to Os - American Chemical Society

Article pubs.acs.org/JPCB

Distance Dependence of Intrahelix RuII* to OsII Polypyridyl ExcitedState Energy Transfer in Oligoproline Assemblies1 M. Kyle Brennaman, Cavan N. Fleming, Cheryl A. Slate, Scafford A. Serron, Stephanie E. Bettis, Bruce W. Erickson, John M. Papanikolas, and Thomas J. Meyer* Department of Chemistry, University of North Carolina at Chapel Hill, Chapel Hill, North Carolina 27599, United States S Supporting Information *

ABSTRACT: Energy transfer between the metal-to-ligand charge transfer (MLCT) excited states of [Pra [M II(bpy) 2(4-Me-4′(-N(H)CO)bpy)](PF6) 2 units ([Pra(MIIbpy2(mbpy)]2+: MII = RuII or OsII, bpy = 2,2′-bipyridine, mbpy = 4′-methyl-2,2′bipyridine-4-carboxamido, Pra = 4-MII-L-proline) linked covalently to oligoproline assemblies in room temperature acetonitrile occurs on the picosecond−nanosecond time scale and has been time-resolved by transient emission measurements. Three derivatized oligoprolines, [CH3-CO-Pro6-Pra[OsII(bpy)2(mbpy)]2+-Pro2-Pra[RuII(bpy)2(mbpy)]2+Pro2-Pra[RuII(bpy)2(mbpy)]2+-Pro6-Glu-NH2]6+ (ORR-2, Pro = L-proline and Glu = glutamic acid); [CH3-CO-Pro6-Pra[OsII(bpy)2(mbpy)]2+-Pro3-Pra[RuII(bpy)2(mbpy)]2+Pro3-Pra[RuII(bpy)2(mbpy)]2+-Pro6-Glu-NH2]6+ (ORR-3); and CH3-CO-Pro6-Pra[OsII(bpy)2(mbpy)]2+-Pro5-Pra[RuII(bpy)2(mbpy)]2+-Pro5-Pra[RuII(bpy)2(mbpy)]2+Pro6Glu2-NH2]6+ (ORR-5), were prepared by using solid-phase peptide synthesis. Given the helical nature of the resulting assemblies and the nature of the synthesis, composition, length, and loading pattern are precisely controlled in the assemblies. In acetonitrile, they adopt a proline I helical secondary structure, confirmed by circular dichroism, in which the appended chromophores are ordered in well-defined orientations and internuclear separation distances although helix formation for ORR-2 is incomplete. Quantitative comparison of oligoproline ground-state absorption and steady-state emission spectra to those for the constituents, [Boc-Pra[MII(bpy)2(mbpy)]2+OH](PF6)2 (Boc = Nα-(1,1-dimethylethoxycarbonyl), shows that following RuII light absorption, RuII* undergoes facile energy transfer resulting in sensitization of OsII. Sensitization efficiencies are 93% for ORR-2, 77% for ORR-3, and 73% for ORR-5. Picosecond-resolved emission measurements reveal complex, coupled dynamics that arise from excited-state decay and kinetically competitive −RuII*−RuII− → −RuII−RuII*− energy transfer migration/exchange and downhill −RuII*−OsII → −RuII−OsII* energy transfer. These processes were modeled simultaneously to extract rate constants for RuII* → RuII energy-transfer migration, kRu*−Ru, and RuII* → OsII energy transfer, kRu*−Os. For ORR-2, kRu*−Ru = 2.9 × 107 s−1 and kRu*−Os = 3.4 × 108 s−1. For ORR-3, kRu*−Ru = 1.2 × 107 s−1 and kRu*−Os = 1.3 × 108 s−1. For ORR-5, kRu*−Ru = 3.6 × 106 s−1 and kRu*−Os = 5.8 × 107 s−1, all in acetonitrile at 22 °C. The data were analyzed by assuming Dexter energy transfer with the Franck−Condon factors arising from intramolecular structural and medium changes evaluated by use of an emission spectral fitting procedure. Fits of the data to the Dexter mechanism were consistent with the predicted distance dependence of energy transfer.

■

INTRODUCTION Spatially directed energy transfer following light absorption is at the heart of the light-harvesting antenna of photosynthesis.2−4 Transporting energy over significant distances at rapid rates is dependent both on driving force and optimal placement of chromophores. A variety of strategies have been explored to model and mimic the photosynthetic antenna behavior, for example, based on spatially ordered arrays in dendrimers, polymers, MOFs, and peptides.5−18 Of these, peptides prepared by stepwise site-by-site synthesis are appealing in offering synthetic control both in composition and in relative positioning of light-absorbing units due to the modular, stepwise approach offered by the solid-phase peptide synthetic technique. For oligoprolines there is the added advantage of controlling secondary structure with nine or more residues naturally adopting helical structureseither proline I (Pro I) or proline © XXXX American Chemical Society

II (Pro II) (Figure 1). Proline I is a right-handed helix consisting of local cis-amide bonds and is favored in nonpolar solvents. Proline II adopts a left-handed helical structure in polar solvents, notably water, with a repeat spacing of 9.4 Å versus 6.3 Å in Pro I. Oligoprolines have been used successfully to explore spatially controlled redox splitting20−23 and the distance dependence of electron transfer in the oligoprolines CH3CO-Pro6-Pra(PTZ)Pron-Pra(RuIIb2m)2+-Pro6-NH2 with PTZ 3-(10H-phenothiazine-10)propanoyl and (RuIIb2m)2+ bis(4,4′-diethylamide-2,2′bipyridine)(4-methyl,4′-carboxylate,2,2′-bipyridine)ruthenium(II) dication with n = 2−5 (Figure 2). In this series, the distance dependences of reductive quenching of the metal-toReceived: January 10, 2013 Revised: April 18, 2013

A

dx.doi.org/10.1021/jp400312f | J. Phys. Chem. B XXXX, XXX, XXX−XXX

The Journal of Physical Chemistry B

Article

Circular Dichroism. Circular dichroism (CD) spectra were recorded with an AVIV model 62DS CD spectrophotometer. The ellipticity θobs of solutions of the peptides at various temperatures in HPLC-grade water and spectral-grade acetonitrile was measured at integral values of wavelength over the range of 190−250 nm using an averaging time of 4 s. The CD data are reported as mean residue ellipticity ([θ], in deg cm2 dmol−1) calculated by the equation [θ](λ,T) = θobs(λ,T)/(Cdn), where θobs(λ,T) is the observed ellipticity (in degrees) at a given wavelength and temperature T, C is the oligoproline concentration (in mol/L), d is the optical path length (in mm), and n is the number of residues in the oligoproline. Synthesis. [Boc-Pra(OsIIbpy2mbpy)−OH](PF6)2. (2S,4S)α N -(1,1-Dimethylethoxycarbonyl)-4-(bis(2,2′-bipyridine)(4′methyl-2,2′-bipyridine-4-carboxamido)osmium(II)-L-proline bis(hexafluorophosphate). The starting materials, OsII(bpy)2(Cl)2 and 4′-methyl-2,2′-bipyridine-4-carboxylic acid (m-bpy−OH), were assembled as described previously25 to form [OsII(bpy)2(m-bpy−OH)](PF6)2. The derivatized complex was coupled to Boc-Pra-OCH3 by using standard peptide coupling chemistry to afford the methyl ester derivative in 51% yield. The ester (244 mg, 0.198 mmol, 1.0 equiv) was dissolved in 3:1 (v/v) CH3OH/H2O (12 mL) with stirring. Saponification was achieved by adding solid LiOH and stirring the mixture overnight at room temperature. Methanol was removed by rotary evaporation. More water (25 mL), several drops of HPF6, and NH4PF6 were added and the aqueous mixture was extracted with CH2Cl2 until the aqueous layer was colorless. The combined organics were dried over Na2SO4 and the solvent was removed by rotary evaporation. The crude complex was purified by ion-exchange chromatography on a sulfopropyl-Sephadex column (Sephadex-SP C-25 resin, 40− 120 μm, Pharmacia Fine Chemicals) using a step gradient from 0.01 M NH4Cl to 0.03, 0.05, 0.07, and 0.1 M NH4Cl. The dark green-black band eluting with 0.1 M salt was collected and treated with several drops of HPF6 and with NH4PF6. The aqueous mixture was extracted with CH2Cl2 until the aqueous layer was colorless. The combined organics were dried over Na2SO4 and the solvent was removed by rotary evaporation. The black residue was dried overnight under high vacuum to give 224 mg of the protected amino acid [Boc-Pra(OsII(bpy)2(m-bpy−OH)](PF6−)2 in 93% yield. FAB-MS (calcd for C42H42N8O5Os [M − 2PF6)+]: m/z 930.2893) m/ z 930.2887; 1H NMR (200 MHz, CD2Cl2) δ 1.30 (2 s, 9 H, (CH3)3C), 2.04 (m, 1 H, CβH), 2.44 (m, 1 H, CβH), 2.64 (s, 3 H, m-CH3), 3.61 (m, 2 H, 2 x CδH), 4.33 (m, 1 H, CαH), 4.60 (m, 1 H, CγH), 7.21 (d, 1 H, m5′), 7.36 (m, 5 H, 4b5 and m5), 7.49 (m, 1 H, m6′), 7.59 (m, 4 H, 4b6), 7.71 (d, 1 H, NγH), 7.85 (m, 5 H, 4b4 and m6), 8.40 (m, 5 H, 4b3 and m3′), and 8.69 ppm (br s, 1 H, m3); 13C NMR (400 MHz, CD2Cl2) δ 162.41, 158.75, 151.25, 151.13, 150.98, 150.84, 150.77, 150.63, 149.93, 138.06, 137.93, 137.81, 137.73, 129.99, 129.94, 129.08, 128.95, 128.87, 126.26, 126.22, 125.01, 124.94, 124.78, 122.36, 50.58, 37.83, 36.12, 33.20, 33.14, 32.54, 32.40, 32.33, 30.53, 30.42, 30.08, 29.88, 29.75, 28.33, 26.82, 23.09, 21.32, and 14.28 ppm. [Boc-Pra(RuIIbpy2mbpy)−OH](PF6)2. (2S,4S)-Nα-(1,1-Dimethylethoxycarbonyl)-4-(bis(2,2′-bipyridine)(4′-methyl-2,2′-bipyridine-4-carboxamido)ruthenium(II)- L -proline bis(hexafluorophosphate). The Ru analogue of the Os complex described above was prepared by an analogous procedure given previously.25,26

