Doublet Spin State Mediated Photoluminescence Upconversion in Organic Radical Donor-Triplet Acceptor Dyads

1 Doublet Spin State Mediated Photoluminescence Upconversion in Organic Radical Donor-Triplet Acceptor Dyads Joseph M. O’Shea‡, Young Ju Yun‡, Abdelqader M. Jamhawi‡, Francesca Peccati§,ƒ, Gonzalo Jiménez-Osés§,ƒ,*, and A. Jean-Luc Ayitou‡,* ‡ Department of Chemistry, University of Illinois Chicago, Chicago, IL 60607, United States. § Center for Cooperative Research in Biosciences (CIC bioGUNE), Basque Research and Technology Alliance (BRTA), Bizkaia Technology Park, Building 800, 48160 Derio, Spain. ƒ Ikerbasque, Basque Foundation for Science, 48013 Bilbao, Spain. *Corresponding Authors’ e-Mails: [email protected] & [email protected] Keywords: Doublet-to-Triplet Energy Transfer, Donor-Acceptor Systems, Photon Upconversion 2 ABSTRACT: Donor-acceptor dyads are promising materials for improving triplet-sensitized photon upconversion due to faster intramolecular energy transfer (ET), which unfortunately competes with charge transfer (CT) dynamics. To circumvent the issue associated with CT, we propose a novel purely organic donor-acceptor dyad, where the CT character is confined within the donor moiety. In this work, we report the synthesis and characterization of a stable organic radical donortriplet acceptor dyad (TTM–Cz–Per) consisting of the acceptor perylene (Per) linked to the donor (4-N-carbazolyl-2,6-dichlorophenyl)bis(2,4,6-trichlorophenyl)methyl radical (TTM–Cz). Upon red-excitation of TTM–Cz–Per, the doublet emission of the donor (TTM–Cz) is significantly quenched, and a recorded delayed emission centered at ca. 490 nm was attributed to the fluorescence emission from the Per acceptor. Time-resolved transient absorption spectroscopy suggests doublet-to-triplet energy transfer (DTET) dynamics from the donor to the acceptor as the time constant τ for the donor transient species decreases from 21.47 ns for the TTM–Cz sensitizer to 8.73 ns for TTM–Cz–Per dyad. This process is accompanied by the appearance of a long-lived component with τ = 97.06 ns, which we ascribe to the triplet transient of the acceptor Per. Furthermore, computational results indicate that the DTET is intramolecular as computed spin densities of the quartet state show unpaired electrons of ρ ≈ 1 on the TTM-Cz donor and of ρ ≈ 2 on the acceptor Per. The present study highlights the possibility to employ doublet chromophoric systems for light-harvesting and energy upconversion, which can be further tailored for several optoelectronic applications. 3 1. INTRODUCTION: Triplet-triplet annihilation photon upconversion (TTA-UC) is a nonlinear photophysical process that converts low-energy photons into higher-energy photons via the “fusion” of two triplet excitons into one high-energy excited singlet and a ground state singlet.1 Compared to other upconversion methods, TTA-UC is attractive due to its ability to operate with low excitation intensities and noncoherent light sources.2 It is because of this that TTA-UC has been tailored for applications such as solar energy conversion3–7 and photocatalysis.8–14 Conventionally, TTA-UC is achieved through the use bichromophoric systems consisting of a triplet donor/sensitizer and acceptor/emitter. Upon photoirradiation, the donor is excited from its singlet ground state (S0) to its singlet excited state(s) before undergoing intersystem crossing (ISC) to populate its triplet manifolds. Through a Dexter energy transfer (ET) mechanism15, more specifically a triplet-triplet energy transfer (TTET), the donor transfers its excitation energy to the acceptor, resulting in the donor returning to its ground state and the acceptor being promoted to its triplet excited state. Two triplet-excited acceptors can then undergo TTA to produce one higher energy singlet-excited acceptor and one ground-state acceptor. The singlet excited acceptor then undergoes radiative decay to its ground