Excited-State Proton Transfer and Formation of the Excited Tautomer

Jul 19, 2016 - ... Aqueous Sugar and Alcohol Solutions Investigated by Fluorescence Spectroscopy. Young-Ho Hong , Byeong-Seo Cheong , Han-Gook Cho...
0 downloads 0 Views 2MB Size
Subscriber access provided by UNIV OF TECH SYDNEY

Article

Excited-state Proton Transfer and Formation of the Excited Tautomer of 3-hydroxy-pyridine-dipicolinium Cyanine Dye Ori Green, Ron Simkovitch, Luís Pinto da Silva, Joaquim C.G. Esteves da Silva, Doron Shabat, and Dan Huppert J. Phys. Chem. A, Just Accepted Manuscript • DOI: 10.1021/acs.jpca.6b04666 • Publication Date (Web): 19 Jul 2016 Downloaded from http://pubs.acs.org on July 20, 2016

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 A 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 42

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

Excited-state Proton Transfer and Formation of the Excited Tautomer of 3-hydroxy-pyridine-dipicolinium Cyanine Dye Ori Greena, Ron Simkovitcha, Luís Pinto da Silvab, Joaquim C.G. Esteves da Silvab,c, Doron Shabata and Dan Hupperta* a

Raymond and Beverly Sackler Faculty of Exact Sciences, School of Chemistry, Tel Aviv University, Tel Aviv 69978, Israel

b

Centro de Investigação em Química, Departamento de Química e Bioquímica, Faculdade de Ciências da Universidade do Porto, R. Campo Alegre 687.

c

Centro de Investigação em Química, Departamento de Geociências, Ambiente e

Ordenamento do Território, Faculdade de Ciências da Universidade do Porto, R. Campo Alegre 687. *Corresponding author: Dan Huppert E-mail: [email protected] Phone: 972-3-6407012 Fax: 972-3-6407491

1 Environment ACS Paragon Plus

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

Abstract Steady-state and time resolved fluorescence techniques and theoretical calculations, were employed to study the photoprotolytic properties of a newly synthesized photoacid 3-hydroxy-pyridine-dipicolinium cyanine (HPPC) dye. This dye is similar to Quinone cyanine 9, which we have previously studied and is the strongest photoacid currently synthesized. In this compound, we found that several proton transfer phenomena occur after excitation. We found that the excited-state proton transfer (ESPT) rate in water is ultrafast with kPT≈1.5×1012s-1. In methanol and ethanol the rate is slower by about five and six times respectively. The fluorescence spectrum of HPPC in water consists of three bands with maxima at 520, 600 and 665nm, whereas in monlos and other protic solvents the fluorescence spectrum consists only of two emission bands at 530nm and ~700nm. We assign the emission bands of HPPC at 520nm to the protonated form and the 700nm band in monols and 665nm in water to the deprotonated form. The 600nm band that is the most intense band in the fluorescence spectrum of HPPC in water we assign to the tautomeric form in which the proton is attached to the pyridine's nitrogen atom. Based on density functional calculations, we suggest that in water the proton transfer process to the pyridine's nitrogen atom occurs in a stepwise manner via a two water molecule bridge.

2 Environment ACS Paragon Plus

Page 2 of 42

Page 3 of 42

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

Introduction Many chemical compounds are unstable in their first electronically excited state. A large family of compounds undergo proton or hydrogen transfer in their excited states (ES). There are at least four types of proton transfer reactions in the ES. The first class of compounds undergo an excited-state intermolecular proton-transfer reactions (ESPT).1-15 The proton is transferred to a nearby solvent molecule that is hydrogen bonded to the photoprotolytic active molecular site. These molecules are termed photoacids. These molecules are weak acids in their ground state and have pKa values in the range of 5-10. Their acidity changes in the excited state and their pKa* is lower by about 7-11 orders of magnitude. The second class of photo-protolytic compounds undergo an intramolecular proton transfer (ESIPT).16-38 These compounds consist of a functional group that serves as a proton donor and it is situated in close proximity to a second group that accepts the hydrogen atom (proton) in the excited state. The proton donor is a hydroxyl group and the proton acceptor is a heterocyclic nitrogen atom or a carbonyl. The third type of compounds is photobases. In their ES they are stronger bases then in their ground state and they accept a proton from the solvent or excess proton in solution. Their ES photobasicity can be easily observed in solvents that are weak organic acids, or in aqueous solutions containing strong acids. In some compounds like 7-hydroxy-coumarin dyes, 1-naphthol and 1-naphthol sulfonate derivatives the ES reaction with an excess proton in aqueous solution leads to a protonation process accompanied by a nonradiative process and as a consequence the protonation reaction rate can be observed by measuring the excited state lifetime. The fourth type of photo-protic reaction occurs in bifunctional compounds. These compounds consists of both photoacidic and photobasic functional groups. These groups are in close proximity, one is a proton donor and the other one is a proton acceptor. In the excited-state they tend to form a zwitterion in which the proton donor group is negatively charged after losing the proton and the basic group that receives the proton is positively charged. This zwitter-ion form is termed also the tautomer form. In a recent work we studied the photoacidity, photobasicity and the formation of the tautomeric form of 7-hydroxy-coumarin.39 The tautomeric form emission band is observed at long wavelengths and the band peak is at 485nm, 3 Environment ACS Paragon Plus

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

whereas the protonated and deprotonated forms emission bands have a maxima at 380nm and 450nm respectively. We found that at neutral pH water-monol mixtures, the ESPT process to the solvent is faster than the formation rate of the tautomer. We concluded that the protonation of the coumarin carbonyl and the formation of the tautomer is a rather slow process. In the current work we study the photoprotolytic processes of a newly synthesized cyanine dye shown in Scheme 1, the 3-hydroxy-pyridine-dipicolinium cyanine (HPPC) dye.

