key: cord-0105004-xubxdq4k authors: Holmes, Joseph; Sushma, Arathi Anil; Tsvetkova, Irina B.; Schaich, William L.; Schaller, Richard D.; Dragnea, Bogdan title: Ultrafast Collective Excited State Dynamics of a Virus-supported Fluorophore Antenna date: 2022-02-03 journal: nan DOI: nan sha: 630d365fc8eba16d8cf7272f4aa2ab0719e0b269 doc_id: 105004 cord_uid: xubxdq4k Radiation brightening was recently observed in a multi-fluorophore-conjugated brome mosaic virus (BMV) particle, at room temperature under pulsed excitation. Based on its nonlinear dependence on the number of fluorophores, the origins of the phenomenon were attributed to a collective relaxation. However, the mechanism remains unknown. We present ultrafast transient absorption and fluorescence spectroscopic studies which shed new light on the collective nature of the relaxation dynamics in such radiation-brightened, multi-fluorophore particles. Our findings indicate that the emission dynamics is consistent with a superradiance mechanism. The ratio between the rates of competing radiative and non-radiative relaxation pathways depends on the number of fluorophores per virus. We also discuss the evidence of coherent oscillations in the transient absorption trace from multi-fluorophore conjugated which last for $sim100$s of picoseconds, at room temperature. The findings suggest that small icosahedral virus shells provide a unique biological scaffold for developing non-classical, deep subwavelength light sources, and may open new realms for the development of photonic probes for medical imaging applications. picoseconds, at room temperature. The findings suggest that small icosahedral virus shells provide a unique biological scaffold for developing non-classical, deep subwavelength light sources, and may open new realms for the development of photonic probes for medical imaging applications. High-contrast luminescent nanoprobes enable a myriad applications including biological detection, 1 therapeutics, 2,3 sensing, 4,5 optogenetics, 6, 7 and anti-counterfeiting. 8, 9 For the vast majority of current probes, radiation is the result of random, spontaneous relaxation. Consequently, emission dynamics obeys the classical exponential decay. 10 Since background emission has similar dynamics, time-domain background removal to improve contrast is seldom a viable option. Increasing the number of emitters per nanoprobe to augment brightness generally results in self-quenching due to inter-emitter distances becoming short enough ( 5 nm) for efficient resonant energy transfer to occur. Both challenges could be addressed by constructing a multi-emitter nanoprobe with correlated, non-classical emission. In this letter we present experimental evidence for this behavior, which occurs after pulsed excitation of a dense array of hundreds of fluorescent dyes, deterministically-arranged on a 28 nm diameter icosahedral virus template. Instead of the several ns exponential decay expected from individual fluorophores in solution, emission from a multi-fluorophore virus particle, at saturation coverage, occurs as a short burst of ≈ 20 ps. Peak intensity is attained at ≈ 25 ps after the ultrafast excitation pulse. Instead of being nearly quenched like under cw excitation, the estimated quantum yield in burst-mode is comparable to that of free, individual fluorophores. These unusual characteristics occur at room temperature, in a biocompatible setting. Therefore, such viromimetic probes are promising to overcome some of the current limitations of classical biophotonic probes and open new venues in fluorescence imaging. Radiation brightening from dye-conjugated fluorescent virus-like particles (fVLPs) was first reported by Tsvetkova et al . 11 It was found that when the fluorophores are conjugated with reactive residues of the brome mosaic virus (BMV) capsid interfaces, emission by the complex is strongly accelerated with respect to that of the free dye. 11 The brightening effect was found to be a nonlinear function of N , the average number of fluorophores per particle. Thus, the origins of the phenomenon were attributed to a collective effect. 11, 12 However, emission dynamics, which carries potential clues about the mechanism, remained unknown. To obtain additional information about the mechanism of radiation brightening in fVLPs, we performed measurements of the emission and the excited state relaxation dynamics. In this work, a fluorescein-derived dye, Oregon Green TM 488, was covalently bound to the BMV capsid via NHS ester labelling of exposed lysines. 