Improved spectral coverage and fluorescence quenching in donor-acceptor systems involving indolo[3-2-b]carbazole and boron-dipyrromethene or diketopyrrolopyrrole

Received Date : 28-Sep-2014

Revised Date : 29-Dec-2014

Accepted Date : 03-Feb-2015

Article type : Special Issue Research Article

Improved spectral coverage and fluorescence quenching in donor-

acceptor systems involving Indolo[3-2-b]carbazole and Boron-

†

dipyrromethene (BODIPY) or Diketopyrrolo-pyrrole (DPP)

1

2

1

Adis Khetubol , Sven Van Snick , Melissa. L. Clark , Eduard Fron , Eduardo Coutiño-

1

2

González , Arvid Cloet , Koen Kennes , Yuliar Firdaus , Maarten Vlasselaer , Volker

2

1*

Leen , Wim Dehaen , Mark Van der Auweraer

1

.

KULeuven, Chemistry Department, Molecular Imaging and Photonics, Celestijnenlaan 200F, B2404,

2

3

001, Leuven, Belgium, . KULeuven, Chemistry Department, Molecular Design and Synthesis,

Celestijnenlaan 200F, B2404, 3001, Leuven, Belgium

This article has been accepted for publication and undergone full peer review but has not been

through the copyediting, typesetting, pagination and proofreading process, which may lead to

differences between this version and the Version of Record. Please cite this article as doi:

1

0.1111/php.12437

*

Corresponding author e-mail: Mark.Vanderauweraer@chem.kuleuven.be (Mark Van der Auweraer)

†

This manuscript is part of the Special Issue dedicated to the memory of Michael Kasha

ABSTRACT

A novel π-conjugated triad and a polymer incorporating indolo[3,2-b]-carbazole (ICZ) and 4,4-

difluoro-4-bora-3a,4a-diaza-s-indacene (BODIPY) were synthesized via a Sonogashira coupling.

Compared to the parent BODIPY the absorption and fluorescence spectrum were for both

compounds broader and red-shifted. The red-shift of the fluorescence and the decrease of the

fluorescence quantum yield and decay time upon increasing solvent polarity were attributed to the

formation of a partial charge transfer state. Upon excitation in the ICZ absorption band the ICZ

fluorescence was quenched in both compounds mainly due to energy transfer to the BODIPY moiety.

In a similar ICZ-π-DPP polymer, (where DPP is diketopyrolopyrrole), a smaller red shift of the

absorption and fluorescence spectra compared to the parent DPP was observed. A less efficient

quenching of the ICZ fluorescence in the ICZ-π-DPP polymer could be related to the unfavorable

orientation of the transition dipoles of ICZ and DPP. The rate constant for energy transfer was for all

compounds an order of magnitude smaller than predicted by Förster theory. While in a solid film of

the triad a further red-shift of the absorption maximum of nearly 100 nm was observed, no such shift

was observed for the ICZ-π-BODIPY polymer.

INTRODUCTION

In the search for alternative energy sources organic material based photovoltaics (OPVs) have been

successfully explored as a viable alternative to Si-based solar cells for some applications (1,2,3). The

driving force for this attention devoted to OPVs, based on both small molecules and polymers, is

largely determined by their advantageous properties such as, low weight, easy processability, high

absorption coefficients.

The obvious advantages notwithstanding, to date, OPVs suffer from a suboptimal power conversion

efficiency and operational lifetime, limiting their large scale introduction to the market. To improve

the device efficiencies, several strategies have been proposed and investigated with varying success

(4,5,6,7,8,9). One such strategy is to improve the photocurrent generated by the device which

depends among other factors on the efficiency of light absorption, and exciton dissociation (3,4). To

increase the efficiency for light absorption, a better coverage of the terrestrial solar emission

spectrum is necessary (10,11,12,13,14,15). This can be achieved in donor-acceptor (D-A)

combinations where the absorption spectrum is shifted to longer wavelengths corresponding to a

smaller optical band gap (16,17,18,19,20,21). Furthermore, using such conjugated D-A combinations

where the HOMO and LUMO are confined to different parts of the chromophore, can increase the

efficiency of exciton dissociation and decrease that of geminate electron hole recombination

(

16,17,18,19,20). In D-A polymers the donor and acceptor are often connected by an olefinic or

acetylenic linker acting as a charge mediator, however, acetylenic spacers have been shown to yield

higher power conversion efficiencies for bulk heterojunctions (22,23).

Indolo[3,2-b]-carbazoles (ICZs) with different substitution patterns and polymers thereof have been

investigated for use in organic electronic applications such as organic field effect transistors (OFETs)

(

24,25,26,27,28), organic light emitting diodes (OLEDs) (29,30,31,32), organic photovoltaics (OPVs)

33,34,35), and dye sensitized solar cells (DSSCs) (36). We recently reported a short and high yielding

(

alternative route towards 2,8-difunctionalized indolo[3,2-b]carbazole which has demonstrated to be

a versatile building block (37). Due to its relatively large optical band gap, large molar extinction

coefficient, relatively long singlet decay time and a moderate fluorescence quantum yield (38), ICZ

can be considered as a suitable donor in a system involved electron or energy transfer, when linked

to an appropriate acceptor building block.

We employed in this work 4,4-difluoro-4-bora-3a,4a-diaza-s-indacene (BODIPY) as the acceptor due

to its excellent properties such as high molar extinction coefficient, excellent thermal and

photochemical stability, and good solubility (39). Moreover, our building block synthesis allows us to

develop many different BODIPY analogues with emission wavelengths tunable from 500 to over 700

nm, by attaching supplementary units at different positions (40,41,42,43,44,45,46).

As a comparison, diketopyrrolo pyrrole (DPP) is also employed as the acceptor moiety. DPP has been

demonstrated to be a good co-monomer for thiophene based D-A conjugates and polymers as it

allows for the synthesis of small molecules and copolymers combining a broad absorption, in the

visible with a small optical band gap and, high thermal stability and charge carrier mobility

(

47,48,49,50,51). We hereby report on the synthesis and spectroscopic properties of small molecule

and polymeric D-A materials using indolo[3,2-b]-carbazole (ICZ) and 4,4-difluoro-4-bora-3a,4a-diaza-s-

indacene (BODIPY) or diketopyrrolo-pyrrole (DPP), to be used as an efficient light harvesting system

in both in solution and in films.

MATERIALS AND METHODS

Materials: All solvents, used for spectroscopic studies are of spectroscopic grade and were

purchased from Sigma-Aldrich and Acros. They were used without a further purification. The

compounds ICZ 1 (37), BODIPY 1 (52), BODIPY 2 (52), and DPP 2 (53) were prepared according to

literature methods.