Figure 1. Structures of the polyproline helices, proline I (left of dashed line, labeled Pro I) and proline II (right of dashed line, labeled Pro II). Adapted with permission from ref 19. The two center images show the helices from an axial, end-on perspective. The separation distance for a full turn of each type of helix is shown with reference to the two outer images which depict a side view of the polyproline helices.

Figure 2. Computer-generated structure of the RuII(bpy)-PTZ oligoproline assembly CH3CO-Pro6-Pra(PTZ)-Pro5-Pra(RuIIb2m)2+Pro6-NH2, with permission from ref 21.

ligand charge transfer (MLCT) excited state of the amidelinked chromophore, [Ru(bpy)2(4-Me-4′(−N(H)CO)bpy)]2+, by the appended phenothiazine derivative, −RuIII(bpy−)2+*− (Pro)n−PTZ− → −RuII(bpy−)+−(Pro)n−PTZ+−, and back electron transfer, −Ru I I (bpy − ) + −(Pro) n −PTZ + − → −RuII(bpy)2+−(Pro)n−PTZ−, were explored by varying the number of intervening proline spacers.21,24 Here we present the design, synthesis, and the dynamics of intra-assembly energy transfer in three helical oligoproline assemblies based on the MLCT chromophores [Ru(bpy)2(4Me-4′(-N(H)CO)bpy)]2+ and [Os(bpy)2(4-Me-4′(-N(H)CO)bpy)]2+, [CH3−CO-Pro6-Pra[RuII(bpy)2(mbpy)]2+-PronPra[RuII(bpy)2(mbpy)]2+-Pron-Pra[OsII(bpy)2(mbpy)]2+-Pro6Glux-NH2]6+ (n = 2, x = 1 for ORR-2; n = 3, x = 1 for ORR-3; n = 5, x = 2 for ORR-5) (Figure 3). The goal of the study was to investigate the distance dependence of two energy-transfer events: energy-transfer migration/exchange, −RuII*−(Pro)n− RuII− → −RuII−(Pro)n−Ru2+*− and −RuII*−(Pro)n−OsII− → −RuII−(Pro)n−OsII*− energy transfer.

■

MATERIALS AND METHODS H NMR spectra were recorded on a 200 MHz AC200 Bruker spectrometer. 13C NMR spectra were obtained on a 400 MHz Varian XL 400 spectrometer. Thin layer chromatography was performed on Bakerflex alumina plates. Mass spectra were obtained at the North Carolina State University Spectrometry Laboratory for Biotechnology. Positive-ion fast-atom bombardment mass spectra (FAB-MS) were recorded with a JEOL HX110HF mass spectrometer. Positive-ion electrospray ionization mass spectra (ESI-MS) were recorded with a Micromass Quattro II Electrospray Triple Quadrupole mass spectrometer.

1

B



Article

Figure 3. Structures of the oligoproline assemblies [CH 3-CO-Pro6-Pra[RuII(bpy)2(mbpy)]2+-Pron-Pra[RuII(bpy)2(mbpy)]2+-Pron-Pra[OsII(bpy)2(mbpy)]2+-Pro6-Glu-NH2]6+ (n = 2, 3). For n = 5, the structure is [CH3-CO-Pro6-Pra[RuII(bpy)2(mbpy)]2+-Pro5-Pra[RuII(bpy)2(mbpy)]2+-Pro5-Pra[OsII(bpy)2(mbpy)]2+-Pro6-Glu2-NH2]6+.

ORR-5. [CH 3 -CO-Pro 6 -Pra(Os II bpy 2 mbpy)-Pro 5 -Pra(RuIIbpy2mbpy)-Pro5-Pra(RuIIbpy2mbpy)-Pro6-Glu2-NH2]6+. This 27-residue acetylated peptide amide was assembled by manual solid-phase peptide synthesis on a 0.12 mmol scale by using Boc-Glu(Obz)−OH (Peninsula), Boc-Pro-OH (Bachem), Boc-Pro-Pro-OH (Bachem), the Ru and Os chromophoric modules, and methylbenzhydrylamine (MBHA) resin from Applied Biosystems. The amino acids were coupled for ∼24 h by using the Boc-amino acid (1.1 equiv), BOP (1.1 equiv), HOBt (1.1 equiv), NMM (2.1 equiv), and DMAP (01. equiv) in CH2Cl2. All other amino acids were coupled for 1 h in DMF using 4.0 equiv of the Boc-amino acid and the same ratio of coupling reagents as above with respect to the amino acid. Bocamino acids coupled to an N-terminal redox residue were double coupled, as were each of the last three Boc-amino acids. Prior to cleavage, the resin-bound peptide was N-terminally acetylated with 20% acetic anhydride in CH2Cl2. The peptide was cleaved from half of the peptide resin by using a mixture of trifluoroacetic acid (7.48 mL), bromotrimethylsilane (1.35 mL), and thioanisole (1.2 mL). The crude peptide was purified by preparative reversed phase HPLC on a butyl-silica column (Vydac C4) eluted over 60 min with a linear gradient of 27−

34% acetonitrile in 0.08% trifluoroacetic acid/water. Fractions containing the ORR peptide were freed of solvent to give the pure peptide as a brown powder (16 mg). ESI-MS calculated for [C233H269N49O35Ru2Os]6+ = m/z 4708.3 Da, found m/z 4707.1 Da. ORR-2. [CH 3 -CO-Pro 6 -Pra(Os II bpy 2 mbpy)-Pro 2 -Pra(RuIIbpy2mbpy)-Pro2-Pra(RuIIbpy2mbpy)-Pro6-Glu-NH2]6+. This 20-residue acetylated peptide amide was assembled by manual solid-phase peptide synthesis on a 0.1 mmol scale by using the same procedures and conditions described above for the synthesis of ORR-5. The crude peptide was prepared from Boc-Glu(OBz), Boc-Pro-Pro, and the Ru and Os chromophoric modules. Cleavage from the resin was achieved with bromotrimethylsilane, TFA, and thioanisole followed by purification by reversed-phase HPLC on a butyl-silica column (Vydac C4) eluted over 60 min with a linear gradient of 22− 36% acetonitrile in 0.08% TFA/water gave purified, fluffy brown powder (51 mg). ESI-MS calculated for [C198H220N42O26Ru2Os]6+ = 3996.53 Da, found 3995.99 Da. ORR-3. [CH 3 -CO-Pro 6 -Pra(Os II bpy 2 mbpy)-Pro 3 -Pra(RuIIbpy2mbpy)-Pro3-Pra(RuIIbpy2mbpy)-Pro6-Glu-NH2]6+. This peptide assembly was synthesized in the same manner and C