11 Maximum labeling density was ∼ 300 dyes/virus 11 (see Figure SI1 ). Figure 1A shows a molecular model of BMV, with the dye-accessible external and internal lysines colored in red and green respectively. 12 Here "internal" means located between the lumenal or outer capsid surfaces. The minimum nearest neighbour distance between lysines was estimated to be 1.9 nm. 11 To follow the excited state dynamics of coupled fluorophores in fVLPs, we performed pump-probe femtosecond, transient absorption (TA) spectroscopy. Differential absorption spectra of isolated fluorophores and fVLP samples with an average number of 278 dyes per particle (hereby called BMV-OG278) are presented as a function of the wavelength and timedelay in Figure 1B and Figure 1C , respectively. The ultrafast excitation pulse at 488 nm was provided by an optical parametric amplifier pumped by a regenerative amplifier (∼ 120 fs pulse width). In both the control and BMV-OG278 sample, two intense negative signals are evident at very early times in the transient spectra. One appears at 570 − 580 nm and it can be attributed to Raman scattering of water. 13 Since this event is simultaneous with the excitation pulse, its appearance defines the zero delay between pump and probe pulses. The second, early negative signal is due to the coherent interaction between pump and probe pulses, which leads to stimulated Raman amplification appearing at 650-750 nm. 14 Both Raman signals vanish within 0.4 ps from the pump pulse. A spectral region of particular interest is 500 − 530 nm, where ground state bleach (GSB) is expected. Indeed, both sample and control exhibit a prominent feature in this spectral region. However, there are also some stark differences between the two, Figure 1B and Figure 1C : In particular, the excited state decay is much faster for the BMV-OG278 sample than for the free dye solution. Minima in ∆A spectra were at 498 nm for free dye and 504 nm for VLP-bound fluorophores, and are consistent with the peak wavelength in steady-state absorption spectra ( Figure SI1 ). The TA spectral evolution of the free dye solution and BMV-OG278 samples depends on pump energy, Figures SI3-7. The 5 − 6 nm shift between free and bound fluorophores is attributable to the difference in dielectric constant between protein and water. For quantitative time-domain analysis and comparison, TA spectra were integrated from 520 to 530 nm and fitted with an exponential model from 400 fs to 2.6 ns ( Figure SI8-12 ). For analysis, the wavelength and time-delay range were chosen past the ground state bleach minimum to avoid artifacts from pump laser scattering as well to remove coherent artifacts which arise around the zero delay time where the pump and probe beams are temporally overlapping. 15, 16 Exponential fit parameters (amplitude and decay time) were obtained by means of nonlinear least squares, and the estimates of the errors from the model were computed from the sample covariance matrix. Best fit kinetic models for the transients were found to be predominantly mono-exponential for the free fluorophore and bi-exponential for the virus-bound fluorophore. Normalized amplitudes (A i ) and decay times (τ i ) obtained from the fit procedure are presented as a function of N in the plots in Figure 2 , with N = 0 corresponding to free fluorophores in solution. At low power, the excited state lifetime (τ 1 ) of free dyes measured by TA was in average Figure 2C , red curve). At high labeling density, the τ 1 is reduced by a factor of 6 when compared to the free fluorophore. In previous fluorescence lifetime experiments at similar excitation fluence and N the fluorescence lifetime was short, and the photon counts were quenched. 11 A possible explanation is that at low pump power, only a few fluorophores are excited while the majority are in the ground state and there is a high probability for homo-resonance energy transfer (RET) to occur, with the result of an increase in fluorescence quenching. In any event, we note a correlation between the evolution of the TA τ 1 evolution and that of the fluorescence decay measured previously. As the pump fluence increases, τ 1 decreases for the free dye, possibly due to photobleaching or to an increase in inter-system crossing and triplet state formation. 