Synthesis of the ICZ-π-BODIPY triad (triad 1), a round-bottomed flask was loaded with BODIPY 1 (150

mg, 0.42 mmol), ICZ 1 (181.5 mg, 0.29 mmol), CuI (2 mg, 2.5 mol%), Pd(PPh ) (2 mg, 1 mol%) and

3

4

flushed with Ar, THF (1 mL) and iPr NEt (5 mL) were then added. This mixture was heated to reflux for

2

one hour and then allowed to cool to room temperature. The reaction mixture was dissolved in DCM

and run through a celite pad. The organic layer was concentrated and the resulting solid was purified

by silica gel chromatography (CH Cl :petroleum ether 1:1) to yield pure product of triad 1 as a dark

2

purple powder (24% yield). mp: >300°C; δ (300 MHz; CDCl ) 0.88 (10 H, m), 1.05-1.25 (8 H, m), 1.47

H

3

(

4 H, m), 2.45 (6 H, s), 2.58 (6 H, s), 2.60 (6 H, s), 3.78 (4 H, t, J 8.1), 6.21 (2 H, s), 6.62 (2 H, s), 7.13 (2

13

H, s), 7.21 (2 H, d, J 4.05), 7.47 (2 H, d, J 9.78), 7.66 (10 H, m); No adequate C spectrum could be

obtained for the triad using the 75 MHz or the 100 MHz probe.

Scheme 1

Synthesis of the ICZ–π–BODIPY polymer (polymer 1), a round-bottomed flask was loaded with

BODIPY 2 (50 mg, 0.1 mmol), ICZ 1 (62.4 mg, 0.1 mmol), CuI (1 mg, 2.5 mol%), Pd(PPh ) (1 mg, 1

3

4

mol%) and flushed with Ar, THF (2 mL) and iPr NEt (1 mL) were then added. The reaction mixture was

2

heated to reflux and allowed to stir overnight. The resulting dark-blue solution was concentrated

under reduced pressure, providing a dark-blue powder which was purified by Soxhlet extraction,

firstly with CH OH, then acetone, then hexane and finally THF. A black powder with M of 4.4 kg/mol

3

n

and PDI of 2.5 was obtained when the THF fraction was concentrated under reduced pressure. Yield:

6%. δ (300 MHz; CDCl ) 0.80-1.35 (20 H, bs), 2.20 (9 H, bs), 2.59-2.74 (6 H, bs), 3.76 (4 H, bs), 6.51

4

H

3

(

2 H, bs), 6.92 (4 H, bs), 7.16 (4 H, bs), 7.61 (10 H, bs).

Scheme 2

Synthesis of the ICZ–π–DPP polymer (polymer 2), a round-bottomed flask was charged with ICZ 1

(

100 mg, 0.16 mmol), DPP1 (98 mg, 0.16 mmol), CuI (1.2 mg, 0.006 mmol), Pd(PPh ) Cl (2.3 mg,

3

2

0

.003 mmol) and flushed with Ar, THF (5 mL) and iPr NH (5 mL) were then added, the reaction

2

mixture was heated to reflux and stirred overnight. The resulting dark-red solution was concentrated

under reduced pressure, providing a dark-red powder which was purified by Soxhlet extraction with

CH OH, acetone, hexane and finally CHCl . A dark-red powder with M of 3.6 kg/mol and PDI of 1.5

3

n

was obtained when the CHCl fraction was concentrated under reduced pressure. Yield: 46%. δ (300

3

H

MHz; CDCl ) 0.87 (10 H, bs), 1.16-1.24 (8 H, bs), 1.55 (4 H, bs), 3.79 (4 H, bs), 6.68 (2 H, bs), 7.25 (4 H,

3

bs), 7.53-7.84 (20 H, bs).

Scheme 3

Synthesis of DPP 1, a round-bottomed flask was loaded with DPP 2 (1200 mg, 3,9 mmol), K CO (2209

2

3

mg, 16 mmol) and DMF (25 mL). The reaction set-up was flushed with Ar and heated to 140°C, then

hexyl bromide (2.24 mL, 16 mmol) was added in one portion. The reaction was allowed to stir

overnight at 140°C and subsequently poured into ice-water (500 mL). The resulting precipitate was

filtered and washed with CH OH. The compound was obtained in 33% yield as a dark-red solid. Mp:

3

1

76.2- 177.3 °C; δH (300 MHz; CDCl ) 0.83 (6 H, t, J 6.2), 1.20-1.23 (12 H, m), 1.51-1.54 (4 H, m), 3.72

3

(

4 H, t, J 7.5), 7.62-7.68 (8 H, m); δC (75 MHz; CDCl ) 13.9, 22.4, 26.3, 29.4, 31.2, 41.8, 109.9, 125.8,

3

1

26.9, 130.1, 132.2, 147.4, 162.4. MS(EI): m/z 468([M]+,100%).

Scheme 4

Experimental details: Steady-state absorption and corrected photoluminescence (PL) spectra

were recorded with a Lambda 40 (Perkin-Elmer) spectrophotometer and Horiba Jobin Yvon Fluorolog

FL3-22 fluorometer, respectively. The fluorometer allowed us to correct the spectra for the

wavelength dependence of the sensitivity of the detection system and of the intensity of the

excitation beam.

The fluorescence decays were determined by the single photon timing (SPT) method. The excitation

wavelengths of 410 nm and 488 nm were obtained using the second harmonic of a Ti:sapphire laser

(

Tsunami, Spectra Physics). The Ti:sapphire laser was pumped by an intracavity frequency doubled

Millenia laser (Spectra Physics). For the excitation at 550, 635 nm and 644 the output of the an OPO

GWU), pumped by the Ti:sapphire laser, was frequency doubled. The detection system consisted of a

subtractive double monochromator (9030DS, Sciencetech) and a microchannel plate photomultiplier

R3809U, Hamamatsu). Fluorescence decays were recorded under magic angle polarization. A time-

(

to-amplitude converter (TAC), and an analog-to-digital converter (ADC) on a PC board (PICOQUANT)

were used to obtain the fluorescence decay histograms in 4096 channels with time increments of 2-5

ps/channel. The fluorescence decays were globally analyzed as a sum of exponentials with the time-

2

resolved fluorescence analysis (TRFA) software. The quality of the fit was controlled by χ < 1.1 and

visual inspection of the residuals (54).

The relative fluorescence quantum yields (QYs) of the solution samples upon excitation at 410 and

5

15 nm were determined using a solution of perylene in toluene and rodamine 6G in water as a

reference respectively. The molar extinction coefficients (ε ) were determined by averaging the

A

values obtained from the diluted solution prepared at three different concentrations ranging from

-

6

-5

1

×10 to 1×10 mol/L.

The film samples were prepared by spin-coating at a rotational speed of 1000 rpm from a solution in

toluene of the compound (the concentration was 5 mg/mL) on a clean glass substrate. By this

approach films with a thickness of about 100 nm could be obtained.

RESULTS

Triad 1 and polymer 1 in solution

The excited states of triad 1 and polymer 1, in order to create a charge-transfer excited state as an

effective mean to decrease the optical band gap, resulting in a bathochromic shift of the absorbance,

we linked BODIPY 1 to ICZ 1 through the acetylenic π-spacers present on ICZ 1 (See 2.1.1 and scheme

1

), resulting in triad 1. The normalized UV-Vis absorption spectra of a diluted solution in chloroform

of ICZ 1, BODIPY 1 (52) and triad 1 and of ICZ 1, BODIPY 2 (52) and polymer 2 are illustrated in Fig.