Scheme 1: molecular structure of 3-hydroxy-pyridine-dipicolinium cyanine (HPPC) dye

The main finding of the current study is that the ESPT rate constant in water of HPPC is ultrafast and has a rate constant kPT= 1.5×1012s-1, whereas in methanol and ethanol the kPT values are five and six times smaller than in water. We explain the results as follows: In water, the main photoprotolytic process is the formation of the tautomeric form whereas in methanol and ethanol this process is rather slow and the main process is an ESPT to the solvent. The proton is transferred from the 3-hydroxyl group to the pyridine's nitrogen atom via a water bridge of two water molecules. Scheme 2 shows the three possible forms of HPPC both in the ground and excited state. NRONROH +HNRO-

(N)

(A)

(T)

Scheme 2: Possible forms of HPPC

4 Environment ACS Paragon Plus

Page 4 of 42

Page 5 of 42

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

Materials and methods Synthesis All reactions requiring anhydrous conditions were performed under an Argon atmosphere. All reactions were carried out at room temperature unless stated otherwise. Chemicals and solvents were either A.R. grade or purified by standard techniques. Thin layer chromatography (TLC): silica gel plates Merck 60 F254: compounds were visualized by irradiation with UV light. Flash chromatography (FC): silica gel Merck 60 (particle size 0.040-0.063 mm), eluent given in parentheses. High pressure liquid chromatography (HPLC): C18 5u, 250x4.6mm, eluent given in parentheses. Preparative HPLC: C18 5u, 250x21mm, eluent given in parentheses. 1HNMR spectra were measured using Bruker Avance operated at 400MHz as mentioned. All general reagents, including salts and solvents, were purchased from Sigma-Aldrich. The synthesized product compound was purified by HPLC to 99% purity. Abbreviations.

ACN- Acetonitrile, DCM- Dichloromethane, DMAP- 4-Dimethylaminopyridine, DMF- N,N'-Dimethylformamide, EtOAc- Ethylacetate, Hex- n-Hexanes, MeOHMethanol, THF- Tetrahydrofurane, TFA- Trifluoroacetic acid, Et3N- Triethylamine, EtOH- Ethyl alcohol, NaOAc- Sodium acetate, Ac2O –Acetic anhydride, AcOH – Acetic acid, NaOAc –Sodium Acetate, NHS- N-Hydroxysuccinimide, DCC- N,N'Dicyclohexylcarbodiimide. Synthesis Procedure

Dialaldehyde 1a40 (0.067 mmol), piperidine (0.133 mmol) and picolinium iodoide2 9 (0.133 mmol) were dissolved in EtOH (1ml). The reaction mixture stirred for 60

5 Environment ACS Paragon Plus

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

minutes at 80oC and monitored by RP-HPLC (grad. 10%-90 ACN in water, 20min). After completion, the reaction mixture was concentrated by evaporation under reduced pressure. The crude product was diluted with 1:1:0.1 H2O: ACN: AcOH, and purified by preparetive RP-HPLC (grad. 10%-90 ACN in water, 20min) to give Compound 1; Hydroxy-pyridine-dipicolinium Cyanine Dye (HPPC). 1H NMR (400MHz, MeOD): δ = 8.78-8.72 (4H, m), 8.29-8.18 (5H, m), 8.06 (1H, d, J=15.9 Hz), 7.90 (1H, d, J=15.9 Hz), 7.82 (1H, d, J=15.8 Hz), 7.62 (1H, d, J=8.48 Hz), 7.35 (1H, d, J=8.4 Hz), 4.32 (3H, s), 4.29 (3H, s). MS (ES+ ): m/z calc. for C21H21N3O2: 330.2 [M]+. Fluorescence methods The time-resolved fluorescence was measured by the up-conversion technique. The fluorescence of 3-hydroxy-pyridine-dipicolinium cyanine was studied in several polar, protic and aprotic solvents at room temperature. The laser used for the fluorescence up-conversion was a cavity-dumped Ti:sapphire femtosecond laser (Mira, Coherent), which provides short, 120fs, pulses at about 800nm. The cavity dumper operated with a low repetition rate of 800kHz. The up-conversion system is a commercial system (FOG-100, CDP, Russia). The samples were excited by the SHG pulses of

8mW at wavelengths of 390-420nm on average. The time response of the

up-conversion system is determined by the Raman-Stokes line of water, red-shifted by 3600cm–1. We found that the full-width at half-maximum (fwhm) of the signal is 240fs. To avoid photo-degradation, samples were placed in a rotating optical cell, rotating at a frequency of 10Hz. Sample degradation was minimal and did not affect the profile of the signal decay. Experiments were carried out on solutions at concentrations of about 0.5mM. The steady-state fluorescence and excitation spectra were measured by a Horiba Jobin Yvon FluoroMax-3 fluorescence spectrofluorometer. The absorption spectra were measured by a Cary 5000 spectrometer. Quantum mechanical Methods All calculations were carried out with the use of the Gaussian 09 program package.41 The ground-state (S0) geometries of the studied species of HPPC were calculated at the ωB97XD/6-31G(d,p) level of theory.42 The confirmation that these structures were indeed minima or transitions states in their potential-energy surfaces

6 Environment ACS Paragon Plus

Page 6 of 42

Page 7 of 42

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

(PES) was obtained in subsequent vibrational-frequency calculations, at the same level of theory. The proton-transfer reaction was modeled by performing an intrinsic-reactioncoordinate (IRC) calculation, in which the transition state is connected to the reactants and products.43 These calculations were performed at the ωB97XD/6-31G(d,p) level of theory.42 The S0 PES was re-evaluated by performing single-point energy calculations at the ωB97XD/6-31+G(d,p) level of theory. The first (S1) and second (S2) singlet excited-state PES were evaluated by calculating the respective FranckCondon curves. These are modeled by calculating the time-dependent (TD) ωB97XD/6-31+G(d,p) vertical-excitation energies at the respective S0 structures. Such alternative strategy arises from difficulty in theoretically following an ESPT reaction in the excited state. The choice of an effective reaction coordinate is far from trivial, as the motion of the proton is accompanied by a significant electronic rearrangement. It is normally coupled to other degrees of freedom and is associated with problems related to classical or quantum treatment of its dynamics.44-46 Moreover, the inability to analytically resolve second derivatives of excited-state energy severally impairs the modelling of ESPT reactions entirely on the excited state. The IRC-path was obtained by constructing a water bridge made of two explicit solvent molecules, which connect the hydroxyl group and the central pyridine’s nitrogen heteroatom. In order to see if the stability of the “two-water” bridge throughout the PT reaction was accurate, or only an artifact of the absence of other explicit solvent molecules, we have also used a quantum/molecular mechanics (QM/MM) approach via the ONIOM method to assess the effect of a more complete solvation shell. The complexes formed between the “two-water” bridged hydroxyl and tautomer species and 80 explicit water molecules were obtained by using the ONIOM method47. The “two-water” bridged structures were included in the High layer, while the 80 water molecules were included in the Low layer. Geometry optimizations were made in vacuo with mechanical embedding, with the High level modelled at the ωB97XD/6-31G(d,p) level of theory, while the Low level was treated with the UFF force field48 . Partial atomic charges were assigned using the QEq formalism for all atoms49. Single point energy calculations were made with electronic embedding at the ωB97XD/6-31+G(d,p):UFF, with implicit solvation. Water was the 7 Environment ACS Paragon Plus