17 Indeed, some photobleaching was observed in amount of 10 − 20% between the first and second run for each sample. However, at high labeling density ( N 200) τ 1 increases from ≈ 600 ps at lowest pump fluence up to ≈ 2000 ps for highest pump fluence. Thus, it appears that the nonradiative rate, which was deemed responsible for the initial shortening of the lifetime at low fluences and high density, 11 slows down to a value that is observed at lower labeling densities. The second fit component (parameters, A 2 and τ 2 ) dominates BMV-OG dynamics at short times suggesting a new relaxation channel that only operates in the multi-fluorophore VLP. The associated amplitude A 2 is significantly greater than A 1 , Figure There are qualitative differences between ASE and SF in both time and frequency domain. In the frequency domain, SF(SR) emission occurs in a broader window 21 while in ASE, spectral narrowing is expected as pump intensity increases. 22 We have not observed spectral narrowing in our experiments. 11 In the time domain, the SF pulse is sharp and develops after a time delay. 23 In ASE, the time delay is vanishingly small, the output pulse is longer than in SF and noisy, and the pulse duration is insensitive to dephasing factors (e.g. local viscosity and temperature). [24] [25] [26] From previous work 11, 12 we know that the radiation brightening from multi-fluorophore VLPs is very sensitive to the local environment, unlike ASE. Therefore, the significant shortening of excited state lifetime observed in TA and streak camera experiments, the delayed intense pulse in fluorescence emission, the broad fluorescence spectrum, the sensitivity to the fluorophore local environment, all point to a superradiance-like mechanism responsible for radiation brightening. We discuss a simple superradiance model in the SI, which illustrates the burst emission. An intriguing feature here is the fact that the phenomenon occurs at room temperature. SR emission dynamics was experimentally studied first in atomic gases at very low temperatures. 27 Later, it was detected in low-temperature solids. [28] [29] [30] [31] [32] Excitonic superradiance was studied extensively in molecular aggregates, at cryogenic temperatures as well. 33, 34 Only recently SR was shown to occur at room temperature in photosynthetic complexes. 35, 36 The structured environment within the photosynthetic complex is believed to promote collective fluorophore behavior since rigidity and proper orientation can alleviate thermal dephasing. 37 In support of this idea, alterations in the relative position and orientation of fluorophores in the protein pockets of an artificial photosynthetic system were found to strongly affect the excited state dynamics of protein-bound fluorophores. 38, 39 However, it is worth emphasizing that interfluorophore interactions in light harvesting complexes and in most molecular J-aggregates are characterized by strong coupling. [40] [41] [42] By comparison, the nearest-neighbor fluorophore distances on BMV-OG are much longer than that of molecular aggregates, which precludes electron tunnelling. In our case, collective emission of radiation arise from weak coupling at room temperature. Until now, this coupling range has not been given much attention. Dipole-dipole coupling is considered detrimental to the formation of coherent multi-fluorophore states. 27 Could the virus template play a role in coupling? The original model of Dicke superradiance (see SI) does not take into account the spatial organization or orientation of the fluorophores. However, from previous work it was clear that the nature of the template is important for the radiation brightening effect. 11,12 Below we provide further evidence for it from a study of relaxation dynamics. We performed additional TA experiments with a set-up in which pump and probe beams were counter-propagating, at a small angle. This set-up change allowed us to reduce laser scattering artifacts and measure the excited state dynamics at high laser powers with the probe wavelength closer to ground state bleach peak position λ = 500 − 520 nm. The higher signal-to-noise ratio afforded by this setup change has allowed us to make an interesting observation. Figure 4 presents normalized absorbance difference curves obtained for free dye and BMV-OG with N ranging from 200 to 300 dyes/virus. The data were tail-fitted with double exponential decay and the residuals from the fitting are presented on the top panels of the graph. At high dye loads, oscillations in TA can be observed, even in the raw data but most prominently in the residuals. Oscillating TA signals