Proton-Transfer-Induced Fluorescence in Self-Assembled Short

Subscriber access provided by LUNDS UNIV

A: Spectroscopy, Molecular Structure, and Quantum Chemistry

Proton-Transfer Induced Fluorescence in Self Assembled Short Peptides Sijo K Joseph, Natalia Kuritz, Eldad Yahel, Nadezda Lapshina, Gil Rosenman, and Amir Natan J. Phys. Chem. A, Just Accepted Manuscript • Publication Date (Web): 12 Feb 2019 Downloaded from http://pubs.acs.org on February 12, 2019

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 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 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.

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 25 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

Proton-Transfer Induced Fluorescence in Self Assembled Short Peptides. Sijo K. Joseph,†,‡ Natalia Kuritz,† Eldad Yahel,† Nadezda Lapshina,† Gil Rosenman,† and Amir Natan∗,†,‡ †Department of Physical Electronics, Tel-Aviv University, Tel-Aviv 69978, Israel. ‡The Sackler Center for Computational Molecular and Materials Science, Tel-Aviv University, Tel-Aviv 69978, Israel. E-mail: [email protected]

Abstract We employ Molecular Dynamics (MD) and Time-Dependent Density Functional Theory (TDDFT) to explore the fluorescence of hydrogen-bonded dimer and trimer structures of Cyclic FF (Phe- Phe) molecules. We show that in some of these configurations a photon can induce either an intra-molecular proton transfer, or an inter-molecular proton transfer that can occur in the excited S1 and S2 states. This proton transfer, taking place within the hydrogen bond, leads to a significant red-shift that can explain the experimentally observed visible fluorescence in biological and bioinspired peptide nanostructures with a β-sheet biomolecular arrangement. Finally, we also show that such proton transfer is highly sensitive to the geometrical bonding of the dimers and trimers, and that it occurs only in specific configurations allowed by the formation of hydrogen bonds.

1

ACS Paragon Plus Environment

The Journal of Physical Chemistry 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

Introduction The discovery of the Green Fluorescent Protein (GFP) in jellyfish bio-luminescence by Osamu Shimomura has opened the door to new technologies of bio-compatible markers that can be used in-vitro and in-vivo [1–3]. In general, both GFP and fluorescence have important technological applications, such as monitoring gene expression and protein localization in living organisms, as well as detection of migration and invasion of cancer cells in vivo [4]. Fluorescence can also shed light on the structure of some important biological systems, such as amyloid fibrils, that play a critical role in the Alzheimer disease mechanism [5–13], or natural spiders’ silk [14–17], as well as within nanotechnology applications such as the field of peptide nanophotonics [18]. In recent years there have been many observations of an emergent fluorescence phenomenon in self- assembled short peptide aggregates [19–23], and in peptide and protein fibers, including the important case of amyloid fibrils [9–13]. It has been observed that the excitation of such molecular structures in the UV spectrum can yield fluorescence in the visible range, i.e. the emission of red, green and blue colors. It is generally agreed that the secondary structure, and, specifically, the formation of structures such as β-sheets that are rich with hydrogen bonds (HB), play a critical role in the appearance of visible range fluorescence [24–26]. In particular, this phenomenon is less widely observed or does not exist in the monomers or in other forms of secondary structures that are based on the same monomers [27]. In 2004 Shukla et al. [28] suggested that HBs may led new delocalized electronic states that might be related to the appearance of fluorescence. Pinotsi et al. [29] and Chan et al. [10] have shown that fluorescence is related to protein aggregation in amyloid fibrils, and to the formation of β-sheet regions that are rich with HBs. Recently, fluorescence in the visible range was demonstrated for various types of self-assembled fibers of short peptides structures, that are also supported with a β-sheet like organization [9–13, 19, 27, 30–34]. In Ref [35], it is demonstrated that the intrinsic blue fluorescence of Human Serum Albumin (HSA) exclusively originates from the β-sheet rich oligomeric interfaces, it is observed that excita2


Page 2 of 25

Page 3 of 25 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


tion wavelength-dependent fluorescence originates mainly due to the presence of various oligomers such as dimer, trimer, tetramer, and high order oligomers. In Ref [36], hydrogen bonding interactions of hybrid peptide-water clusters is explored theoretically using Force Fields and the role of nuclear quantum effects (NQE) is studied. They had indicated the importance of quantum mechanical effects of hydrogen bonding in peptide-water and peptide-peptide interactions. In Ref [37], the photo-induced electron transfer is indicated as a mechanism behind the fluorescence in certain amino-acids. A. A. Hassanali [38] el. al. had explored the zwitterionic form of amino acids, they have observed charge transfer effects between both the hydrophobic and hydrophilic groups and the surrounding water. Excited state intra-molecular proton transfer (ESIPT) is a known mechanism that can lead to fluorescence in many molecules [39–43]. Proton transfer between different amino acids is also considered to be the leading mechanism in GFP [1, 44]. Pinotsi et al. [25] and Grisanti et al. [11] have shown, using Molecular Dynamics (MD) and Time-Dependent Density Functional Theory (TDDFT), that an inter-molecular proton transfer can happen along the HB between the N terminal and the C terminal of model amyloid fibril proteins in a crystalline structure. They have also shown that this proton transfer affects the ground and excited state properties and, hence, could be the main factor behind the fluorescence in such structures. In this work, we explore cyclic Phe-Phe (Cyclic-FF) peptide dimers and trimers in the ground and excited state geometries, in order to evaluate the absorption and emission spectra, respectively. We use MD and TDDFT [45] calculations for this analysis. We show that in some geometries of dimers and trimers, intra-molecular and inter-molecular proton transfer through an HB can happen in the excited state. We show that this leads to a dramatic change in the optical properties of the dimers and trimers, leading to a large red-shift of the electronic gap. This shift can explain the appearance of fluorescence in the visible range. While the calculations are focused on Cyclic-FF, the mechanism of proton transfer is not directly related to the aromatic side groups and is probably of a general nature. Therefore, we expect to find similar behavior in aggregates of other peptides.

