochlorin macrocycle. As already noted, the theoretical and experimental spectra in Fig. 2 match each other reasonably well up until ~870 nm. Past this wavelength the experimental spectrum increasingly deviates from the calculated spectrum. We believe that the main reason for this discrepancy is the omission of combination modes in the model. This results in some overestimation of the FC factor values corresponding to absorption at very long wavelengths. Yet another shortcoming of the theory is too small a model system. The MeBChlide a surrounded with larger number of TEA molecules may produce a number of low-frequency modes, which, when coupled to the Qy electronic transition, may similarly affect the calculated FC coupling strengths.

9

ACS Paragon Plus Environment

The Journal of Physical Chemistry Letters

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 10 of 17

Figure 3. Distribution in energy of the vibrational modes in the ground electronic state of the MeBChlide a - TEA complex used in the calculations: 315 fundamental modes on the left hand side and all the modes (fundamental (highlighted in red) plus their 20 overtone modes (black)), which have energy below 3000 cm-1 on the right hand side. The inset highlights energy range of the 70 lowest-energy fundamental modes mainly responsible for the long-wavelength absorption tail, see text for details.

These and other improvements of the theory are in order, becoming a subject matter of one of our forthcoming publications. Yet these future advances will hardly overturn the main findings of the present work that (i) the solvated BChl a at ambient (physiological) temperature has an extremely long Qy absorption tail, which (ii) quasi-exponentially decreases towards the red side

10

ACS Paragon Plus Environment

Page 11 of 17

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry Letters

of the spectrum, and that (iii) this tail is mostly contributed by the absorption from up to 20 thermally populated (hot) overtone levels of roughly 70 intra-molecular normal vibrational oscillators in the ground electronic state. The knowledge of the precise shape and extension of spectral bands of BChl a is important, both fundamentally and practically, for better understanding of photophysics and photochemistry of this important molecule for life as well as for a full comprehension of the energy transfer and relaxation properties within or between the bacterial

photosynthetic

complexes

by

generalized

exciton

or

Förster

mechanisms,

respectively.15,16

Experimental and computational methods. Bacteriochloropyll a molecules from Sigma were without further purification dissolved in TEA (Sigma) to the optical density of about 0.2 at the Qy absorption maximum in a 1 mm path length cuvette. This corresponds to about 2.8 x 10-5 M (or 3.8 x 10-6 mol/mol) concentration of BChl a.17 The solvent was dried by storing on molecular sieves (3Å 1/16 by Wako) for at least 24 h prior to use. The measurements with spectral resolution of 1.5 nm were carried out at ambient temperature of 22±1 ºC using a home-built microspectroscopy system18 composed of an Olympus IX-71 inverted microscope furnished with a Mitutoyo 10x long-working-distance objective (NA = 0.28) and an Andor Shamrock 303i spectrometer equipped with an Andor spectroscopic camera iDus 420. The absorption spectrum was measured applying a standard tungsten halogen illumination lamp. The fluorescence spectrum was measured using a 594 nm He-Ne laser (Melles-Griot 25-LYR-173-230) as the excitation source reflected by a 594 nm dichroic mirror; the same mirror together with a 650 nm longpass filter (Andor) were used to clean the fluorescence from laser scattering. The measured spectrum was corrected for the spectral sensitivity of the detection system. 11

ACS Paragon Plus Environment

The Journal of Physical Chemistry Letters

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 12 of 17

The source for fluorescence excitation measurements was a Model 3900S continuous wave Ti: sapphire laser (tuning range 700-1020 nm, bandwidth < 0.5 cm-1) pumped by an 8W Millennia Nd:YAG laser (all Spectra Physics). The excitation was delivered to the system via a 400 µm fiber (Thorlabs) with collimators at each end. Additional two lenses in telescope arrangement were used to tune the beam collimation and excitation spot size. The excitation power applied was adjusted to keep the recorded signal in the linear spectroscopy range; actual power values varied from 1 µW to 2 mW, depending on the wavelength. Given the spot diameter of 200 µm, the excitation density at the focal plane did not exceed 0.2 W/cm2, or