previously observed in photosynthetic complexes at room temper- In conclusion, the symmetric BMV capsids provides a biological scaffold for a multifluorophore array that acts an antenna with accelerated emission rate at room temperature. We have employed ultrafast, time-resolved transient absorption and fluorescence spectroscopy to provide further evidence for the mechanism of radiation brightening. Taken Electron-transparent samples were prepared by placing the dilute sample (10 µL) onto a carbon-coated copper grid. After 10 mins, excess solution on the grid was removed with filter paper. The grid was stained with uranyl acetate (10 µL of 2% solution) for 10 minutes and the excess solution was removed by blotting with filter paper. The sample was then left to dry for several minutes. Images were acquired at an accelerating voltage of 80 kV on a JEOL JEM 1010 Transmission electron microscope and analyzed with the ImageJ Processing Toolkit (see SI2). The sample was excited by a pulse generated through an optical parametric amplifier using a Ultrafast transient absorption measurements were carried out using an output of regeneratively amplified Ti:sapphire laser (800 nm, 120 fs, 2 kHz repetition rate) which was split into two beams. The first beam, containing 10% power, was focused into a sapphire crystal to generate a white light continuum (440-750 nm), which serves as the probe laser. The other beam, containing 90% of the power, was sent into an optical parametric amplifier to generate the 480 nm pump beam. After passing through a depolarizer, the pump beam is focused and overlapped with the probe beam at the sample (focal diameter being 300 µm). Evolution of the difference spectra from 900fs to 2600ps are shown in figures SI3-7. Femtosecond-TA data in this set of plots were pre-processed following a penalized least squares method that involved a third order penalty. ? We note that this method has been used previously by other groups to remove artifact contributions and as a baseline correction method in the treatment of ultrafast spectrokinetic data. ? ? ? The break in curves is due to removal of pump laser scattering. The major spectral features observed in the evolution-associated difference spectra (SI3- To examine the time-dependent GSB decay, we integrated the unprocessed difference spectra obtained by femtosecond-TA over the 520 − 530 nm region to extract a single kinetic trace at the central wavelength of 525 nm. Due to pump-pulse related artifacts, which are observed at very early times, the spectrokinetic data were fit from 400 fs to 2.6 ns using an exponential model. Each exponential term was given an individual amplitude A i and a characteristic time constant τ i -lifetime: Following the fitting procedure, individual amplitudes were subsequently normalized by the sum of A i 's to reveal their overall fractional contribution to the recovery kinetics. Consider a collection of N identical 2-level entities, each with a ground state, |g , and an excited state, |e . We assume they are fixed rigidly in place (to avoid recoil effects) and are much closer to each other than the wavelength of the radiation they can emit, but not so close that direct interactions (e.g. Van der Waals couplings) become significant. The initial condition at time t = 0s that all the entities are in the excited state, a symmetric arrangement. Here we assume that the spontaneous emission process is much more rapid than nonradiative decays and neglect the latter. Each entity is dipole coupled in the same way to the electromagnetic field so the perturbation term in the Hamiltonian is symmetric in its action. As Dicke first described, ? the system should spread down a "ladder" of purely symmetric states. We write for each rung of the ladder the state |N, s where 0 ≤ s ≤ N is the number of entities in the ground state at that rung. One does not need to specify which entities are in their ground state because |N, s formally includes all possible ways of distributing s ground states among the N entities. These collective states are readily orthonormalized. The time dependent wavefunction for the collection of entities may be written as and the focus is on the N + 1 occupational probabilities Π s (t) = |α s (t)| 2 , which satisfy, s Π s (t) = 1. The initial condition has Π s (0) = δ s,0 and the system decays towards Π s (∞) = δ s,N with N photons emitted. For intermediate times one needs to solve the coupled equations ? which describe how Π s changes due to decays