The Journal of Physical Chemistry

chosen solvent and was modelled with the SMD solvation model. The reaction field was computed separately in each sub-calculation always using the cavity of the real system. The validity of constructing Franck-Condon curves arises from the notion that the absorption process can provide valuable insight into fluorescent properties of a molecule. It should be noted that the absorption spectrum often reflects the main features of the emissive state, differing only by a red shift that occurs as a result of geometrical relaxation of the molecule50-52 and so, this approach has been used routinely with good results in the study of ESPT reactions.53-61 All optimization, frequency and IRC calculations were performed in vacuo, while all single-point calculations were made in implicit water, which was modelled by the SMD model.62 More specifically, the solvent was modelled by a Polarizable Continuum Model (PCM) using the integral equation formalism variant, and with radii and non-electrostatic terms for Truhlar and coworkers’ SMD solvation model.62 ωB97XD is a long-range-corrected functional and was chosen because of good results in local n→π*, π→π*, charge-transfer and Rydberg states.63 Moreover, this functional also includes an empirical dispersion-correction term, which may be important for correctly modelling this ESPT reaction. Results Figure 1a shows the UV-vis absorption spectrum of 3-hydroxy-pyridine-dipicolinium cyanine (HPPC) dye in ethanol and water at neutral pH.

a)

b)

1.0

Bulk H2O

1.0

Bulk H2O

30µM HCl

0.8

60µM HCl 148µM HCl 1025µM HCl

89µM HCl 442µM HCl 1888µM HCl

Bulk ethanol

0.8

369nm 420nm

0.6

Abs

Abs

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 8 of 42

0.4

0.6 0.4

357nm 402nm

0.2

0.2

0.0

0.0 250

300

350

400

450

500

550

600

250

300

Wavelength (nm)

8 Environment ACS Paragon Plus

350

400

450

500

Wavelength (nm)

550

600

Page 9 of 42

c)

1.0 0.8

Abs

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

0.6

Bulk H2O

29µM NaOH

60µM NaOH 138µM NaOH 235µM NaOH 330µM NaOH 412nm

89µM NaOH 187µM NaOH 282µM NaOH

0.4

487nm

0.2 420nm

0.0 250

300

350

400

450

500

550

600

Wavelength

Figure 1: Steady-state UV-vis absorption spectrum of HPPC in a. water and ethanol b. at various HCl concentrations c. at various NaOH concentrations

In water three absorption bands are observed in the spectral range of 325-600nm, whereas in ethanol only two absorption bands are observed. The bands in the UV region are assigned the S0S1 and S0S2 transitions of the protonated NROH form. Both the absorption bands in the UV region are blue shifted in water with respect to their position in monols. The spectral position of the additional band that appears only in water is at around 500nm. We assign this band to the anionic form NRO-. The relative band intensity is only ~7% of the band with a maximum at 400nm. Another possibility for the origin of this band would be the tautomeric form +HNRO-. However, the theoretical calculations results show that the S0 NROH→+HNROreaction is not energetically favorable, which indicates that the tautomer should not be found in ground state. Moreover, the theoretical excitation energy of NRO- agrees with experiment much more than the calculated value obtained for +HNRO-. Figure 1b and c show the absorption spectra of HPPC in water at acidic and basic pH respectively. In acidic media up to about pH=2, the relative intensity of the two absorption bands in the spectral region 325-500nm is almost constant. At basic solutions we observe a large change in the spectrum due to the ground state proton transfer reaction: NROH + B −  → NRO − + BH

(1) As the pH increases the deprotonated form fraction in the solution increases whereas that of the protonated form absorption decreases. At pH above pH≈10 the spectrum is nearly independent of further increase in the base (NaOH) concentration. The position

9 Environment ACS Paragon Plus

The Journal of Physical Chemistry

of the weak band at 500nm we assign to the anionic form in neutral pH aqueous solution. Figure 2a and b show on linear and semilogarithmic scales the normalized steady-state fluorescence spectrum of HPPC in water, methanol and water : methanol mixtures. a)

bulk MeOH

b)

xMeOH=0.84, xwater=0.16

xMeOH =0.72, xwater=0.28

xMeOH=0.84, xwater=0.16

xMeOH =0.72, xwater=0.28

xMeOH=0.56, xwater=0.44

xMeOH=0.46, xwater=0.54

xMeOH =0.39, xwater=0.61

xMeOH=0.56, xwater=0.44

xMeOH=0.46, xwater=0.54

xMeOH =0.39, xwater=0.61

xMeOH=0.34, xwater=0.66

xMeOH=0.27, xwater=0.73

xMeOH =0.22, xwater=0.78

xMeOH=0.34, xwater=0.66

xMeOH=0.27, xwater=0.73

xMeOH =0.22, xwater=0.78

xMeOH=0.21, xwater=0.79

xMeOH=0.16, xwater=0.84

xMeOH =0.11, xwater=0.89

xMeOH=0.21, xwater=0.79

xMeOH=0.16, xwater=0.84

xMeOH =0.11, xwater=0.89

xMeOH=0.04, xwater=0.96

bulk MeOH

bulk water

xMeOH=0.04, xwater=0.96

1.0

bulk water

1

Norm. Signal

0.8

Norm. Signal

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 42

0.6

0.4

0.1

0.01

0.2

0.0 450

500

550

600

650

700

750

1E-3 450

Wavelength (nm)

500

550

600

650

700

750

Wavelength (nm)

Figure 2: Normalized steady-state fluorescence spectrum of HPPC in water, methanol and water : methanol mixtures, excited at 420nm a. linear scale. b. semilogarithmic scale