3



Methods and computational protocol MD procedure: The MD simulations were performed using QMDFF, which is a general quantum mechanically derived force field [46] based on ORCA [47] DFT calculations. For the ORCA DFT calculations we employed the B3LYP functional with the Grimme dispersion correction (B3LYPD3) and the TZVPP basis set. Based on these calculations, the QMDFF program was used to produce a force field. The molecular dynamics simulations were performed within the framework of QMSIM, which is a molecular dynamics simulation program of nonperiodic and periodic systems developed by S. Grimme. In the MD procedure we simulated a few (2-4) molecules in a cubic box (34.2Å×34.2Å×34.2Å) with periodic boundary conditions. We simulated the dynamics of typically 3-5 cycles of 1000 time steps at a high temperature (480K) and several cycles of 1000 time steps at room temperature (300K), where each time step was 1fs in duration. TDDFT linear response calculations: We used TDDFT [45, 48] calculations to investigate the excited state electronic properties of Cyclic-FF monomers, dimers, and trimers. Linear response theory with the Tamm-Dancoff Approximation (TDA) was used to evaluate the excited state energies and oscillator strengths (|S(λ)|2 ) [49–51]. The absorption and emission spectra were determined from the computed oscillator strength. We determined the first 40 excited electronic states for all the geometries in this study. The oscillator strengths were convoluted with a Lorentzian curve to obtain the desired continuous spectra. We used the 6-311G** basis-set with the B3LYP functional and Grimme dispersion Van der Waals corrections (B3LYP-D3) [52]. Geometry relaxation in excited state: The excited state minimum energy geometry was determined from excited state energy Hessian calculations, as implemented in the Q-Chem code [53]. Protocol: The protocol for finding candidate dimer and trimer structures was the following: (1) Several initial configurations for dimers and trimers were produced, either manually or from MD simulations (described above); (2) Each configuration was further geometrically relaxed to its ground state energy minimum using DFT with the B3LYP-D3 functional and the 6-311G** basis set; (3) The absorption spectrum was calculated with linear response TDDFT (described

4


Page 4 of 25

Page 5 of 25 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


above); (3) The minimum energy geometries for the first and second excited state (S1 and S2) were calculated, using the B3LYP-D3 functional and the 6-311G** basis set; (4) The emission spectrum was calculated with linear response TDDFT for the excited states minimum energy geometries.

Results and discussion In this section we present the results for Cyclic-FF dimers and trimers.

Cyclic-FF Dimer We used the MD procedure as well as a manual guess structure, for the Hydrogen-bonded dimers’ initial configurations, which were further relaxed with DFT. The energy dependence of the MD procedure of the Hydrogen-bonded dimers’ initial configurations, at each time step, is shown in Fig. 1. We also show the two low energy configurations that were eventually selected for the TDDFT analysis. Table 1: Binding energies for different dimer geometries Cyclic-FF Dimer Ground State Geometry MD Geometry 4606 MD Geometry 6941 Manual guess

Cyclic-FF Dimer Ground State Energy in Hartree (Ha) -1914.097 -1914.096 -1914.108

Binding Energy per Molecule in Hartree (Ha) -0.0201 -0.0197 -0.0259

The results of the ground state geometry optimization are shown in Table 1. The calculated Cyclic-FF Monomer ground state energy was −957.028 Hartree (Ha). According to Table 1, the most stable configuration that we obtained was based on the initial estimation, with a binding energy of 0.0259 Hartree (0.7048eV), or 16.2525 Kcal/mol, per monomer. We observed two HBs in the optimized dimer configuration. Hence, the binding energy for a single HB is 8.1263 Kcal/mol. This value is close to the reported value of 5 − 8 Kcal/mol for a single HB in β-sheet structures [54]. The MD generated structures are slightly higher in

5



0.1

Total Energy(eV)

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 6 of 25

0.05

0

-0.05

0

2000

4000 6000 Simulation Step

8000

Figure 1: Energy as a function of the MD simulation step for the Hydrogen-bonded dimers. Two of the lowest energy structures, which were selected for further analysis, are marked by the black vertical arrows, and their configurations are shown with a ball and stick atomistic model.

Figure 2: The ground state geometry, similarly the first and second excited state geometries of the Cyclic-FF Dimer, are shown. Here, (a) shows the ground state geometry, (b) and (c) show the inter-molecular proton transfer and intra-molecular proton transfer from the ground state to the first excited state S1 , respectively, whereas (d) refers to the inter-molecular proton transfer from the ground state to the second excited state S2 . In all three cases, the ground state C=O bond is transformed into a C-O-H bond via an intra- or inter-molecular proton transfer from a nearby N-H bond. The dashed line rectangle shows the location of the HB.

