Is π-Stacking Prone To Accelerate Singlet–Singlet Energy Transfers

Mar 23, 2018 - π-Stacking is the most common structural feature that dictates the optical and electronic properties of chromophores in the solid stat...
0 downloads 0 Views 3MB Size
Article Cite This: Inorg. Chem. XXXX, XXX, XXX−XXX

pubs.acs.org/IC

Is π‑Stacking Prone To Accelerate Singlet−Singlet Energy Transfers? Di Gao, Shawkat M. Aly, Paul-Ludovic Karsenti, and Pierre D. Harvey* Département de Chimie, Université de Sherbrooke, Sherbrooke, Quebec J1K 2R1, Canada S Supporting Information *

ABSTRACT: π-Stacking is the most common structural feature that dictates the optical and electronic properties of chromophores in the solid state. Herein, a unidirectional singlet−singlet energytransfer dyad has been designed to test the effect of π-stacking of zinc(II) porphyrin, [Zn2], as a slipped dimer acceptor using a BODIPY unit, [bod], as the donor, bridged by the linker C6H4C CC6H4. The rate of singlet energy transfer, kET(S1), at 298 K (kET(S1) = 4.5 × 1010 s−1) extracted through the change in fluorescence lifetime, τF, of [bod] in the presence (27.1 ps) and the absence of [Zn2] (4.61 ns) from Streak camera measurements, and the rise time of the acceptor signal in femtosecond transient absorption spectra (22.0 ps), is faster than most literature cases where no π-stacking effect exists (i.e., monoporphyrin units). At 77 K, the τF of [bod] increases to 45.3 ps, indicating that kET(S1) decreases by 2-fold (2.2 × 1010 s−1), a value similar to most values reported in the literature, thus suggesting that the higher value at 298 K is thermally promoted at a higher temperature.



INTRODUCTION Solid-state packing of conjugated polymers and small molecules plays a significant role in bulk heterojunction solar cells by influencing the morphology, which in turn promotes a better charge migration.1,2 Recent advances were made in porphyrincontaining materials and π,π-stacking, leading to J-aggregates, has been proposed to be responsible for better power conversion efficiency (PCE), namely by exhibiting a broad and red-shifted absorption band approaching the near-IR region and thus prone to better light harvesting process.3−5 A few small molecules and conjugated polymers containing porphyrins and boron−dipyrromethene groups (BODIPY) have shown great promises for bulk heterojunction solar cells.6−8 The main arguments to combine the two chromophores together was the suspected efficient antenna effect (i.e., singlet energy transfer) between the two, and the formation of J-aggregates of the porphyrin units thus potentially promoting communication between them. Although it is logical to rationalize that more absorption in the near-IR region promotes the capture of more photons, it is not clear whether an efficient energy transfer can really occur when J-aggregates are formed due to the complexity of the solid state environment. Instead, a molecular model carrying a well-defined J-aggregate of porphyrins, exhibiting more or less the same structural features of what is generally encountered in the literature, attached to an antenna, is more appropriate to conduct such an endeavor. Dyad 1 (Chart 1) exhibits a slipped dimer equivalent to a Jaggregate containing two zinc(II)porphyrins, [Zn2], acting as the energy acceptor, and a BODIPY, [bod], as the energy donor, linked by a C6H4CCC6H4 bridge. The presence of methyl groups at the β-positions of [bod], renders the dihedral angle between the [bod] and C6H4 planes nearly perpendicular, thus minimizing π-conjugation. Indeed, dyad 1 exhibits the © XXXX American Chemical Society

fastest (equal with another example) rate for singlet energy transfer, kET(S1), in this category of dyads [Zn]-C6H4C CC6H4-[bod] (mono zinc(II)porphyrin).