1

5

(a) and Fig. 1(b) respectively. The absorption spectrum of triad 1 in chloroform shows a maximum at

29 nm (See Fig. 1(a)), corresponding to a bathochromic shift of 32 nm or 0.15 eV compared to the

parent BODIPY 1 for which the absorption has a maximum at 497 nm. With the goal to further shift

the absorption spectrum of triad 1 towards longer wavelengths we synthesized a ICZ-π-BODIPY

polymer (polymer 1) from BODIPY 2 and ICZ 1 (See 2.1.2 and scheme 2) using a Sonogashira coupling.

Compared to the triad a further broadening and a further red shift of the BODIPY centered absorption

band to 611 nm were observed for a solution of polymer 1 in chloroform (See Fig. 1 (b)). This is red

shift amounts to 75 nm (0.28 eV) compared to the parent BODIPY 2 with an absorption maximum at

5

36 nm. Compared to BODIPY 1 (See Fig. 1(a)) the shift is 114 nm or 0.46 eV. As the spectra of

BODIPY 2 are quite strongly red shifted due to the presence of two iodine substituents it makes more

sense to compare the spectra of polymer 1 to those of BODIPY 1 rather than to those of BODIPY 2.

Furthermore the features of the absorption band of these complexes differ significantly from those of

BODIPY 1 and BODIPY 2. The bands are broadened and the maximums no longer correspond to the 0-

0

vibronic band as in their parent BODIPYs. For the triad 1 the 0-0 vibronic band can be attributed to

the shoulder around 560 nm. However in contrast to triad 1 it is no longer possible to discern the

position of the 0-0 vibronic transition from the features of the absorption band of polymer 1. As the

maximum of the absorption bands of triad 1 and polymer 1 does not coincide with the 0-0 vibronic

transition the red shift of the 0-0 transition is probably even larger than the shift of the maxima. For

both triad 1 and polymer 1 similar features and a similar position of the maximum are observed in

toluene, while in THF a small blue shift of the absorption band was observed compared to chloroform

or toluene. The estimated values of molar extinction coefficient at the absorption maximum of the

4

-1

4

-1

triad 1 and polymer 1 in chloroform amount to 7.5±0.5×10 M cm and 7.0±0.5×10 M cm at 530

and 611 nm respectively.

Excitation of the triad 1 at 515 nm in the BODIPY centered absorption band in toluene yields

fluorescence with a maximum around 650 nm (see Fig. 2 (a)) and a fluorescence quantum yield of 8±1

%

. In contrast to what is normally observed for BODIPYs (40,41,42,43,44,45,46,55) the transition is

characterized by a large width of the absorption and emission bands and a large Stokes shift 0.44 eV.

In chloroform and THF the emission maximum is shifted to 725 and 730 nm, and the fluorescence

quantum yield (Φ ) is decreased to 0.5±0.1 and 0.20±0.05 %, respectively.

f

Upon excitation at 550 nm, i.e. in the BODIPY centered absorption band the fluorescence decay of

triad 1 in toluene at 650 and 720 nm (Figure 3) could be fitted globally to a sum of two exponentials

(

Eq. 1) with amplitudes (A ) and decay times (τ) given in Table 1.

i











t

I

(

t

)

=

A exp −

(1)



i



^τi

Although the bi-exponential fluorescence decay suggests the presence of two different excited

species the amplitudes of both components depend only marginally on the analysis wavelength. This

suggests either that one of the excited species is not emitting or that both species have a similar

emission spectrum. The average fluorescence decay time, <τ>, of the emission at 650 nm calculated

using equation 2 amounts to 1.17 ns.

^Ai^τ

i



τ =

^A

i

(

2)

The fluorescent rate <k > constant calculated from the fluorescence decay time and the fluorescent

f

7

-1

quantum yield using equation 3 amounted to 6.9x10 s which is a three to four times smaller than

the values generally obtained for a BODIPY in apolar solvents (39,40,41,42,43,44,45,46,52).

Φ _f

τ

k_f

=

(

3)

In chloroform and THF the fluorescence decays had to be analyzed as sum of three exponentials.

However, for the component with the shortest decay time a decay time of only 20 ps was recovered

which is beyond the time resolution. As this component was no longer observed when the sample in

chloroform was excited at 410 nm (See SI) it was attributed to scattered light and its presence was

neglected in the further data analysis which yielded the values of amplitudes and decay time shown

in table 1. Upon increasing the solvent polarity from toluene over chloroform to THF both decay

times as well as the average decay time calculated for the emission at 650 nm, which is, in contrast to

what is observed in toluene, very close to τ , decrease. The amplitude of the component with the

1

longest decay time is significantly smaller than in toluene and shows a small decrease upon increasing

the emission wavelength from 650 to 720 nm. Also the value of <k > is decreased upon increasing the

f

solvent polarity (See table 1).

As illustrated in Fig. 4 (a) excitation of the polymer 1 at 600 nm, in the BODIPY centered absorption

band, yields in toluene a fluorescence spectrum with a maximum around 680 nm. The fluorescence

quantum yield obtained upon excitation at 515 nm amounts to 5.5±0.5 % (Fig. 4 (a)). In chloroform

and THF the emission maximum is shifted to 700 nm and 705 nm respectively, while the fluorescence

quantum yield for excitation at 515 nm has decreased to 1.2±0.1 and 0.49±0.05 % respectively.

Upon excitation at 635 nm the fluorescence decay of polymer 1 in toluene at 670 and 720 nm could

be analyzed as sum of two exponentials (Table 2). While the value of short decay time, τ , is close to

1

that observed for triad 1, the long decay time, τ , is significantly smaller. The amplitudes of both

2

decay times are close to those observed for triad 1 and do not depend upon the emission wavelength

while also the value of <k > is close to that observed for triad 1. In chloroform and THF the

f

fluorescence decays have to be analyzed as a sum of three exponentials while the shortest decay

time of respectively 22 and 9 ps is much shorter than the time resolution of the set-up. As in analogy

to triad 1, this component was no longer observed when the sample in chloroform was excited at 410

nm (See SI) it was attributed to scattered light and its presence was neglected in the further data

analysis (See table 2). Upon increasing the solvent polarity for toluene over chloroform to THF the

short decay time, τ , as well as the average decay time calculated for the emission at 670 nm, which

1

is, in contrast to what is observed in toluene, very close to τ , decrease while the long decay time

1

changes much less. In analogy to what was observed for triad 1 the amplitude of the component with

the longest decay time is significantly smaller than in toluene and shows a small decrease upon

increasing the emission wavelength from 670 to 720 nm. Also the value of <k > is decreased upon

f

increasing the solvent polarity (see table 2).