The fluorescence spectrum in neat methanol solution of HPPC consists of two main emission bands when excited at 420 nm the neutral form. The short wavelength band with a maximum at about 530nm, which we assign to the neutral NROH form and the deprotonated anionic NRO- form has a band with a maximum at ~700nm. In water at neutral pH the emission spectrum consists of three emission bands. The first is a very weak band with a peak at ~520nm assigned to the NROH form. The second is a strong intensity band with a maximum at ~600nm that we assign to the tautomeric form +HNRO-. The third band is a weaker band than the main band and is observed with a band peak at ~665nm and is assigned to the deprotonated form NRO-. The dip in the spectrum at ~650nm is an instrumental artifact. When we compare the fluorescence spectrum of HPPC in water and methanol we conclude that in water both the NROH and the NRO- bands shift to the blue. The steady-state fluorescence bands of the NROH and NRO- forms of HPPC in water-methanol mixtures shift gradually to the blue as the mole fraction of water increases, the relative intensity of the NROH decreases with the increase of water -

content in the mixture. The relative fluorescence intensity I FNRO / I FNROH is larger by a

10 Environment ACS Paragon Plus

Page 11 of 42

factor of ~20 in water than in methanol because of the large reduction of the NROH fluorescence in water. If the fluorescence lifetime τF of HPPC NRO- and NROH forms in methanol and water are similar

τ F = kF−1

(2a)

kF = kr + knr

(2b)

Then we can qualitatively conclude that the excited state proton transfer rate to the solvent in water is about 20 times larger than in methanol. We also notice that there exists an iso-emissive region at 525-540nm between the NROH and the +HNROspectrum of water and water-methanol solutions up to xH2O ≈0.6. It means that in water rich mixtures the NROH form converts to the tautomer that subsequently dissociates and converts to the NRO- deprotonated form. Figure 3 shows the time-resolved fluorescence of HPPC in water, monitored at several wavelengths in the spectral region of 480-600nm along with the instrument response (IRF) measured by monitoring the water Raman signal.

480nm 510nm 540nm 600nm

a) 1.0

490nm 520nm 560nm IRF

500nm 530nm 580nm

b)

480nm 510nm 540nm 600nm

1.0

0.8

Norm. Signal

0.8

Norm. Signal

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

600nm

0.6

600nm

0.4 0.2

0.2

0.0

480nm

0

1000 2000 3000 4000 5000 6000 7000

Time (fs)

0.0 0

400

800

1200

1600

500nm 530nm 580nm

0.6

0.4

480nm

490nm 520nm 560nm IRF

2000

Time (fs)

11 Environment ACS Paragon Plus

The Journal of Physical Chemistry

c)

480nm 510nm 540nm 600nm

1

490nm 520nm 560nm IRF

500nm 530nm 580nm

600nm

Norm. Signal

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 42

0.1

480nm

0.01 0

5000

λ 10000

15000

20000

25000

Time (fs)

Figure 3: Time-resolved fluorescence of HPPC in water, monitored at several wavelengths in the spectral region of 480-600nm along with the instrument response (IRF) a. Linear short scale b. Linear long scale c. Semilogarithmic scale

The signals were acquired by the fluorescence up-conversion system (see Materials and methods section). The fluorescence average decay time τ av = ∫ I F ( t ) dt strongly depends on the monitored wavelength. We used a multiexponential fit function convoluted with the IRF to fit the fluorescence experimental signals. The amplitudes and time-constants of the fits are given at Table 1 Table 1: Fitting parameters a of the time-resolved fluorescence of HPPC in H2O and D2O with the use of a multiexponential function b,c solvent

H2O

D2O

wavelength (nm)

a1

τ1(fs)

a2

τ2(fs)

a3

τ3(fs)

a4

τ4(ps)

480

0.66

120

0.28

400

0.05

2300

0.01

40

490

0.57

120

0.35

370

0.07

2150

0.01

40

500

0.47

120

0.43

400

0.09

2150

0.01

60

510

0.39

120

0.46

400

0.14

2150

0.01

60

520

0.35

120

0.44

400

0.19

2150

0.02

60

530

0.3

120

0.45

400

0.23

2150

0.02

60

540

0.21

130

0.48

460

0.28

2150

0.03

60

560

0.12

130

0.49

600

0.31

2400

0.08

60

580

0.04

130

0.48

700

0.32

2400

0.16

60

600

-

-

0.39

830

0.31

2800

0.3

60

480

0.71

160

0.21

730

0.08

3400

0.001

80

490

0.61

180

0.29

740

0.1

3400

0.008

80

500

0.52

180

0.33

770

0.15

3400

0.007

100

12 Environment ACS Paragon Plus

Page 13 of 42

510

0.43

200

0.37

800

0.2

3500

0.008

100

520

0.39

200

0.32

800

0.27

3400

0.02

100

530

0.28

230

0.31

810

0.38

3400

0.03

100

540

0.26

260

0.27

850

0.43

3400

0.04

100

560

0.09

260

0.25

850

0.55

3600

0.11

100

a

The error in all amplitude parameters is ±10%, the error in the time constants is ±15% for the second and third time components.

b.

Full-width at half-maximum of the system response was τ0=240fs 4

c.

y(t ) = ∑ ai exp(−t / τ i ) i =1

Figure 4 shows the fluorescence up-conversion signals of HPPC in water at the long wavelength spectral region 600-700nm.

H2O 620nm

H2O 640nm

H2O 680nm

H2O 700nm

H2O 660nm

b)

1.0

1.0

0.8

0.8

Norm. Signal

Norm. Signal

a)

0.6 0.4 0.2

H2O 620nm

H2O 640nm

H2O 680nm

H2O 700nm

0.6 0.4

0.0 0

500

1000

1500

2000

0

Time (fs)

c)

H2O 660nm

0.2

0.0

H2O 620nm H2O 680nm

H2O 640nm H2O 700nm

H2O 660nm

0.1

0.01 0

400

800

1200

1000 2000 3000 4000 5000 6000 7000

Time (fs)

1

Norm. Signal

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

1600

2000

Time (fs)

13 Environment ACS Paragon Plus

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

Page 14 of 42

Figure 4: Time-resolved fluorescence of HPPC in water, monitored at several wavelengths in the spectral region of 600-700nm a. Linear short scale b. Linear long scale c. Semilogarithmic scale