Article

the appended complexes have a similar registry along the helical axis relative to each other. For ORR-3, adjacent chromophores are out of registry by ∼60° relative to each other. Oligoproline Conformation by Circular Dichroism. The solution conformations of ORR-2, ORR-3, and ORR-5 were examined by circular dichroism (CD) measurements in several solvents.19 The proline I helical structure exhibits a distinctive CD signature with a negative feature at 199 nm, a strong positive feature at 215 nm, and a weak negative feature at 230 nm. The proline II helix also exhibits an easily recognizable CD spectrum with a strong negative feature at 205 nm and a weak positive feature at 266 nm. In water, the CD spectra of these assemblies were consistent with proline II. When the proline II form was dissolved in CH3CN, the solvent used for photophysical studies, CD spectra with time showed that ORR-3 and ORR-5 slowly changed from proline II to proline I. For ORR-2 in CH3CN, CD spectra recorded over a 20 h period showed that the peptide changed conformation but that the conformational change was not complete. The intensity of the positive feature at 215 nm, which is characteristic for the proline I helix, was much lower than expected consistent with an equilibrium and the presence of both types of helices or of incomplete helix formation. Presumably steric congestion between the [M(bpy)2(4-Me4′(-N(H)CO)bpy)]2+ chromophoric units, with molecular diameters of ∼13 Å, inhibits complete isomerization to proline I for ORR-2. These CD results along with the results of molecular modeling of the proline helices suggests that adjacent chromophores are spaced ∼12.5 Å apart for ORR-5 and 6.3− 9.4 Å for ORR-2 and 8.4 Å for ORR-3. UV−Visible Spectra. Figure 4 shows UV−visible absorption spectra for ORR-2 in CH3CN compared to model

with similar yields as for ORR-2. ESI-MS calculated for [C209H238N44O28Ru2Os]6+ = 4206.77 Da, found 4184 Da. Photophysical Measurements. Acetonitrile (Burdick and Jackson) used in the photophysical measurements was either used as received or distilled over CaH2. Steady-state emission spectra were recorded on a SPEX Fluorolog-212A photoncounting spectrofluorimeter and were corrected for the instrument response. Optically dilute samples (Aλ(excitn) < 0.12) were either freeze−pump−thaw degassed to 10−6 Torr or argon-sparged for at least 45 min prior to use. Time-resolved emission measurements were conducted by time-correlated single photon counting (TCSPC). The apparatus consisted of a mode-locked Nd:YAG laser (Coherent, Antares 76-s) whose frequency-tripled output was used to synchronously pump a single jet dye laser (Coherent, 700 series) circulating Stilbene 3 laser dye solution. The dye laser output at 430 nm was cavity dumped to produce ∼10 ps pulses. The repetition rate of the dye laser was selected as either 475 kHz (for measurements at 780 nm) with an average power of ∼1 mW, or 190 kHz (for measurements at 640 nm) with an average power of 350 μW. Once through an iris, the 2 mm diameter, unfocused laser beam illuminated a 10 mm path length cuvette. The intensity of the incident light was varied by use of neutral-density filters mounted between the sample and the laser beam. Sample was collected at 90° relative to laser excitation, focused into a single grating monochromator (CVI, Digikrom 240), and subsequently delivered to a cooled, multichannel plate-photomultiplier tube (MCP-PMT) (Hamamatsu, R3809U-51). The signal from the MCP-PMT was amplified (Philips Scientific 6954 preamplifier) prior to being fed into a constant fraction discriminator (CFD) (Tennelec, TC454) whose output served as the start pulse for the time-toamplitude converter (TAC) (Tennelec, TC864). The stop pulse in the timing scheme was obtained by focusing ∼10% of the excitation beam into a Si:PIN photodiode. The photodiode pulse was sent into a variable delay box, then to the CFD, and finally to the TAC. The TAC’s output was sent to a multichannel analyzer (Tennelec, PCA-multiport) which was interfaced to a PC. The instrument response of the apparatus was 80 ps at the fwhm.

■

RESULTS Oligoproline Design. The three [CH3-CO-Pro6-Pra[RuII(bpy)2(mbpy)]2+-Pron-Pra[RuII(bpy)2(mbpy)]2+-Pron-Pra[OsII(bpy)2(mbpy)]2+-Pro6-Glux-NH2]6+ oligoprolines were designed and prepared with three separate sets of proline spacers between sequential units in the −RuII−(Pro)n−RuII− (Pro)n−OsII− arrays with n = 2, x = 1 (ORR-2); n = 3, x = 1 (ORR-3); or n = 5, x = 2 (ORR-5). The oligoprolines were designed with six proline residues at each terminus to induce proline helix formation. As noted in the Introduction, oligoprolines with nine or more residues fold into proline I or proline II helices depending on solvent.27 Our experiments were conducted in acetonitrile with the proline I helix dominant. In this structure, there are cis-peptide bonds and a right-handed twist with 3.3 residues per turn, 1.9 Å per residue (Figure 1). Adjacent chromophoric sites form a dihedral angle of ∼60° with the helical axis. ORR-5 was designed with two glutamic acid residues at the C terminus for possible attachment to oxide surfaces through the carboxylic acid side chains but the internal peptide helical structure is the same. ORR-2 and ORR-3 have only a single glutamic acid residue at its C terminus. For ORR-2 and ORR-5,

Figure 4. Absorption spectra of ORR-2 and the component complex salts [Pra(RuIIbpy2mbpy)](PF6)2 and [Pra(OsIIbpy2mbpy)](PF6)2 scaled by factors of 2/3 and 1/3, respectively, in acetonitrile at room temperature. Also shown is the sum (○) of scaled model complex spectra and their match with the spectrum of ORR-2.

complexes Boc-Pra(Ru I I bpy 2 mbpy) 2 + and Boc-Pra(OsIIbpy2mbpy)2+. The spectra are dominated by dπ(MII) → π*(bpy) metal-to-ligand charge transfer (MLCT) absorptions in the visible and ligand-localized π → π* absorptions in the UV. The spectrum of the oligoproline assembly is essentially the sum of the spectra (±5%) of the components. CW Emission. Emission spectra for Pra(OsIIbpy2mbpy)2+* and Pra(RuIIbpy2mbpy)2+* were fit by application of a singleD



Article

mode Franck−Condon spectral fitting analysis,28 eq 1, resulting in the parameters in Table 1. 3 ⎧⎛ ⎪ E0 − νℏω ⎞ ⎛ S ν ⎞ ⎡⎢ ⎟ ⎜ ⎟exp −4 E0 ⎠ ⎝ ν! ⎠ ⎢⎣ ν=0 ⎪ ⎩⎝ 5

I (ν ̅ ) =

∑ ⎨⎜

⎛ ν − E + νℏω ⎞2 ⎤⎫ ⎪ 0 ⎟ ⎥⎬ ln(2)⎜ ̅ Δν1/2 ⎠ ⎥⎦⎪ ⎝ ̅ ⎭

(1)

Table 1. Emission Spectral Fitting Parameters from a Franck-Condon Analysis, Eq 1, of the Emission Spectral Profiles of OsII and RuII Complexes in Acetonitrile at Room Temperature (See Text) complex II

2+

[Pra(Os bpy2mbpy)] [Pra(RuIIbpy2mbpy)]2+

E0 (cm−1)