The intramolecular energy transfer in triad 1 and polymer 1, figure 2 (b) shows the fluorescence

spectrum of triad 1 in toluene upon excitation at 410 nm in the S →S absorption band (the S →S

0

2

0

1

transition on the ICZ moiety). This spectrum displays a maximum around 650 nm and a fluorescence

quantum yield of 9±1 %, which are both comparable to the data obtained for direct excitation of the

BODIPY moiety. Furthermore around 430-440 nm a very weak secondary maximum with a

fluorescence quantum yield of 0.08±0.02 % is observed. The latter is attributed to residual emission

of the ICZ moiety and suggests incomplete quenching of the ICZ emission by the BODIPY. As the

fluorescence quantum yield of free ICZ 1 in toluene amounts to 30±3 %, this means according to Eq. 4

that the quenching is at least for 99.7 % efficient.

(

4)

In the equation above Φ_D_Ais the fluorescence quantum yield of the ICZ emission in the triad or

polymer, Φ the fluorescence quantum yield of ICZ 1 and Φ the fluorescence quantum yield of

D

q

quenching. As the fluorescence quantum yield of the long wavelength emission (650 nm) is within

experimental error the same for excitation at 410 and 510 nm, the quenching can be attributed

nearly exclusively to energy transfer. This is confirmed by coincidence of the absorption spectrum of

triad 1 in toluene with the excitation spectrum of the emission at 700 nm in toluene (See SI, Figure S2

(

a)). If a significant fraction of the excited ICZ moieties were quenched by another process the relative

intensity of the band around 400 nm (compared to that around 600 nm) should be smaller in the

excitation than in the absorption spectrum. Also in chloroform and THF the emission spectra

obtained upon excitation at 410 nm consisted of a long wavelength band and a band around 430-440

nm. The long wavelength band was, identical to that obtained upon excitation at 515 nm and was

attributed to a transition mainly located on BODIPY moiety while the short wavelength band was

attributed to a transition localized on the ICZ 1 moiety, considering the similarity of its features and

position with those of the emission spectrum of ICZ derivatives. The short wavelength band is

decreased in intensity in chloroform and THF but to a much smaller extent than the long wavelength

band. In chloroform and THF the quenching of ICZ emission occurs with an efficiency of around 98.7

%

.

Figure 4 (b) shows the emission spectra of polymer 1 obtained in toluene upon excitation at 410 nm,

i.e. in the S →S absorption band of the ICZ moiety. The long wavelength band of this emission

0

1

spectrum has a maximum around 680 nm and a fluorescence quantum yield of 5.2±0.5 %,

comparable to those observed for direct excitation of the BODIPY moiety. Furthermore around 430-

4

40 nm a very weak secondary maximum attributed to residual emission of the ICZ moiety is

observed with a fluorescence quantum yield of 0.4±0.04 % (Fig. 4 (b)). As in triad 1 these data suggest

incomplete quenching of the ICZ emission by the BODIPY moiety The quantum yield of fluorescence

quenching amounts to at least 98.7 %, which is slightly less than found for triad 1. In chloroform and

THF similar results were obtained, corresponding to a fluorescence quenching efficiency of 98.3 and

9

9.1% respectively. When comparing the ratio’s of the amplitude of the 400 nm and 600 nm band in

the absorption spectrum and the excitation spectrum of the fluorescence at 720 nm (See SI, Figure S2

b)), both obtained in toluene, one observes that at 400 nm this ratio is 18±2 % smaller for the

(

excitation spectrum relative to the absorption spectrum. This means that the 18±2 % of the

quenching does occur by another process than energy transfer to the BODIPY centered transition.

In order to obtain more information on the energy transfer the fluorescence decay of the ICZ-

centered emission was determined by single photon timing (SPT) in the triad 1, polymer 1, and the

model compound ICZ 1. The ICZ-centered state was excited at 410 nm and the decay of its emission

was recorded at 460 nm (See Fig. 5). The fluorescence decays were analyzed as a sum of exponentials

(Eq. 1). For ICZ 1 the fluorescence decay could be analyzed as a sum of two exponentials (Table 3) for

both solvents. One should note that the data in Table 3 indicate that the slow decaying component,

corresponding to A τ Aτ + A τ , contributes to 97 and 94 % of the total emission in toluene and

2

1

2 2

chloroform respectively. A triple exponential decay was required to fit the data for triad 1 and

polymer 1 (Table 3). For triad 1 and polymer 1 the recovered values of the shortest decay time, τ₁,

-

2

-2

range from 1.9x10 to 7.4x10 ns, which is very close to or below the time resolution of the set-up

-

2

(

5x10 ns). A tri-exponential decay would mean that there are three distinct excited species.

However neither for triad 1 nor for polymer 1 there is an obvious reason to have three distinct

excited state populations decaying with a different rate. This means that the triple exponential decay

most likely represents a continuous distribution of decay times roughly ranging from τ to τ .

1

3

Although in this approach it is not possible to attribute a physical meaning to the individual decay

times, these parameters however can be used to calculate the average decay time <t> (See Eq. 2) of

the excited state of the ICZ moiety in triad 1 and polymer 1

The range of decay times is situated at smaller values than the average decay time of the model

compound ICZ 1. This observation, which is reflected in values of the average decay times, suggests

that the ICZ fluorescence is quenched by the BODIPY or DPP moieties. As analysis of the stationary

photoluminescence data already suggested that this quenching is nearly exclusively due to energy

transfer from the excited state situated on the ICZ moiety to a state mainly situated on the BODIPY

moiety, the distribution of decay times most probably represents a distribution of rate constants for

energy transfer (vide infra). The recovered values of the amplitudes suggest that the distribution is

biased strongly towards the shorter decay times, i.e. larger rate constants for energy transfer for triad

1

and polymer 1. Similar to Eq. 4, the efficiency of quenching of the ICZ fluorescence by a transition

mainly situated on the BODIPY or DPP moiety can be estimated by Eq. 5.

^τDA

Φ_q_τ= 1−

(5)

τ _D

Φ_q_τis the quenching efficiency. <τ > represents the amplitude-weighted average fluorescence

D

decay time recorded from a solution of ICZ 1, and <τ > is the amplitude-weighted average

DA

fluorescence decay time obtained for the ICZ centered emission of triad 1, or polymer 1. In the same

approach the average rate constant for quenching <k > is given by Eq. 6.

qτ

1

k_q_τ

=

−

(6)

τ _D_A

τ _D

9

-1

The values of <k > obtained in this way range from 1.6x10 (chloroform) to 2.2x10 s (toluene) and

qτ

9

-1

from 1.1x10 (toluene) to 4.8x10 s (chloroform) for triad 1 and polymer 1 respectively. Based on

8

-1

the values of

obtained from a solution of ICZ 1, rate constants (k = 1/<τ >) of 1.6×10 s and

d

D

8

-1

2

.3×10 s can be estimated for the decay of the singlet excited state of ICZ 1 in absence of quenching

in toluene and chloroform respectively. Using the quenching efficiencies obtained by Eq. 4, the

quenching rate constants <k_q_,_s_t_a_t> (Table 3) can be estimated using equation 7.