6


Page 7 of 25 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


Table 2: Proton-transferred dimer ground and excited state energy gap Estimated Dimer Geometry

Ground State Energy in Ha

Excited State Energy in Ha

Energy Gap in eV (∆En = En − E0 )

Ground State Optimized Geometry First Excited State Optimization (Inter-Molecular Proton Transfer) (Geometry-I) First Excited State Optimization (Intra-Molecular Proton Transfer) (Geometry-II) Second Excited State Optimization (Inter-Molecular Proton Transfer)

E0 = −1914.1082

E1 = −1913.9135 E2 = −1913.9132

∆E1 = 5.298 ∆E2 = 5.306

E0 = −1913.9854

E1 = −1913.9790 E2 = −1913.9642

∆E1 = 0.174 ∆E2 = 0.577

E0 = −1914.0421

E1 = −1913.9464 E2 = −1913.9077

∆E1 = 2.604 ∆E2 = 3.657

E0 = −1913.9846

E1 = −1913.973 E2 = −1913.972

∆E1 = 0.315 ∆E2 = 0.343

energy and have lower dimer binding energies, with the difference being around 2KB T at room temperature. We first analyzed the behavior of the manually generated structure under light excitation. It is evident from Table 2 that the first and second excited states (S1 and S2 ) are degenerate. We performed the following minimization processes for the geometry of the excited state: for the S1 first excited state we performed one minimization process by first operating with the smaller 321G basis set and then continuing with the 6-311G** basis. This two-step procedure led to an inter-molecular proton transfer state with a gap of 0.17eV; we call this "Geometry-I" and describe it in Fig. 2b. In addition, we performed another minimization process for the S1 state, where we started from the ground state geometry and did the excited state geometry minimization with the 6-311G** basis from the start. This led to another energy minimum, which we call "Geometry-II" for further reference ; this geometry is shown in Fig. 2c. In this case, we got an intramolecular proton transfer state, leading to a gap of 2.6eV. Geometry-I is 0.8eV lower in excited state energy (E1 ) in comparison to Geometry-II and hence we expect it to happen more. Here, there could be some barriers that did not let the system relax into Geometry-I when operating on the larger basis initially. However, if we also consider quantum mechanical effects [55–57], we can assume that the inter-molecular proton transferred excited state will be preferred. Note that, taking into account the 7



quantum nature of the proton, there will be some quantum mechanical probability for the barrier crossing via several quantum mechanical effects such as zero point motion or quantum tunneling. Recently there has been much interest in such nuclear quantum effects (NQE) in the atomistic simulations of chemical, biological and materials systems [58]. NQE where shown to affect the optical properties of 9-Methylguanine, Ellipticine and certain aromatic molecules [59–61]. The role of NQE in hydrogen bonded molecules was demonstrated in ref [36]. In addition to minimization of geometry in the S1 state, we performed minimization of geometry in the S2 state; here, we started from the ground state geometry and performed the minimization with the 6-311G** basis. The resulting minimum is again an inter- molecular proton transfer state - similar to Geometry-I of the S1 minimization. The observed gap in this minimum was found to be 0.33eV. In Fig. 2(a), it can be observed that a Hydrogen atom is attached to the Nitrogen atom of the bottom molecule (see the small box around the Hydrogen), while in the Geometry-I excited state (Fig. 2b), this Hydrogen atom is transferred to the nearby Oxygen atom of the second molecule, resulting in the formation of a C-O-H bond. This inter-proton transfer can also be interpreted as a transition of the dimer from the Lactam form to the Lactim form [62–65]. Geometry-II (Fig2c) also shows a proton transfer, but now it is an intra-proton transfer. The change in the Hydrogen atom location leads to a high reduction of the de-excitation energy, which results in a significant red-shift in the emitted wavelength. A similar proton transfer mechanism is also demonstrated for the S2 excited state minimization, as shown in Fig. 2d. The proton transfer is illustrated further in Fig. 3a, where the distances between the Hydrogen atom and the neighbor Oxygen and Nitrogen atoms are shown as a function of the S2 energy minimization step. We also show in the SI that the HB nitrogen and oxygen atoms get closer in the middle of proton transfer, from around dN O ' 2.8Å to dN O ' 2.4Å, after the transfer is completed, their distance is restored. Another important phenomenon, is that of electronic charge transfer between the monomers, this phenomenon can become important also in the presence of solvent molecules which can participate in the charge transfer [66], Jong et al. [67] have also

8


Page 8 of 25

Page 9 of 25 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


recently explored the role of NC termini interactions on the optical properties of aggregates of a six-chain amino acid peptide. They have shown experimentally and theoretically that a significant charge transfer mechanism between the C and N terminals of the peptides can be related to the optical band gap reduction in this system. In our case we have no solvents in the simulation, we show in the SI that there is no significant charge transfer between the monomers for S0 -S1 in the ground nuclear geometry, however, in the proton transfer geometry all states show a significant charge transfer of ∼0.8e for S0 and of ∼0.2e for S1 -S5 . Fig. 3b shows the ground and excited states’ energies as a function of the minimization step; while there are some jumps in the energy, it shows a large decrease in the S2 and S1 excited states in parallel to the proton transfer. At the same time, the ground state energy, S0 , increases, leading to a very small or even a zero gap. The minimization becomes noisy after the proton transfer as the S2 , S1 and S0 states become almost degenerate, which makes the Born-Oppenheimer approximation less justified. Furthermore, TDDFT itself is well known to be unreliable close to conical intersections because double- and triple-excitations would be required to accurately describe such (low energy) excitations. This calls for a higher level of theory simulation for this system.