S

12930 15660

0.58 0.72

Δν̅1/2 (cm−1) 1900 1920

ℏω (cm−1) 1250 1350

In this equation, I(ν)̅ is the emission intensity at energy ν̅ in cm−1, E0 is the energy spacing between the v* = 0 level for the average mode in the excited state and v = 0 level in the ground state. S is the electron−vibrational coupling constant or Huang−Rhys factor, which is a measure of excited-state distortion. Δν̅1/2 is the full width at half-maximum for a single vibronic component and ℏω is the quantum spacing for the average vibrational mode. Prior to fitting, emission spectra were converted to units of quanta per second versus wavenumber following the procedure of Parker and Rees.29 Fits were performed by holding ℏω constant at 1350 cm−1 for [Pra(Ru I I bpy 2 mbpy)] 2 + * and 1250 cm − 1 for [Pra(OsIIbpy2mbpy)]2+*. The free-energy content of the MLCT excited state above the ground state, ΔGES°, calculated from E0 + λ0 (see Discussion), is 2.14 eV for −RuII* and 1.80 eV for −OsII*. From this, the driving force, ΔG°′, for RuII* → OsII energy transfer is calculated to be −0.34 eV. Figure 5 shows the emission spectrum of ORR-2 at room temperature in CH3CN and, for comparison, spectra of [Pra(RuIIbpy2mbpy)](PF6)2 and [Pra(OsIIbpy2mbpy)](PF6)2 with excitation at 430 nm. There are three contributors to emission from the oligoprolines: unquenched RuII*, sensitized OsII*, and directly excited OsII*. The contribution from direct OsII* emission was deleted from the overall emission spectrum by quantitative subtraction of the absorbance-scaled [Pra(OsIIbpy2mbpy)]2+* emission by application of eqs 2a−2c. The efficiency of RuII* → OsII energy transfer, ηen, was determined by using eqs 2a−2c to fit the ORR emission spectrum as a linear combination of the appropriately scaled [Pra(RuIIbpy2mbpy)]2+* and [Pra(OsIIbpy2mbpy)]2+* emission spectra. Idirect Os = (A ORR /A Os in ORR )IOs

Figure 5. Emission spectrum for ORR-2 compared to appropriately scaled spectra for the model complexes [Pra(RuIIbpy2mbpy)](PF6)2 and [Pra(OsIIbpy2mbpy)](PF6)2 in acetonitrile at room temperature. Also shown is the sum (○) of the scaled model spectra, which matches that of the peptide within experimental error (±5%). Excitation was at 430 nm, consistent with the time-resolved emission data.

arise from direct excitation of OsII rather than energy-transfer sensitization to form OsII*. The quantities ARu, AOs, and AORR are the absorbances of Ru(II), Os(II), and the assembly. The fractions of unquenched RuII* sites and sensitized OsII* sites are denoted by αOs and αRu. The emission profile corrected for direct OsII excitation can be reproduced by a superposition of scaled RuII* and OsII* spectra to give ηen = 0.93 for ORR-2, ηen = 0.77 for ORR-3, and ηen = 0.73 for ORR-5. For these assemblies, −RuaII−RubII−OsII−, quantitative emission comparisons show that photoexcitation of RuII results in energy transfer to the lower energy OsII trap site by RuII* → OsII energy transfer. This follows from the observation of enhanced OsII* emission compared to Pra[OsII(bpy)2(mbpy)]2+. Sensitization in the assemblies could occur by excitation and energy transfer from the adjacent −RuII− site, −RubII*−OsII−, by long-range transfer, −RuaII*− RubII−OsII− → −RuaII−RubII−OsII*−, or by RuaII*−RubII− energy migration/exchange followed by sensitization, −RuaII*− RubII−OsII− → −RuaII−RubII*−OsII− → −RuaII−RubII− OsII*−. The CW emission data do not distinguish between these possibilities but they can be delineated by analysis of the time-resolved emission data. Emission Decay. Time-correlated single photon counting (TCSPC) was used to evaluate the kinetics of energy transfer from RuII* to RuII or OsII following 430 nm excitation. Figure 6

(2a)

(IORR /A ORR ) − Idirect os = αRu(IRu/ARu ) + αOs(IOs/A Os) (2b)

ηen = αOs/(αRu + αOs)

(2c)

In eqs 2a−2c, IORR, IRu, IOs, and Idirect Os are the integrated emission profiles for the assembly, [Pra(RuIIbpy2mbpy)]2+*, and [Pra(OsIIbpy2mbpy)]2+* scaled by their absorbances, respectively. Idirect Os is the portion of IOs that is calculated to E



Article

Figure 7. Time-resolved emission monitored at 780 nm from ORR-2 in argon-deaerated, room temperature acetonitrile following 430 nm excitation. The shaded region shows the contribution to the emission decay due to direct excitation of OsII and RuII.

Figure 6. Time-resolved emission from ORR-2 in argon-purged, room temperature acetonitrile following 430 nm excitation. Also shown is the transient decay for [Pra(RuIIbpy2mbpy)](PF6)2 scaled to match the data past 500 ns for unquenched RuII* and the result when the curve was subtracted from the mixed peptide data (○). The emission monitoring wavelength was 640 nm.

reaching a maximum by 9 ns, at 12 ns for ORR-3 and at 9 ns for ORR-5. These observations are consistent with contributions from both direct and energy-transfer sensitization. Part of the instrument-limited rise time (fwhm = 80 ps) is due to direct excitation of OsII which is a competitive light absorber at the 430 nm excitation wavelength. The extent of direct OsII excitation was determined to be 30%, 26%, and 29% for ORR-2, ORR-3, and ORR-5, respectively, by quantitatively comparing the 780 nm emission of the ORR peptide with the [Pra(OsIIbpy2mbpy)]2+* model after correction for integrated irradiance and OsII absorbance. There is a contribution to the transient decay data at 780 nm from the low-energy side of the RuII* emission. Its relative contribution can be determined because of the large difference in decay rates between RuII* and OsII* with all emission past ∼500 ns apparently due to RuII* decay from an impurity. The RuII* contribution to 780 nm emission was evaluated by scaling the kinetic trace monitored at 640 nm, where RuII* is the sole emitter, to match the data past 500 ns where only RuII* emission is detected. The sum of the scaled emission contributions from RuII* and direct OsII excitation is shown as the shaded region in Figure 7. The dynamics of RuII* decay and OsII* formation and subsequent decay are coupled and the data were treated accordingly by applying a rate law matrix approach. Monitoring at 640 nm provides transient information about RuII* kinetics and at 780 nm, information about all three contributors as described above. The reactions in Scheme 1 were considered in modeling the kinetic data with emission−time traces at 640 and 780 nm fit simultaneously. The kinetic scheme included three distinct pathways for sensitizing OsII to OsII*: (1) −RuaII−RubII*− OsII− → −RuaII−RubII−OsII*− energy transfer, (2) −RuaII*−

shows the RuII* emission decay in ORR-2 monitored at 640 nm. The emission profile exhibits two distinct kinetic signaturesone complex and one first orderwith the distribution dependent on time scale. For the first 500 ns, RuII* decay is complex. For the TCSPC experiments described here, no pulse energy dependence was observed over a factor of 3 in incident irradiance pointing to single-photon excitation of the assembly under the conditions of the experiment. The absence of multiphoton effects confirms that the complex kinetic behavior arises from kinetically competitive energy transfer involving the OsII chromophore. By 500 ns after the laser pulse, the 640 nm emission intensity drops below ∼10% and the emission−time traces follow firstorder kinetics with a rate constant, k = 7.9 × 105 s−1 (τ = 1260 ns), consistent with decay from unquenched [Pra(RuIIbpy2mbpy)]2+*. As shown in Figure 3, the contribution from unquenched RuII* emitters can be scaled and subtracted from the overall 640 nm emission−time trace leaving only the quenching dynamics. This comparison suggests that, by 500 ns after the laser pulse, intra-assembly dynamical processes are complete and the remaining RuII* excited states are not involved in energy transfer. The appearance of the smallamplitude first-order decay component at longer times is presumably due to a small amount of unquenched impurity. Qualitatively similar behavior was observed at 640 nm for ORR-3 and ORR-5. Deconvolution of the excited-state emission−time profiles was complicated by contributions from the trace-emitting impurities noted above. These presumably arise from oligoproline fragments without OsII quenching sites. The appearance of emitting impurities in these samples, even from HPLC purified samples, is an almost unavoidable complication. They contribute disproportionately to the observed emission because they are unquenched and long-lived. Under the conditions of these experiments, intrinsic emission from assembly RuII* sites is largely quenched. Transient emission decay monitored at 780 nm for ORR-2 following 430 nm excitation (Figure 7) is dominated by OsII*. The majority (∼68%) of the maximum emission intensity is observed promptly, yet the emission continues to grow,

Scheme 1. Energy-Transfer Kinetic Scheme

F



Article

RubII−OsII− → −RuaII−RubII−OsII*− long-range energy transfer, and (3) −RuaII*−RubII−OsII− → −RuaII−RubII*−OsII− energy migration followed by RubII* → OsII energy transfer. Fits to the data using a model based upon all of the reactions shown in Scheme 1, including all three OsII* sensitization pathways, were nonconvergent and unsuccessful. Rather, the data can be fit, eqs 3−6, by including only two processes for OsII* sensitizationone fast and one slowwhich leaves no basis for including a third component. The kinetic scheme ultimately used to model the data included fast RubII* → OsII energy transfer and RuaII* → RubII energy migration/exchange with subsequent OsII sensitization from RubII*. With ΔG°′ = −0.34 eV for energy transfer and ΔG°′ = 0 for energy migration/exchange, energy transfer is expected to be more rapid. Considering the difference in driving force between RuII* → RuII energy migration and RuII* → OsII energy transfer and given the observed distance dependence from the data shown here, it is expected that long-range RuaII* → OsII energy transfer is an important and independent process yet kinetically indistinguishable from RuaII* → RubII energy migration, suggesting similar rate constants for these two processes.