^Φq

1

k_q_,_s_t_a_t

=

∗

(7)

1

− Φ_qτ _D

10

10 -1

The values of <k_q_,_s_t_a_t> obtained in this way range from 5.3x10 (toluene) to 1.7x10 s (chloroform)

10

10 -1

and from 1.2x10 (toluene) to 1.3x10 s (chloroform) for triad 1 and polymer 1 respectively. The

quenching efficiencies for triad 1 and polymer 1, estimated on the basis of the time-resolved

fluorescence experiments (Φ ), are smaller (5 to 10 %) than those estimated from the fluorescence

qτ

quantum yields (Eq. 4) in both toluene and chloroform while on the other hand the values of <k_q_,_s_t_a_t

>

determined form Eq. 7 are about one order of magnitude larger than those of <k > determined form

qτ

Eq. 6. A reason for this discrepancy is that the way the average decay time is calculated emphasizes

long decay times and hence the tail of the distribution of the quenching rate constants. This means

that such procedure to estimate an average decay time can only be used if the different decay times

do, contrary to the present case, differ less than one order of magnitude. Furthermore the values

found for the shortest decay time, τ , are subject to a large relative error as they are close to or

1

beyond the time resolution of the set-up. One cannot exclude that at short times the decay is faster

-

2

than the time-resolution of our set-up (5.0x10 ns) and that static quenching occurs (56,57). This

could also explain the large discrepancy between the values of <k > found for triad 1 and polymer 1

qτ

in the two solvents.

Polymer 2 in solution

The excited states of polymer 2, in the previous sections the use of a BODIPY as an electron

accepting building block in the synthesis of the D-A conjugated systems has been investigated. It has

been demonstrated that diketopyrrolopyrrole (DPP) is also a good co-monomer for thiophene based

D-A polymers (47,50,58) and therefore we synthesized an alternating copolymer (polymer 2) of DPP

and ICZ 1 through the use of the acetylenic π-spacers.

Figure 6 shows the normalized absorption spectra of a dilute solution in chloroform of ICZ 1, DPP 1

and polymer 2. Polymer 2 shows an absorption maximum at 505 nm with a molar extinction

4

-1

coefficient (for one ICZ-DPP dyad) of ~ 4.5×10 M cm , corresponding to a bathochromic shift of 25

nm (0.13 eV) relative to the parent DPP 1. As previously described, the absorption band at 426 nm is

attributed to the ICZ moiety. In analogy to the observations made for the triad and polymer 1 a blue

shift by ~ 7 nm of the absorption band is observed in THF.

The normalized photoluminescence (PL) spectrum of polymer 2 in toluene upon excitation at 515 nm,

i.e. in the DPP centered band, shows fluorescence with a maximum around 577 nm, corresponding to

a Stokes shift of 72 nm (0.31 eV), and a shoulder around 625 nm. The distance between both bands is

0

.16 eV which can correspond with the frequency of a carbon - carbon vibration of the ring. This

would indicate that the bands at 577 and 625 nm correspond to the 0-0 and 0-1 vibronic transitions.

The fluorescence quantum yield (Ф ) amounts to 65.2 % in toluene. In chloroform and THF the

f

maximum of the fluorescence spectrum obtained upon excitation at 515 nm shifts to 700 and 675 nm

while the fluorescence quantum yield is decreased to 3.9±0.4 and 1.9±0.2 % respectively and the

vibrational fine structure is lost.

Upon excitation at 488 nm, i.e. in the DPP centered absorption band the fluorescence decay of

polymer in toluene at 600 (Fig. 8), 650 and 700 nm could be fitted globally to a sum of two

exponentials (Eq. 1) with amplitudes (A ) and decay times (τ) given in Table 4.

i

Although the bi-exponential fluorescence decay suggests the presence of two different excited

species the amplitudes of both components depend only marginally on the analysis wavelength. This

suggests either that one of the excited species is not emitting or that both species have a similar

emission spectrum. The average fluorescence decay time, <τ>, of the emission at 650 nm calculated

using equation 2 amounts to 2.42 ns. The fluorescent rate constant <k > calculated from the

f

8

-1

fluorescence decay time and the fluorescent quantum yield using equation 3 amounted to 2.7x10 s ,

which is of the order of magnitude of the values observed for DPP (59).

In chloroform and THF the fluorescence decays have to be analyzed as a sum of four exponentials

while the shortest decay time of respectively 13 and 12 ps is much shorter than the time resolution of

the set-up. In analogy to triad 1 and polymer 1, this component, which was not observed upon

excitation at 410 nm (See SI), was attributed to scattered light. Hence its presence was neglected in

the further data analysis (Table 4). Upon increasing the solvent polarity from toluene over chloroform

to THF the decay times τ and τ decrease, in analogy to the results obtained for triad 1 and polymer

1

2

1

. However in contrast to what was observed for triad 1 and polymer 1 a third component with a

shorter decay time (τ ) is necessary to obtain a good fit. In analogy to what was observed for triad 1

3

and polymer 1 the amplitude of the component with the longest decay time (τ ) is significantly

2

smaller than in toluene and shows a small decrease upon increasing the emission wavelength from

6

00 to 700 nm. Also the value of the average decay time <τ>, calculated for the emission at 600 nm

and <k > are decreased upon increasing the solvent polarity (Table 4).

f

The intramolecular energy transfer in polymer 2, upon excitation at 410 nm (See Fig. 7 (b)) the

fluorescence quantum yield for emission from the DPP moiety amounts to 47±5 % in toluene which is

smaller than the value (65.2±0.065) found upon excitation at 515 nm, while that of the ICZ emission

amounts to 3.6±0.4 %. According to equation 4 this means that the ICZ emission is quenched for at

least 88.2 %, which is about 10 % smaller than the values found for triad 1 and polymer 1, probably

by energy transfer to a DPP centered excited state. When comparing the ratio’s of the amplitude of

the 400 nm and 600 nm band in the absorption spectrum and the excitation spectrum of the

fluorescence at 650 nm (See SI, Figure S2 (c)), both obtained in toluene, one observes that at 400 nm

this ratio is 37±4 % smaller for the excitation spectrum relative to the absorption spectrum. If all

quenching would occur by energy transfer this ratio should only by 12±4 % smaller for the excitation

than for the absorption spectrum (as the quenching efficiency is 88 %). This means that the 26±3 % of

the ICZ moieties are quenched by another process than energy transfer to the DPP centered excited

state or that 26/0.88 = 30±3 % of the quenching does occur by another process than energy transfer

to the DPP centered transition. This agrees with the observation that the fluorescence quantum yield

of the DPP centered emission is 18±2 % smaller upon excitation at 410 nm compared to excitation at

5

15 nm. In order to obtain more information on the energy transfer the fluorescence decay of the

ICZ-centered emission of polymer 2 was determined by single photon timing (SPT). The ICZ-centered

state was excited at 410 nm and the decay of its emission was recorded at 460 nm (See Fig. 5). The

fluorescence decays were analyzed as a sum of three exponentials (Eq. 1) yielding the results shown

in Table 3. While for the triad 1 and polymer 1 the component with the shortest decay time (τ ) had a