EXPERIMENTAL SECTION

Materials. Compounds 4 and 5 were synthesized according to literature procedures.9,10 All reagents were used as received. Toluene and tetrahydrofuran were distilled from benzophenone and sodium. Triethylamine was distilled over calcium hydride. Flash column chromatography was performed on silica gel 60 (230−400 mesh). Compound 2. First, 6mL of freshly distilled toluene and 3 mL of dry DMF were added to a mixture of 4 (15 mg, 0.01 mmol), [(4bromophenyl)ethynyl]triisopropylsilane (17 mg, 0.05 mmol), Cs2CO3 (6 mg, 0.02 mmol) and Pd(PPh3)4 (1 mg, 0.001 mmol) under an Ar atmosphere. The mixture was heated at 100 °C for 16 h and the solvent was evaporated under vacuum. The solid was passed a column chromatography (silica, CH2Cl2/hexane, 1:3) to get a brown powder. 0.015 mL of tetrabutylammonium fluoride was added to a DCM solution containing the resulting powder. The mixture was stirred 10 min and the solvent was evaporated. The solid was passed through a column chromatography (silica, CH2Cl2/hexane, 1:3) yielding a brown powder (9 mg, 62%). 1 H NMR (300 MHz, CD2Cl2, 25 °C): δ = 9.39 (brs, 1H, H-meso), 8.99 (brs, 4H, H-β), 8.71−8.53 (m, 10H, H-β), 8.42 (brs, 1H, H−Ar), 8.27 (s, 2H, H-β), 7.76 (brs, 1H, Ar−H), 7.55 (brs, 2H, Ar−H), 7.32−7.23 (brs, 4H, Ar−H), 7.01 (brs, 1H, Ar−H), 6.82 (brs, 1H, Ar−H), 3.87−3.86 (brs, 8H), 3.49 (d, 1H), 1.98 (brs, 8H), 1.54 (m, 24H), 1.33 (s, 16H), 0.91 ppm (brs, 12H). MALDITOF: m/z: calcd for C92H96N8Zn2 1443.63 [M+H]+, found 1444.66. Compound 1. A sample of 2 (7 mg, 0.005 mmol), 5 (5 mg, 0.01 mmol) were weighed into a 25 mL round-bottom flask, which was pump-purged three times with Ar. Dry, degassed toluene/TEA (3 mL, 5:1) was added to the flask, and the mixture was bubbled with Ar for Received: December 3, 2017

A

DOI: 10.1021/acs.inorgchem.7b03050 Inorg. Chem. XXXX, XXX, XXX−XXX

Article

Inorganic Chemistry Chart 1. Structures of the Investigated Dyad 1 and the Model Compounds 2 and 3

Scheme 1. Synthesis of 1_3a

R = n-octyl. Reagents and conditions: (i) a) CsCO3, Pd(PPh3)4, 4-BrC6H4CCSi(i-Pr)3, DMF, 100 °C; b) TBAF, DCM; (ii) Pd2(dba)3, AsPh3, toluene/Et3N, 35 °C.

a

7.5 Hz, 6H). MALDI-TOF: m/z: calcd for C31H31BF2N2 480.25, found 480.23[M], 461.23[M−F]. Instruments. 1H Nuclear Magnetic Resonance (NMR) spectra were recorded on a Bruker Avance 300 Ultrashield NMR spectrometer. Chemical shifts are given in ppm relative to residual peaks of chloroform (δ = 7.26 ppm), or dichloromethane (δ = 5.32 ppm). The absorption spectra were measured on a Hewlett-Packard 8452A diode array spectrometer (0.1 and 1 s integration time at 77 and 298 K, respectively). Steady-state emission spectra acquired both on Edinburgh Instrument FLS980 phosphorimeter equipped with single monochromator and Quanta Master-400 phosphorimeter from Photon Technology International (PTI). Fluorescence lifetimes were measured with the FLS980 spectrometer using a picosecond (ps) pulsed diode laser as source of excitation (fwhm ≈ 90 ps). Measurements of the time-resolved fluorescence were acquired by time-correlated single photon counting (TCSPC) where scan is subdivided into decay and instrumental response function (IRF) measurements. Lifetimes were extracted using Exponential Reconvolution Fit analysis software provided by Edinburgh Instruments. Fast fluorescence dynamics were monitored using a Streak camera (AxisTRS, Axis Photonique Inc.) with about 7−10 ps resolution where the excitation is the second harmonic generation (SHG) of a Soltice (Spectra-Physics) Ti−sapphire laser (λex = 495 nm; fwhm ≈100 fs; pulse energy = 0.1 mJ per pulse, rep. rate = 1 kHz; spot size ≈2 mm). Time constants associated with the decay curves were extracted by global analysis of the corresponding fluorescence decay curves using the Glotaran (http://glotaran.org) analysis program. The femtosecond (fs) transient absorption (TA) measurements were acquired on a homemade setup with a Ti−sapphire source (Soltice, Spectra Physics). The measurements were made using the output of an OPA (OPA-800 CF, Spectra-Physics) operating at λex = 425 nm, a pulse width of ∼90

