Nanoparticle Size-Dependent Singlet Fission and Exciton Dynamics in Amorphous TIPS-Pentacene

Abstract

Aqueous nanoparticle (NP) dispersions are commonly used as model systems for the spectroscopic study of singlet exciton fission (SF) in acenes such as 6,13-(triisopropylsilylethynyl)­pentacene (TIPS-Pn). However, the potential for particle size effects to complicate interpretation of results in such model systems is generally ignored. In this work, we study amorphous TIPS-Pn NP dispersions prepared by the re-precipitation method over a range of particle sizes. Time-resolved fluorescence and femtosecond transient absorption spectroscopies show that exciton dynamics in these systems depend significantly upon particle size. Kinetic analysis reveals that SF becomes slower at smaller NP sizes, while triplet exciton decay (through both correlated triplet pair relaxation and geminate triplet–triplet annihilation) accelerates. These significant size-dependent effects are ascribed to increased morphological disorder within smaller NPs, weakening the intermolecular couplings which control SF and triplet migration. A non-radiative singlet quenching channel separate from SF is also identified, which has not been previously reported for NPs of SF-capable chromophores. This non-radiative singlet decay becomes a significant relaxation pathway at small particle sizes, substantially reducing SF yields. Interestingly, exciton kinetics in the largest NPs considered here (81 nm diameter) approach those of bulk amorphous TIPS-Pn. This work demonstrates that particle size effects are significant for small NPs of SF chromophores and must be accounted for in order to accurately model bulk materials with such NP dispersions.

What Next for Singlet Fission in Photovoltaics? The Fate of Triplet and Triplet-Pair Excitons

“…SF yields ϕ SF and ϕ SF ′ for (a) amorphous TIPS-Pn NPs as a function of NP diameter and (b) PMMA/TIPS-Pn composite NPs as a function of PMMA/TIPS-Pn mass ratio, where a higher mass ratio corresponds to greater average separation between TIPS-Pn molecules. Data are taken from refs and . In these systems, T 1 is indistinguishable from 1 (T···T).…”

“…Given the lack of delayed fluorescence in systems in which SF is exothermic (such as pentacene), it is hard to verify this assertion via the presence of phenomena such as the quantum beating observed in tetracene and rubrene. , It has been suggested that spin decoherence may be faster in more disordered systems, so the assumption that spin decoherence should be on a ns−μs time scale may not be reasonable for all systems. Regardless, it is expected that the 1 (T···T) state is electronically and chemically indistinguishable from free triplets, and as such, these states are grouped together in many studies. − In experiments that probe spin resonance, such as magnetic-field dependence on delayed fluorescence or electron paramagnetic resonance spectroscopy, it is possible to infer the presence of 1 (T···T) rather than T 1 +T 1 if spin coherence is observed at a time when 1 (TT) can be ruled out. However, this information is absent in other methods commonly used to study SF (such as transient absorption spectroscopy) or in systems where there is no delayed fluorescence (such as exothermic SF systems), so the convention has become somewhat inconsistent.…”

“…However, this information is absent in other methods commonly used to study SF (such as transient absorption spectroscopy) or in systems where there is no delayed fluorescence (such as exothermic SF systems), so the convention has become somewhat inconsistent. Some studies have assumed that triplet-like signals appearing on the ps–ns scale were 1 (T···T), ,, while others have assigned them as T 1 +T 1 and noted that there may still be spin coherence present. ,, To avoid confusion, we suggest more clarity and transparency about what is known in these systems when making any assignments.…”

“…Miyata et al have stated that the rate of SF should be defined by when 1 (T···T) rather than 1 (TT) is formed, and we extend this thinking to calculations for SF efficiency or yields. We suggest the previous convention in our group of defining the singlet fission yield strictly as the formation of T 1 , , or since the 1 (T···T) state is often electronically indistinguishable from free triplets, The efficiency of the first step of singlet fission, i.e., 1 (TT) formation, can be expressed separately as where including [T 1 ] and [ 1 (T···T)] accounts for any pairs that have already separated. For efficient SF, both of these yields should be close to 2.…”

“…Figure shows ϕ SF and ϕ SF ′ for a series of 6,13-bis­(triisopropylsilylethynyl)­pentacene (TIPS-Pn) NPs of varying morphology and demonstrates the importance of distinguishing between 1 (TT) and T 1 excitons. In these systems, 1 (TT) and T 1 are spectrally distinct, so they can be separated and their respective concentrations determined to calculate the yields in eqs and . , In both studies, the yields show that the efficiency of the entire SF process, i.e., the formation of T 1 or 1 (T···T) (ϕ SF ), is much lower than the efficiency of the formation of 1 (TT) (ϕ SF ′ ). If solely ϕ SF ′ were considered in these studies, the SF efficiency would be significantly overestimated.…”