In eqs 4a−4d, K is the rate law matrix, C0 is the initial concentration matrix, G gives the eigenvalues of K, P gives the eigenvectors of K, Q is defined above, and C([X],t) is the timedependent concentration matrix for states Xi. The rate constants k1 and k2 for excited-state decay were held fixed at the known values for [Pra(RuIIbpy2mbpy)]2+* (8.0 × 105 s−1, τ = 1260 ns) and [Pra(OsIIbpy2mbpy)]2+* (1.7 × 107 s−1, τ = 58 ns), respectively. The time evolution of [Ru*a ], [Ru*b ], [Os*], and [GS] are given by C([Ru*a ],t), C([Ru*b ],t), C([Os*],t), and C([GS],t), respectively. The 640 nm transient was modeled by combining the matrix solution with the firstorder decay of adventitious RuII* as described above. For time-dependent emission decay monitored at 640 nm, I640nm(t), is given by eq 5. I640nm(t ) = A1 × (C([Ru*a ], t ) + C([Ru*b], t )) + A 2 × e−k1t

A1 and A2 are scaling factors for the contributions from RuII* excited states involved in the energy-transfer processes of Scheme 1 or from those that decay independently with the characteristic RuII* lifetime. The time-dependent emission profile at 780 nm, I780 nm(t), is given by eq 6.

d[Ru*a ] = −kRu *−Ru[Ru*a ] − k1[Ru*a ] + kRu *−Ru[Ru*b] dt

I780nm(t ) = A3 × C([Os*], t ) + A4 × I640nm(t )

(3a)

(3c)

d[GS] = k 2[Os*] + k1[Ru*a ] + k1[Ru*b] dt

(3d)

■

DISCUSSION The experiments described here complement those from the earlier study on the distance dependence of electron transfer in oligoproline assemblies.21 As noted above, the appeal of the oligoproline approach is control of composition and secondary

Simultaneous solution of the differential equations with matrix methods,30 gives the solution in eqs 4a−4d. −(kRu *−Ru + k1) kRu *−Ru

0

0

kRu *−Ru

0 −(kRu *−Ru + kRu *−Os + k1)

0

0

kRu *−Os

−k 2 0

k1

k1

k2

K=

⎡[Ru*]⎤ ⎢ a ⎥ ⎡ 0.31 ⎤ ⎢[Ru*]⎥ ⎢ 0.31 ⎥ b ⎥ ⎥=⎢ C0 = ⎢ ⎢[Os*]⎥ ⎢ 0.38 ⎥ ⎢ ⎥ ⎢⎣ 0 ⎥⎦ ⎢⎣[GS] ⎥⎦ exp(G1t ) 0 Q=

(4a)

(4b)

0

0

0

exp(G2t ) 0

0

0

0

exp(G3t ) 0

0

0

0

C([X], t ) = PQP−1C0

0

(6)

A3 is the relative amplitude for OsII* emission. The RuII* contribution at 780 nm is included by adding the product of the scaling factor A4 and I640 nm(t). After first normalizing the 780 and 640 nm transients to 1, the curves were simultaneously fit by using eqs 5 and 6 by adjusting the four A scaling factors and two rate constants, kRu*−Ru and kRu*−Os. Comparison of the best fit kinetic analysis to experiment is shown in Figure 8. The corresponding best fit parameters are summarized in Table 2.

d[Ru*b] = kRu *−Ru[Ru*a ] − kRu *−Ru[Ru*b] − kRu *−Os[Ru*b] dt − k1[Ru*b] (3b) d[Os*] = −k 2[Os*] + kRu *−Os[Ru*b] dt

(5)

exp(G4t )

Figure 8. Time-resolved emission monitored at 640 nm (gray circles) and 780 nm (red circles) from ORR-2 in argon-deaerated, room temperature acetonitrile following 430 nm excitation. Overlaid are the best fits to the 640 nm data (dark red line) and the 780 nm data (blue line) from eqs 5 and 6 and the parameters in Table 2.

(4c) (4d) G



Article

ORR-5. A correction for impurity emission was not required for ORR-3, implying that this sample was free of unbound [RuII(bpy)3]2+ emitting impurity. The experimental observations are consistent with extensive energy-transfer quenching in the assemblies and with the results of the analysis of the time-dependent and steady-state emission data. Application of a matrix analysis to the time-dependent data, corrected for impurity emission, and the mechanism in Scheme 1 with overlapping rate constants for RuII*→ RuII energy migration/self-exchange and long-range RuII*→ OsII energy transfer provided satisfactory fits to the data. The model utilized the independently measured lifetimes for the [BocPra(RuIIbpy2(mbpy)]2+* and [Boc-Pra(OsIIbpy2(mbpy)]2+* models allowing the emission−time traces to be deconvoluted into separate rate constants for RuII* → RuII energy migration/ self-exchange and RuII* → OsII energy transfer. Energy-Transfer Dynamics. Inspection of the rate constant data in Table 2 for ORR-2, ORR-3, and ORR-5 reveals a distance dependence for both energy migration and energy transfer as shown by the dependence on the number of intervening proline spacers between chromophores. From these data, kRu*−Ru decreases by a factor of ∼8 over the series from 2 to 5 proline spacers from 2.9 × 107 s−1 (ORR-2) to 1.2 × 107 s−1 (ORR-3) to 3.6 × 106 s−1 (ORR-5). Similarly, kRu*−Os decreases by a factor of ∼6 from 3.4 × 108 s−1 (ORR-2) to 1.3 × 108 s−1 (ORR-3) to 5.8 × 107 s−1 (ORR-5). Using these data as a guide, the validity can be assessed of ignoring long-range RuaII* → OsII energy transfer yet including RuaII* → RubII* energy migration. For ORR-5, the separation distance between RubII and OsII is similar to that between RuaII and OsII for ORR-2. The rate constants kRu*−Ru for ORR-2 and kRu*−Os for ORR-5 are of similar magnitude. This similarity suggests that the isoenergetic energy migration between adjacent RuII sites and the energetically favored yet remote RuaII* → OsII energy transfer are kinetically indistinguishable. The scheme presented here which includes energy migration overlapped with long-range energy transfer is the only scheme that yielded successful fits and physically meaningful parameters. The latter two are presumed to actively contribute to the observed kinetics and our analysis suggests that these have a similar distance dependence. From the CD results, in CH3CN, ORR-2, ORR-3, and ORR5 adopt the proline I helical structure. Based on this structure, calculated internuclear, center-to-center through-space and through-bond separation distances are listed in Table 3. In calculating average through-bond distances, the shortest through-bond distance through a single proline unit on a

Table 2. Kinetic Parameters Obtained by Fitting Emission Decay Profiles to Eqs 5 and 6a

ORR-2 ORR-3 ORR-5

kRu−Ru, s−1 (τRu*−Ru, ns)

kRu*−Os, s−1 (τRu−Os, ns)

A1

A2

A3

A4

2.9 × 10 (35) 1.2 × 107 (87) 3.6 × 106 (280)

3.4 × 10 (3.0) 1.3 × 108 (7.8) 5.8 × 107 (17)