10 min, Pd2(dba)3 (1 mg, 0.001 mmol) and AsPh3 (3 mg, 0.01 mmol) were added to the mixture and the reaction mixture was stirred at 35 °C under argon for 12 h. The solvent was evaporated under vacuum and the crude solid was passed through a column chromatography (silica, CHCl3/hexane, 1:1) to afford a brown compound (7 mg, 76%). 1 H NMR (300 MHz, CD2Cl2, 25 °C): δ = 9.45 (dd, 1H, J = 1.4, 7.8 Hz, H-meso), 9.03 (s, 1H, Ar−H), 9.00 (d, 2H, J = 4.7 Hz, H-β), 8.69 (m, 8H, H-β), 8.55 (t, 4H, J = 5.1 Hz, H-β), 8.50 (dd, 1H, J = 1.6, 7.9 Hz, Ar−H), 8.28 (d, 2H, J = 4.5 Hz, H-β), 7.95 (d, 2H, J = 8.2 Hz, Ar−H), 7.85 (dd, 1H, J = 1.6, 7.6 Hz, Ar−H), 7.61−7.57 (m, 2H, Ar− H), 7.49 (d, 2H, J = 8.2 Hz, Ar−H), 7.37−7.29 (m, 2H, Ar−H), 7.24 (d, 1H, J = 6.5 Hz, Ar−H), 7.01 (dd, 1H, J = 6.9, 8.3 Hz, Ar−H), 6.83 (d, 1H, J = 8.0 Hz, Ar−H), 3.91−3.67 (brs, 8H), 2.56 (s, 6H), 2.42 (q, 4H), 2.00 (brs, 8H), 1.72−1.64 (m, 16H), 1.36−1.32 (m, 30H), 1.07 (t, 6H, J = 7.5 Hz), 0.95−0.83 ppm (m, 12H). MALDI-TOF: m/z: calcd for C115H121BF2N10Zn2 1821.84 [M+H]+, found 1823.03 [M +H]+, 1804.04 [M+−F]. Compound 3. A sample of phenylacetylene (8 μL, 0.08 mmol), 5 (20 mg, 0.04 mmol) were weighed into a 25 mL round-bottom flask which was then pump-purged three times with argon. Dry, degassed toluene/triethylamine (3 mL, 5:1) was added to the flask, and the mixture was bubbled with argon for 10 min, then Pd2(dba)3 (5 mg, 0.006 mmol) and AsPh3 (12 mg, 0.04 mmol) were added to the mixture and the reaction mixture was stirred at 35 °C under argon for 12 h. The solvent was evaporated under vacuum and the crude solid was passed a column chromatography (silica, CH2Cl2/hexane, 1:3) to afford violet compound (11 mg, 60%). 1 H NMR (300 MHz, CDCl3, 25 °C): δ = 7.66 (d, 2H, J = 8.3 Hz, Ar−H), 7.57 (dd, 2H, J = 4.8, 7.4 Hz, Ar−H), 7.38 (m, 3H, Ar−H), 7.29 (d, 2H, J = 8.4 Hz, Ar−H), 2.54 (s, 6H), 2.31 (q, 4H, J = 7.5 Hz), 1.34 (s, 6H), 0.99 ppm (t, J = B

