Absolute measurements of the triplet-triplet annihilation rate and the charge-carrier recombination layer thickness in working polymer light-emitting diodes based on polyspirobifluorene

The triplet exciton densities in electroluminescent devices prepared from two polyspirobifluorene derivatives have been investigated by means of time-resolved transient triplet absorption as a function of optical and electrical excitation power at 20 K. Because of the low mobility of the triplet excitons at this temperature, the triplet generation profile within the active polymer layer is preserved throughout the triplet lifetime and as a consequence the absolute triplet-triplet annihilation efficiency is not homogeneously distributed but depends on position within the active layer. This then gives a method to measure the charge-carrier recombination layer after electrical excitation relative to the light penetration depth, which is identical to the triplet generation layer after optical excitation. With the latter being obtained from ellipsometry, an absolute value of 5 nm is found for the exciton formation layer in polyspirobifluorene devices. This layer increases to 11 nm if the balance between the electron and the hole mobility is improved by chemically modifying the polymer backbone. Also, and consistent with previous work, triplet diffusion is dispersive at low temperature. As a consequence of this, the triplet-triplet annihilation rate is not a constant in the classical sense but depends on the triplet excitation dose. At 20 K and for typical excitation doses, absolute values of the latter rate are of the order of ${10}^{-14}{\mathrm{cm}}^3{\mathrm{s}}^{-1}$.

Figure 1

Chemical structures of the investigated polymers. Above: Repeat unit of the polyspirobifluorene homopolymer; below: the TAD-type hole transport moiety that is randomly incorporated in the above polymer backbone for the copolymer.

Figure 2

Some spectra of the copolymer including transient triplet absorption (a), phosphorescence (b), prompt fluorescence (c), and absorption spectra (d). The dashed lines indicate the optical excitation and the transient absorption probe energy, respectively.

Figure 3

(Color) Two typical time-resolved induced absorption datasets for the homopolymer during optically ($30\mathrm{mW}∕{\mathrm{cm}}^2$, black) and electrically (6 V, green) exciting the device for 1 ms. In order to better evaluate the initial time region, the insets show the same data in a semilogarithmic fashion. In part (a), simulations according to the homogeneous triplet generation model, Eq. (2), are included as red and blue solid lines. Similarly, part (b) contains fits to the inhomogeneous triplet generation model, Eq. (5). For the latter case, the obtained initial slopes and the steady-state values are shown as dotted and dashed straight lines, respectively.

Figure 4

Double logarithmic presentation of the copolymer parameter $a=\lambda {\gamma }_{\mathrm{TT}}∕\epsilon$ for optical and electrical excitation vs the corresponding triplet generation flux as derived from simulations of the induced triplet absorption dependencies such as shown in Fig. 3. The solid lines represent least-square fits to both excitation modes. For guidance, the $x$ axis roughly covers the range from 1.5 to $30\mathrm{mW}∕{\mathrm{cm}}^2$ for optical and 5 to 7.7 V for electrical excitation.

Figure 5

Double logarithmic dependencies of the fluorescence intensity (during the pulse) and the phosphorescence intensity (100–800 ms after the pulse) as a function of the optical excitation pulse length at an excitation dose of $30\mathrm{mW}∕{\mathrm{cm}}^2$ for the copolymer. The solid lines are guide lines with slopes as indicated.

Figure 6

Wavelength-dependent light penetration depth ${\lambda }^{\mathrm{opt}}$ of the copolymer as obtained from ellipsometry (scatter plus line) together with the thin-film absorption spectrum (solid line).