state, emitting an upconverted photon (Figure 1a). Given the nature of this process, several aspects need to be considered when trying to maximize upconversion quantum yield (ΦUC). In most cases, the chromophoric samples used in TTA-UC are carefully deaerated to prevent quenching by molecular oxygen.16 Alternatively, the chromophoric mixtures can be frozen in rigid matrices to reduce diffusional quenching. Not only does quenching by molecular oxygen decrease the efficiency of the underlying photophysical processes, viz., TET, TTA, and ΦUC, but the quenching could also generate singlet oxygen (1O2) which can compromise acene acceptors (e.g. 4 anthracenes) present in the sample, limiting the use of some of those chromophores under ambient conditions.17 Figure 1. Jablonski diagrams illustrating a) traditional TTA-UC systems using separate donors and acceptors, b) charge transfer in prior dyads made by our lab, and c) our hypothesis of TTM– Cz based dyads for TTA-UC. Energy S0 S1 T1 Donor Acceptor S0 T1 S1 ISC Abs TTET UC Em TTA ×2 Fl Phos a) Classical TTA-UC using donor and acceptor systems (S0)D(S0)A Abs Em 1CT*3CT* ISC CR Prior work from our group using donor-acceptor dyadsb) (D1)LE(S0)A Abs Em DTET ×2 TTA UC Em This Workc) N NN N Mn+ Mettaloporphyrin sensitizer N N C8H17 C8H17 S S QDM 9,10-Diphenylanthracene Perylene Trip–QDM–Per Trip–QDM–DPA N N S S R N N S S R Cl Cl Cl Cl Cl Cl Cl Cl N TTM–Cz–Per Cl Cl Cl Cl Cl Cl Cl Cl N TTM–Cz–DPA EnergyEnergy CR & (S1)D(S0)A (T1)D(S0)A (S0)D(T1)A Em UC Em TTA ×2 (S0)D(S1)A (D0)(S0)A (D1)CT(S0)A (D0)(T1)A (D0)(S1)A CR 5 Furthermore, for an efficient Dexter ET to occur, there needs to be maximum wavefunction overlap between the donor and the acceptor, as well as the metastable triplet-excited acceptors when they undergo self-quenching. Hence, to achieve efficient ET and TTA, one must select donors and acceptors with sufficiently long triplet-state lifetimes so that diffusional-based ET can occur;18 however, this can also be addressed by tethering the donor and acceptor(s) together to create donor-acceptor (D-A) dyads. In recent years, several groups have reported D-A dyads that exhibit different degrees of complexities and photophysics.19–22 D-A dyads allow for rapid intramolecular ET between the donor and acceptor, which improves the efficiency of said ET. Prior work in our lab on D-A dyads for TTA-UC has revealed, however, that in addition to intramolecular ET, charge transfer (CT) can also occur between the donor and acceptor (Figure 1b).23–25 In this picture, the ET process is hampered by the CT dynamic, affecting the overall TTAUC process and lowering the ΦUC. In the present work, we hypothesize that the issue of CT competing with ET in D-A dyads could be nullified if the CT state were contained within the donor fragment. With this in mind, (4N-carbazolyl-2,6-dichlorophenyl)bis(2,4,6-trichlorophenyl)methyl radical (TTM–Cz) appeared as a promising donor for D-A dyads. First reported in 2006 by the Juliá group,26 TTM–Cz is a stable radical adduct of the tris(2,4,6-trichlorophenyl)methyl radical (TTM) and carbazole (Cz). Previous studies on TTM–Cz and its derivatives led to the postulation that in the excited state, TTM–Cz can either emit (D1®D0) from its locally excited (LE) state or undergo structural reorganization following a CT, in the form of a twisted intramolecular charge transfer (TICT),27 from the Cz moiety to the trivalent carbon atom in the TTM moiety (Figure 2a).26,28–30 For the above reasons, we concluded that TTM–Cz is a suitable donor that could be used to prevent photoinduced CT in a D-A dyad used for TTA-UC. Furthermore, TTM–Cz is also attractive for 6 circumventing the well-known singlet-to-triplet ISC. With TTM–Cz, one can directly perform doublet-to-triplet energy transfer (DTET), eliminating the need for ISC and the energy loss that comes with it (Figure 1c). Although the circumvention of ISC does not directly impact ΦUC, it allows