1

decay time very close to or below the time resolution of the set-up the τ was three to ten times

1

larger for polymer 2. Furthermore this component had a considerably smaller amplitude (A ). As in

1

analogy to triad 1 and polymer 1 there is no obvious reason to have three distinct excited state

populations decaying with a different rate the triple exponential decay most likely also represents a

continuous distribution of decay times roughly ranging from τ to τ . As for triad 1 and polymer 1 the

1

3

fit parameters can however be used to calculate the average decay time <τ>, the efficiency of the

quenching (Φ ) and the average quenching rate constant (<k >) using Eq. 2, 5 and 6 respectively (cfr.

qτ

8

-1

8

-1

supra). In this way values of Φ and <k > of 58 % and 59 %, and 2.2x10 s and 3.3x10 s were

qτ

obtained in toluene and chloroform respectively (Table 3). Considering the stationary fluorescence

data this quenching is to a large extent due to energy transfer from the excited state situated on the

ICZ moiety to a state mainly situated on the BODIPY moiety. In contrast to what was observed for

triad 1 and polymer 1 where the recovered values of the amplitudes suggest that the distribution is

biased strongly towards the shorter decay times, i.e. larger rate constants for energy transfer, the

distribution is more symmetric and uniform for polymer 2. Furthermore the recovered values of the

decay times recovered for polymer 2 are significantly longer than those recovered for triad 1 or

polymer 1. This is reflected in the value of Φ which are 40 % smaller than those found for triad 1

qτ

and polymer 1 while the values of <k > are five to fifteen times smaller than those found for triad 1

qτ

9

-1

and polymer 1. The quenching rate constant <k_q_,_s_t_a_t>, calculated using Eq. 7 amounts to 1.2x10 s in

toluene which is an order of magnitude smaller than the values of <k_q_,_s_t_a_t> found for triad 1 and

polymer 1 in toluene. In analogy to what was observed for triad 1 and polymer 1 the value estimated

for <k > for polymer 2 in toluene is five times smaller than that estimated for <k_q_,_s_t_a_t>, hence also the

qτ

value of Φ in toluene is 30 % smaller than that of Φ .

qτ

q

Films of triad 1 and Polymer 1

As the spectra of conjugated polymers often differ significantly in neat films (used in OLEDS or OPVs)

when compared to a dilute solution, we also investigated the spectroscopic properties of triad 1 and

polymer 1 in a thin spin-coated film. Figure 9 illustrates normalized absorption spectra obtained from

a solution and a spin-coated film of the triad 1 (a) and those of the polymer 1 (b). In contrast to the

solution samples, a broader absorption spectrum with the maximum shifted to 621 nm, which

corresponds to a shift of 0.35 eV compared to a solution in toluene, observed for the film of triad 1.

Furthermore this band has a tail extending to 1000 nm. In addition a shoulder between 550 and 570

nm is observed. The film of polymer 1 has no shift of the absorption peak. The spectrum is only

slightly broadened compared to the solution which could be attributed to an increased degree of

disorder (60).

The fluorescence spectrum of the films of triad 1 and polymer 1 consists of a broad band with

maximum at respectively 708 and 740 nm (See Fig. 10). Compare to a solution in toluene the

emission is shifted over 58 nm (0.16 eV) and 60 nm (0.15 eV) for triad 1 and polymer 1 respectively.

In the film the Stokes shift amounts to 87 nm (0.25 eV) and 129 nm (0.35 eV) for triad 1 and polymer

1

respectively.

DISCUSSION

The properties of the lowest excited state For triad 1 and polymer 1 both the shift of the 0-0

transition of the long wavelength absorption band compared to respectively BODIPY 1 and 2 as well as

the significant increase of the width of this band suggest a partial delocalization of the lowest excited

singlet state over the ICZ moiety. The small blue shift of the absorption band in THF can be attributed to

decreased polarizability of the solvent (55). The absence of a red shift of the absorption maximum

upon increasing the solvent polarity suggests that the ground state has little to no permanent dipole

moment. The less intense absorption signal at 430 nm can be attributed to a transition residing mainly on

the ICZ chromophore. The much smaller red shift of this transition compared to the parent ICZ is probably

due to the fact that it is oriented along the short axis (6-12 direction) of the ICZ and hence less influenced

by the conjugation.

The increase of the Stokes shift as well as the major red shift of the emission of triad 1 and polymer 1

upon increasing the solvent polarity suggest a significant dipole moment and hence charge transfer

δ+

δ-

character of the excited state, ICZ -BODIPY (Scheme 5) (55). A similar behavior could be observed

when BODIPY dyes are conjugated to an electron donor such as in 2-anilinosubstituted BODIPY (39).

The Stokes shift observed for polymer 1 in toluene (0.21 eV) is much smaller than the one observed

for the triad (0.44 eV) which suggests a less extensive change in geometry upon excitation. This

correlates with more extensive delocalization leading to a reduced exciton phonon coupling. One

should note that for polymer 1 the solvent dependence of the Stokes shift and the emission

maximum are smaller than observed for triad 1 suggesting a smaller excited state dipole moment. As

each BODIPY is now conjugated on both sides to an ICZ moiety, charge transfer from both ICZ

moieties to the BODIPY could lead to a quadrupolar rather than a dipolar structure.

The analysis of the fluorescence decays of the long wavelength emission shows the presence of two

excited species. Either one of those species is non emitting or both species have a nearly identical

emission spectrum in toluene as the values of amplitudes of both components of the decay are

within experimental error identical at 650 (for triad 1) or 670 nm (for polymer 1) and 720 nm. If one

assumes a single emitting species the interconversion between both species is reversible leading to a

bi-exponential decay (61). As photo-excitation (or energy transfer from the ICZ centered excited state

(

cfr. infra) will form the fluorescent species the rate for interconversion can be approximated by the

−

1

−1

inverse of the short decay time, τ₁, while the inverse of the long decay time, τ ₂, is a weighted

average of the inverse decay times of both species (61,62,63). In this framework, the ratio of the pre-

exponential factors then reflects to some extent the position of the equilibrium between both states.

A likely candidate for the nonfluorescent species is a completely charge separated state (electron

+

-

transfer state) as ICZ - BODIPY (Scheme 5) with a radical cation on the ICZ moiety and a radical anion

on the BODIPY moiety. Such process was also observed for a bisthiophene substituted peryleneimide

(64) and rigid perylene end-capped pentaphenylenes (65), although due to a stronger interaction

between the electron donor and acceptor this CT-state was still emitting in these systems. Due to the

spatial separation and limited overlap of the wavefunctions of the radical cation and the radical anion

such state will have a very small transition dipole towards the ground state (66,67). It will

furthermore be more stabilized than the partially charge separated state upon increasing the solvent

polarity which could explain why the amplitude of the slow decaying component decreases strongly

upon increasing the solvent polarity. The decrease of the long decay time, τ , upon increasing the

2

solvent polarity is hence due to the stabilization of the charge transfer state, decreasing the energy

difference with the ground state and the increase of the reorganization energy (67,68,69,70). While

for triad 1 the quenching of the ICZ-emission leads nearly exclusively to the formation of the emitting

δ+

δ-

partial charge separated state, ICZ - BODIPY , the excitation spectra suggest that in polymer 1 the

quenching of the ICZ occurs partially by another mechanism, which could be the direct formation of

+

-

the fully charge separated state, ICZ - BODIPY .