1.48 1.58 1.32

0.09 0.05 0.17

1.41 1.84 1.67

0.30 0 0.40

7

8

a

Excitation at 430 nm in room temperature CH3CN. Data treatment is described in the text with monitoring at 780 nm (OsII*) and 640 nm (RuII*). Uncertainties are ±10%.

structure based on proline I and proline II helices. This provides a basis for controlling average distances for electron and energy transfer systematically by controlling the number of intervening proline spacers between the active elements on the assemblies. Synthesis and characterization of the three-member set ORR-2, ORR-3, and ORR-5, with the digits indicating the number of proline spacersfollowed straightforwardly from earlier syntheses of related oligoproline assemblies. Excitedstate lifetimes were evaluated by the time-correlated singlephoton-counting technique with a time window extending to 2 μs. The solvent of choice for the measurements, CH3CN, favors the compact proline I helical structure, note Figure 1. In this right-handed helical structure, there are local cis-amide bonds with a repeat spacing of 6.3 Å along the helix (Figure 1). For ORR-3 and ORR-5 prepared and purified under aqueous conditions, addition to CH3CN results in relatively slow, hours, but complete conversion to proline I as shown by CD measurements. However, for ORR-2, CD measurements over time revealed incomplete conversion from the proline II to proline I. This result is apparently due to local steric congestion between the closely spaced [Ru(bpy)2(4-Me-4′(-N(H)CO)bpy)]2+ and [Os(bpy)2(4-Me-4′(-N(H)CO)bpy)]2+ groups along the polyproline chain which causes a local disruption in the extended helix. The choice of oligoprolines for the current study, based on the −RuIIa2+−(Pro)n−RuIIb2+−(Pro)n−OsII− motif, with n = 2, 3, 5, was designed to explore a double distance dependence with a “two for the price of one” strategy. Analysis of emission− time profiles included contributions from both RuII* → RuII energy migration/self-exchange with ΔG° = 0 and RuII* → OsII energy transfer with ΔG°′ = −0.34 eV. In all three assemblies, but especially for ORR-2, the extent of emission quenching, compared to emission from the proline model [Boc-Pra(RuIIbpy2mbpy)]2+*, was extensive, ranging from 93% for ORR-2 to 73% for ORR-5, with the observed photophysics dominated by net intra-assembly RuII* → OsII transfer. Deconvolution of the excited-state emission−time profiles was complicated by contributions from trace, nonquenched, emitting impurities, presumably oligoproline fragments without OsII quenching sites. As noted in Results, the appearance of emitting impurities in these samples, even from HPLC-purified samples, is an almost unavoidable complication. Intrinsic emission from assembly RuII* sites is largely quenched and even small amounts of unbound [RuII(bpy)3]2+ emitting impurity contributes a disproportionate amount to the observed emission. It was possible to remove the contribution from emitting impurities by utilization of the scaling analysis described in Results. This analysis, based on emission at 640 nm, showed that ORR-2 emission included an impurity contribution of ≤5% from unquenched RuII*, and 11% in

Table 3. Energy-Transfer Matrix Elements and Structural Parameters for Energy Transfer in the Oligoproline Assemblies in CH3CN at 22 °C array

d (Å)a

no. of proline spacersb

r (Å)c

VRu*−Os (cm−1)d

ORR-2 ORR-3 ORR-5

6.3 8.4 12.5

2 3 5

8.6 12.9 21.5

9 6 4

a

Through-space separation distance between the centers of the MII complexes calculated from the proline helix structure; note Figure 1. b Number of proline spacers between chromophore-derivatized prolines. cThrough-bond separation distance through the proline spacer. dUncertainties are ±10%. H



Article

bond-for-bond basis of 4.3 Å was used, calculated from literature values for individual bond lengths.21,31 Use of the through-space distance for ORR-2, assuming the proline I structure, is problematic since the CD measurements revealed only partial proline II to proline I conversion in acetonitrile. The distances for ORR-3 and ORR-5 are also local averages based on the CD-defined secondary structures. These values reflect a distribution of local rotational conformers with slight differences in the energy-transfer distance among them. Energy-Transfer Mechanism. Energy transfer between polypyridyl excited-state donor and ground-state acceptor are induced by perturbations between the two. The perturbation operator, given by H′ = H′c + H′e, includes contributions from both “Coulomb” and exchange interactions. The Coulomb term, H′c, describes Förster-type energy transfer which arises from a long-range electromagnetic interaction with the electric dipole field in the excited donor inducing a dipole oscillation in the ground-state acceptor. Förster energy transfer varies as 1/ R6. The exchange perturbation, H′e, induces “Dexter” energy transfer which occurs by transfer of the excited-state dipole to the ground-state acceptor and varies exponentially as 1/ R.1,28,32,33 Although the data are limited, and the accuracy of the through-space distance for n = 2 problematic, given incomplete proline I formation, the observed decrease in kEnT with internuclear separation distance, R, is qualitatively consistent with the expected exponential dependence for Dexter energy transfer (Figure 9). The distance dependence for the Dexter energy-transfer rate constant, kEnT, is usually expressed as in eq 7, with L the effective tunneling distance between the donor and acceptor which is generally equal to the sum of average Bohr radii for the donor and acceptor.32,34 k0 is the rate constant in the limit that R = 0, and γ = 2R0/L. R0 is the intermolecular separation distance at which spontaneous excited-state decay, with lifetime τD, and energy transfer are equally probable. The distance dependence can also be written in the form commonly used for electron transfer shown in eqs 8 and 9. In these equations, βen is the energy transfer damping factor, given by βen = 2/L. kEnT,0 is the energy-transfer rate constant at the close contact donor−acceptor distance, d0, for through-space (eq 8), or r0 for through-bond (eq 9) energy transfer.35−37 ⎧ ⎛ ⎛ 2R ⎞ 1 R ⎞⎫ kEnT = k 0 exp⎜ − ⎟ = exp⎨γ ⎜1 − ⎟⎬ ⎝ L ⎠ τD R 0 ⎠⎭ ⎩ ⎝

(7)

kEnT = kEnT,0 exp{−βen(d − d0)}

(8)

kEnT = kEnT,0 exp{−βen(r − r0)}

(9)

⎪

⎪

⎪

⎪

Figure 9. Energy-transfer rate constant, kRu*−Os, as a function of through-space (A) and through-bond (B) separation distances. Solid curves show fits to eqs 8 and 9, respectively, with parameters given in Table 4.

Table 4. Dexter Mechanism Fit Parameters from Eqs 8 and 9 for Energy Transfer in the Oligoprolines in CH3CN at 22 °C through-spacea Ru*−Ru EnT Ru*−Os EnT

kEnT,0 (s−1)

βen

kEnT,0 (s−1)

βen

Fcalc

− 3.4 × 108

− 0.39

− 3.4 × 108

− 0.19

1.8 × 10−6 1.3 × 10−4

a Parameters derived from fits to eq 8 using the through-space separation distances of Table 3. bParameters derived from fits to eq 9 using the through-bond separation distances of Table 3.

Rate constants for Ru * → Os energy transfer are shown plotted against through-space, d−d0, and through-bond, r−r0, intermolecular separation distances in Figure 9, with the distances calculated as described previously.38−40 The data were fit by assuming the Dexter distance dependence in eqs 8 and 9 which neglects the expected small distance dependence of the solvent reorganization energy. Attempted fits to the 1/R6 dependence for Förster transfer were inadequate to fit the data. Results of the fits to eqs 8 and 9 are shown in Table 4 for both the energy-transfer rate constants at close contact, kEnT,0, and βen. The data set is small, and again, the through-space distance for n = 2 questionable, but an exponential fall-off in ken is suggested by the data. In the earlier electron-transfer study,21 II

through-bondb

II

it was concluded that electron transfer occurred through space and not through the proline bonding network. Given our limited data set here, there is not enough data to make that distinction for energy transfer. Similar analysis for RuII* → RuII energy migration is not presented because of its apparent overlapping contribution with long-range OsII* → RuII energy transfer which was difficult to disentangle. Energy-Transfer Analysis. In the weak coupling limit, assuming the validity of the Born−Oppenheimer and Condon approximations, the rate constant for energy transfer, kEnT, is given by eq 10. I


The Journal of Physical Chemistry B kEnt =

2π 2 V Fcalc ℏ

Article

RuII* → RuII self-exchange ignored because of possibly complicating kinetic overlap with long-range energy transfer. The results of these calculations are plotted in Figure 10 and

(10)

In this equation, V is the energy transfer matrix element and Fcalc the Franck−Condon vibrational overlap factor. Fcalc is determined by the degree of structural change between excited and ground states at the two sites undergoing energy transfer. It includes all contributions to the energy-transfer barrier from intramolecular structural changes and dipole orientational changes in the surrounding medium. The Franck−Condon factor can be evaluated from the normalized spectral band shapes, for emission by the donor f D(E), and absorption by the acceptor FA(E) (eq 11).28,34 The determination of FA(E) for the RuII and OsII acceptors is complicated because the singlet → triplet absorptions for these chromophores are overshadowed in intensity by overlapping singlet → singlet absorptions.