the process to be more energy efficient, which is important to the central idea of upconverting low-energy photons. To date, DTET using TTM-based radicals has been demonstrated by several groups. Duan et al. used TTM–Cz as a donor to upconvert red light to blue or cyan light31 and in a similar manner, Yang et al. exploited the doublet photophysics of a triphenylamino-TTM derivative to achieve near-infrared TTA-UC with rubrene and perylene.32 Additionally, Friend et al. showed that rapid DTET can occur between TTM–Cz and anthracenes when linked together through the Cz moiety.33 Figure 2. a) Optical transitions (absorption & emission) of TTM–Cz featuring CT characteristics in its excited doublet state. b) Doublet donor-acceptor dyads TTM–Cz–Per and TTM–Cz–DPA depicting the doublet energy donor (perylene = Per and 9,10-diphenylanthracene = DPA). Herein, we report the synthesis and photophysical characterization of TTM–Cz based D-A dyads TTM–Cz–Per and TTM–Cz–DPA that use perylene (Per) and 9,10-diphenylanthracene (DPA) as the acceptors, respectively (Figure 2b). Using spectroscopic techniques and Cl Cl Cl Cl Cl Cl Cl Cl N D1 hν (580–690 nm) hν’ (600–900 nm) TTM–CzLE TTM–CzCT and Cl Cl Cl Cl Cl Cl Cl Cl N TTM–Cz D0 Cl Cl Cl Cl Cl Cl N Cl Cl – Cl Cl Cl Cl Cl Cl Cl Cl N TTM–Cz–Per Cl Cl Cl Cl Cl Cl Cl Cl N TTM–Cz–DPA Energy Donor (sensitizer) Energy Acceptors (emitters) b)a) 7 computational tools, we aimed to determine the nature of the electronic interactions between TTM–Cz and the polyaromatic (PAH) acceptors (Per and DPA). Our investigations revealed that in the case of TTM–Cz–Per, there is rapid DTET from TTM–Cz to Per that is insensitive to molecular oxygen. Furthermore, without adding free acceptor chromophores, TTM–Cz–Per by itself is capable of performing red-to-blue TTA-UC regardless of the polarity of the media used or whether or not molecular oxygen is present. To the best of our knowledge, this is the first such example of an all-in-one D-A dyad that possesses such characteristics. 8 2. MATERIALS & METHODS All commercially obtained reagents and solvents were utilized without further purification. All spectroscopy measurements were performed using spectroscopy-grade solvents. EPR spectra were recorded on a Bruker® EMX EPR spectrometer. 2.1. Synthesis Scheme 1. Synthesis of TTM–Cz based dyads TTM–Cz–Per and TTM–Cz–DPA. Note: For Method B, TTM–Cz–Br was obtained by purification of the first step using 3-Bromocarbazole (similar to Method A), resulting in a mixture of the radical and nonradical precursor. This is why the Suzuki coupling gave a mixture of the radical and nonradical precursors. The synthesis and characterization of TTM–Cz, TTM–Cz–Per, and TTM–Cz–DPA, as well as their corresponding precursors, are detailed in the Supporting Information (Schemes S1–S8). Two different methods were employed for synthesizing the TTM–Cz compounds, as depicted in p-chloranil THF, rt, dark, 1 h R N H Cs2CO3 + DMF, 160 °C overnight R = H: Cz Cl Cl Cl Cl Cl Cl Cl Cl Cl TTM R = Br: Cz–Br R = Per: Cz–Per Method A Toluene/H2O, 90 °C overnight Method B Cl Cl Cl Cl Cl Cl Cl Cl PAH N Cl Cl Cl Cl Cl Cl Cl Cl PAH N + Cl Cl Cl Cl Cl Cl Cl Cl Br N + TTM–Cz–Br PAH Bpin PAH–Bpin p-chloranil THF, rt, dark, 1 h Cl Cl Cl Cl Cl Cl Cl Cl R N R = H: TTM–Cz (10%) R = Br: TTM–Cz–Br (11%) R = Per: TTM–Cz–Per Cl Cl Cl Cl Cl Cl Cl Cl PAH N PAH = Per: TTM–Cz–Per (3%) PAH = DPA: TTM–Cz–DPA (90%) Mixture of radical and nonradical precursor Bu4NOH (1 M, MeOH) THF. rt, dark, 5 h Bu4NOH (1 M, MeOH) THF. rt, dark, 5 h Pd(PPh3)4 K2CO3 Cl Cl Cl Cl Cl Cl Cl Cl R N Cl Cl Cl Cl Cl Cl Cl Cl R N + Mixture of radical and nonradical precursor 9 Scheme 1. Method A follows previously reported protocols to prepare TTM–Cz,26,28 where radical TTM was linked to Cz in the presence of a base via a radical-mediated nucleophilic aromatic substitution. This produced a mixture of radical TTM–Cz and the nonradical counterpart which must be further