Figure 3: S2 minimization: (a) OH and NH distances as a function of minimization step; (b) ground and a few excited state energies as a function of minimization step. The full minimization graph is shown in the SI. The fluorescence mechanism can be interpreted as follows. First, the dimer absorbs a UV-light 9



photon, transitioning to a higher electronic excited state, but still without nuclear motion, hence staying in the ground state geometry. Then, the molecule’s nuclei relax their energy via a protontransfer mechanism in the excited electronic state. It can be seen that the proton transfer leads to a significant change in both the S0 and Sn electronic states, and that the gap is significantly reduced. Once the proton-transferred geometry is reached, the gap between the excited state and the ground state is significantly lower, i.e. possibly in the visible, or even the IR, range. Hence, the dimer makes a transition from the proton-transferred excited state to the ground state (i.e. Sn → S0 ), yielding fluorescence at a much longer wavelength. The resulting calculated band gap, in both the S1 → S0 and S2 → S0 inter-molecular proton transfer states, is in the IR range and not in the visible range, the size of gap (0.17eV) is in fact close to zero and might suggest a possible non-radiative decay. However, there could be several possible ways to get a gap in the visible range. First, the TDDFT calculation might have under-estimated the gap; we show in the SI that if we calculate the proton-transfer geometry with the range separated CAM-B3LYP [68] hybrid functional, we get a signficantly larger gap of almost 2eV. Another effect which we did not consider is the possible influence of solvent molecules, the reason was that we tried to mimic experimental setup which was dry, however, if we consider water solvent with a Polarizable Contiuum Model (PCM) [69, 70] we get a similar increase in the gap in the proton transfer geometry (results are shown in the SI). Second, there could be two additional mechanisms that are possible; (a) a similar transition of a higher state Sn → S0 that also has a proton transfer, such transitions can also explain excitation dependent fluorescence, where the fluorescence wavelength depends on the excitation wavelength (b) a delayed fluorescence mechanism [71], i.e. the electron is first excited to S1 , but might climb to higher electronic states during the geometrical relaxation. Those mechanisms certainly require additional research, both experimentally and theoretically. This reduction in de-excitation energy is a requirement for the fluorescence, but it does not predict the emission intensities in the desired range. However, it is possible to argue that once the molecule has relaxed to the excited state geometry minimum, the only requirement for the fluorescence to happen is that the electronic transition to the ground state is allowed. The intensity of the

10


Page 10 of 25

Page 11 of 25 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


0.035 Absorption Emission

0.03 0.025 0.02 0.015 0.01 0.005 0 150

200

250

300

350

400

450

500

550

600

Figure 4: The absorption (dashed blue) and emission (solid red) spectra of the Cyclic-FF Molecule, computed with B3LYP-D3/6-311G**. The solid curve shows the emission spectra in a protontransferred excited state geometry; a significant shift to the visible range is obtained compared to absorption spectra. Here |S(λ)|2 is the dimensionless oscillator strength.

11



Figure 5: The absorption and emission Spectra of Cyclic-FF MD selected dimers, computed with B3LYP-D3/6-311G**: (a) Absorption spectrum for geometries 4606 (dashed blue) and 6941 (solid red) - atoms are in their ground state minimum energy geometry ; (b) Emission spectra for geometries 4606 (dashed blue) and 6941 (solid red) - atoms are in the S1 excited state minimum energy. Both geometries do not have a proton transfer in their excited state and, while their emission spectra differ from their absorption spectra, the difference is much smaller in comparison to the proton-transfer case. emission is determined by the quantum mechanical transition probability, which depends on the oscillator strength. The oscillator strength and the corresponding absorption and emission spectra are plotted in Fig. 4. The absorption spectra are computed with ground state optimized geometry while the emission spectra are computed using the first excited state optimized geometry. We observe that the emission spectra shift and change considerably due to the proton-transfer mechanism. The shift in the spectrum covers a significant part of the visible region with considerable quantum mechanical transition amplitude. A detailed examination of the emission spectrum reveals that the absorption peak at 220nm has disappeared, while the next new peaks appear at 297nm, 423nm, 443nm, and 532nm, it is important to note that the two large emission peaks (297nm and 423nm) are not a split of the original absorption peak but new and significantly higher excited states that became more acvtive in the proton transfer geometry. An additional important question is that of yield; the S0 → S2 oscillator strength in the ground state geometry, is only ∼1/40 relative to the S0 → S5 which is the largest absorption peak, hence suggesting that the effect can be small. We next examined two other dimer configurations, namely the MD4606 and MD6941 geome12


Page 12 of 25

Page 13 of 25 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


tries, that were selected from the MD simulation. Those configurations are higher in energy by around 2KB T per monomer at room temperature. Neither configurations showed a proton transfer in the excited state minimization. The calculated absorption and emission spectra for those structures are shown in Fig. 5. The absorption spectrum resembles that of the manual guess geometry, whereas the emission spectrum is significantly different. The MD4606 geometry retains the main absorption peak at 220nm and shows a new peak at 270nm with a shallow shoulder between 300 to 400nm, here, while there was no proton transfer, the geometry change was sufficient to cause a change in the electronic states and some smaller band reduction. On the other hand, the MD6941 geometry shows a small red-shift for the peak at 220nm and a much weaker peak at 400nm. Overall, it is evident that the manual guess configuration, which manifested a proton transfer in the excited state, had a significant gap reduction and a large red-shift of the whole spectrum. In contrast, the other two, MD selected, configurations showed a much less significant shift. This demonstrates that the proton-transfer happens only at certain configurations of the HB between the monomers. A more thorough analysis is needed to estimate the yield of this effect at room temperature or, equivalently, to estimate the percentage of dimers that are in a configuration that allows a proton transfer and fluorescence.