∫

fD (E)FA(E) E4

δE

(11)

In the limit of a single coupled high- or medium-frequency mode of quantum spacing ℏω and electron−vibrational coupling constant S, with low-frequency modes and solvent treated classically, Fcalc is given by eq 12.28 In this equation, n* and m are the vibrational quantum numbers for the excitedstate donor and ground-state acceptor, respectively. The SD and ℏωD values are associated with the donor (D) while SA and ℏωA are associated with the acceptor (A). ⎛ ⎞1/2 1 Fcalc = ⎜ ⎟ ⎝ 4πλADkBT ⎠

∞

∞

∑ ∑ exp(−SD) n*= 0 m = 0

⎛ S n * ⎞⎛ S m ⎞ exp( −SA )⎜ D ⎟⎜ A ⎟ ⎝ n*! ⎠⎝ m! ⎠ ⎛ (ΔG° + λ + mℏω + n*ℏω )2 ⎞ AD D A ⎟ exp⎜ − 4λADkBT ⎠ ⎝

(12) Figure 10. Energy-transfer matrix element, VRu*−Os, as a function of through-space (A) and through-bond (B) separation distances. Values are given in Table 3.

λAD is the total classical reorganization energy given by λA + λD and is related to the bandwidth parameter, Δν̅1/2, obtained from emission spectral fitting by eq 13 with kB the Boltzmann constant. 2 (Δν1/2 ̅ ) = 16kBTλ ln 2

listed in Table 3. Their magnitudes are comparable to values found for energy transfer between other RuII and OsII polypyridyl complexes.43−47 As for kEnT, the observed decrease in V with separation distance is qualitatively consistent with the expected exponential dependence for Dexter energy transfer (Figure 10).

(13)

ΔG° is the free energy change for energy transfer as calculated by eq 14. ΔG° = (E0,A + λA ) − (E0,D + λD)

■

(14)

All of these parameters for RuII* and OsII* are available by Franck−Condon analysis of the RuII* and OsII* emission spectra as described in Results. Application of the single, average mode approximation has been shown to be reasonably accurate for Os(bpy)32+* and related OsII* excited states based on the results of a mode-by-mode analysis by evaluation of resonance Raman excitation profiles.41,42 Results of emission spectral fitting for [Boc-Pra(RuIIbpy2(mbpy)]2+* and [BocPra(OsIIbpy2(mbpy)]2+* are summarized in Table 1 and Fcalc values in cm−1 are listed in Table 4. With experimental values for the energy-transfer rate constants and spectral evaluation of Fcalc, it is possible to use eq 10 to calculate the energy-transfer matrix element, V, as a function of distance for RubII* → OsII energy transfer with

ASSOCIATED CONTENT

S Supporting Information *

Absorption spectrum of ORR-5, Figure S1. Comparison of emission spectra for ORR-5 and constituent chromophores, Figure S2. Time-resolved emission from ORR-3 (Figure S3) and ORR-5 (Figure S4) monitored at 640 and 780 nm including kinetic fits. Comparison of time-resolved 780 nm emission from the three peptide assemblies, Figure S5. This material is available free of charge via the Internet at http:// pubs.acs.org.

■

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. J



Article

Notes

Structure of a Coiled Coil Peptide Scaffold. Inorg. Chem. 2012, 51, 11324−11338. (19) Slate, C. A. “Synthesis of Designed Helical Proteins”. Ph.D. dissertation, The University of North Carolina at Chapel Hill, 1997. (20) McCafferty, D. G.; Friesen, D. A.; Danielson, E.; Wall, C. G.; Saderholm, M. J.; Erickson, B. W.; Meyer, T. J. Photochemical Energy Conversion in a Helical Oligoproline Assembly. Proc. Natl. Acad. Sci. U.S.A. 1996, 93, 8200−8204. (21) Serron, S. A.; Aldridge, W. S.; Fleming, C. N.; Danell, R. M.; Baik, M. H.; Sykora, M.; Dattelbaum, D. M.; Meyer, T. J. Evidence for Through-Space Electron Transfer in the Distance Dependence of Normal and Inverted Electron Transfer in Oligoproline Arrays. J. Am. Chem. Soc. 2004, 126, 14506−14514. (22) Slate, C. A.; Striplin, D. R.; Moss, J. A.; Chen, P.-Y.; Erickson, B. W.; Meyer, T. J. Photochemical Energy Transduction in Helical Proline Arrays. J. Am. Chem. Soc. 1998, 120, 4885−4886. (23) Striplin, D. R.; Reece, S. Y.; McCafferty, D. G.; Wall, C. G.; Friesen, D. A.; Erickson, B. W.; Meyer, T. J. Solvent Dependence of Intramolecular Electron Transfer in a Helical Oligoproline Assembly. J. Am. Chem. Soc. 2004, 126, 5282−5291. (24) Aldridge, W. S., III; Hornstein, B. J.; Serron, S. A.; Dattelbaum, D. M.; Schoonover, J. R.; Meyer, T. J. Synthesis and Characterization of Oligoproline-Based Molecular Assemblies for Light Harvesting. J. Org. Chem. 2006, 71, 5186−5190. (25) McCafferty, D. G.; Bishop, B. M.; Wall, C. G.; Hughes, S. G.; Mecklenberg, S. L.; Meyer, T. J.; Erickson, B. W. Synthesis of Redox Derivatives of Lysine and Their Use in Solid-Phase Synthesis of a Light-Harvesting Peptide. Tetrahedron 1995, 51, 1093−1106. (26) Peek, B. M.; Ross, G. T.; Edwards, S. W.; Meyer, G. J.; Meyer, T. J.; Erickson, B. W. Synthesis of Redox Derivatives of Lysine and Related Peptides Containing Phenothiazine or Tris(2,2′-bipyridine)Ruthenium(II). Int. J. Pept. Protein Res. 1991, 38, 114−123. (27) McCafferty, D. G.; Slate, C. A.; Nakhle, B. M.; Graham, H. D.; Austell, T. L.; Vachet, R. W.; Mullis, B. H.; Erickson, B. W. Engineering of a 129-Residue Tripod Protein by Chemoselective Ligation of Proline-II Helices. Tetrahedron 1995, 51, 9859−9872. (28) Ito, A.; Meyer, T. J. The Golden Rule. Application for Fun and Profit in Electron Transfer, Energy Transfer, and Excited-State Decay. Phys. Chem. Chem. Phys. 2012, 14, 13731−13745. (29) Parker, C. A.; Rees, W. T. Correction of Fluorescence Spectra and Measurement of Fluorescence Quantum Efficiency. Analyst 1960, 85, 587−600. (30) Berberan-Santos, M. N.; Martinho, J. M. G. The Integration of Kinetic Rate Equations by Matrix Methods. J. Chem. Educ. 1990, 67, 375. (31) Bowen, H. J. M.; Sutton, L. E. Table. In Tables of interatomic distances and configuration in molecules and ions; Sutton, L. E., Jenkin, D. G., Mitchell, A. D., Cross, L. C., Eds.; Chemical Society: London, 1958. (32) Olaya-Castra, A.; Scholes, G. D. Energy Transfer from FörsterDexter Theory to Quantum Coherent Light-Harvesting. Int. Rev. Phys. Chem. 2011, 30, 49−77. (33) Meyer, T. J. Photochemistry of Metal Coordination Complexes: Metal to Ligand Charge Transfer Excited States. Pure Appl. Chem. 1986, 58, 1193−1206. (34) Förster, T. 10th Spiers Memorial Lecture. Transfer Mechanisms of Electronic Excitation. Discuss. Faraday Soc. 1959, 27, 7−17. (35) Dexter, D. L. A Theory of Sensitized Luminescence in Solids. J. Phys. Chem. 1953, 21, 836−850. (36) Newton, M. D.; Sutin, N. Electron Transfer Reactions in Condensed Phases. Annu. Rev. Phys. Chem. 1984, 35, 437−480. (37) Oevering, H.; Paddon-Row, M. N.; Heppener, M.; Oliver, A. M.; Cotsaris, E.; Verhoeven, J. W.; Hush, N. S. Long-Range Photoinduced Through-Bond Electron Transfer and Radiative Recombination via Rigid Nonconjugated Bridges: Distance and Solvent Dependence. J. Am. Chem. Soc. 1987, 109, 3258−3269. (38) Cowan, P. M.; McGavin, S. Structure of Poly-L-Proline. Nature 1955, 176, 501−503.