oxidized (radicalized) to afford the desired radical TTM–Cz. Method B consists of Suzuki coupling of mono-brominated TTM–Cz–Br to borylated PAH (either Per or DPA) followed by oxidation (radical reaction) to give the desired radical dyads TTM–Cz–Per and TTM–Cz–DPA. 2.2. Computational Methods Ground state geometry optimizations and spin density calculations of the (D0)(S0)A and (D0)(T1)A states were performed at the UB3LYP/6-31G(d,p) level of theory.34 Vertical excitations to the (D1)LE(S0)A state and the corresponding spin densities were computed with time-dependent density-functional theory (TD-DFT) using the Tamm-Dancoff approximation35 and UCAMB3LYP/6-31G(d,p)36 after reoptimization at the same level of theory, as range separated hybrid functionals have shown high accuracy in the prediction of CT bands.37 Solvation effects were included using the IEF-PCM model for hexane and THF. Further details are provided in the Supporting Information. 2.3.Photophysical Method & Photoluminescence Upconversion Procedure UV–vis absorption spectra were recorded on an Ocean Optics® QEPRO-ABS spectrometer using a DH-mini light source. Emission spectra were recorded on an Edinburgh Instruments FLS1000 photoluminescence spectrometer. For power dependent measurements, the noncoherent Xe lamp of the FLS1000 spectrometer was used as the light source and a pinhole with a diameter 16 Cz–Per as the intrinsic emission of the TTM–Cz moiety completely vanished. These findings are in agreement with those of the Juliá group,28 where emission intensity was greater in nonpolar solvents than in polar media, and electron-donating substituents on the Cz moiety caused a redshift in the emission band. This supports the hypothesis that there is an intramolecular CT from the Cz moiety to the radical-centered carbon atom of TTM; despite this, the significant decrease in intensity of the intrinsic emission of TTM–Cz in TTM–Cz–Per can also be partly attributed to quenching of the doublet excited state of TTM–Cz via 2(TTM–Cz)*→Per DTET to produce the 3(Per)* species. In addition to the red photoluminescence, a blue-green emission centered at ca. 490 nm was detected for samples of TTM–Cz–Per, and this emission was present regardless of solvent polarity or the presence of oxygen, as shown in Figures 6b and 6d. Figure 7a reveals that the intensity of this higher-energy emission band quickly diminished as the solution concentration of TTM–Cz– Per was increased. The reduction in emission intensity can be ascribed to the secondary inner filter effect, as this emission overlaps with the absorption profile of TTM–Cz–Per. When the solution concentration was kept constant while varying the power density of the excitation light, the intensity of the emission showed a quadratic behavior, as shown in Figures 7b and 7c. This is characteristic of the non-linearity of the TTA-UC process.5,7,14,18,38–40 Based on this finding, we attributed the emission band at ca. 420–600 nm to the delayed fluorescence of Per as the result of TTA between two 3(Per)* species of TTM–Cz–Per. 17 Figure 7. a) Delayed emission/fluorescence of Per at varying concentrations of TTM–Cz–Per; inset shows the integrated intensity of the emission at various concentrations. b) Power densitydependent delayed fluorescence of Per from red-excitation (lExc = 620 nm) of TTM–Cz–Per. c) Quadratic response of the delayed fluorescence of Per vs. excitation power density for solutions of TTM–Cz–Per under both ambient and oxygen-free conditions. d) Absorbance at 376 nm (inset: 610 nm) for toluene solutions of TTM–Cz and TTM–Cz–Per that were continuously irradiated with red light for 24 hours (λExc = 620 nm, power = 2.23 mW, [TTM–Cz] = 38 μM, [TTM–Cz– Per] = 22 μM). Note: For the power density-dependent measurements, a noncoherent Xe lamp was used as the excitation source with a 3 mm diameter pinhole directly in front of the sample. Power density was adjusted using a variable neutral