Cyclic-FF Trimers In this section we examined larger aggregates of three molecules (i.e. trimers) of the Cyclic-FF peptide. We conducted MD simulations and selected the two lowest-energy configurations as initial configurations for the DFT and TDDFT calculations. Figure 6 shows the MD simulation energy versus time, as well as the location of the structures that were selected - MD2284 and MD4004. The selected structures were further relaxed to their energy minimum using ab-initio DFT B3LYPD3/6-311G** calculations. The results for the binding energies of the resulting structures are given in Table 3. The binding energy of the lowest energy structure (i.e. structure MD4004) was found to be 0.024Ha (0.65eV ) per monomer, equivalent to 15.06Kcal/mol, which is slightly less than the lowest binding energy that was found for the dimer. The MD2284 structure binding energy 13



was found to be 0.021Ha (0.57eV ) and about 3KB T higher at room temperature. Next, we calculated the excited state properties using the same procedure that we used for the dimer structures. The MD4004 geometry showed a proton transfer in the S1 excited state and a significant change in the emission spectrum in comparison to the absorption spectrum - the latter has a single large peak at 220nm, whereas the former shows stronger peaks at 300nm and 400nm, as well as some weaker peaks at 470nm, 560nm and 650nm. The second geometry configuration, MD2284, does not show a proton transfer state and its calculated emission spectrum is only slightly shifted from the absorption spectrum, both showing a similar structure with a single peak at around 220nm. The absorption and emission spectra for both geometries are shown in Fig. 7. Table 4 shows the band gap reduction for both geometries.

0.05

Total Energy (eV)

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 14 of 25

0

-0.05

-0.1

0

2000

4000

6000

8000

Simulation Step Figure 6: Cyclic-FF trimer MD geometries are shown. The different molecular configurations’ energy is shown as a function of the MD simulation step. Two of the lowest energy configurations were selected for further analysis and marked with black dotted arrows. The selected configurations are the 2284th and 4004th geometries, respectively.

14


Page 15 of 25 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


Table 3: Trimer ground state total and binding energies Cyclic-FF Trimer Ground State Geometry MD Geometry 2284 MD Geometry 4004

Cyclic-FF Trimer Ground State Energy in Hartree (Ha) -2871.1479 -2871.1569

Binding Energy per Molecule in Hartree (Ha) -0.0210 -0.0240

Figure 7: Absorption and emission spectra for Cyclic-FF trimers: (a) Absorption spectrum of geometries 2284 (dashed blue) and 4404 (solid red), atoms are in the ground state geometry energy minimum; (b) Emission spectrum of geometries 2284 (dashed blue) and 4404 (solid red), atoms are in the S1 excited state geometry energy minimum. Geometry 4404, which shows a proton transfer in S1 , also exhibits a significant red-shift in the spectrum, Geometry 2284, which does not have a proton transfer, has an emission spectrum that is very close to that of the absorption. Table 4: Trimer ground and excited state energy gap (No proton-transfer & proton-transfer) MD Selected Trimer Geometry Ground State Optimized Trimer Geometry (2284) First Excited State Optimization Trimer Geometry (2284) (No-Proton Transfer) Ground State Optimization Trimer Geometry (4004) First Excited State Optimization Trimer Geometry (4004) (Inter-Molecular Proton Transfer)

Ground State Energy (E0 (Ha )) E0 = −2871.1479

Excited State Energy (En (Ha )) E1 = −2870.9600 E2 = −2870.9574

Energy Gap in eV (∆En = En − E0 ) ∆E1 = 5.11 ∆E2 = 5.18

E0 = −2871.1498

E1 = −2870.9619 E2 = −2870.9593

∆E1 = 5.11 ∆E2 = 5.18

E0 = −2871.1569

E1 = −2870.9631 E2 = −2870.9626

∆E1 = 5.27 ∆E2 = 5.29

E0 = −2871.0747

E1 = −2871.0695 E2 = −2871.0262

∆E1 = 0.14 ∆E2 = 1.32

15



Summary We demonstrated that an excited state inter-molecular proton transfer can occur in dimers and trimers of cyclic FF molecules. This proton transfer takes place within the HBs between the monomers, and is the basis for a large red- shift in the emission spectrum. This red-shift change in the emission properties can explain the phenomenon of fluorescence in the visible range, which was recently observed in UV-light excitation measurements of Cyclic-FF molecule aggregates. The HBs are highly important for this phenomenon, due to two reasons. Firstly, these bonds stabilize the formation of dimers and trimers, and probably also for larger molecule aggregates [27]. Secondly, HBs enable proton transfer without barrier crossing, along the HB, resulting in a modified geometry with lower energy in the excited state. With the B3LYP-D3 functional, we got a very small electronic gap of 0.17eV in the proton transfer geometry, this suggests a far IR emission or even a non-radiative decay. However, calculation of the same geometry with the CAM-B3LYP range separated hybrid functional showed an electronic gap of almost 2eV, demonstrating that this geometry electronic properties are sensitive to the choice of functional. The effect of implicit solvent inclusion was similar. While the proton transfer geometry is preferred in energy for several excited states, there can be barriers and hence the inclusion of nuclear quantum effects might be important to show such a transfer in more states. This calls for future calculations with additional functionals and perhaps also a higher level of theory. To check further the question of yield, additional geometries and electronic transitions should be checked. Another mechanism which we did not consider is that of electronic charge transfer, both the B3LYP and CAM-B3LYP calculations did not show a significant charge transfer in the ground state geometry, however, the proton transfer geometry showed a significant charge transfer in both the ground and excited states as is shown in the SI. Such an effect of proton transfer induced fluorescence was also demonstrated in other molecular systems [26, 42]. Our analysis also shows that the mechanism of proton transfer is highly dependent on the formation of specific HBs between the monomers. Although we demonstrated this phenomenon with a di-peptide that possesses phenyl side groups, the mechanism is clearly 16


Page 16 of 25

Page 17 of 25 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


dependent on the HBs and not on the aromatic nature of the side groups. As this mechanism is of a quite general nature we expect to find it also in aggregates of other peptides. Finally, we analyzed only the transitions that happen from S1 and S2 to S0 , excitation to higher states, Sn , could be accompanied also by a proton transfer and hence could lead to excitation dependent fluorescence.

