Article pubs.acs.org/JPCA
Photochemical Degradation of Various Bridge-Substituted FluoreneBased Materials Björn Kobin, Sandra Behren, Beatrice Braun-Cula, and Stefan Hecht* Department of Chemistry & IRIS Adlershof, Humboldt-Universität zu Berlin, 12489 Berlin, Germany S Supporting Information *
ABSTRACT: Photochemical degradation is an important issue to be overcome in advancing the lifetime of fluorenecontaining conjugated polymers. In order to optimize the inertness of the materials, a quantitative measure for the efficiency of degradation is needed. Here, we introduce a method to measure a relative quantum yield of the photochemical degradation by monitoring the kinetics of the process by means of UV/vis spectroscopy and liquid chromatography (LC) techniques. This method is employed to a set of differently substituted 2,7-diphenylfluorenes, serving as model compounds for polyfluorene materials. Our measurements show that the quantum yield changes by orders of magnitude upon varying the bridge substituents and that altered kinetics indicate changing degradation mechanisms.
■
INTRODUCTION
should prove viable to the design of long-lived blue emitter materials. In this contribution, we describe the synthesis and characterization of a variety of differently substituted fluorene derivatives and compare their stability toward photochemical degradation in solution in a quantitative manner. At this point we want to emphasize that the term “stability” is somewhat misleading as it refers to a thermodynamic property of the ground state,38 while photodegradation is an inherently kinetic phenomenon, in which the availability of reaction pathways originating from an electronically excited state determines the outcome of the photochemical reaction. Thus, we prefer to use the term “inertness” as it refers to the lack of reactivity (and not necessarily stability).39 This does not mean that we reject the validity of the laws of thermodynamics, but minimization of the ground state Gibbs free energy of the system does not constitute the main driving force. On a more quantitative level, the rate of a photoreaction is dependent on the intensity of absorbed light, and hence the sensitivity toward photodegradation is most accurately described by a reaction quantum yield. Since the absolute quantum yield of the unwanted photodegradation process is typically very low and thus difficult to measure, we chose to determine relative quantum yields, which are normalized to the most inert, i.e., least reactive, compound in the investigated series. In order to facilitate the kinetic analysis of degradation, we decided to use small molecules that carry only one fluorene unit and hence only one methylene bridge in organic solution under ambient conditions. To shift the optical transitions to lower energies, the fluorene
Over the past years there has been a lot of research in the field of organic electronics devoted toward exploring specific advantages of organic materials associated with their potentially low-cost manufacturing, light weight, and flexibility, while at the same time trying to overcome their disadvantages related to the efficiency and longevity of organic-based devices.1 Regarding the aspect of enhancing lifetime, this has been a particularly challenging topic in the context of absorbing and emitting devices, such as organic photovoltaic cells and organic light emitting diodes (OLEDs). In the latter case especially the lifetime of blue emitting OLEDs is still limited.2 With regard to conjugated polymers as blue emitting materials in OLEDs, polyfluorenes (PFs) as well as ladder-type poly(p-phenylene)s (LPPPs)3 have extensively been exploited,4,5 but suffer from relatively fast (photo)degradation. In PFs the degradation process is typically accompanied by an emerging green emission, which has been attributed to either fluorenone defects, hydroxyl (chain end) defects, cross-linking, or excimer formation.6−22 Most studies propose fluorenone type defects, which may be formed thermally23,24 as well as photochemically, by direct excitation25,26 or via singlet oxygen mediated oxidation,27,28 and typically occur at synthetic defect sites. Clearly, the substitution in the bridging benzylic 9-position of the fluorene moiety plays a crucial role in the degradation process. This is exemplified by the observation that introduction of aryl moieties in this position, in particular as the spiro-bifluorene motif, renders the respective polymers more stable toward thermal and photochemical degradation.29−37 However, no systematic study comparing the influence of a larger variety of substituents in the 9-position of a fluorene-type backbone has been reported thus far and yet © XXXX American Chemical Society
Received: February 29, 2016 Revised: June 8, 2016
A
DOI: 10.1021/acs.jpca.6b02127 J. Phys. Chem. A XXXX, XXX, XXX−XXX
Article
The Journal of Physical Chemistry A core was substituted by two terminal phenyl units in 2- and 7positions, therefore giving rise to the investigated, differently substituted 2,7-diphenylfluorene (DPF) derivatives (Scheme 1). The methylene bridge was symmetrically substituted with a Scheme 1. Chemical Structures of the Investigated Fluorene Derivatives
variety of groups, including methyl (DPF-Me2), hydrogen (DPF),40 phenyl (DPF-Ph2), spiro-bifluorene (DPF-sp), fluorine (DPF-F2), and oxo (DPF-O)40,41 substituents.