density filter. Using a relative method, we estimated the ΦUC ≈ 0.012% for a sample of OD = 0.021 excited with the non-coherent xenon lamp at λExc = 610 nm (for more details, see Supporting Information, 18 Figure S18). Based on the above result, we postulated that as the TTM–Cz fragment of the dyad was selectively excited at 610 nm, the Per acceptor was promoted to the triplet excited state via DTET, i.e. (D1)TTM–Cz (S0)Per→(D0)TTM–Cz (T1)Per. This is also supported by the fact that the intensity of the red emission from TTM–Cz–Per is significantly reduced compared to that of TTM–Cz which would be the result of the DTET process quenching the doublet excited state of the TTM–Cz moiety. Since there is no noticeable difference in upconversion efficiency between nonpolar and polar solvents, we hypothesize that the DTET could occur from either the LE or CT state of TTM–Cz. Also, it is noteworthy to mention that the upconversion efficiency appears to be negligibly affected by oxygen, a finding that is unprecedented for TTA-UC. In addition, TTA-UC occurred without adding free Per acceptors to the samples. This is another unprecedented finding because, to our knowledge, all previous reports on D-A dyads for TTA-UC required the addition of a free acceptor for the upconversion process to be observed.20,22,23,25,41 Further, as tabulated in Table 1, entries 1 and 3, we observed that TTM–Cz–Per has similar emission decay kinetics at 690 nm as TTM–Cz in toluene. This can be explained by the fact that although the two components exhibit different multiplicities (doublet and triplet), they act as a two-in-one chromophore. On the other hand, we observed that the kinetics associated with the emission decay in the control sample of TTM–Cz + Per (Table 1) appeared longer, indicating that in the TTM–Cz–Per dyad, the DTET is primarily intramolecular. Also, in both TTM–Cz–Per and TTM–Cz + Per, the emission decay at 475–480 nm is longer in deaerated solutions when compared to solutions under ambient conditions. This is not surprising as in deaerated samples, quenching of the T1 state of Per is precluded. 19 We also investigated the stability a toluene solution of TTM–Cz–Per under continuous irradiation (at 620 nm, 2.23 mW) over a period of 24 h. As shown in Figure 7d, this dyad is noticeably stable than TTM–Cz. This result agrees with previous findings by Albrecht et al., who demonstrated that additional substitution on the carbazole moiety with a phenyl ring significantly enhances photostability.42,43 3.3.Transient Absorption Spectroscopy To further explore the DTET process and understand the nature of the excited state and/or transients, we performed femtosecond and nanosecond transient absorption (fs-TA and ns-TA, respectively) spectroscopy on oxygen-free solutions of TTM–Cz, TTM–Cz–Per, and a mixture of TTM–Cz and Per (1:2.4 molar ratio). The combined fs-ns-TA results are shown in Figure 8 (see Supporting Information Figures S20 and S21 for the individual fs-TA and ns-TA data, respectively). The doublet transient of TTM–Cz, 2(TTM–Cz)*, absorbs from 500 to ca. 900 nm with a maximum at ca. 570 nm (Figures 8a,b). A negative absorption band appearing between 600 and 750 nm is the combination of the ground-state bleach and stimulated emission, centered at ca. 630 and ca. 690 nm, respectively. Since it was hypothesized that the doublet excited state of TTM– Cz features LE and CT states26,28–30, we ascribed the maximum at ca. 570 nm to be the LE D1→Dn transition and the broad absorption at ca. 850 nm to be associated with the CT transient. The LE transient band was fitted with a biexponential function to produce two lifetime constants, τ1 = 152 ps and τ2 = 21.47 ns; whereas the CT transient generated a monoexponential decay with τ = 24.39 ns (Figure 8c and Table 2). Interestingly, these transient species are also present in the mixture of TTM–Cz + Per; however, they have much shorter lifetime values, with τ1 = 139 