DOI: 10.1021/acs.inorgchem.7b03050 Inorg. Chem. XXXX, XXX, XXX−XXX

Article

Inorganic Chemistry fs, repetition rate = 1 kHz, spot size ∼500 μm, a white light continuum generated inside a sapphire window and a custom-made dual CCD camera of 64 × 1024 pixels sensitive between 200 and 1100 nm (S7030, Spectronic Devices). The results were also globally analyzed with the Glotaran program, permitting to extract a sum of independent exponentials,

I(λ , t ) = C1(λ) e−t1/ τ |C 2(λ) e−t2 / τ | + ... that fits the whole 3D transient map. The quantum yield measurements were carried out at 298 K in 2MeTHF using tetraphenylporphyrin, H2TPP (ΦF = 0.11)11 as references. Four different solutions for both sample and reference were prepared in glovebox under argon atmosphere (O2 < 13 ppm) and used for the measurements. Concentrations were adjusted so as to have an absorbance of ∼0.05 at the excitation wavelength and absorption spectra recorded five times for accuracy and error minimization. Computations. All density functional theory (DFT) and time dependent density functional theory (TD-DFT) calculations were performed with Gaussian 0912 at the Université de Sherbrooke with the Mammouth supercomputer supported by Le Réseau Québécois De Calculs Hautes Performances. The DFT geometry optimizations as well as TD-DFT calculations13−21 were carried out using the B3LYP method. A 6-31g* basis set was applied to C, H, N, O, B, F atoms in porphyrins, VDZ (valence double ζ) with SBKJC effective core potentials were used for all Zn atoms.22−27 All calculations were carried out in a THF solvent field. The calculated absorption spectra were obtained from GaussSum 2.1.28

Figure 1. Geometry optimization of 1 (vide inf ra). The dihedral angles are [Zn]-C6H4, 57.51°, C6H4−C6H4, 2.76°, C6H4-[bod], 88.47°, [Zn][bod], 27.09°, center-to-center distance = 17.76 Å, Cmeso-Cmeso = 4.014 Å, and smallest C···Cmeso separation near the C6H4CCC6H4 linker = 4.644 Å. A contact between the ortho-H of the C6H4 ring and the β-H of the upper [Zn] is obvious. Color code: C, dark gray; H, pale gray; N, purple; B, pink; F, green.



Spectroscopy. Bolze et al. demonstrated that the use of DPB as a spacer between the two porphyrin units induces a broadening of both the absorption and fluorescence band.35 This feature is well illustrated in compounds 1 and 2 (Figure 2). This spectral feature indicates the presence of [Zn]···[Zn] interactions, which in the context of this work, means that [bod] can capture a photon and then can relay this excitation energy to two [Zn] chromophores ([Zn2]), instead of one ([Zn]), forming together a single entity. This single entity “behavior” was recently corroborated using DFT computations by Steer and collaborator who concluded that there is a large degree of MO coupling between the two [Zn] units.36 The presence of a significant MO coupling in 1 was also verified by DFT computations in order to address whether the increase in average [Zn]···[Zn] distance (i.e., from 3.80 Å (X-ray)30−33 to an average of 4.3 Å (DFT computations) affects this trait (see below). The positions of the 0−0 absorption and fluorescence peaks in the model compounds 2 and 3 confirm that [bod] and [Zn2] act as S1 energy donor and acceptor, respectively. The fluorescence quantum yield, ΦF, of 3 is 87% (uncertainties ±10%; Table 1) meaning that the energy donor exhibits a rather intense fluorescence. The emission spectra of 1 exhibits both fluorescence’s ([bod] and [Zn2]). The total ΦF value for 1 is 2.9% (this value represents the sum of both fluorescence bands) and indicates efficient S1 energy transfer [bod]* → [Zn2] since there is an obvious decrease in ΦF values going from 3 to 1. This efficiency is corroborated by the comparison of the excitation and absorption spectra as they exhibit a good match when monitored in the emission band of the [Zn2] chromophore. The spectral overlap between the fluorescence of [bod] and the absorption band of [Zn2] is significant as the fluorescence band of [bod] is completely integrated within the absorption band of [Zn2], meaning that the J-integral, one key parameter of both the Forster resonance energy transfer (FRET)37 and Dexter mechanisms of energy transfer (double electron exchange),38 is maximized (see description in Figures S10