The red shift of the long wavelength absorption band and the fluorescence spectrum in toluene of

polymer 2 suggests that, in analogy to triad 1 and polymer 1, the orbitals involved in this transition

are to some extent delocalized between the ICZ and the DPP moieties, although to a smaller extent

than in polymer 1. The less extensive delocalization, suggested by the smaller red shift of the

absorption band versus that of DPP 1, is confirmed by the observation that while the average

fluorescent rate constant of triad 1 and polymer 1 is much smaller than that of the corresponding

BODIPY model compounds that of polymer 2 is close to the values that can be calculated for a

diketopyrrolo-pyrrole (59). The broadening of the DPP centered absorption band and the better

vibrational fine structure of the fluorescence spectrum (compared to the absorption spectrum)

suggest a change in molecular geometry, e.g. planarization, upon excitation. The blue shift of the long

wavelength absorption band in THF can, in analogy to triad 1 and polymer 1, be attributed to

decreased polarizability of THF compared to chloroform or toluene (55). The increase of the Stokes

shift as well as the major red shift and loss of vibrational fine structure of the emission of polymer 2

upon increasing the solvent polarity suggest that the formation of an excited state with a significant

dipole moment and charge transfer character (55). In analogy to the results obtained for triad 1 and

polymer 1 the analysis of the time-resolved fluorescence experiments suggests that this state with a

+

-

partial charge transfer is in equilibrium with a non-fluorescent electron transfer state, ICZ - DDP

Scheme 5). In analogy to what is observed for polymer 1, the fluorescence excitation spectra show

that for polymer 2 the quenching of the ICZ emission leads only partially to the population of the

(

δ+

δ-

fluorescent partial charge separated state, ICZ -BODIPY . This could suggest that for polymer 2 this

+

-

fully charge separated state, ICZ - DDP , can also be populated directly by the quenching of the ICZ

centered excited state. The larger amplitude of the component with decay time τ indicates that this

2

electron transfer state is less stabilized versus the fluorescent partial charge transfer state than in

triad 1 or polymer 1 (61,62,63). The more complex decays observed in chloroform or THF compared

to triad 1 or polymer 1, are probably due to the presence of oligomers with different chain lengths.

One should note that also a bisthiophene substituted perylenimide yielded two different electron

transfer states in polar solvents (64).

The intramolecular energy transfer For triad 1, polymer 1, and polymer 2 we observed that the

rate constant for energy transfer from the ICZ centered excited sate to the excited state centered on

the BODIPY or DPP moiety determined for stationary fluorescence experiments, <k_q_,_s_t_a_t>, is always

significantly larger than that obtained from time-resolved experiments, <k >. As the way the average

qτ

fluorescence decay time <τ> is calculated emphasizes long decay times and hence the tail of the

distribution of the quenching rate constants, the values of <k > are probably underestimated. For

qτ

triad 1 and polymer 1 the inability to measure accurately components of the decays with a decay

time less than the time resolution of our set-up for SPT (static quenching) (56,57) can lead to a

further underestimation of <k >. This argument does however not hold for polymer 2 where the

qτ

shortest decay times of 270 and 210 ps in toluene and chloroform respectively are clearly much

larger than the time resolution of the SPT set-up. In spite of this discrepancy it is remarkable that

both approaches yield for polymer 2 values of the rate constant for quenching of the ICZ emission

that is an order of magnitude smaller found for triad 1 or polymer 1. The much smaller values of <k_q>

and <k_q_,_s_t_a_t> observed for polymer 2 were initially unexpected as the overlap between the long

wavelength absorption and the fluorescence of the ICZ chromophore is similar for the three

compounds.

When an excited state (in this case the BOPDPY centered partial charge transfer state) is formed

starting from another excited state (in this case the ICZ-centered excited state) by energy transfer or

complex formation (61,62,63) one expects that the fluorescence decay of the former one can be

described by a difference of two exponentials with equal amplitude, i.e. the fluorescence decay

should be characterized by a growing-in. However upon excitation at 410 nm the fluorescence decay

δ+

δ-

δ+

δ-

of the emission of neither ICZ -BODIPY (triad 1 or polymer 1) nor ICZ -DDP (polymer 2)

contained components with a negative amplitude. While for triad 1 and polymer 1 the failure to

observe such component could be attributed to the fact that its decay time (i.e. the growing in time

constant) is beyond the time resolution of the set-up, this argument does not hold for polymer 2. At

least for the latter compound the failure to observe a growing in of the fluorescence is possibly due

to the overlap of the absorption spectra of the ICZ-centered excited and the DPP centered partial

charge transfer excited state or the formation of the partial charge transfer state (corresponding to

the DPP centered emission) from the fully charge separated state (dashed lines in scheme 5). One

should note that also in other systems characterized by intramolecular excitation transfer (71,72) no

component with negative amplitude could be observed for the acceptor emission.

In order to get further insight in the values of <k > and <k_q_,_s_t_a_t> (which we attributed to energy

qτ

transfer), we estimated the rate constants for Förster type energy transfer (73,74,75) between the

transition centered on the ICZ chromophore and that centered on the BODIPY or DPP chromophore.

The characteristic distance for energy transfer, R , is given by Eq. 8,

0

1

2

−4

0

f

6

R = 0.021085

(

Jκ n Φ

)

(8)

0

where J is the spectral overlap integral, κ the orientation factor, n the refractive index of the medium,

0

(

which is 1.496 for toluene and 1.446 for chloroform), and Φ _f, is the fluorescence quantum yield of

the ICZ chromophore (0.30±0.03 and 0.16±0.02 in toluene and chloroform respectively). The spectral

-

1

-1

4

overlap integral J in M cm nm is given by Eq. 9,

4

J = ε _A

(λ

)F _D

(λ

)

λ dλ

(9)



-

1

-1

where ε (λ) is the molar extinction coefficient of the acceptor (M cm ), F (λ) is the area normalized

A

D

-

1

fluorescence spectrum of the donor (nm ). The orientation factor κ is given by Eq. 10,

κ

=

cosα

−

3cosΘ_DcosΘ_A

(10)

where α is the angle between the transition dipoles of donor and acceptor, and Θ and Θ are the

A

D

angles between the transition dipoles of each the donor and acceptor and the vector linking their

center. For the triad 1 and polymer 1, α, Θ and Θ amount to 20°, 22° and 42° respectively yielding -

A

D

2

1

.13 and 1.27 for κ and κ (Fig. 11).