ps and τ2 = 13.65 ns for the LE transient, and τ = 13.93 ns for the CT transient (Figures 8g–i). For the TTM–Cz + 20 Per sample, we attributed the shorter lifetimes to be a result of 2(TTM–Cz)*→Per DTET leading to the quenching of the excited state of TTM–Cz. Figure 8. (a & b) Transient absorption map and spectra for TTM–Cz and the corresponding kinetic traces (c). (d & e) Transient absorption map and spectra for TTM–Cz–Per and the corresponding kinetic traces (f). (g & h) Transient absorption map and spectra for TTM–Cz + Per (1:2.4 molar ratio) and the corresponding kinetic traces (i). Note: Insets for the kinetic traces are the kinetic traces with a logarithmic time scale. pump @ 610 nm in deaerated toluene (OD = 0.4 @ 610 nm) pump @ 610 nm in deaerated toluene (OD = 0.4 @ 610 nm) a) b) c) d) e) f) g) h) i) pump @ 610 nm in deaerated toluene (OD = 0.54 @ 610 nm) τ1, 570 = 152 ps τ2, 570 = 21.47 ns τ1, 490 = 13.40 ns τ2, 490 = 5.98 µs τ1, 570 = 8.73 ns τ2, 570 = 97.06 ns τ850 = 24.39 ns τ810 = 4.31 ns τ1, 570 = 139 ps τ2, 570 = 13.65 ns τ850 = 13.93 ns 21 Furthermore, an additional long-lived transient species with absorption centered at 490 nm can be attributed to the triplet transient of Per, 3Per*. This species decays with a biexponential kinetic with time constants τ1 = 13.40 ns and τ2 = 5.98 μs (Figures 8g–i and Table 2). The τ1 component was assigned to the residual doublet transient of TTM–Cz, which overlaps with the spectrum of 3Per*. This is supported by the fact that τ1 at 490 nm is almost equal to τ2 at 570 nm for the decay of the LE transient of TTM–Cz in the TTM–Cz + Per sample. The long-lived component with τ2 is attributed to the decay of 3Per*, as relaxation to the S0 ground state is a spin-forbidden process. Table 2. Transients time constants for TTM–Cz, TTM–Cz–Per, and TTM–Cz + Per (1:2.4 molar ratio). Compound λAbs (nm) τ1 (ns) τ2 (ns) TTM–Cz 570 0.152 (25.3%) 21.47 (74.7%) 850 24.39 - TTM–Cz–Per 570 8.73 (81.1%) 97.06 (18.9%) 810 4.31 - TTM–Cz + Per 490 13.40 (49.5%) 5980 (50.5%) 570 0.139 (38.3%) 13.65 (61.7%) 850 13.93 - For TTM–Cz–Per, the transient absorption band stretched from 500 to 900 nm with apparent maxima at ca. 570 and 810 nm (Figures 8d,e). As in TTM–Cz, the transient species with absorptions at 570 and 810 nm can be attributed to the LE and CT states of the TTM–Cz moiety, respectively. However, it is likely that the absorption feature of the triplet transient of Per, 3Per*, partially overlaps with the broad peak near 570 nm. Expectedly, the fitting of the band centered at ca. 570 nm gave a biexponential kinetic with time constants τ1 = 8.73 ns and τ2 = 97.06 ns (Figure 22 8f and Table 2). The τ1 component was attributed to the decay of the LE transient of TTM–Cz component in the dyad, which, like TTM–Cz + Per sample, has a shorter lifetime due to DTET from TTM–Cz to Per. The long-lived τ2 component is likely the decay of 3Per*; however, it is worth noting that the lifetime is significantly shorter than what was observed in the mixture of TTM–Cz + Per. This can be explained by a likely intramolecular back energy transfer from 3Per* to TTM–Cz. Furthermore, the decay of the CT transient band at ca. 810 nm was fitted with a monoexponential function to produce a time constant of τ = 4.31 ns, which was attributed to the residual of the CT transient from the TTM–Cz component. Furthermore, the LE transient species from TTM–Cz and TTM–Cz–Per exhibit unrelated characteristics that deserve to be mentioned. In TTM–Cz, the transient species grows rapidly before decaying biexponentially, whereas in TTM–Cz–Per, within the 350-fs time resolution of the instrument, the transient species has an initial rapid growth followed by a quick/sharp decay (Figure 8f). This decay is followed by a slower growth that turns into a complete biexponential decay (See Supporting Information, Figures S20 & S21). We postulate that this complex dynamic is likely due to the LE state of