Acknowledgments We would like to thank the Israel Ministry of Science, Technology and Space for financial support.

References (1) Tsien, R. Y. The green fluorescent protein. Annu. Rev. Biochem 1998, 67, 509–544. (2) Wolfbeis, O. S. Fluorescence Spectroscopy; Springer-Verlag: Berlin, 1993. (3) Chalfie, M.; Tu, Y.; Euskirchen, G.; Ward, W.; Prasher, D. Green fluorescent protein as a marker for gene expression. Science 1994, 263, 802–805. (4) Hoffman, R. M. In Vivo Cellular Imaging Using Fluorescent Proteins; Humana Press, Springer: New York, 1993. (5) Serpell, L. C. Alzheimer’s amyloid fibrils: structure and assembly. Biochim. Biophys. Acta Molecular Basis of Disease 2000, 1502, 16–30. (6) Dobson, C. M. The structural basis of protein folding and its links with human disease. Philos. Trans. R. Soc. Lond. B Biol. Sci 2001, 356, 133–145. (7) Hamley, I. Peptide Fibrillization. Angew. Chem. Int. Ed. 2007, 46, 8128–8147. (8) Lu, J.-X.; Qiang, W.; Yau, W.-M.; Schwieters, C. D.; Meredith, S. C.; Tycko, R. Molecular structure of β-amyloid fibrils in Alzheimer’s disease brain tissue. Cell 2013, 154, 1257–1268.

17



(9) del Mercato, L. L.; Pompa, P. P.; Maruccio, G.; Torre, A. D.; Sabella, S.; Tamburro, A. M.; Cingolani, R.; Rinaldi, R. Charge transport and intrinsic fluorescence in amyloid-like fibrils. Proc. Natl. Acad. Sci. U.S.A 2007, 104, 18019–18024. (10) Chan, F. T.; Schierle, G. S. K.; Kumita, J. R.; Bertoncini, C. W.; Dobson, C. M.; Kaminski, C. F. Protein amyloids develop an intrinsic fluorescence signature during aggregation. Analyst 2013, 138, 2156–2162. (11) Grisanti, L.; Pinotsi, D.; Gebauer, R.; Schierle, G. S. K.; Hassanali, A. A. A computational study on how structure influences the optical properties in model crystal structures of amyloid fibrils. Phys. Chem. Chem. Phys. 2017, 19, 4030–4040. (12) Sharpe, S.; Simonetti, K.; Yau, J.; Walsh, P. Solid-State NMR characterization of autofluorescent fibrils formed by the elastin-derived peptide GVGVAGVG. Biomacromolecules 2011, 12, 1546–1555. (13) Xue, C.; Lin, T. Y.; Chang, D.; Guo, Z. Thioflavin T as an amyloid dye: fibril quantification, optimal concentration and effect on aggregation. R. Soc. Open. Sci. 2017, 4, 160696. (14) Rammensee, S.; Slotta, U.; Scheibel, T.; Bausch, A. R. Assembly mechanism of recombinant spider silk proteins. Proc. Natl. Acad. Sci. U.S.A. 2008, 105, 6590–6595. (15) Du, N.; Liu, X. Y.; Narayanan, J.; Li, L.; Lim, M. L. M.; Li, D. Design of Superior Spider Silk: From Nanostructure to Mechanical Properties. Biophys. J. 2006, 91, 4528–4535. (16) Keten, S.; Xu, Z.; Ihle, B.; Buehler, M. J. Nanoconfinement controls stiffness, strength and mechanical toughness of β-sheet crystals in silk. Nat. Mater. 2010, 9, 359. (17) Keten, S.; Buehler, M. J. Asymptotic Strength Limit of Hydrogen-Bond Assemblies in Proteins at Vanishing Pulling Rates. Phys. Rev. Lett. 2008, 100, 198301. (18) A.Handelman,; Lapshina, N.; Apter, B.; Fainberg, B.; Rosenman, G. Peptide Integrated Optics. Adv. Mater. 2018, 30, 1705776. 18


Page 18 of 25

Page 19 of 25 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


(19) Liu, K.; Zhang, R.; Li, Y.; Jiao, T.; Ding, D.; Yan, X. Tunable Aggregation-Induced Emission of Tetraphenylethylene via Short Peptide-Directed Self-Assembly. Adv. Mater. Interf. 2017, 4, 1600183. (20) Channon, K. J.; Devlin, G. L.; MacPhee, C. E. Efficient Energy Transfer within SelfAssembling Peptide Fibers: A Route to Light-Harvesting Nanomaterials. J. Am. Chem. Soc. 2009, 131, 12520–12521. (21) Hong, Y.; Meng, L.; Chen, S.; Leung, C. W. T.; Da, L.-T.; Faisal, M.; Silva, D.-A.; Liu, J.; Lam, J. W. Y.; Huang, X.; Tang, B. Z. Monitoring and Inhibition of Insulin Fibrillation by a Small Organic Fluorogen with Aggregation-Induced Emission Characteristics. J. Am. Chem. Soc. 2012, 134, 1680–1689. (22) Mei, J.; Hong, Y.; Lam, J. W. Y.; Qin, A.; Tang, Y.; Tang, B. Z. Aggregation-Induced Emission: The Whole Is More Brilliant than the Parts. Adv. Mater. 2014, 26, 5429–5479. (23) Mei, J.; Leung, N. L. C.; Kwok, R. T. K.; Lam, J. W. Y.; Tang, B. Z. Aggregation-Induced Emission: Together We Shine, United We Soar! Chem. Rev. 2015, 115, 11718–11940. (24) Cui, G.; Lan, Z.; Thiel, W. Intramolecular Hydrogen Bonding Plays a Crucial Role in the Photophysics and Photochemistry of the GFP Chromophore. J. Am. Chem. Soc. 2012, 134, 1662–1672. (25) Pinotsi, D.; Grisanti, L.; Mahou, P.; Gebauer, R.; Kaminski, C. F.; Hassanali, A.; Kaminski Schierle, G. S. Proton Transfer and Structure-Specific Fluorescence in Hydrogen BondRich Protein Structures. J. Am. Chem. Soc. 2016, 138, 3046–3057. (26) Gil, M.; Kijak, M.; Piwo´nski, H.; Herbich, J.; Waluk, J. Non-typical fluorescence studies of excited and ground state proton and hydrogen transfer. Methods Appl. in Fluoresc. 2017, 5, 014007.