For polymer 2 α, Θ and Θ amount to 43°, 83° and 40° based on the orientation of the transition

A

D

dipole in the DPP chromophore as established by Mizuguchi et al. (see SI, Fig. S3) (58). This approach

2

yielded 0.45 and 0.20 for κ and κ for polymer 2 (Fig. 11). In these estimations a planar structure was

assumed for the molecules. Deviations from this planar structure will lead to a decrease of κ² as

discussed below. Using equation 8 to 10 the values of R given in Table 5 are obtained.

0

We can also write the equation for k_F_ö_r_s_t_e_r, the rate constant for energy transfer by the Förster

mechanism as

6

R₀

k_F_ö_r_s_t_e_r= k_d

(11)

6

R

-

1

Where k = 1/<τ > (s ) is the rate constant for the decay of the fluorescence of ICZ 1, and R the

d

D

8

-1

distance between the centers of the donor and the acceptor. Using values of 1.60x10 s and

8

-1

2

.28x10 s for k in respectively toluene and chloroform and values of 1.26 and 1.46 nm for R (for

d

respectively triad 1 or polymer 1 and polymer 2) we obtain the values of R and k

given in Table

0

Förster

5

. The values of R were determined from the structures in scheme 1, 2 and 3.

First of all, we note that the calculated values of k_F_ö_r_s_t_e_rare systematically about a factor 10 to 40

larger than the fastest rate constants for quenching of the ICZ emission observed experimentally

-

1

(

which we put equal to τ ) in the time resolved experiments. A similar discrepancy (except for triad 1

in CHCl ) is observed when we consider the values <k_q_,_s_t_a_t> inferred from the stationary fluorescence

3

experiments using Eq. 4 and 7. The analysis of the nature of the fluorescence decays indicates that

besides a large fraction of molecules with a fast decay time of the ICZ emission there is always a small

fraction present with a decay time close to that of unquenched ICZ. While for triad 1 and polymer 1

this fraction is no more than a few percent, and could perhaps be attributed to an impurity not

containing a BODIPY chromophore, this fraction amounts to 35 to 40 % for polymer 2. Considering

the fact that the NMR spectra do not suggest the presence of such an impurity for any of the three

compounds, it is highly unlikely that the fraction of molecules of polymer 2 with a long fluorescence

decay time correspond to an impurity. The latter argument as well as the systematic difference

between the values of the calculated rate constants, k_F_ö_r_s_t_e_rand those of the experimentally estimated

-

1

rate constants, τ₁or <k_q_,_s_t_a_t>, which mutually agree, indicates that the discrepancy is not due to the

experimental uncertainty or artifact and must have a more fundamental reason.

Equations 7 to 10 are based on a point dipole approximation for the transition dipoles of the donor

and acceptor. It has been shown that when the size of the chromophores is no longer negligible

compared to the distance between the centers of donor and acceptor, the Coulomb matrix element,

which is the perturbation responsible for Förster type energy transfer, is overestimated

(76,77,78,79,80) by this point dipole approximation. If, for example a center to center distance of 1

nm is used, the matrix element is overestimated by about a factor three which means that the rate

constant, which is proportional to the square of this matrix element, is overestimated by a factor of

ten. Taking this overestimation into account, the values of k_q_,_F_ö_r_s_t_e_restimated from the absorption

spectra of the transitions centered at BODIPY or DPP and the emission spectra of the ICZ centered

-

1

transitions, become similar to the experimental values of τ₁or <k_q_,_s_t_a_t>. Furthermore one should

note that the ratio of the values of k_q_,_F_ö_r_s_t_e_rfor triad 1 or polymer 1 to those of polymer 2 resembles

-

1

the ratio of the values of τ₁or <k_q_,_s_t_a_t> of triad 1 or polymer 1 to those of polymer 2. When

considering the calculations for k_q_,_F_ö_r_s_t_e_rit is clear that this difference is mainly due to ratio of the

2

value of <κ > for triad 1 or polymer 1 to that of polymer 2.

One has also to take into account that the energy transfer is not purely driven by Coulomb interaction

(

73,74,75), but that configuration interaction with charge transfer configurations (81,82,83) as well as

exchange interactions (84) also provide a contribution to the rate constant for energy transfer. The

latter contributions require orbital overlap between both chromophores which will occur mainly

through the acetylenic linker (through-bond) (82,85). Although the spectral shift of the BODIPY

centered band in triad 1 or polymer 1 and the DPP centered band in polymer 2 versus respectively

bands of the parent BODIPY or DPP molecules, suggests orbital interaction between both

chromophores, the energy transfer rates and their difference between triad 1 or polymer 1 and

polymer 2 are in good agreement with those predicted for Förster transfer when the point dipole

approximation is replaced by a more realistic extended dipole approximation (81,82,83). This agrees

with earlier observation for intramolecular energy transfer between porphyrines connected by a

phenanthroline (71). If other transfer mechanisms would have a major contribution this would only

lead to a larger discrepancy between the observed and calculated rates for energy transfer. Also the

delocalization of the wavefunctions of the ICZ and BODIPY or DPP chromophores over the acetylenic

linkers or beyond would lead to a limited enhancement of the coupling and hence to an increase of

k_F_ö_r_s_t_e_rwhich would increase the discrepancy with the experimental results (86,87).

For triad 1, polymer 1 and polymer 2 the analysis of the fluorescence decays showed a distribution of

decay times of the ICZ emission reflects a distribution of rate constants for energy transfer. For

energy donor and acceptor linked by a rigid acetylenic spacer, rotation of the donor and acceptor

around this linker is the only structural parameter for which a relatively broad distribution can be

expected. In triad 1 and polymer 1 a rotation of the BODIPY moiety around the acetylenic spacer will

2

only lead to modest changes in α and Θ and hence in κ (see Fig. 11 (a)). However, in polymer 2 a

A

rotation of the DPP moiety around the acetylenic spacer (see Fig. 9 (b)) will lead to major changes in

2

α and hence in κ . Equation 10 predicts (keeping Θ at 83 ° and Θ at 40° respectively) that κ and

D

A

hence k_q_,_F_ö_r_s_t_e_rwould equal zero for polymer 2 when α would become equal to 73°. Furthermore, one

should take into account that for all compounds leaving the coplanar configuration would also reduce

the orbital overlap between the ICZ and BODIPY (or DPP) moieties and hence the contribution of

exchange or charge transfer interactions to the rate constants for energy transfer.

Films of triad 1 and Polymer 1 The red shift of the maximum of the absorption band of triad 1 in

the film compared to a solution in toluene suggests exciton interaction in a J-type packing or a highly

Article Doi

Improved spectral coverage and fluorescence quenching in donor-acceptor systems involving indolo[3-2-b]carbazole and boron-dipyrromethene or diketopyrrolopyrrole

DOI: 10.1111/php.12437

Source and publish data:

Authors:

Article abstract of DOI:10.1111/php.12437

Full text of DOI:10.1111/php.12437

Products guided by the article

R&D Labs maybe for 266357-54-4

Relevant to this article

Hot Product