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS This material is based upon work supported as part of the UNC EFRC: Center for Solar Fuels, an Energy Frontier Research Center funded by the U.S. Department of Energy, Office of Science, Office of Basic Energy Sciences, under Award No. DESC0001011.

■

REFERENCES

(1) Dr. Bruce Wayne Erickson, Professor of Chemistry, at UNC-CH passed away November 10, 1998, yet we are delighted to include him as a co-author here for work contributed before his passing. (2) Cheng, Y. C.; Fleming, G. R. Dynamics of Light Harvesting in Photosynthesis. Annu. Rev. Phys. Chem. 2009, 60, 241−262. (3) Scholes, G. D.; Fleming, G. R. Energy Transfer and Photosynthetic Light Harvesting. In Adventures in Chemical Physics: A Special Volume of Advances in Chemical Physics; Berry, R. S., Jortner, J., Eds.; John Wiley and Sons, Inc.: Hoboken, NJ, 2005; Vol. 132, pp 57−129. (4) Sener, M. K.; Schulten, K. Physical Principles of Efficient Excitation Transfer in Light Harvesting In Energy Harvesting Materials; Andrews, D. L., Ed.; World Scientific: Singapore, 2012; pp 1−26. (5) Andrews, D. L. Light Harvesting in Dendrimer Materials: Designer Photophysics and Electrodynamics. J. Mater. Res. 2012, 27, 627−638. (6) Andrews, D. L.; Li, S.; Rodriguez, J.; Slota, J. Development of the Energy Flow in Light-Harvesting Dendrimers. J. Chem. Phys. 2007, 127, 134902. (7) Balzani, V.; Bergamini, G.; Ceroni, P.; Marchi, E. Designing Light Harvesting Antennas by Luminescent Dendrimers. New J. Chem. 2011, 35, 1944−1954. (8) Balzani, V.; Ceroni, P.; Maestri, M.; Vicinelli, V. Light-Harvesting Dendrimers. Curr. Opin. Chem. Biol. 2003, 7, 657−665. (9) Bandy, T. J.; Brewer, A.; Burns, J. R.; Marth, G.; Nguyen, T.; Stulz, E. DNA as Supramolecular Scaffold for Functional Molecules: Progress in DNA Nanotechnology. Chem. Soc. Rev 2011, 40, 138−148. (10) Barbieri, A.; Ventura, B.; Ziessel, R. Photoinduced EnergyTransfer Dynamics in Multichromophoric Arrays Containing Transition Metal Complexes and Organic Molecules. Coord. Chem. Rev. 2012, 256, 1732−1741. (11) Bottari, G.; Trukhina, O.; Ince, M.; Torres, T. Towards Artificial Photosynthesis: Supramolecular, Donor-Acceptor, Porphyrin- and Phthalocyanine/Carbon Nanostructure Ensembles. Coord. Chem. Rev. 2012, 256, 2453−2477. (12) Dutta, P. K.; Varghese, R.; Nangreave, J.; Lin, S.; Yan, H.; Liu, Y. DNA-directed Artificial Light-Harvesting Antenna. J. Am. Chem. Soc. 2011, 133, 11985−11993. (13) Fujitsuka, M.; Tetsuro, M. Delocalization and Migration of Excitation Energy and Charge in Supramolecular Systems. In Supramolecular Photochemistry; Ramamurthy, V., Inoue, Y., Eds.; Wiley: New York, 2011; pp 517−546. (14) Garo, F.; Haner, R. A DNA-based Light-Harvesting Antenna. Angew. Chem., Int. Ed. Engl. 2012, 51, 916−919. (15) Kent, C. A.; Liu, D.; Ma, L.; Papanikolas, J. M.; Meyer, T. J.; Lin, W. Light Harvesting in Microscale Metal-Organic Frameworks by Energy Migration and Interfacial Electron Transfer Quenching. J. Am. Chem. Soc. 2011, 133, 12940−12943. (16) Lee, C. Y.; Farha, O. K.; Hong, B. J.; Sarjeant, A. A.; Nguyen, S. B. T.; Hupp, J. T. Light-Harvesting Metal-Organic Frameworks (MOFS): Efficient Strut-to-Strut Energy Transfer in Bodipy and Porphyrin-based MOFS. J. Am. Chem. Soc. 2011, 133, 15858−15861. (17) Nantalaksakul, A.; Reddy, D. R.; Bardeen, C. J.; Thayumanavan, S. Light Harvesting Xendrimers. Photosynth. Res. 2006, 87, 133−150. (18) Wilger, D. J.; Bettis, S. E.; Materese, C. K.; Minakova, M.; Papoian, G. A.; Papanikolas, J. M.; Waters, M. L. Tunable Energy Transfer Rates via Control of Primary, Secondary, and Tertiary K



Article

(39) Sasisekharan, V. Structure of Poly-L-Proline. II. Acta Crystallogr. 1959, 12, 897−903. (40) Traub, W.; Shmueli, U. Structure of Poly-L-Proline I. Nature 1963, 198, 1165−1166. (41) Graff, D.; Claude, J. P.; Meyer, T. J. Calculation of Rate Constants from Spectra. Nonradiative Decay and Electron Transfer. In Electron Transfer Reactions: Inorganic, Organometallic, and Biological Applications; Isied, S., Ed.; American Chemical Society: Washington, DC, 1997; Vol. 11, pp 183−198. (42) Thompson, D. G.; Schoonover, J. R.; Timpson, C. J.; Meyer, T. J. Time-Dependent Raman Analysis of Metal-to-Ligand Charge Transfer Excited States: Application to Radiative and Nonradiative Decay. J. Phys. Chem. A 2003, 107, 10250−10260. (43) Harriman, A.; Romero, F. M.; Ziessel, R.; Benniston, A. C. Triplet Energy Transfer within Closely Spaced Positional Isomers of Ru/Os Polypyridine-Based Heterodiads. J. Phys. Chem. A 1999, 103, 5399−5408. (44) Shaw, J. R.; Sadler, G. S.; Wacholtz, W. F.; Ryu, C. K.; Schmehl, R. H. Toward the Development of Supramolecular Metal Complex Light Harvesting Arrays: Factors Affecting Photoinduced Energy Transfer in Bimetallic Complexes. New J. Chem. 1996, 20, 749. (45) De Cola, L.; Balzani, V.; Barigelletti, F.; Flamigni, L.; Belser, P.; Zelewsky, A.; Frank, M.; Vogtle, F. Photoinduced Energy and Electron Transfer Processes in Supramolecular Species, Tris(bipyridine) Complexes of Ruthenium(II)/Osmium(II), Ru(II)/Ru(III), Os(II)/ Os(III), and Ru(II)/Os(III) Separated by a Rigid Spacer. Inorg. Chem. 1993, 32, 5228−5238. (46) Diring, S.; Ziessel, R.; Barigelletti, F.; Barbieri, A.; Ventura, B. A Pre-organised Truxene Platform for Phosphorescent [Ru(bpy)2] and [Os(bpy)2] Metal Centres: A Clear-Cut Switch from Förster- to Dexter-Type Energy-Transfer Mechanism. Chem.Eur. J. 2010, 16, 9226−9236. (47) Grosshenny, V.; Harriman, A.; Hissler, M.; Ziessel, R. Intramolecular Triplet Energy Transfer in Alkyne-Bridged Ru-Os Multinuclear Complexes: Switching Between Dipole-Dipole and Electron-Exchange Mechanisms. J. Chem. Soc., Faraday Trans. 1996, 92, 2223−2238.

L


to Os - American Chemical Society

Recommend Documents