19



(27) Handelman, A.; Kuritz, N.; Natan, A.; Rosenman, G. Reconstructive phase transition in ultrashort peptide nanostructures and induced visible photoluminescence. Langmuir 2016, 32, 2847–2862. (28) Shukla, A.; Mukherjee, S.; Sharma, S.; Agrawal, V.; Kishan, K. R.; Guptasarma, P. A novel UV laser-induced visible blue radiation from protein crystals and aggregates: scattering artifacts or fluorescence transitions of peptide electrons delocalized through hydrogen bonding? Arch. Biochem. Biophys. 2004, 428, 144–153. (29) Pinotsi, D.; Buell, A. K.; M, D. C.; Kaminski Schierle, G. S.; Kaminski, C. F. A Label-Free, Quantitative Assay of Amyloid Fibril Growth Based on Intrinsic Fluorescence. ChemBioChem 2013, 14, 846–850. (30) Amdursky, N.; Molotskii, M.; Aronov, D.; Adler-Abramovich, L.; Gazit, E.; Rosenman, G. Blue Luminescence Based on Quantum Confinement at Peptide Nanotubes. Nano Lett. 2009, 9, 3111–3115. (31) Amdursky, N.; Koren, I.; Gazit, E.; Rosenman, G. Adjustable Photoluminescence of Peptide Nanotubes Coatings. J. Nanosci. Nanotechnol. 2011, 11, 9282–9286. (32) Castillo, J.; Andersen, K.; Svendsen, W. In Biomaterials Science and Engineering; Pignatello, R., Ed.; InTech, 2011. (33) Gan, Z.; Xu, H. Photoluminescence of Diphenylalanine Peptide Nano/Microstructures: From Mechanisms to Applications. Macromol. Rapid Commun. 2017, 38, 1700370. (34) Apter, B.; Lapshina, N.; Handelman, A.; Fainberg, B. D.; Rosenman, G. Peptide Nanophotonics: From Optical Waveguiding to Precise Medicine and Multifunctional Biochips. Small 2018, 1801147. (35) Bhattacharya, A.; Bhowmik, S.; Singh, A. K.; Kodgire, P.; Das, A. K.; Mukherjee, T. K.

20


Page 20 of 25

Page 21 of 25 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


Direct Evidence of Intrinsic Blue Fluorescence from Oligomeric Interfaces of Human Serum Albumin. Langmuir 2017, 33, 10606–10615. (36) Jong, K.; Ansari, N.; Grisanti, L.; Hassanali, A. Understanding the quantum mechanical properties of hydrogen bonds in solvated biomolecules from cluster calculations. J. Mol. Liq. 2018, 263, 501 – 509. (37) Prasad, S.; Mandal, I.; Singh, S.; Paul, A.; Mandal, B.; Venkatramani, R.; Swaminathan, R. Near UV-Visible electronic absorption originating from charged amino acids in a monomeric protein. Chem. Sci. 2017, 8, 5416–5433. (38) Ulman, K.; Busch, S.; Hassanali, A. A. Quantum mechanical effects in zwitterionic amino acids: The case of proline, hydroxyproline, and alanine in water. J. Chem. Phys. 2018, 148, 222826. (39) Bi, X.; Liu, B.; McDonald, L.; Pang, Y. Excited-State Intramolecular Proton Transfer (ESIPT) of Fluorescent Flavonoid Dyes: A Close Look by Low Temperature Fluorescence. J. Phys. Chem. B 2017, 121, 4981–4986. (40) Stasyuk, A. J.; Chen, Y.-T.; Chen, C.-L.; Wu, P.-J.; Chou, P.-T. A new class of N-H excitedstate intramolecular proton transfer (ESIPT) molecules bearing localized zwitterionic tautomers. Phys. Chem. Chem. Phys. 2016, 18, 24428–24436. (41) Padalkar, V. S.; Seki, S. Excited-state intramolecular proton-transfer (ESIPT)-inspired solid state emitters. Chem. Soc. Rev. 2016, 45, 169–202. (42) Sytnik, A.; Kasha, M. Excited-state intramolecular proton transfer as a fluorescence probe for protein binding-site static polarity. Proc. Natl. Acad. Sci. U.S.A. 1994, 91, 10757. (43) Jagadesan, P.; Eder, G.; McGrier, P. L. The excited-state intramolecular proton transfer properties of three imine-linked two-dimensional porous organic polymers. J. Mater. Chem. C 2017, 5, 5676–5679. 21