The signals at this spectral region show a signal rise which is longer in time than what is expected from the finite time resolution because of the IRF width. The signal rise strongly depends on the monitored wavelength. At long wavelength the rise time is much longer than at short wavelengths and the amplitude of the rise component is much larger. We used the following procedure to fit the long wavelength λ≥620nm signals. At these wavelengths three fluorescence bands NROH, +HNRO- and NRO- overlap and each band's contribution to the signal depends on the monitored wavelength. The overall time-resolved fluorescence signal measured at a certain wavelength λ, y(λ,t) is given by: 3

y (λ , t ) = ∑ ai (λ ) yi (t )

(3)

i =1

Where ai(λ) and yi(t) are the amplitude and the time-resolved fluorescence of each of the three forms of the HPPC molecule. At λ=620nm the major contribution is that of the zwitterion whereas at λ=680nm the major contribution is that of the deprotonated NRO- form. The signal at λ=560nm has already a mixture of the three bands. We used the weight sum of the signal at 560nm and at 680nm to fit the signals at λ≥600nm, as seen in table S2 of the SI. y (λ , t ) = a (560nm ) ⋅ y (560nm, t ) + [1 − a (560nm)] ⋅ y (680nm, t ) We assume that because of the fluorescence band overlap the signal at 680nm has also a contribution of the signal at 560nm with amplitude of a(560nm)=0.2 and larger contribution of the zwitterion. We get good fits to the long wavelength signals (λ≥600nm) with amplitude a(560nm) that ranges from the signal at λ=600nm with a(560nm)=0.735 and is reduced at λ=680nm to a(560nm)=0.2. Table S2 also provides the amplitudes for several signals of HPPC measured at λ≥600nm in D2O. Figure 5 shows the HPPC time-resolved spectra at several times constructed from the time-resolved fluorescence signals shown in Figure 3 and Figure 4.

14 Environment ACS Paragon Plus

Page 15 of 42

NROH (520nm)

0 fs 215 fs 425 fs 685 fs 1085 fs 1285 fs 1685 fs

12000

10000

t 8000

Signal

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

6000 +

HNRO- (600nm) NRO-+H3O+ (665nm)

4000

2000

t 0 500

550

600

650

700

Wavelength (nm)

Figure 5: Time-resolved spectra of HPPC in water, shown at several times and constructed from the time-resolved fluorescence spectra of Figure 3 and Figure 4.

The intensity of the NROH form fluorescence band at 520nm drops at a fast rate and forms the +HNRO- zwitterion and the anionic NRO- forms. The first step is a large fluorescence red shift of the NROH band with a time constant of ~120fs. In the discussion section we provide an explanation to this ultrafast time component. At intermediate times of few hundreds of femtoseconds, the protonation of the pyridine's nitrogen atom occurs with a band peak at 600nm. The +HNRO- band at 600nm equilibrates at longer times with the deprotonated form NRO-, whose band peak is at 665nm. The three fluorescence bands are broad and therefore the overlap between +

HNRO- and NRO- is large. The fluorescence cross section of the NROH form is

much larger than that of the +HNRO- or the NRO- bands. From the fluorescence upconversion signal intensity at t=tmax we deduce that the NRO- fluorescence cross section is about an order of magnitude smaller than that of the NROH form. Kinetic isotope effect Figure 6 shows on linear and semilogarithmic scales the fluorescence up-conversion signals of HPPC in D2O at short wavelength region of 500-580nm and figure S1 shows the signals at the long wavelength spectral region of 600-680nm.

15 Environment ACS Paragon Plus

The Journal of Physical Chemistry

D2O 500nm

D2O 520nm

D2O 560nm

D2O 580nm

D2O 540nm

b)

1.0

1.0

0.8

0.8

Norm. Signal

Norm. Signal

a)

0.6 0.4 0.2

D2O 500nm

D2O 520nm

D2O 560nm

D2O 580nm

D2O 540nm

0.6 0.4 0.2

0.0

0.0 0

500

1000

1500

2000

2500

0

5000

Time (fs)

c)

D2O 500nm

D2O 520nm

D2O 560nm

D2O 580nm

10000

15000

20000

25000

Time (fs)

D2O 540nm

1

Norm. Signal

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 16 of 42

0.1

0.01 0

5000

10000

15000

20000

25000

Time (fs)

Figure 6: Time-resolved fluorescence of HPPC in D2O, monitored at several wavelengths in the spectral region of 500-580nm a. Linear short scale b. Linear long scale c. Semilogarithmic scale

The decay of the signals measured at 500nm and 520nm (see Figure 6c) are nonexponential and the τav is about 1.2ps. The signals at λ≥540nm show a significant rise component followed by a fast and slow decay components. These signals of λ≥540nm hint that they consist of at least two fluorescence contributions of two forms of HPPC. At short wavelengths of 540,580,600 nm the tautomeric form fluorescence band overlaps strongly with the neutral form fluorescence spectrum. At longer wavelength λ≥620nm the fluorescence signal consists of contributions of three bands, the "N", "T" and the "A" bands (see Scheme 2). Figure 7 shows the time-resolved fluorescence of HPPC in H2O and D2O at several wavelengths.

16 Environment ACS Paragon Plus

Page 17 of 42

a)

b) 1.0

1.0 H2O 500nm

0.6 0.4

D2O 520nm

0.8

Norm. Signal

Norm. Signal

H2O 520nm

D2O 500nm

0.8

0.2

0.6 0.4 0.2

0.0

0.0 0

1000 2000 3000 4000 5000 6000 7000

0

Time (fs)

c)

1000 2000 3000 4000 5000 6000 7000

Time (fs)

d) 1.0

1.0 H2O 540nm

H2O 560nm

D2O 540nm

0.6 0.4

D2O 560nm

0.8

Norm. Signal

0.8

Norm. Signal

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

0.2

0.6 0.4 0.2

0.0

0.0 0

1000 2000 3000 4000 5000 6000 7000

Time (fs)

0

1000 2000 3000 4000 5000 6000 7000

Time (fs)

Figure 7: Time-resolved fluorescence of HPPC in H2O and D2O, monitored at several wavelengths in the spectral region of 500-560nm.