■
RESULTS AND DISCUSSION Synthesis and Structural Characterization. The target compounds were synthesized by Suzuki cross-coupling42 of phenylboronic acid with the different central 2,7-dibromofluorene units. The latter ones included 2,7-dibromo-9,9difluorofluorene, the central building block for DPF-F2, which was synthesized according to a simplified protocol by Katzenellenbogen,43 while 2,7-dibromo-9,9-diphenylfluorene was prepared starting from 2,7-dibromofluorenone via a sequence of ring opening, esterification, phenyl lithium addition, and Friedel−Crafts alkylation according to slightly modified literature procedures.30,44 More information about synthesis and characterization can be found in the Supporting Information. In the case of four derivatives we were able to obtain crystals suitable for single crystal X-ray structural analysis, which provided an accurate depiction of the molecular geometry in the solid state (Figure 1). DPF-Me2 and DPF-sp crystallize in Pna21 (orthorhombic), DPF-F2 in P21/m (monoclinic), and DPF-O46 in Cmc21 (orthorhombic) space groups. None of the structures incorporate solvent molecules. DPF-F2 and DPF-O possess mirror symmetry with respect to the orientation of the phenyl units. DPF-F2 exists in two different conformations in the crystal structure. The twist of the planes of the two terminal phenyl moieties with regard to the central fluorene plane amounts to 27.08(7)° and 31.43(7)° in the case of DPF-F2 and 35.93(5)° for DPF-O and ranges between 32.22(10)° and 37.28(9)° in the case of DPF-Me2 and 40.27(6)° and 43.84(6)° for DPF-sp, which shows that the latter two molecules are not symmetrical with regard to the rotation of the terminal phenyl substituents, presumably due to packing effects. Based on the relative orientation of the aromatic planes DPF-F2 can be considered the most planar of these structures in the solid state
Figure 1. Molecular structures in the crystalline solid of DPF-Me2, DPF-sp, DPF-F2, and DPF-O. For DPF-F2 only one of the two different molecular structures is shown. Atomic displacement parameter (ADP) plots45 show atomic positions as ellipsoids drawn at the 50% probability level.
and DPF-sp the least planar one. From the above comparison it becomes apparent that the phenyl rings are twisted somewhat out of plane most likely to reduce steric repulsion of the ortho hydrogen atoms, similarly to the classical biphenyl twist. This twist slightly reduces the overlap across the phenyl−fluorene connection and hence limits π-conjugation in the solid. However, due to the low barrier of rotation this effect should be negligible in solution. Optical Properties. The absorption and emission spectra of all investigated compounds have been recorded in ethyl acetate solution at room temperature (Figure 2, Table 1), and the corresponding absorption and emission maxima as well as emission quantum yields are summarized in Table 1. Additional data regarding the use of other organic solvents are compiled in the Supporting Information. The all-hydrocarbon compounds (DPF, DPF-Me2, DPF-Ph2, and DPF-sp) show more or less the same behavior in absorption: There are two maxima or one maximum and one shoulder between 310 and 330 nm. The first maximum is slightly stronger than the second one, except for DPF-sp. Here, the stronger absorption in the second maximum can be attributed to the additional absorption of the spiro-bifluorene unit. The shape of the emission spectrum is even more similar: B
DOI: 10.1021/acs.jpca.6b02127 J. Phys. Chem. A XXXX, XXX, XXX−XXX
Article
The Journal of Physical Chemistry A
Figure 2. Absorption (solid line) and normalized emission (dotted) spectra of the DPFs in ethyl acetate, ≈10−5 mol L−1: DPF (red), DPF-Me2 (black), DPF-Ph2 (blue), DPF-sp (magenta), DPF-F2 (green), and DPF-O (yellow).
a low energy emission related to that transition with a maximum around 540 nm and a ΦPL of about 0.04. Photochemical Degradation. The experiments were carried out in ethyl acetate as the solvent because it does not undergo solvent-specific reactions, such as chloroform and methylene chloride, and degradation occurs faster than in toluene, thereby facilitating kinetic studies.28 First insight into the photodegradation of the different derivatives was gained by monitoring the absorption spectra upon prolonged irradiation with UV light (300−400 nm) as illustrated for DPF and DPFMe2 in Figures 3a and 3b. In both cases the main absorption band decreases while a low-energy tail and some blue-shifted band are building up. In the case of DPF-Me2 the photodegradation is initially relatively slow, but then the process clearly accelerates (Figure 3c). The respective extinction difference (ED) plot47 (Figure 3a, inset) does not show a linear relationship, which provides a clear indication that no uniform photochemical reaction, converting the substrate (A) directly into one product (B), is taking place and instead several parallel/serial reactions seem to occur. The observed acceleration effect may be related to an autocatalytic degradation process involving singlet oxygen, as we have recently proposed for methylated ladder-type oligo-pphenylenes.28 In later stages of the degradation process (not shown here) the absorption broadens and evenly decreases upon further irradiation. UPLC-UV/vis-absorption analysis (Figure 3e) of degraded solutions shows a large variety of products. This is in agreement with the ED plots, demonstrating the large variety of involved photoreactions. The two aryl substituted compounds DPF-Ph2 and DPF-sp also show the same behavior (see Figures S4.3 and S4.4). In the case of the nonmethylated derivative DPF the conversion is much faster. The reaction starts rapidly, decelerates over time, and stops more or less at a certain point (Figures 3b and 3d). The ED plot (Figure 3b, inset) clearly shows a linear relationship, and linear fitting provides lines, whichmore or lesscross the origin. This kind of kinetics may be related to a first-order reaction, as expected for a uniform photoreaction (A → B). Furthermore, it was found that the final absorption spectrum (red line in Figure 3b) is reminiscent of the absorption spectrum of DPF-O, which was unambiguously confirmed by UPLC analysis (Figure 3f). The degradation of the fluorinated derivative DPF-F2 seems even faster, also giving DPF-O as reaction product. However, the ED plot (see Figure S4.2), which is not strictly linear, as well as the initial increase and subsequent decrease of the band at 3 eV
Table 1. Absorption and Emission Data of Investigated Compounds in Solutiona compound
λmax/nm
εmax/103, L mol−1 cm−1 (λ/nm)
DPF
322
45 (322)
DPF-Me2
325
48 (325)
DPF-Ph2
313, 330
39 (330)
DPF-sp
309, 332
42 (309)
DPF-F2 DPF-O
317 286, 322, 335, 431
45 (317) 19 (322)
λem/ nm
λex/ nm
ΦPL (CH2Cl2)
352, 369 356, 372 357, 374 357, 374 396 538
322
0.5
325
0.8
330