(44) Fang, C.; Frontiera, R. R.; Tran, R.; Mathies, R. A. Mapping GFP structure evolution during proton transfer with femtosecond Raman spectroscopy. Nature 2009, 462, 200. (45) Runge, E.; Gross, E. K. Density-functional theory for time-dependent systems. Phys. Rev. Lett. 1984, 52, 997. (46) Grimme, S. A General Quantum Mechanically Derived Force Field (QMDFF) for Molecules and Condensed Phase Simulations. J. Chem. Theory Comput. 2014, 10, 4497–4514. (47) Neese, F. The ORCA program system. Wiley Interdiscip. Rev. Comput. Mol. Sci. 2012, 2, 73–78. (48) Ullrich, C. A. Time-Dependent Density-Functional Theory; Oxford University Press: Oxford, 2012. (49) Casida, M. E. Recent Advances in Density Functional Methods; World Scientific, 2011; pp 155–192. (50) Casida, M.; Huix-Rotllant, M. Progress in Time-Dependent Density-Functional Theory. Annu. Rev. Phys. Chem. 2012, 63, 287–323. (51) Hirata, S.; Head-Gordon, M. Time-dependent density functional theory within the TammDancoff approximation. Chem. Phys. Lett. 1999, 314, 291–299. (52) Grimme, S.; Antony, J.; Ehrlich, S.; Krieg, H. A consistent and accurate ab-initio parametrization of density functional dispersion correction (DFT-D) for the 94 elements H-Pu. J. Chem. Phys 2010, 132, 154104. (53) Shao, Y.; Gan, Z.; Epifanovsky, E.; Gilbert, A. T.; Wormit, M.; Kussmann, J.; Lange, A. W.; Behn, A.; Deng, J.; Feng, X.; et.al, Advances in molecular quantum chemistry contained in the Q-Chem 4 program package. Mol. Phys. 2015, 113, 184–215. (54) Sheu, S.-Y.; Yang, D.-Y.; Selzle, H.; Schlag, E. Energetics of hydrogen bonds in peptides. Proc. Natl. Acad. Sci. U.S.A. 2003, 100, 12683–12687. 22


Page 22 of 25

Page 23 of 25 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


(55) Jose, D.; Datta, A. Tunneling Governs Intramolecular Proton Transfer in Thiotropolone at Room Temperature. Angew. Chem. Int. Ed. 2012, 51, 9389–9392. (56) Cha, Y.; Murray, C.; Klinman, J. Hydrogen tunneling in enzyme reactions. Science 1989, 243, 1325–1330. (57) Schreiner, P. R. Tunneling Control of Chemical Reactions: The Third Reactivity Paradigm. J. Am. Chem. Soc. 2017, 139, 15276–15283, PMID: 29028320. (58) Markland, T. E.; Ceriotti, M. Nuclear quantum effects enter the mainstream. Nat. Rev. Chem. 2018, 2, 0109. (59) Law, Y. K.; Hassanali, A. A. Role of Quantum Vibrations on the Structural, Electronic, and Optical Properties of 9-Methylguanine. J. Phys. Chem. A 2015, 119, 10816–10827. (60) Sappati, S.; Hassanali, A.; Gebauer, R.; Ghosh, P. Nuclear quantum effects in a HIV/cancer inhibitor: The case of ellipticine. J. Chem. Phys. 2016, 145, 205102. (61) Law, Y. K.; Hassanali, A. A. The importance of nuclear quantum effects in spectral line broadening of optical spectra and electrostatic properties in aromatic chromophores. J. Chem. Phys. 2018, 148, 102331. (62) Minkin, V. I.; Olekhnovich, L. P. .; Zhdanov, Y. A. Molecular Design of Tautomeric Compounds; D. Reidel Publishing Company: Dordrecht, Holland, 1988. (63) Antonov, L., Ed. Tautomerism, Methods and Theories; Wiley-VCH Verlag GmbH & Co: Weinheim, Germany, 2014. (64) Valters, R. E.; Flitsch, W.; Katritzky, A. R. Ring-Chain Tautomerism; Plenum Press: Spring Street, New York, 1985. (65) Peng, C. S.; Tokmakoff, A. Identification of Lactam-Lactim Tautomers of Aromatic Heterocycles in Aqueous Solution Using 2D IR Spectroscopy. J. Phys. Chem. Lett. 2012, 3, 3302– 3306. 23



(66) Hassanali, A. A.; Zhong, D.; Singer, S. J. An AIMD study of CPD repair mechanism in water: role of solvent in ring splitting. J. Phys. Chem. B 2011, 115, 3860–3871. (67) Jong, K.; Azar, Y. T.; Grisanti, L.; Stephens, A. D.; Jones, S. T.; Credgington, D.; Kaminski, G.; Hassanali, A. Low Energy Optical Excitations as an Indicator of Structural Changes Initiated at the Termini of Amyloid Proteins. 10.26434/chemrxiv.7215065.v1 2018, (68) Yanai, T.; Tew, D. P.; Handy, N. C. A new hybrid exchange–correlation functional using the Coulomb-attenuating method (CAM-B3LYP). Chem. Phys. Lett. 2004, 393, 51–57. (69) Lange, A. W.; Herbert, J. M. Polarizable continuum reaction-field solvation models affording smooth potential energy surfaces. J. Phys. Chem. Lett. 2009, 1, 556–561. (70) Lange, A. W.; Herbert, J. M. A smooth, nonsingular, and faithful discretization scheme for polarizable continuum models: The switching/Gaussian approach. J. Chem. Phys. 2010, 133, 244111. (71) Mamada, M.; Inada, K.; Komino, T.; Potscavage, W. J.; Nakanotani, H.; Adachi, C. Highly Efficient Thermally Activated Delayed Fluorescence from an Excited-State Intramolecular Proton Transfer System. ACS Cent Sci. 2017, 3, 769–777.

24


Page 24 of 25

Page 25 of 25 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


Graphical TOC Entry

25


Proton-Transfer-Induced Fluorescence in Self-Assembled Short

Recommend Documents