Each panel shows the HPPC signal in both H2O and D2O at a particular wavelength, at short wavelengths in the spectral region of 500-560nm. The fluorescence decay time of HPPC in D2O is about 2.5 times longer than in H2O. The proton/deuteron is first transferred from HPPC to a hydrogen bonded H2O or D2O molecule, and then by a water bridge of two water molecules it is transferred to the pyridine nitrogen to form the "tautomeric" form. The KIE of the total process is about 2.5. The signals shown in Figure 7 are also shown on a semilogarithmic scale in figure S2 in the SI. Figure S3 in the SI shows the fluorescence up-conversion signals of HPPC in both H2O and D2O at longer wavelengths of 620-680nm. Both H2O and D2O signals show rise components at short times followed by decay components with long decay times. The D2O signals show much longer rise time than that of HPPC in H2O. Figure 8a and b show the time-resolved fluorescence of HPPC in methanol at several wavelengths in the spectral region of 500-700nm.

17 Environment ACS Paragon Plus

The Journal of Physical Chemistry

a)

MeOH 500nm MeOH 560nm

MeOH 520nm MeOH 580nm

MeOH 540nm

MeOH 600nm MeOH 660nm

b)

1.0

1.0

0.8

0.8

Norm. Signal

Norm. Signal

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

0.6 0.4 0.2

Page 18 of 42

MeOH 620nm MeOH 680nm

MeOH 640nm MeOH 700nm

0.6 0.4 0.2

0.0

0.0 0

1000

2000

3000

4000

5000

0

1000

Time (fs)

2000

3000

4000

5000

Time (fs)

Figure 8: Time-resolved fluorescence of HPPC in methanol, monitored at several wavelengths in the spectral region of a. 500-600nm b. 600-700nm and shown on a linear scale

The signal at short wavelengths up to 560nm (Figure 8a) show a fast decay time. The steady-state fluorescence spectra of HPPC in methanol consists of two major emission bands. The short wavelength band peak is at 530nm and this fluorescence band is assigned to the neutral NROH band. The long wavelength band maximum is at 700nm and is assigned to the deprotonated anionic form the NRO-. The tautomeric HPPC fluorescence band may be present in methanol but if it exists then it's intensity is much smaller than in water and it is buried because of the overlap between the two broad emission bands of the NRO- and the NROH forms. Figure 8b shows the timeresolved fluorescence of HPPC at long wavelengths 600-700nm. The signals show at relatively short times rise components, which at long times are followed by long decay components. Figure 9 shows the time-resolved fluorescence of HPPC in methanol and methanol-d measured at several wavelengths in the short wavelength spectral region of 500-560nm.

18 Environment ACS Paragon Plus

Page 19 of 42

a)

1.0

MeOH 500nm MeOD 500nm

b) Norm. Signal

Norm. Signal

1.0

MeOH 520nm MeOD 520nm

0.8

0.8 0.6 0.4

0.6 0.4 0.2

0.2

0.0

0.0 0

5000

10000

15000

20000

0

25000

5000

c)

1.0

MeOH 540nm MeOD 540nm

10000

15000

20000

25000

Time (fs)

Time (fs)

d)

1.0

MeOH 560nm MeOD 560nm

0.8

Norm. Signal

0.8

Norm. Signal

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

0.6 0.4

0.6 0.4 0.2

0.2 0.0

0.0 0

5000

10000

15000

20000

25000

0

5000

Time (fs)

10000

15000

20000

Time (fs)

Figure 9: Time-resolved fluorescence of HPPC in MeOH and MeOD, monitored at several wavelengths in the spectral region of 500-560nm.

Each panel in the figure shows the signal of HPPC in both methanol and methanol-d at a particular wavelength. As seen in the figure the fluorescence average decay time τav in methanol-d is larger than in methanol-h. This fact implies that there exists a kinetic isotope effect (KIE). The effective kinetic isotope effect is deduced from the decay times of the signals at 500nm and 520nm is about 1.7±0.2. This KIE is distinctively smaller than that of H2O/D2O which is 2.5±0.2. The reason for a smaller kinetic isotope effect on methanol/methanol-d may arise from nonradiative fluorescence decay of the neutral NROH form which may have a decay time of the same order as the proton transfer reaction. In such a case, the fluorescence decay rate constant of the NROH form is given by three contributions kF = kr + knr + kPT . If knr is H

D

of the order of kPT then the effective KIE obtained by kF / kF depends also on the nonradiative process.

19 Environment ACS Paragon Plus

25000

The Journal of Physical Chemistry

Figure S4 shows the time-resolved fluorescence of HPPC in methanol and methanol-d measured at several long wavelengths in the spectral region of 620680nm. The decay time of the signal at these wavelengths are relatively long. The signal at short times shows a rise with wavelength dependent amplitude and time constant. The signal rise at 640nm in methanol and methanol-d is about 1250fs for methanol and 2500fs for methanol-d, whereas at 680nm the rise times are 4ps and 8ps for methanol and methanol-d respectively. HPPC spectroscopy in trifluoroethanol (TFE) Figure 10 shows the normalized steady-state fluorescence of HPPC in TFE and TFE: ethanol mixtures. Bulk TFE

xTFE=0.62 xEtOH=0.38

xTFE=0.89 xEtOH=0.11

xTFE=0.55 xEtOH=0.45

xTFE=0.8 xEtOH=0.2

xTFE=0.49 xEtOH=0.51

xTFE=0.73 xEtOH=0.27

xTFE=0.45 xEtOH=0.55

xTFE=0.67 xEtOH=0.33

TFE and high EtOH Conc.

1.0 0.8

Norm. Signal

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 20 of 42

0.6 0.4 0.2 Bulk TFE

0.0 450

500

550

600

650

700

750

Wavelength (nm)

Figure 10: Normalized steady-state fluorescence spectrum of HPPC in TFE and TFE: ethanol mixtures, excited at 420nm shown on a linear scale.

In TFE the HPPC spectrum consists of a single fluorescence band, that of the NROH form with a maximum at ~525nm. We therefore conclude that ESPT from the hydroxyl group of HPPC in TFE does not occur within the excited state lifetime of the ROH lifetime. As ethanol is added to the TFE solvent an ESPT process takes place. At about 50% mol ratio of TFE-ethanol mixture the NRO- band and NROH band intensities are about equal.

20 Environment ACS Paragon Plus

Page 21 of 42

Figure 11 shows the time-resolved fluorescence of HPPC in TFE and ethanol measured at several wavelengths by the fluorescence up-conversion technique. a)

b)

1.0 EtOH 480nm TFE 480nm

0.6 0.4

1.0 EtOH 500nm TFE 500nm

0.8

Norm. Signal

Norm. Signal

0.8

0.2

0.6 0.4 0.2

0.0

0.0 0

5000

10000

15000

20000

25000

0

5000

Time (fs)

c)

e) EtOH 520nm TFE 520nm

Norm. Signal

Norm. Signal

15000

20000

25000

0.6 0.4

1.0 0.8

0.2

EtOH 540nm TFE 540nm

0.6 0.4 0.2

0.0

0.0 0

5000

10000

15000

20000

25000

0

5000

Time (fs)

g)

1.0 0.8

10000

15000

20000

25000

Time (fs)

EtOH 560nm TFE 560nm

0.6 0.4 0.2

1.0 0.8

Norm. Signal

f)

10000

Time (fs)

1.0 0.8

Norm. Signal

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

EtOH 580nm TFE 580nm

0.6 0.4 0.2

0.0

0.0 0

5000

10000

15000

20000

25000

0

5000

Time (fs)

10000

15000

20000

25000

Time (fs)

Figure 11: Time-resolved fluorescence of HPPC in EtOH and EtOD, monitored at several wavelengths in the spectral region of 500-600nm

Each of the panels in the figure shows the fluorescence up-conversion signal of HPPC in both ethanol and TFE. At short wavelengths, 480nm and 500nm, both ethanol and TFE signals show an instantaneous signal rise followed by a decay. In ethanol the

21 Environment ACS Paragon Plus

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

decay is much faster than in TFE. At longer wavelengths λ≥520nm the signal of HPPC in ethanol shows a fast decay with a time constant that depends on the monitored wavelength. The average decay time is determined by the ESPT process, by the solvent dynamics and by nonradiative decay. In TFE the HPPC signal shows a signal rise followed by a long decay time. At 600nm the main rise time component is about 10ps and the decay time is about 200ps. The fluorescence signal of HPPC in TFE differs from that in ethanol because of the lack of the ESPT in TFE. The solvation dynamics of coumarin-30 dye (shown in scheme S1) in ethanol and TFE is similar. Figure S5 in the SI shows the fluorescence up-conversion signals of coumarin 30 in both solvents at several wavelengths. As seen in the figure the single fluorescence band of coumarin 30 shows a time dependent red band shift. The solvation dynamics of coumarin dyes (coumarin 153) was measured and analyzed by Maroncelli and coworkers.64 The solvation dynamics show a complex decay pattern of several time components with short (100fs) and longer time constants. As seen in figure S5 of the SI, the fluorescence of coumarin 30 dye in both TFE and ethanol shows indeed a dynamic red shift of the emission band in both solvents with about the same time constants at λ≥480 and somewhat faster solvent response for TFE at shorter wavelengths. To summarize: the comparison of the signals of HPPC in TFE and ethanol with that of coumarin 30 in the same solvents shows that in the coumarin 30 dye, in which the ESPT process does not occur, the differences between the fluorescence up-conversion signals in ethanol and TFE are not as drastic as for HPPC in these solvents. This difference arises from the ultrafast ESPT process that takes place in HPPC in ethanol. Computational calculations results In order to further characterize HPPC and its photoprotolytic cycle, we have proceeded to its computational study in aqueous solution. The Franck-Condon properties of the neutral, tautomeric and anionic species of HPPC, in implicit water, are presented in Table 2. These three species were also studied in the presence of two explicit water molecules, besides employing an implicit solvation model for water, and their Franck-Condon properties are presented in Table 3. We have used two water molecules because, as will be shown in the text, this number of solvent molecules is needed for an efficient PT from the hydroxyl group to the pyridine's nitrogen atom of HPPC. 22 Environment ACS Paragon Plus

Page 22 of 42

Page 23 of 42

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

Table 2: Excitation energies (Eex, in eV), transition types and major orbital contribution corresponding to the three studied HPPC, in the absence of explicit water. Species Neutral

Anion

Tautomer

Eex

Transition

Orbital Contribution

3.37

S0→S1

HOMO→LUMO

3.53

S0→S2

HOMO→LUMO(+1)

2.28

S0→S1

HOMO→LUMO

3.19

S0→S2

HOMO→LUMO(+1)

2.37

S0→S1

HOMO→LUMO

3.19

S0→S2

HOMO→LUMO(+1)

The Franck-Condon properties of these three species present some similarities with the S1 state being composed mostly by a HOMO → LUMO transition, while the S2 state is composed by a HOMO → LUMO(+1) excitation. Moreover, all these excitations have π→π* character. Our results are in line with the absorption bands in the UV region seen in Figure 1. We can see that the theoretical values for the S0→S2 transition differs only 0.00-0.06 eV from experiment (depending of the presence or absence of explicit water molecules), while for the S0→S1 transition the difference is between 0.24-0.28 eV, which is still within the typical error of density functionals. It should be noted that the inclusion of explicit water molecules increases the agreement of theory with experiment. Table 3: Excitation energies (Eex, in eV), transition types and major orbital contribution corresponding to the three studied HPPC, in the presence of explicit water. Species Neutral

Anion

Tautomer

Eex

Transition

Orbital Contribution

3.33

S0→S1

HOMO→LUMO

3.48

S0→S2

HOMO→LUMO(+1)

2.42

S0→S1

HOMO→LUMO

2.91

S0→S2

HOMO→LUMO(+1)

2.39

S0→S1

HOMO→LUMO

3.15

S0→S2

HOMO→LUMO(+1)

23 Environment ACS Paragon Plus

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

The assignment of the absorption band with a peak at 480 nm is more complex. If we only consider the energy values of the S0→S1 transition of both the tautomeric and anionic HPPC forms, in the absence of explicit solvation, we would attribute this band to the tautomer (a difference of 0.21 eV in comparison to 0.30 eV of the anionic species). However, the inclusion of explicit solvation increases significantly the energy of the S0→S1 transition of the anion (by 0.14 eV), while that of the tautomer remains basically the same. This increase in energy leads us to assign the anionic species to the 480nm band in the experimental spectrum. Thus, this finding coupled with the one presented ahead in the manuscript, in which we demonstrate that the tautomer should not be formed in the S0 state, allow us to attribute the 480 nm absorption band to the anionic HPPC species. This large effect of explicit water in the Franck-Condon state of the anionic state can be attributed to changes in electron density upon excitation. In Figure 12 are plotted the barycenters of the electron density of this species, upon excitation, and in the presence/absence of explicit solvation. We can see that in the absence of explicit water, excitation triggers electron transfer from the central pyridine ring to both the conjugate hydrocarbon chains, connecting the pyridine to the two picolinium rings of HPPC. Addition of explicit water changes the symmetric electron transfer to only one of the hydrocarbon chains.

Figure 12: Barycenters of the negative (blue region) and positive (green region) parts

of the electron density, upon excitation, of anionic HPPC. Figure 12a refers to the absence of explicit water molecules, while Figure 12b refers to the presence of two explicit solvent molecules. The next step in the calculations is to check the possible formation of the tautomeric HPPC species, by a water-mediated PT from the hydroxyl group of the central pyridine ring to the nitrogen heteroatom present in the same moiety (Figure

24 Environment ACS Paragon Plus

Page 24 of 42

Page 25 of 42

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

13). This reaction consists of a PT from the hydroxyl group by a two-molecule water bridge, which will propagate the proton towards the pyridine nitrogen. This first step will form a geminate ion pair, and from it the donated proton can either diffuse to the bulk solvent (forming the anionic species), or it can recombine with the nitrogen heteroatom (leading to the formation of the tautomer). The reaction was modelled by performing an IRC calculation, which started from the transition state towards the reactants and products. The reaction was followed until structures with zero imaginary frequencies, signaling the presence of minima in the PES of the water-mediated PT reaction, were found (Figure 13). This reaction is not expected to occur in the S0 state, due to a 20.9 kcal mol-1 activation barrier and due to the fact that the solvated hydroxyl species is more stable than the tautomer by 3.8 kcal mol-1. So, we do not expect the tautomer species to be formed in the S0 state, in aqueous solution, given the energetic unaffordability of this reaction. Thus, we attribute the 480 nm absorption band to anionic HPPC (and not to the tautomeric form), which should only be formed by a PT reaction between HPPC and a hydroxyl anion (at basic pH). This is not the case of a PT reaction in the S1 PES. The activation barrier in S1 is only of 4.3 kcal mol-1, while the solvated tautomer is more stable than hydroxyl HPPC by 17.1 kcal mol-1. Thus, our theoretical calculations indicated that this ESPT reaction is indeed energetically favorable, which explains the photoacidity of HPPC. Moreover, these results allows us to assign the fluorescence band that peaks at ~600 nm (in aqueous solution) to the tautomer species.

25 Environment ACS Paragon Plus

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

Figure 13: Potential energy curves, as a function of intrinsic reaction coordinates, of

the water-mediated conversion of neutral HPPC into its tautomeric species.

It should be noted that upon excitation, the S1 state of the neutral species is very close in energy to its S2 state (by about 4.2 kcal/mol-1). Thus, S1→S2 internal conversion is possible. However, from that point onwards the energy difference increases significantly (up to 20.7 kcal mol-1), which decreases the possibility of S1→S2 internal conversion during the ESPT reaction. It should be noted that some doubts regarding if the “two-water” bridge is indeed stable throughout the reaction, or if results only as an artifact due to the absence of other explicit water molecules. To assess this, we have added 80 explicit water molecules to the IRC-obtained structures of the hydroxyl and tautomer species complexed with the “two-water” bridge (Figure S6, with the methodological details described in the Quantum Mechanical Methods subsection and also found in the Supporting Information). The geometry optimization of these complexes revealed that the “two-water” bridge is maintained for both structures, even in the presence of other explicit water molecules, thus supporting its stability. Moreover, calculations regarding the S0 stability of the hydroxyl and tautomer species in these structures revealed that the latter species is more stable than the former one by only -0.4 kcal mol-1. This is a small difference for the relative stability calculated above, of 3.8 kcal

26 Environment ACS Paragon Plus

Page 26 of 42

Page 27 of 42

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

mol-1, which indicates that the solvation effect made by the 80 explicit water molecules has only a small effect in the hydroxyl and tautomer species. Figure 14 shows the changes in length between two different bonds, during the formation of the tautomeric species: that between the hydroxyl group’s oxygen and the hydroxyl group’s hydrogen (OOH-HOH), and that between the central pyridine’s nitrogen and the hydrogen of a nearby water molecule (N-HWat). It can be seen that the first major change is an increase in OOH-HOH length, which is not followed by NHWat. Only after the formation of the geminate ion pair does N-HWat decreases significantly, but at this point the increase in OOH-HOH is much smaller. This indicates that this is a stepwise reaction mechanism, and not a concerted one, since first there is a clear rupture of the OOH-HOH bond and only then does the new N-HWat covalent bond begin to be formed. This reaction mechanism is in line with that found experimentally for another intramolecular ESPT-based tautomerism reactions.65

Figure 14: Changes in bond length of OOH-HOH and N-HWat, as a function of intrinsic

reaction coordinate. Main findings 1. Excited-state proton transfer (ESPT) occurs in HPPC in protic solvents like water, methanol, ethanol and propanol. 2. The ESPT rate of HPPC depends on the solvent. In water the ESPT rate is less than 1ps whereas in methanol and ethanol it is about 3ps and 6ps respectively. 3. A kinetic isotope effect (KIE) is observed. In water it is about 2 and it reduces in methanol to about 1.7 whereas in ethanol it is only about 1.25 because of a competing nonradiative process.

27 Environment ACS Paragon Plus

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

4. The reduced KIE in methanol and ethanol with respect to water indicates that the nonradiative rate of the NROH form of HPPC in methanol and ethanol is of the order of the ESPT process. 5. In water solution the major ESPT process protonates the pyridine's nitrogen atom (see Scheme 2) via a water bridge of two water molecules and the predominant excited state form of HPPC is that of the tautomer. Discussion The photoacids family of compounds can be divided into four groups.66 The weak photoacids with pKa*≥0.4 are incapable of transferring a proton within the excited state lifetime (up to 10ns), to methanol or ethanol. Examples of weak photoacids are the phenol (pKa*=3.4), 2-naphthol (pKa*=2.7) and 8-hydroxy-1,3,6pyrenetrisulfonate (HPTS) with pKa*=1.4. In water the ESPT rate constant of weak photoacids is close to linear with the free energy ∆G of the reaction. The ESPT rate to water ranges between 2×107s-1 (phenol- pKa*=3.4) and 2×1010s-1 for 2-naphthol-6,8disulfonate with pKa*≈0.7. The ESPT rate to methanol and ethanol of weak acids is about three orders of magnitude smaller than in water and kr>>kPT. For stronger photoacids with pKa*