RESULTS AND DISCUSSION Synthesis. The synthetic route toward cofacial-special pair is presented in Scheme 1. Compound 4 has been synthesized according to the literature by the Suzuki-coupling reaction adapting the reactions for R = n-octyl (see the Experimental Section for detail).9,10 After the Suzuki-coupling with 4BrC6H4CCSi(i-Pr)3 and deprotection step i), compound 2 was obtained. Bodipy 529 was linked onto the artificial special pair 2 by the Sonogashira coupling reaction with the AsPh3/ Pd2(dba)3 as the catalyst. For comparison purposes, the corresponding antenna model 3 (bridge-[bod]) was also prepared by Sonogashira reaction using the same synthetic approach (step ii)). The compounds were characterized by 1H NMR spectroscopy and mass spectrometry (Figures S1−S7). Computations. In the absence of X-ray structure, geometry optimization was performed using DFT calculations (B3LYP) to extract the relevant structural parameters (Figure 1). Often, the [Zn] chromophore used in the literature is flanked by aromatic groups at the meso-position of the macrocycle. For biphenylene-containing slipped dimers, these aromatics induce obvious steric hindrance between the ortho-H of the C6H4 ring and the β-H of the upper [Zn]. This situation is desired in this work since there is a need for an accurate mimic of the [Zn]··· [Zn] contacts. Indeed, the separation between the C-atoms in the DPB spacer (diphenylene) linking the porphyrin macrocycle is in general 3.80 Å (X-ray data).30−33 The geometry optimization of 1 exhibits a distance of 3.86 Å, which compares reasonably. Similarly, in the absence of substrate between the two macrocycles, literature X-ray data indicates that the CmesoCmeso is, again, 3.80 Å,30−33 but in the presence of a substrate, this distance increases to 4.176 Å,34 hence forcing the interplane distance to increase. The steric hindrance between the ortho-H of the C6H4 ring and the β-H of the upper [Zn] also induces an increase in interplane distance and DFT computes a Cmeso-Cmeso of 4.01 Å. Near the contact, the calculated separation is ∼4.64 Å. C

DOI: 10.1021/acs.inorgchem.7b03050 Inorg. Chem. XXXX, XXX, XXX−XXX

Article

Inorganic Chemistry

Figure 2. Absorption (black), excitation (red), and fluorescence (blue) spectra of 1−3 in 2MeTHF at 298 K (left) and 77 K (right).

Table 1. Fluorescence Lifetimes Using TCSPC (in 2MeTHF) at 298 and 77 K 298 K λexa

λem

2 1

395 443

3

477

675 541 667 548

b

77 K χ

τF (ns)

λex

1.089 0.994 0.976 1.075

2.05 0.032f 2.07 4.61

443 395

2

c

a

λemb

χ2

τF (ns)

ΦF (%)

670 534 674 534

1.075 1.198 1.052 1.051

2.48 0.034f 2.51 5.67

d 2.9e

477

87e

Excitation wavelength (nm). Monitoring wavelength (nm). ±5%. Not measured. Quantum yield measured using H2TPP as the comparative standard (λexc = 500 nm). fNot reliable, as the fwhm of the pulse is ∼90 ps. a

b

c

d

e

entity. Compound 3 exhibits the expected τF values for a BODIPY chromophore (4.61 ns) while the τF value for the [bod] chromophore in 1 decreases below the limit of the technique (from 4.61 ns to