to |N, s + 1 and gains from |N, s − 1 . The rate factors, Γ s , are given by where Γ is the decay rate of a single, isolated entity. Note that for s near N/2, Γ s /N Γ ≈ N/4. The rate of photon emission is W (t) = s Γ s Π s (t). Figure S13 (A) shows the evolution/spreading of the occupational probabilities for the case of N = 5 coupled entities. The photon emission rate for N = 5 collective (red curve) and that of independent emitters (blue curve) is given in Figure S13 An overview of nanoparticles commonly used in fluorescent bioimaging Near-infrared fluorescent probes in cancer imaging and therapy: an emerging field Near infrared fluorescence imaging in nano-therapeutics and photo-thermal evaluation Fluorescent probes for sensing and imaging Fluorescent probes for bioimaging of potential biomarkers in Parkinson's disease Fluorescent biosensors for neurotransmission and neuromodulation: engineering and applications. Frontiers in cellular neuroscience New red-fluorescent calcium indicators for optogenetics, photoactivation and multi-color imaging Photoluminescent and chromic nanomaterials for anticounterfeiting technologies: recent advances and future challenges Tunable and ultra-stable UV light-switchable fluorescent composites for information hiding and storage Molecular Fluorescence: Principles and Applications Radiation Brightening from Virus-like Particles Subset of Fluorophores Is Responsible for Radiation Brightening in Viromimetic Particles Raman spectra in spectrofluorimetry Artifacts in femtosecond transient absorption spectroscopy Picosecond coherence coupling in the pump and probe technique Elimination of the coherent artifact in a pump-probe experiment by directly detecting the background-free diffraction signal Photobleaching of organic fluorophores: quantitative characterization, mechanisms, protection. Methods and Applications in Fluorescence Dicke Superradiance in Solids Brief comment: Dicke superradiance and superfluorescence find application for remote sensing in air How to Recognize Lasing Coherence in Spontaneous Radiation Processes Amplified spontaneous emission and external signal amplification in an inverted medium Transition from Superfluorescence to Amplified Spontaneous Emission Laser characteristics of KCL: O-2 Quantum-statistical analysis of superfluorescence and amplified spontaneous emission in dense media Influence of collisional dephasing processes on superfluorescence Superradiance: An Essay on the Theory of Collective Spontaneous Emission Ultrashort radiation of biexcitonic superfluorescence from high-density assembly of semiconductor quantum dots Single-photon superradiance from a quantum dot Superfluorescence from lead halide perovskite quantum dot superlattices Monodisperse long-chain sulfobetaine-capped CsPbBr3 nanocrystals and their superfluorescent assemblies Superradiant Emission from Coherent Excitons in van Der Waals Heterostructures Superradiance and Exciton (De)localization in Light-Harvesting Complex II from Green Plants? Evidence for wavelike energy transfer through quantum coherence in photosynthetic systems Superradiance and Exciton Delocalization in Bacterial Photosynthetic Light-Harvesting Systems Superradiance of bacteriochlorophyll c aggregates in chlorosomes of green photosynthetic bacteria Correlated protein environments drive quantum coherence lifetimes in photosynthetic pigment-protein complexes Manipulating excited-state dynamics of individual light-harvesting chromophores through restricted motions in a hydrated nanoscale protein cavity Exploiting chromophore-protein interactions through linker engineering to tune photoinduced dynamics in a biomimetic light-harvesting platform Superradiance in Molecular Aggregates Superradiance in molecular H aggregates. Physical review letters Superradiance of J-aggregates: correspondence between an infinite disordered chain and a regular finite chain. Low Temperature Physics Long-lived quantum coherence in photosynthetic complexes at physiological temperature Quantum coherence in photosynthesis Normal mode calculations of icosahedral viruses with full dihedral flexibility by use of molecular symmetry Low Frequency Mechanical Modes of Viral Capsids: An Atomistic Approach All-atom molecular dynamics of the HBV capsid reveals insights into biological function and cryo-EM resolution limits Quantitative mass imaging of single biological macromolecules The Molecular Probes Handbook : a guide to fluorescent probes and labeling technologies