TTM–Cz being populated, followed by a rapid intramolecular DTET to Per and the growth of 3Per*, which formed simultaneously as the 2(TTM–Cz)* decayed. It is worth noting that the DTET dynamic was not clearly observed for either TTM–Cz–Per or TTM–Cz + Per (decay of doublet transient of 2(TTM–Cz)* accompanied by growth of 3Per*); this is most likely due to the overlap between the transient bands of TTM–Cz and the triplet transient spectrum of Per. Alternatively, it is possible that DTET may occur at a rate beyond the resolution of the instrumentation used. 23 3.4. Computational Investigations To ascertain the spin state and overall spin density within the TTM–Cz–Per/DPA dyads, we calculated the spin density maps in the doublet (D0)(S0)A, (D1)LE(S0)A (S = 1/2) and quartet (D0)(T1)A (S = 3/2) states of TTM–Cz, TTM–Cz–Per, and TTM–Cz–DPA as shown in Figure 9. In the (D0)(S0)A state, the spin density maps for all three chromophores have an unpaired electron localized over the TTM fragment. The three compounds share an absorption band at ~400 nm that we attribute to (D1)LE(S0)A and characterize as a local excitation of the common TTM–Cz moiety (see Natural Transition Orbitals in Figure S22). In this state, the unpaired electron delocalizes over the Cz moiety. These results suggest that after excitation to the (D1)LE(S0)A state, these dyads could exhibit the expected intramolecular DTET from TTM–Cz to Per and DPA to generate the corresponding 3Per* and 3DPA*, respectively. This is further corroborated by spin densities computed in the (D0)(T1)A state showing a net density of unpaired electrons of ρ ≈ 1 on the TTMCz donor and of ρ ≈ 2 on the acceptor. Photoluminescence from 1Δg(O2) and/or observation of the delayed fluorescence from the TTM–Cz–Per sample, presumably from the 3Per*, further corroborate the computational results and our hypothesis of intramolecular DTET. 24 Figure 9. Spin density maps of the (D0)(S0)A, (D1)LE(S0)A, and (D0)(T1)A states computed in the gas phase for TTM–Cz, TTM–Cz–Per, and TTM–Cz–DPA. Partial spin densities per fragment (ρ) were normalized so that total densities are ρ = 1 and ρ = 3 for doublet and quartet states, respectively. (D1)LE(S0)A corresponds to the first allowed transition (nonzero oscillator strength): Root = 1, λabs = 428 nm for TTM–Cz; Root = 2, λabs = 433 nm for TTM–Cz–Per; Root = 2, λabs = 431 nm for TTM–Cz–DPA. CONCLUSIONS: The current study demonstrates the first example of all-in-one D-A dyads that can perform TTA-UC under ambient conditions and without the need for additional acceptor/annihilator molecules to be present in the matrix. In the case of the TTM–Cz–Per dyad, detection of 1Δg(O2) emission and upconverted photoluminescence confirmed DTET from the TTM–Cz moiety to the Per acceptor. In addition, the fact that upconverted photoluminescence was seemingly unaffected by solvent polarity suggests that the DTET process can occur from both the LE and CT states of the TTM–Cz moiety. 25 Using time-resolved transient absorption spectroscopy, we observed ultrafast intramolecular interactions between the TTM–Cz donor and the Per acceptor in TTM–Cz–Per. Spin density computations corroborate this finding as the spin density map of the (D0)(T1)A state of TTM–Cz– Per indicates that an unpaired electron becomes delocalized over the Per moiety, which is consistent with Per being in the triplet excited state. On the other hand, the intermolecular DTET in the sample of free TTM–Cz + Per was an inefficient process. The current results highlight the possibility of designing single-component D-A systems for light-harvesting and photon upconversion processes. Furthermore, we demonstrated that it is possible to deconvolute competing ET and CT dynamics in the D-A systems. Currently, we are devising related multichromophoric systems that can potentially be used for other non-linear photon-management processes in next-generation optoelectronic devices. 32 TOC Graphic: