Photophysical properties of borondipyrromethene analogues in solution

Article abstract of DOI:10.1021/jp052626n

The photophysical properties of seven new 8-(p-substituted)phenyl analogues of 4,4-difluoro-3,5-dimethyl-8-(aryl)-4-bora-3a,4a-diaza-s-indacene (derivatives of the well-known fluorophore BODIPY) in several solvents have been studied by means of absorption and steady-state and time-resolved fluorimetry. For each compound, the fluorescence quantum yield and lifetime are lower in solvents with higher polarity owing to an increase in the rate of nonradiative deactivation. Increasing the electron withdrawing strength of the p-substituent on the phenyl group in position 8 also leads to lower fluorescence quantum yields and lifetimes. When the p-substituent on the phenyl group in position 8 is a tertiary amine [8-(4-piperidinophenyl), 8-(4-N,N-dimethylaminophenyl), and 8-(4-morpholinophenyl)], the low quantum yields of these compounds in more polar solvents can be rationalized by the inversion of the energy levels of an apolar, highly fluorescent and a polar, nonfluorescent excited state, where charge transfer from the tertiary amine to the BODIPY unit occurs. These amine analogues can be protonated at low pH in aqueous solution. Fluorescence titrations yielded pK_a values of their conjugate ammonium salts which are in agreement with the electron donating tendency of the amine group: piperidino (4.15) > dimethylamino (2.37) > morpholino (1.47), with the pK_a values in parentheses. The rate constant of radiative deactivation (k_f) is the same for all compounds in all solvents studied (k_f -1.4 × 108 s^-1).

Full text of DOI:10.1021/jp052626n

J. Phys. Chem. A 2005, 109, 7371-7384
7371
Photophysical Properties of Borondipyrromethene Analogues in Solution
Wenwu Qin, Mukulesh Baruah, Mark Van der Auweraer, Frans C. De Schryver, and
Noe1l Boens*
Department of Chemistry, Katholieke UniVersiteit LeuVen, Celestijnenlaan 200F,
3001 HeVerlee (LeuVen), Belgium
ReceiVed: May 19, 2005; In Final Form: June 28, 2005
The photophysical properties of seven new 8-(p-substituted)phenyl analogues of 4,4-difluoro-3,5-dimethyl-
8-(aryl)-4-bora-3a,4a-diaza-s-indacene (derivatives of the well-known fluorophore BODIPY) in several solvents
have been studied by means of absorption and steady-state and time-resolved fluorimetry. For each compound,
the fluorescence quantum yield and lifetime are lower in solvents with higher polarity owing to an increase
in the rate of nonradiative deactivation. Increasing the electron withdrawing strength of the p-substituent on
the phenyl group in position 8 also leads to lower fluorescence quantum yields and lifetimes. When the
p-substituent on the phenyl group in position 8 is a tertiary amine [8-(4-piperidinophenyl), 8-(4-N,N-
dimethylaminophenyl), and 8-(4-morpholinophenyl)], the low quantum yields of these compounds in more
polar solvents can be rationalized by the inversion of the energy levels of an apolar, highly fluorescent and
a polar, nonfluorescent excited state, where charge transfer from the tertiary amine to the BODIPY unit
occurs. These amine analogues can be protonated at low pH in aqueous solution. Fluorescence titrations
yielded pKa values of their conjugate ammonium salts which are in agreement with the electron donating
tendency of the amine group: piperidino (4.15) > dimethylamino (2.37) > morpholino (1.47), with the pKa
values in parentheses. The rate constant of radiative deactivation (kf) is the same for all compounds in all
solvents studied (kf ) 1.4 × 108 s-1).
Introduction
the photophysical properties of the new BODIPY compounds,
we recorded their steady-state absorption and fluorescence
The design and development of fluorescent chemosensors for
the detection of analytically important species is still an
expanding area of research.1 Due to their favorable photophysi-
cal and optoelectronic properties, 4,4-difluoro-4-bora-3a,4a-
diaza-s-indacene2 (difluoroborondipyrromethene, BDP, or BO-
DIPY3) derivatives have become favorite fluorophores in new
fluorescent probes that have applications in many different
areas.4 For example, BODIPY dyes can offer improved spectral
characteristics compared to other conventional dyes when used
in automated DNA sequencing.5 The excellent qualities of
BODIPY comprise high fluorescence quantum yields and
narrow emission bandwidths with high peak intensities, elevated
photostability, relatively high absorption coefficients, and the
extra feature of excitation/emission wavelengths in the visible
region. Moreover, their building block synthesis allows one to
develop many different BODIPY analogues with different
receptor groups and with emission ranges from 500 to over 700
nm. The photophysical properties of several BODIPY deriva-
tives in liquid solution have been studied by means of absorption
and fluorescence techniques.6-8
excitation and emission spectra and collected their time-resolved
fluorescence profiles in several solvents. From these experi-
ments, we could determine the position of the spectra (λabs, λex,
and λem), the full width at half-maximum of the absorption band
(fwhmabs), the fluorescence quantum yields (φf), the rate
constants of radiative (kf) and nonradiative (knr) deactivation,
Stokes shifts (∆νj), fluorescence lifetimes (τ), and molar
absorption coefficients (ꢀ) of the compounds. Furthermore, the
reversible protonation/deprotonation of 5, 6, and 7 in aqueous
solution was also studied.
Experimental Section
Syntheses. General Methods. 1H and 13C NMR spectra were
recorded at room temperature on a Bruker Avance 300 instru-
ment operating at a frequency of 300 MHz for 1H and 75 MHz
1
for 13C. H NMR spectra were referenced to tetramethylsilane
(0.00 ppm) as an internal standard. Chemical shift multiplicities
are reported as s ) singlet, d ) doublet, t ) triplet, q ) quartet,
and m ) multiplet. 13C spectra were referenced to the CDCl3
(77.67 ppm) signal. Mass spectra were recorded on a Hewlett-
Packard 5989A mass spectrometer (E.I. mode and C.I. mode).
High resolution mass data were obtained with a KRATOS
MS50TC instrument. Melting points were taken on a Reichert
Thermovar and are uncorrected.
5-Phenyl-1,9-dimethyldipyrromethene (9b). This condensation
was carried out according to a published procedure9 (Scheme
1). In a 250 mL round-bottomed flask, benzaldehyde (8b) (159
mg, 1.5 mmol) and 2-methylpyrrole (260 mg, 3.21 mmol) were
dissolved in 100 mL of dry dichloromethane under an argon
atmosphere. One drop of trifluoroacetic acid (TFA) was added
In the present work, we synthesized novel fluorescent
compounds with BODIPY as the common fluorophore but with
different para-substituents (R) on the benzene ring at position
8 (Scheme 1). The first BODIPY series includes compounds 1,
2, 3, and 4 (Scheme 1), with the p-substituent (R) ranging from
electron donating (methoxy) to electron withdrawing (cyano).
The second series covers BODIPY analogues 5, 6, and 7 in
which the p-substituent (R) is a tertiary amine. To investigate
* To whom correspondence should be addressed. E-mail: Noel.Boens@
chem.kuleuven.be. Fax: +32-16-327990.
10.1021/jp052626n CCC: $30.25 © 2005 American Chemical Society
Published on Web 08/04/2005

7372 J. Phys. Chem. A, Vol. 109, No. 33, 2005
Qin et al.
SCHEME 1: Synthesis of the BODIPY Derivatives
to the stirred mixture. Stirring was continued overnight at room
temperature. After the disappearance of benzaldehyde [moni-
tored by thin-layer chromatography (TLC)], a solution of
p-chloranil (217.5 mg, 1.5 mmol) in dichloromethane (25 mL)
was added. The mixture was further stirred for 0.5 h and then
quenched with water. The organic layer was separated, and the
water layer was further extracted with dichloromethane. The
combined organic layers were dried over MgSO4 and concen-
trated and purified through column chromatography over neutral
alumina (hexane/ethyl acetate, 7/3, v/v) to obtain a reddish
2H, J ) 8.0 Hz); 6.54 (d, 2H, J ) 3.6 Hz); 6.16 (d, 2H, J )
4.4 Hz); 3.89 (t, 4H, J ) 4.7 Hz); 3.25 (t, 4H, J ) 4.7 Hz);
2.45 (s, 6H). MS (CI) m/z (%): 334 (MH+, 100).
4,4-Difluoro-3,5-dimethyl-8-phenyl-4-bora-3a,4a-diaza-s-in-
dacene (2). A 0.5 mmol (124 mg) portion of dipyrromethene
(9b) was dissolved in 70 mL of dichloromethane. N,N-
Diisopropylethylamine (DIEA, 2.5 mL, 16 mmol) was added,
and the mixture was stirred at room temperature for 20 min
under argon. BF3‚OEt2 (2.5 mL, 20 mmol) was added dropwise,
and stirring was continued for 1 h. The reaction mixture was
washed with water and 2 N NaOH. The aqueous solution was
extracted with dichloromethane, dried over MgSO4, filtered, and
evaporated. The crude material was purified via silica gel
chromatography (ethyl acetate/hexane, 3/7, v/v). Recrystalliza-
tion from hexane gives green crystals in 62% yield (92 mg).
1
brown powder in 39% yield (145 mg). H NMR (CDCl3): δ
7.44 (m, 5H); 6.55 (d, 2H, J ) 4.0 Hz); 6.24 (d, 2H, J ) 4.0
Hz); 2.53 (s, 6H). 13C NMR (CDCl3): δ 158.0, 139.0, 130.8,
130.5, 130.3, 128.6, 119.8, 30.1, 15.3. MS (CI) m/z (%): 249
(MH+, 100).
1
Dipyrromethene analogues 9a, 9c, 9d, 9e, 9f, and 9g were
similarly prepared starting from 4-methoxybenzaldehyde (8a),
methyl 4-formylbenzoate (8c), 4-formylbenzonitrile (8d), 4-di-
methylaminobenzaldehyde (8e), 4-piperidin-1-yl-benzaldehyde
(8f), and 4-morpholin-4-yl-benzaldehyde (8g), respectively.
5-(p-Methoxyphenyl)-1,9-dimethyldipyrromethene (9a). Yield
mp 107 °C. H NMR (CDCl3): δ 7.50 (m, 5H); 6.72 (d, 2H,
J ) 3.6 Hz); 6.29 (d, 2H, J ) 4.4 Hz); 2.67 (s, 6H). 13C NMR
(CDCl3): δ 155.2, 140.4, 138.8, 137.9, 131.2, 129.5, 128.9,
127.9, 117.9, 14.7. LRMS (EI, 70 eV) m/z (%): 297 (M+, 21),
296 (M+, 100), 295 (M+, 33). HRMS (EI+): calcd for
C17H15BF2N2 (M+), 296.12964; found, 296.12875.
1
32%. H NMR (CDCl3): δ 7.7 (s, NH); 7.40 (d, 2H, J ) 8.7
BODIPY derivatives 1, 3, 4, 5, 6, and 7 were prepared using
the same procedure, starting from dipyrromethenes 9a, 9c, 9d,
9e, 9f, and 9g, respectively.
Hz); 6.94 (d, 2H, J ) 8.7 Hz); 6.50 (d, 2H, J ) 4.3 Hz); 6.16
(d, 2H, J ) 4.4 Hz); 3.91 (s, 3H); 2.45 (s, 6H). MS (CI) m/z
(%): 279 (MH+, 100).
4,4-Difluoro-3,5-dimethyl-8-(4-methoxyphenyl)-4-bora-3a,4a-
diaza-s-indacene (1). Recrystallization from hexane/CHCl3
yields red crystals in 82% yield. mp 191 °C. 1H NMR
(CDCl3): δ 7.45 (d, 2H, J ) 8.7 Hz); 7.0 (d, 2H, J ) 8.7 Hz);
6.75 (d, 2H, J ) 3.6 Hz); 6.27 (d, 2H, J ) 4.4 Hz); 3.90 (s,
3H); 2.66 (s, 6H). 13C NMR (CDCl3): δ 156.0, 152.2, 144.2,
134.7, 132.7, 130.3, 122, 3118.9, 111.7, 52.7, 15.2. LRMS (EI,
70 eV) m/z (%): 327 (M+, 23), 326 (M+, 100), 325 (M+, 50),
307 (14), 306 (48), 305 (24). HRMS (EI+): calcd for
C18H17BF2N2O (M+), 326.14020; found, 326.13975.
4,4-Difluoro-3,5-dimethyl-8-(4-carbomethoxyphenyl)-4-bora-
3a,4a-diaza-s-indacene (3). Recrystallization from hexane/
CHCl3 yields green crystals in 67% yield. mp 205 °C. 1H NMR
(CDCl3): δ 8.14 (d, 2H, J ) 8.1 Hz); 7.57 (d, 2H, J ) 8.0
Hz); 6.66 (d, 2H, J ) 3.6 Hz); 6.28 (d, 2H, J ) 3.6 Hz); 3.99
(s, 3H); 2.67 (s, 6H). 13C NMR (CDCl3): δ 166.7, 158.6, 141.3,
138.8, 134.6, 130.7, 129.7, 120.1, 52.8, 30.0, 15.3. LRMS (EI,
70 eV) m/z (%): 353 (M+, 50), 354 (M+, 100), 355 (M+, 15),
333(42), 334(61). HRMS (EI+): calcd for C19H17BF2N2O2
(M+), 354.13511; found, 354.13522.
5-(4-Carbomethoxyphenyl)-1,9-dimethyldipyrromethene (9c).
1
Yield 24%. H NMR (CDCl3): δ 8.09 (d, 2H, J ) 8.0 Hz);
7.53 (d, 2H, J ) 8.0 Hz); 6.39 (d, 2H, J ) 3.6 Hz); 6.16 (d,
2H, J ) 4.3 Hz); 3.97 (s, 3H); 2.46 (s, 6H). MS (CI) m/z (%):
307 (MH+, 100).
5-(4-Cyanophenyl)-1,9-dimethyldipyrromethene (9d). Yield
1
58%. H NMR (CDCl3): δ 7.63 (d, 2H, J ) 7.7 Hz); 7.41 (d,
2H, J ) 7.9 Hz); 6.58 (d, 2H, J ) 3.8 Hz); 6.21 (d, 2H, J )
3.3 Hz); 2.46 (s, 6H). MS (CI) m/z (%): 274 (MH+, 100).
5-(4-N,N-Dimethylphenyl)-1,9-dimethyldipyrromethene (9e).
1
Yield 54%. H NMR (CDCl3): δ 7.38 (d, 2H, J ) 8.0 Hz);
6.73 (d, 2H, J ) 8.8 Hz); 6.60 (d, 2H, J ) 3.6 Hz); 6.17 (d,
2H, J ) 3.6 Hz); 3.04 (s, 6H); 2.45 (s, 6H). MS (CI) m/z (%):
292 (MH+, 100).
5-(4-Pipiridinophenyl)-1,9-dimethyldipyrromethene (9f). Yield
1
64%. H NMR (CDCl3): δ 7.31 (d, 2H, J ) 8.1 Hz); 6.84 (d,
2H, J ) 7.9 Hz); 6.58 (d, 2H, J ) 3.7 Hz); 6.16 (d, 2H, J )
4.1 Hz); 3.19 (m, 4H); 2.45 (s, 6H); 1.72 (m, 4H); 1.59 (m,
2H). MS (CI) m/z (%): 332 (MH+, 100).
5-(4-Morpholinophenyl)-1,9-dimethyldipyrromethene (9g). Yield
71%. H NMR (CDCl3): δ 7.39 (d, 2H, J ) 8.8 Hz); 6.92 (d,
4,4-Difluoro-3,5-dimethyl-8-(4-cyanophenyl)-4-bora-3a,4a-
diaza-s-indacene (4). Recrystallization from hexane/CHCl3
1

Photophysical Properties of BODIPY Analogues
J. Phys. Chem. A, Vol. 109, No. 33, 2005 7373
yields green crystals in 52% yield. mp 193 °C. 1H NMR
(CDCl3): δ 7.79 (d, 2H, J ) 8.0 Hz); 7.6 (d, 2H, J ) 8.0 Hz);
6.61 (d, 2H, J ) 4.4 Hz); 6.30 (d, 2H, J ) 3.7 Hz); 2.67 (s,
6H). 13C NMR (CDCl3): δ 159.3, 139.0, 134.4, 132.4, 131.2,
130.3, 120.5, 114.2, 118.4, 15.3. LRMS (EI, 70 eV) m/z (%):
322 (M+, 30), 321 (M+, 100), 320 (M+, 64), 319 (M+, 22),
302 (32), 301 (82), 299 (25). HRMS (EI+): calcd for
C18H14BF2N3 (M+), 321.12488; found, 321.12519.
4,4-Difluoro-3,5-dimethyl-8-(4-N,N-dimethylaminophenyl)-4-
bora-3a,4a-diaza-s-indacene (5). Recrystallization from hexane/
CHCl3 yields orange red crystals in 86% yield. mp 279 °C. 1H
NMR (CDCl3): δ 7.44 (d, 2H, J ) 8.8 Hz); 6.84 (d, 2H, J )
4.0 Hz); 6.76 (d, 2H, J ) 8.5 Hz); 6.26 (d, 2H, J ) 4.0 Hz);
3.08 (s, 6H); 2.65 (s, 6H). 13C NMR (CDCl3): δ 144.2, 156.0,
152.2, 134.7, 132.7, 130.3, 122.3, 118.9, 111.7, 40.5, 15.2.
LRMS (EI, 70 eV) m/z (%): 340 (M+, 24), 339 (M+, 100),
338 (M+, 45), 319 (41), 318 (18). HRMS (EI+): calcd for
C19H20BF2N3 (M+), 339.17183; found, 339.17184.
4,4-Difluoro-3,5-dimethyl-8-(4-pipiridinophenyl)-4-bora-3a,4a-
diaza-s-indacene (6). Recrystallization from hexane/CHCl3
yields red crystals in 72% yield. mp 280 °C. 1H NMR
(CDCl3): δ 7.41 (d, 2H, J ) 8.8 Hz); 6.95 (d, 2H, J ) 8.8
Hz); 6.82 (d, 2H, J ) 4.4 Hz); 6.26 (d, 2H, J ) 3.6 Hz); 3.32
(t, 4H, J ) 5.5 Hz); 2.65 (s, 6H); 1.66 (m, 6H). 13C NMR
(CDCl3): δ 156.3, 153.4, 144.1, 134.7, 132.6, 130.4, 124.1,
119.1, 114.7, 49.6, 25.9, 24.7, 15.2. LRMS (EI, 70 eV) m/z
(%): 380 (M+, 21), 379 (M+, 100), 378 (M+, 62), 359 (17),
302 (11). HRMS (EI+): calcd for C22H24BF2N3 (M+), 379.20313;
found, 379.20291.
timing method.11 Details of the instrumentation12 and experi-
mental procedures13 used have been described elsewhere.
Briefly, the second harmonic of a Ti:sapphire laser (Tsunami,
Spectra Physics) was used to excite the samples at 488 nm with
a repetition rate of 8.18 MHz. The detection system consisted
of a subtractive double monochromator (9030DS, Sciencetech)
and a microchannel plate photomultiplier (R38090U, Hamamat-
su). A time-correlated single-photon timing PC module (SPC
630, Picoquant) was used to obtain the fluorescence decay
histograms in 4096 channels, with a time increment per channel
of 3.0 ps. In combination with the stepper motor controller
device SM (STP 240, Picoquant GmbH), which sets the
monochromator wavelength, the SPC card enabled us to record
fluorescence decays on the picosecond time scale. Fluorescence
decays at several emission wavelengths were recorded using
10 × 10 mm cuvettes. The absorbance at the excitation
wavelength was always below 0.1. All lifetime measurements
were performed on samples that were degassed by consecutive
freeze-pump-thaw cycles. Histograms of the instrument
response functions (using LUDOX scatterer) and sample decays
were recorded until they typically reached 104 counts in the
peak channel. The total width at half-maximum of the instrument
response function was ∼60 ps.
The fitting parameters were determined by minimizing the
2
global reduced chi-square (øg ):
q
2
øg
)
w (yo - yc)/ν
(1)
∑∑
li li
li
l
i
where the index l sums over q experiments and the index i sums
over the appropriate channel limits for each individual experi-
ment. yoli and ylci denote respectively the observed and calcu-
lated (fitted) values corresponding to the ith channel of the lth
experiment, and wli is the corresponding statistical weight. ν
represents the number of degrees of freedom for the entire
multidimensional fluorescence decay surface.
The statistical criteria to judge the quality of the fit included
both graphical and numerical tests. The graphical methods
covered plots of surfaces (“carpets”) of the autocorrelation
function values versus channel number versus experiment
number and of the weighted residuals versus channel number
versus experiment number. The additional statistical criteria to
judge the quality of the fit are described elsewhere.14
4,4-Difluoro-3,5-dimethyl-8-(4-morpholinophenyl)-4-bora-
3a,4a-diaza-s-indacene (7). Recrystallization from hexane/
CHCl3 yields orange crystals in 93% yield. mp 264 °C. 1H NMR
(CDCl3): δ 7.4 (d, 2H, J ) 8.7 Hz); 6.9 (d, 2H, J ) 8.8 Hz);
6.7 (d, 2H, J ) 4.4 Hz); 6.2 (d, 2H, J ) 3.7 Hz); 3.8 (t, 4H,
J ) 4.7 Hz); 3.2 (t, 4H, J ) 4.7 Hz); 2.6 (s, 6H). 13C NMR
(CDCl3): δ 156.8, 152.9, 149.1, 134.8, 132.4, 130.5, 125.4,
119.3, 114.5, 67.1, 48.5, 15.2. LRMS (EI, 70 eV) m/z (%): 382
(M+, 26), 381 (M+, 100), 379 (M+, 1), 361 (14), 323 (16), 322
(11), 303 (24), 302 (21). HRMS (EI+): calcd for C21H22BF2N3O
(M+), 381.18240; found, 381.18158.
Materials. All solvents (Aldrich, Sigma-Aldrich, Acros
Organics, or Riedel-Dehae¨n) for the spectroscopic measurements
were of spectroscopic grade and were used without further
purification. The chemicals used for the synthesis were of the
best grade available, supplied by Acros Organics or Aldrich,
and used as received. Dichloromethane (p.a.) was dried over
molecular sieves. Boron trifluoride etherate, ∼48% BF3, was
from Acros Organics.
Steady-State Spectroscopy. The absorption measurements
were performed on a Perkin-Elmer Lambda 40 UV-vis
spectrophotometer. Corrected steady-state excitation and emis-
sion spectra were recorded on a SPEX Fluorolog instrument.
For the determination of the relative fluorescence quantum yields
(φf), only dilute solutions with an absorbance below 0.1 at the
excitation wavelength (λex) were used. Rhodamine 6G in H2O
(λex ) 488 nm, φf ) 0.76) was used as a fluorescence standard.10
For the samples that were not degassed, the φf values reported
in this work are the average values of multiple (generally four)
fully independent measurements. For the determination of φf
of samples that were degassed by consecutive freeze-pump-
thaw cycles, only one φf measurement was done. In all cases,
correction for the refractive index was applied.
When the fluorescence decays were monoexponential, the rate
constants of radiative (kf) and nonradiative (knr) deactivation
were calculated from the measured fluorescence quantum yield
(φf) and fluorescence lifetime (τ) according to eqs 2 and 3:
kf ) φf/τ
(2)
(3)
knr ) (1 - φf)/τ
All measurements were done at 20 °C.
Determination of Ka from Fluorimetric Titration. To
determine the ground-state acidity constants (Ka) of (the
ammonium salts of) 5, 6, and 7, the fluorescent indicators were
dissolved in 1.0 M HCl solution in Milli-Q water. Then, 0.1
and 0.01 M solutions of HCl and NaOH were used to adjust
the pH of the solutions. All measurements were performed in
aqueous nonbuffered solutions at 20 °C, using 10 mm optical
path quartz cells.
The expression of the steady-state fluorescence signal (F) as
a function of the ion concentration has been derived by
Kowalczyk et al. for the case of a 1:1 complex between a
fluorescent indicator and an analyte (here H+).15 The expression
Time-Resolved Spectroscopy. Fluorescence decay traces of
the BODIPY compounds were recorded by the single-photon

7374 J. Phys. Chem. A, Vol. 109, No. 33, 2005
Qin et al.
Figure 1. (a) Normalized absorbance and fluorescence emission spectra of 1 in cyclohexane. (b) Enlargement of the main absorbance peak of 1
in several solvents. (c) Enlargement of the main fluorescence emission peak of 1 in several solvents. The excitation wavelengths used in part c are
given in parentheses. The intensities in this figure are normalized to the same value at the wavelength of maximum intensity.
has been extended to the case of an n/1 complex between cation
and indicator (eq 4).16 The Ka values of the ammonium salts of
5, 6, and 7 were determined by fluorimetric titration as a function
of pH using the fluorescence excitation and/or emission spectra
(at least four independent measurements were used to compute
the average Ka value). Nonlinear fitting of eq 4 to the steady-
state fluorescence data (F) recorded as a function of [H+] yields
values of Ka, the fluorescence signals Fmin and Fmax at minimal
and maximal [H+], respectively (corresponding to the uncharged
amine and positively charged ammonium forms of the pH probe,
respectively), and n (the number of protons bound per amine).
Equation 4 assumes that the absorbance of the sample is small
(<0.1) and that the binding of H+ to the probe in the excited
state is negligible.
final curve fittings, from which the estimated values of Ka, Fmin
and Fmax are reported here.
,
Results and Discussion
Spectroscopic Properties of 1, 2, 3, and 4. Figure 1 shows
the absorption and steady-state fluorescence emission spectra
of 1 dissolved in several solvents. The absorption spectra are
of similar shape as those of described BODIPY dyes,6-8 with
a strong absorption band around 500 nm and a shoulder at the
short wavelength side. The absorption spectra are only barely
affected by solvent polarity, with the maximum being slightly
shifted hypsochromically (∼7 nm) when the solvent is changed
from toluene (512 nm) to acetonitrile (505 nm), which is
consistent with the general behavior of BODIPY chromo-
phores.6-8 This shift reflects the polarizability of the solvent
and has also been observed for cyanine dyes17 and aromatic
amines.18 They all show narrow spectral bandwidths with two
absorption maxima: (1) a strong S0-S1 transition with a
maximum appearing at 509 ( 4 nm and (2) the second
maximum, or shoulder on the high energy side, centered at about
480 nm, which is attributed to the 0-1 vibrational transition.
The molar absorption coefficients (ꢀ) for the new BODIPY
analogues are relatively high and lie in the 21 000-65 000 M-1
F
max[H+]n + FminKa
Ka + [H+]n
F )
(4)
Fitting eq 4 to the steady-state fluorescence data (F) with n,
Ka, Fmin, and Fmax as freely adjustable parameters always gave
values of n close to 1, indicating that one proton is bound per
tertiary amine molecule. Therefore, n was kept fixed at 1 in the

Photophysical Properties of BODIPY Analogues
J. Phys. Chem. A, Vol. 109, No. 33, 2005 7375
TABLE 1: Photophysical Properties of BODIPY Compounds 1, 2, 3, and 4 in Several Solvents
λabs
λem
λex
fwhmabs
∆νj
kf
knr
BODIPY
solvent
toluene
CHCl3
(max/nm)
(max/nm)
(max/nm)
(cm-1
)
(cm-1
)
φf
τ (ns)
(109 s-1
)
(109 s-1
)
1
512
511
510
508
507
506
505
513
513
512
510
508
508
507
517
517
515
514
512
511
511
519
519
517
516
513
513
512
524
523
521
521
518
518
517
528
526
524
525
522
521
521
538
538
533
536
533
533
531
543
543
539
540
537
538
537
512
511
510
508
506
505
505
513
513
512
510
508
508
507
517
517
515
514
511
511
511
518
518
517
515
513
513
512
806
769
695
775
781
785
825
836
803
766
809
775
819
819
898
863
869
949
960
960
1000
930
930
898
1056
989
992
992
447
449
414
491
419
458
460
554
482
447
560
528
491
530
755
755
656
799
770
808
739
852
852
789
861
871
906
909
0.52
0.46
0.42
0.36
0.30
0.25
0.21
0.42
0.38
0.29
0.25
0.25
0.19
0.17
0.12
0.095
0.073
0.056
0.053
0.032
0.039
0.076
0.055
0.038
0.037
0.035
0.026
0.030
3.35
3.28
2.87
0.16
0.14
0.15
0.14
0.16
0.20
cyclohexane
THFa
EtOAc
MeOH
MeCN
toluene
CHCl3
2.24
1.87
1.60
2.65
2.65
1.94
0.13
0.13
0.13
0.16
0.14
0.15
0.31
0.40
0.49
0.22
0.23
0.37
2
3
4
cyclohexane
THFa
EtOAc
MeOH
MeCN
1.71
1.46
1.33
0.82
0.15
0.13
0.13
0.15
0.44
0.56
0.63
1.07
toluene
a
CHCl3
cyclohexane
THF
EtOAc
MeOH
MeCN
toluene
CHCl3
cyclohexane
THF
EtOAc
MeOH
MeCN
0.54
0.13
1.71
0.55
0.14
1.68
a The fluorescence decays measured as a function of emission wavelength could only be described globally by a triple-exponential decay law.
cm-1 range. In addition, a considerably weaker, broad absorption
band is found at about 375 nm for 1, and its position is not
appreciably affected by solvent polarity. We attribute this
broader and weaker absorption band to the S0-S2 transition.
Table 1 summarizes the photophysical data of compounds 1-4
in several solvents.
Compound 1 also shows the typical emission features of
BODIPY6-8 (Figure 1), that is, a narrow, slightly Stokes-shifted
band of mirror image shape, a high fluorescence quantum yield
(φf) which decreases (from 0.52 in toluene to 0.21 in acetonitrile)
with increasing solvent polarity, and fluorescence bands that
are hypsochromically shifted (from 524 nm in toluene to 517
nm in acetonitrile) with increasing solvent polarity and decreas-
extinction coefficients. The fluorescence quantum yield (φf) and
lifetime (τ) decrease when the solvent changes from toluene
(0.52 and 3.35 ns) to acetonitrile (0.21 and 1.60 ns), due to an
increase of the nonradiative rate constant (knr) in more polar
solvents. The S1-S0 transition of BODIPY compounds 1-4
resembles strongly that of cyanine dyes. This is no surprise, as
the “core” of the molecule consists of an extended delocalized
π-system with two completely equivalent delocalized structures
A and B (Figure 3a), leading to the absence of bond alternation.
Single-crystal X-ray structural analysis of fused BODIPY
analogues has shown that the average bond length of N3a-C3
and N4a-C5 indicates a pronounced double-bond character in
contrast to the average bond length of N3a-C8a and N4a-
C7a, which is significantly longer.20 Nonetheless, a strong
π-electron delocalization is observed within the central six-
membered ring and both adjacent five-membered rings. This
π-electron delocalization is interrupted between both B-N
bonds. For symmetry reasons, all ring atoms of the BODIPY
chromophore are exactly positioned within a plane. Although
BODIPY resembles monomethine cyanines where the delocal-
ization involves 6 π-electrons distributed over 5 atoms, closer
inspection shows that the actual chromophore consists of 12
π-electrons distributed over 11 atoms, which makes it isoelec-
tronic with heptamethine cyanines.
The absorption and emission spectra of 2, 3, and 4 are, in
general, similar to those of 1 in the solvents used. Both the
long wavelength absorption band and the fluorescence band are
bathochromically shifted when the polarizability of the solvent
is increased.
Figure 4 shows the absorption and fluorescence spectra of 1,
2, 3, and 4 in methanol. A slight bathochromic shift appears to
increase with increasing electron withdrawing power of the
para-substituent (R) (see Scheme 1) in the phenyl group. Indeed,
the hydrogen group (in 2 vs the methoxy group in 1) causes a
spectral shift to the red of 2 and 3 nm in the absorption and
fluorescence bands, respectively, whereas the carbomethoxy
ing polarizability. The linearity of a plot of the absorption (νjabs
)
and fluorescence (νjem) maxima versus f(n2) ) (n2 - 1)/(2n2 +
1) (Figure 2) confirms that van der Waals interactions with a
polarizable solvent can rationalize the solvent dependence of
the excitation energy. If a large difference in permanent dipole
moment would exist between the ground and excited state, the
excitation energy would depend linearly on f(ꢀ) - f(n2) )
[(ꢀ - 1)/(2ꢀ + 1)] - [(n2 - 1)/(2n2 + 1)] rather than on f(n2),
and hence, no linear dependence on f(n2) would be observed
anymore (e.g., the data points for chloroform and toluene would
be situated between those of tetrahydrofuran and cyclohexane).
Also, the independence of the Stokes shift upon the solvent
polarity (Table 1) indicates that the permanent dipole moments
(if any) do not differ between the ground state and the excited
state.19 The absorption and emission spectra display perfect
mirror symmetry, showing that the main bands correspond to
the S0-S1 transition. The fluorescence decay profiles of 1 could
be described by a single-exponential fit (fluorescence lifetimes
(τ) in the 1.60-3.35 ns range) in all of the solvents investigated
except for tetrahydrofuran (THF). In fact, the radiative rate
constant kf ) (0.14 ( 0.01) × 109 s-1 is identical within
experimental error (Table 1). These values suggest that the
oscillator strength is close to 1, corresponding to the large

7376 J. Phys. Chem. A, Vol. 109, No. 33, 2005
Qin et al.
Figure 3. (a) The two main delocalized structures of the BODIPY
fluorophore, including the numbering. (b) Representation of the CT
excited state of compounds 5-7.
Figure 2. Plots of the absorption maxima (νjabs) (a) and fluorescence
emission maxima (νjem) (b) for compounds 1-4 vs (n2 - 1)/(2n2 + 1)
as a function of solvent. From left to right: MeOH, MeCN, EtOAc,
THF, cyclohexane, CHCl3, and toluene.
group (3 vs 1) gives rise to a larger corresponding shift of 5
and 15 nm to lower energies in the absorption and fluorescence
spectra, respectively. The cyano group (in 4) causes an even
larger bathochromic shift of 7 and 20 nm in the absorption and
emission bands, respectively, compared to compound 1. Similar
small bathochromic shifts that increase with the electron
withdrawing strength of the para-substituent (R) in compounds
1-4 are observed in the other solvents. In all cases, the shifts
are smaller in absorption than in emission. The fact that the
absorption and emission bands show a bathochromic shift with
increasing electron withdrawing tendency of the para-substituent
(R) can be rationalized as follows. Ab initio and semiempirical
calculations indicate that the S0-S1 transition is a nearly pure
highest occupied molecular orbital (HOMO)-lowest unoccupied
molecular orbital (LUMO) transition.8 The electron density map
of these states suggests an important increase in the electronic
density at position 8 in the LUMO with respect to the HOMO
state. The higher the electron withdrawing ability of the para-
substituent (R), the better the LUMO is stabilized. On the other
hand, the substituent in position 8 will have little influence on
the HOMO, as the latter orbital has a node at this position.
Hence, an electron withdrawing group in position 8 will decrease
the HOMO-LUMO energy difference. The time-resolved
Figure 4. Normalized absorbance (a) and fluorescence emission (b)
of 1, 2, 3, and 4 in methanol. The intensities are normalized to the
same value at the wavelength of maximum intensity. The inset shows
an enlargement of the main peak. The excitation wavelengths used in
part b are given in parentheses.
emission of 2, 3, and 4 can be described by a monoexponential
decay function in most solvents. In the same solvent, the φf

Photophysical Properties of BODIPY Analogues
J. Phys. Chem. A, Vol. 109, No. 33, 2005 7377
and τ values decrease with increasing electron withdrawing
strength of the para-substituent (R) (Table 1), which is due to
a faster nonradiative decay rather than to a slower radiative
decay. Indeed, the experimental results indicate that the para-
substituent (R) in 2, 3, and 4 does not lead to a significant
change in the value of kf, in analogy to 1. Although both the
fluorescence quantum yield (φf) and lifetime (τ) are highest for
1, the radiative rate constants remain the same [kf ) (0.14 (
0.01) × 109 s-1] for 1-4, irrespective of the nature of the
solvent. The nonradiative rate constants (knr), however, decrease
in going from 4 to 1 according to 4 > 3 > 2 > 1.
The Stokes shift (∆νj), calculated from the wavelengths of
maximum intensity, also increases as with the electron with-
drawing strength of the para-substituent (R). The Stokes shifts
measured for 4 are 2 times larger than those for 1. Moreover,
the Stokes shifts of each compound do not vary much as a
function of the solvent. This suggests that increasing the electron
withdrawing properties would lead to larger geometric rear-
rangements upon excitation. This increased electron phonon
coupling could also be responsible for the increased nonradiative
decay. A possible candidate for the vibration responsible for
the electron phonon coupling could be the rotation of the phenyl
which becomes more coplanar upon excitation, with the effect
being most outspoken for electron withdrawing groups in the
para position.
To conclude, although the photophysical properties of
compounds 1-4 resemble strongly those of cyanine dyes, they
have the advantage that the difluoroboron bridge avoids the cis-
trans isomerism associated with cyanine dyes absorbing in the
same spectral regions and leads in apolar solvents to significantly
higher quantum yields.17,21 The smaller extinction coefficient
and oscillator strength compared to cyanines dyes with the same
number of π-electrons in the polymethine chain can be attributed
to the folding of this polymethine chain, leading to a smaller
transition dipole. Another difference with cyanine dyes is the
presence of a second absorption band 7000-10 000 cm-1 above
the S0-S1 transition. For the latter transition, which will be
discussed more in detail when discussing compounds 5-7, a
bathochromic shift was induced by the electron donating
methoxy group in the para position.
Figure 5. Normalized absorbance (a) and fluorescence emission (b)
of 5 in several solvents. The intensities are normalized to the same
value at the wavelength of maximum intensity. The excitation
wavelengths used in part b are given in parentheses.
one cannot exclude that the weaker S0-S2 transition is now
overlapping with the vibronic progression of the S0-S1 transi-
tion. As the former is characterized by a relatively wide and
structureless band, the general effect will besagreeing with our
observationssan increase of the extinction coefficient of the
first absorption band. Due to the strong overlap with the S1-S0
transition, the maximum of this second band which corresponds
to the energy of the Franck-Condon charge transfer (CT) state
(cfr. infra) is for 5 and 6 hidden under the intense S0-S1
transition, and hence, its maximum cannot be determined
accurately. Therefore, we only observe the red tail of the CT
absorption band. There is no indication that in the solvents used
here the energy of this Franck-Condon CT state is lower than
that of the cyanine-like excited state.
Furthermore, the emission behavior of 5-7 is strongly solvent
dependent. In cyclohexane, BODIPY analogues 5-7 show
emission at about 520 nm shaped like the mirror image of the
absorption spectrum. In cyclohexane, the emission spectra of
5-7 strongly resemble those of 1-4. In analogy to 1-4, the
emission can be attributed to the polymethine chromophore, also
responsible for the lowest energy intense absorption band. The
excited state to which this emission can be attributed will be
called the LE state. In all solvents more polar than cyclohexane,
however, this LE emission is (strongly) quenched and the
formation of a second broad emission band is observed at longer
wavelengths.
Spectroscopic Properties of 5, 6, and 7. The steady-state
absorption and fluorescence emission spectra of 5, 6, and 7 are
depicted in Figures 5, 6, and 7, respectively. The absorption
spectra of 5-7 are apparently almost the same as those of 1-4
(with the exception of a wider base) and are identical with the
well-known absorption pattern for BODIPY derivatives.6-8 In
compounds 5-7, a strong S0-S1 transition with a maximum
around 500 nm and a shoulder at a shorter wavelength is
observed, while the intensity of the absorption band around 350
nm is strongly diminished. The molar absorption coefficients
for BDP derivatives 5-7 are relatively high and lie in the
52 000-95 000 M-1 cm-1 range. While the absorption maxima
show the same solvent dependence as those of compounds 1-4,
closer inspection shows the presence of a tail at the red edge of
the first absorption band in solvents more polar than toluene
for 5 and 6, which becomes the most outspoken in the most
polar solvents acetonitrile and methanol. The red shift of the
whole spectrum including the tail in toluene is due to increased
polarizability. The occurrence of the red tail in more polar
solvents is at the expense of the band observed at 425 nm in
cyclohexane and 450 nm in toluene. For 7, the appearance of
the red tail is less clear. This was not observed for compounds
1-4. Taking into account the red shift of the S0-S2 transition
upon introduction of the electron donating p-methoxy group,
With increasing solvent polarity, the intensity of the long
wavelength emission band decreases, while the maximum of

7378 J. Phys. Chem. A, Vol. 109, No. 33, 2005
Qin et al.
Figure 6. Normalized absorbance (a) and fluorescence emission (b)
of 6 in several solvents. The intensities are normalized to the same
value at the wavelength of maximum intensity. The excitation
wavelengths used in part b are given in parentheses.
Figure 7. Normalized absorbance (a) and fluorescence emission (b)
of 7 in several solvents. The intensities are normalized to the same
value at the wavelength of maximum intensity. The excitation
wavelengths used in part b are given in parentheses.
the band shifts more to the red. In acetonitrile, methanol, and
ethyl acetate, this broad bathochromic emission has such a low
intensity that it can hardly be detected. This red-shifted band
occurs at longer wavelengths for 5 and 6 compared to 7. More
quantitative experiments show that the quantum yield of
fluorescence (φf) in aerated solutions decreases with increasing
polarity of the solvent from cyclohexane to acetonitrile. The φf
values were also determined in deaerated solutions of cyclo-
hexane, toluene, and chloroform and are within experimental
error about the same as those in the corresponding aerated
solutions. This is to be expected because the fluorescence
lifetimes are rather short (see next paragraph), so that oxygen
quenching will have a limited effect. Table 2 summarizes the
photophysical data of BODIPY analogues 5-7 in several
solvents.
The solvent dependence of the long wavelength emission
points to an excited state with a large dipole. This can also
clarify why a broad bell shaped structureless band is observed.
The CT excited state may be represented by the delocalized
structure depicted in Figure 3b.
The same CT state will also be responsible for the short
wavelength absorption band observed for 1-4 in all solvents
and for 5-7 in cyclohexane. This clarifies why this transition
shifts to longer wavelengths from 2-4 over 1 to 5-7 according
to the electron donating character of the para-substituent (R).
This CT state is at lower energy in 5 and 6 compared to 7, as
indicated by the more prominent red tail in the absorption
spectrum and the larger red shift of the fluorescence spectrum
(Figures 5-7). In the absorption spectrum of 7, the absorption
spectrum of the CT state is mainly hidden under that of the LE
state (Figure 7a).
Increasing the solvent polarity also leads to a more prominent
red tail in the absorption spectrum. This indicates that the
transition to the CT state shifts to lower energy upon increasing
the solvent polarity. The dependence of the absorption maximum
on the solvent polarity is given by
2(µG - µE) µG(ꢀ - 1) (µG - µE)(n2 - 1)
νjabs - νj0abs
)
-
4πꢀ0hcF3
2(2n2 + 1)
[
]
2ꢀ + 1
(5)
νjabs and νj0 correspond to the absorption maximum (cm-1) in
abs
a solvent with dielectric constant ꢀ and refractive index n and
the absorption frequency in a vacuum, respectively. ꢀ0, h, c,
and F correspond to the permittivity of a vacuum (8.85 × 10-12
C V-1 m-1), Planck’s constant (6.6 × 10-34 J s), the velocity
of light in a vacuum (3.0 × 1010 cm s-1), and the radius of the
solvent cavity (in m), respectively. µE (C m) and µG (C m) are
the permanent dipole moments of the ground state and the
excited state, respectively. The first term in eq 5 indicates that
the position of the absorption band can depend only on f(ꢀ) when
µG differs from zero. Hence, the red shift of the absorption

Photophysical Properties of BODIPY Analogues
J. Phys. Chem. A, Vol. 109, No. 33, 2005 7379
TABLE 2: Photophysical Properties of BODIPY Compounds 5, 6, and 7 in Several Solvents
λabs
λem
λex
fwhmabs
kf
knr
BODIPY
solvent
toluene
CHCl3
cyclohexane
THF
EtOAc
MeOH
MeCN
(max/nm)
(max/nm)
(max/nm)
(cm-1
)
φfa
φfb
τ1 (ns)
τ2 (ns)
(109 s-1
)
(109 s-1
)
5
509
509
508
506
505
504
503
511
511
509
507
506
505
504
511
511
509
507
506
505
504
580
615
518
509
509
508
934
1136
741
1024
991
1159
1329
890
981
738
979
945
991
1196
769
809
695
897
825
866
949
0.15
0.058
0.37
0.002
0.004
0.15
0.094
0.37
0.028
3.24
a
2.72
0.14
0.14
0.16
0.23
0.29
0.30
6
7
toluene
593
527
519
527
509
511
508
508
0.13
0.077
0.33
0.16
0.090
0.33
0.015
2.31
3.12
3.12
a
CHCl3
cyclohexane
THF
EtOAc
MeOH
MeCN
0.01
0.006
0.005
0.003
0.32
0.11
0.35
0.013
0.008
0.005
0.002
toluene
524
531
520
695
510
513
509
507
0.35
0.10
0.35
0.028
2.15
c
CHCl3
cyclohexane
THF
EtOAc
MeOH
MeCN
a φf values determined for aerated samples. b φf values determined for degassed samples (freeze-pump-thaw cycles). c The fluorescence decays
measured as a function of emission wavelength could only be described globally by a triple-exponential decay law.
maximum of the transition to the CT state with increasing f(ꢀ)
suggests that the ground-state dipole will not be zero and hence
that the polar delocalized structure also contributes to the ground
state. It is however insufficient to shift the Franck-Condon
excited state of the CT transition below that of the LE transition,
even in polar solvents.
Under those conditions, the Franck-Condon LE state will
be, even for the p-aminosubstituted benzenes 5-7 in polar
solvents, below the CT state and the lowest absorption band
will always have a predominant LE character. However, when
after excitation this CT state is reached, it will be stabilized in
solvents more polar than cyclohexane to an extent that its energy
becomes lower than that of the LE, and intramolecular conver-
sion from the LE to the CT state will occur. Hence, compounds
5-7 show dual emission.22
longest emission wavelength (R1/R2 ) -0.198 at 600 nm). At
shorter observation wavelengths (λem), the emission originates
mainly from the LE state, whereas, at longer λem, primarily the
CT state emission is observed (see emission spectra of Figure
5b). This is consistent with a mechanism involving the excitation
of a single ground-state species to a LE state followed by a fast
conversion to a CT state. According to such a kinetic model,
the ratio R1/R2 should be -1 under conditions where the LE is
excited exclusively and only emission from the CT state is
observed.24 This is not yet the case at λem ) 600 nm, which
can be due to either observation of LE emission at 600 nm and/
or direct excitation of the CT state at 488 nm. Actually, the
absorption spectra suggest that for 5-7 the transition to the CT
state is hidden below that of the LE transition. If the conversion
from the LE to the CT state is irreversible, the short decay time
(τ1) can be considered as the lifetime of the LE state which is
substantially shortened by the new decay channel opened. Under
those conditions, the slow decay time (τ2) can be attributed to
the lifetime of the CT state. If the interconversion is reversible,
the interpretation of the decay times is more complex and a
complete analysis generally requires prior knowledge of the
fraction of the excitation leading to LE and CT as well as of
the fraction of the emission due to the LE and CT states or
some a priori knowledge of some rate constants.24,25 Figure 8
displays the fluorescence decay traces at λem ) 525 nm and at
λem ) 600 nm. The curve at 525 nm is clearly a sum of two
exponentials, while the curve at 600 nm is a difference of two
exponentials. The ratio of the pre-exponential factors (R1/R2)
for compound 6 in toluene also decreases with increasing
emission wavelength (λem) (R1/R2 ) 16.38 at 525 nm, R1/R2 )
3.336 at 550 nm, R1/R2 ) 1.376 at 570 nm, and R1/R2 ) 0.010
at 600 nm). For compound 7 in toluene, the variation of R1/R2
(R1/R2 ) 1.087 at 510 nm, R1/R2 ) 1.084 at 520 nm, R1/R2 )
1.054 at 530 nm, and R1/R2 ) 0.905 at 540 nm) as a function
of λem is not as large as that for 5 and 6 in toluene. The small
dependence of R1/R2 on λem for 7 compared to 5 and 6 is to be
expected on the basis of their fluorescence emission spectra
(Figures 5b, 6b, and 7b). In the absorption and emission spectra,
the red tail and shifted band are less pronounced for compound
7 compared to compounds 5 and 6.
Due to the large dipole moment difference between the CT
excited state and the ground state, electron phonon coupling
with solvent phonons, which can act as accepting modes for
internal conversion, will become increasingly important in polar
solvents, leading to a faster decay of the fluorescence.23
In cyclohexane, the emission of 5-7 decays monoexponen-
tially, yielding radiative rate constants of similar magnitude
[kf ) (0.14 ( 0.01) × 109 s-1]. This value for kf was also found
for 1-4. The nonradiative rate constant values (knr) are slightly
higher than those for compound 1. This again suggests that in
cyclohexane the excited state of 5-7 is a pure LE state. In
toluene, the fluorescence decays of 5-7 could only be described
globally by a biexponential decay law with τ1 being in the 15-
28 ps range (fast component) and τ2 being about 3.2 ns (slow
component). The time-resolved fluorescence profiles of 5 in
toluene were measured as a function of emission wavelength
(λem ) 525, 550, 570, and 600 nm) and analyzed globally as
biexponentials with linked decay times τ1 (fast) and τ2 (slow).
Measurements of fluorescence decay traces as a function of λem
in the region 525-600 nm revealed a decrease of the relative
amplitude associated with the fast component (τ1) with increas-
ing detection wavelength (λem). Indeed, the ratio of the
pre-exponential factors (R1/R2) decreases with increasing λem
(R1/R2 ) 5.176 at 525 nm, R1/R2 ) 1.284 at 550 nm, and
R1/R2 ) 0.276 at 570 nm) and even becomes negative at the

7380 J. Phys. Chem. A, Vol. 109, No. 33, 2005
Qin et al.
Figure 8. (left) Experimental fluorescence decay curve (black) of 5 in toluene (λex ) 488 nm, λem ) 525 nm). The best fit (red) via global analysis
with decay times τ1 and τ2 linked over the observation wavelengths (λem) (øg2 ) 1.094) yielded τ1 ) 28 ps and τ2 ) 3.24 ns with R1 ) 12.667 and
R2 ) 2.447. (right) Experimental fluorescence decay curve (black) of 5 in toluene (λex ) 488 nm, λem ) 600 nm). The best fit (red) via global
analysis yielded R1 ) -0.52 and R2 ) 2.622. Also given are the local reduced ø2 values, the weighted residuals (Ri), and the autocorrelation
functions (AC).
Fluorimetric Titrations of 5, 6, and 7. In 5-7, the
intramolecular charge transfer (ICT) can be efficiently blocked
by protonation. In aqueous solution, the effect of protonation
on the emission behavior of 5-7 is very large. Protonation turns
the BODIPY derivatives into highly fluorescent molecules, the
CT emission disappears, and the emission from the LE state
with its characteristically higher fluorescence quantum yield is
re-established. Now, we will examine the spectral changes of
5-7 upon protonation in aqueous (nonbuffered) solution.
Absorption Spectra. The absorption spectra of 5-7 show
interesting changes as a function of pH. From Figure 9a, it is
evident that varying the pH gives rise to several isosbestic points
and is accompanied by a hypsochromic shift of the major
absorption maximum of 5 (from 494 nm at pH 1.09 to 473 nm
at pH 8.60) upon increasing the pH. This blue shift of the major
maximum at higher pH, which will be commented on later, is
accompanied by the appearance of a clear shoulder around 550
nm, which is attributed to a transition to the CT state. Upon
protonation of the aniline moiety, the CT state shifts to higher
energy, as indicated by the appearance of a band with a
maximum at 350 nm, in analogy to 2. This band increases with
decreasing pH. Upon closer inspection of the spectra, one can
see that the transition with a maximum at 494 nm and which is
predominant at low pH is much broader than the absorption
band of the LE transition in 1-4. The red shift of the CT state
of 5 in water is large enough to manifest itself as a separate
shoulder rather than as a red tail in the less polar and hydrogen
bond donating methanol (Figure 5a for 5). As the dipolar
solvating powers of water and polar organic solvents such as
acetonitrilesdescribed by f(ꢀ)sare nearly identical, these results
indicate that hydrogen bond formation plays a large role in the
stabilization of the CT state. For 1-4, it also was found that
the major LE transition undergoes a hypsochromic shift when
going to the more polar and less polarizable solvents acetonitrile
and methanol (see Figure 2) in accordance with their decreased
polarizability. From Figure 9a, it is evident that the absorption
spectra of the neutral amine and the positively charged am-
monium species overlap considerably.
While at low pH the absorption band at 494 nm can be
attributed to the LE transition of BODIPY with a protonated
aniline nitrogen, this band shifts to 473 nm at high pH where
the aniline nitrogen is no longer protonated. This shift is in
agreement with the substituent effects observed for 1-4 because
the LUMO of the BODIPY chromophore is destabilized by
interacting in position 8 with the electron rich aniline moiety.
Analogous effects are found for 6, where the major absorption
maximum shifts hypsochromically from 493 nm at pH 1.12 to
484 nm at pH 7.85. The CT band of the tertiary amine has a
maximum around 550 nm and shifts upon protonation to higher
energy, as indicated by the appearance of a band with a
maximum around 340 nm. While in acetonitrile and methanol
the transition to the CT state of 6 is only visible as a red tail of
the major band (Figure 6a), this transition is shifted sufficiently
to lower energy in water to give rise to a separate shoulder.
The absorption spectrum of compound 7 as a function of pH
is shown in Figure 10a and indicates that the absorption
maximum of 7 remains located at 487 ( 1 nm as a function of
pH. The CT band of 7 in the studied solvents remains largely

Photophysical Properties of BODIPY Analogues
J. Phys. Chem. A, Vol. 109, No. 33, 2005 7381
Figure 9. Absorption spectra (a), fluorescence emission spectra (λex ) 496 nm) (b), fluorescence excitation spectra (λem ) 560 nm) (c), and
fluorescence titration curve (d) of 5 in aqueous nonbuffered solution as a function of pH. The solid line in Figure 9d represents the best fit of eq
4 with n ) 1 to the fluorimetric titration data of 5 obtained from the excitation spectra (Figure 9c, λem ) 560 nm, λex ) 511 nm). Part e shows
normalized absorption and excitation spectra of 5 at pH 1.15, and part f shows the excitation wavelength dependence of φf of 5 at pH 1.15. The
symbols in Figure 9f represent the experimental apparent φf values, and the solid line is the ratio of the fluorescence excitation signal over the
absorbance (Fex/A). The data in Figure 9f are normalized to 1 at 520 nm.
hidden under the major absorption band (Figure 10a), following
the tendency of the absorption spectra of 7 in solvents of
different polarities (Figure 7a)
Fluorescence Spectra. The fluorescence emission and excita-
tion spectra of 5 in aqueous nonbuffered solution are shown as
a function of pH in parts b and c of Figure 9, respectively. By
lowering the pH, the fluorescence signals increase significantly.
Unlike their absorption spectra, the fluorescence emission
maximum at 529 nm and the excitation maximum at 511 nm
remain unchanged while the intensity changes as a function of

7382 J. Phys. Chem. A, Vol. 109, No. 33, 2005
Qin et al.
Figure 10. Absorption spectra (a), fluorescence emission spectra (λex ) 488 nm) (b), and fluorescence excitation spectra (λem ) 560 nm) (c) of
7 in aqueous nonbuffered solution as a function of pH.
TABLE 3: Properties of the BODIPY Derivatives in Aqueous Nonbuffered Solutiona
λ
abs-acid/λabs-base
(max/nm)
λ
em-acid/λem-base
(max/nm)
φf-acid
(pH)
φf-base
(pH)
BDP
pKa
5
6
7
494/473
493/484
488/486
529/529
528/526
531/528
0.03 (1.87)
0.008 (1.38)
0.001 (1.05)
0.006 (3.3)
0.001 (4.32)
0.0002 (3.32)
2.37 ( 0.06
4.15 ( 0.05
1.47 ( 0.03
a Ka values were determined by fitting eq 4 with n ) 1 to the fluorescence data. The fluorescence quantum yields φf-base and φf-acid were
determined at respectively the highest and lowest pH values (in parentheses) used in the fluorimetric titration. The φf values of 5, 6, and 7 were
determined with excitation at 496, 493, and 488 nm, respectively.
pH. The highest φf value (0.03, determined by excitation at 496
nm, Table 3) is found at low pH (between pH 1.09 and 1.87).
At all pH values, the excitation and fluorescence spectra
resemble those of compounds 1-4, suggesting that the fluo-
rescence is due to the LE state. At low pH, the amine function
is protonated and the phenyl moiety becomes substituted with
a weak electron acceptor rather than with an electron donor,
leading to an absorption and emission maximum between that
of 2 and 3. When the ammonium group is deprotonated,
excitation yields with 100% efficiency an ICT state which is
nonfluorescent in water. The absence of pH dependent shifts
of the fluorescence excitation and emission bands may be due
to the extremely low fluorescence quantum yield of the CT
emission in water. This is in agreement with the observations
in other solvents where a decrease of φf of the CT state with
increasing solvent polarity is observed. The binding of H+ by
the tertiary amine suppresses the ICT due to ammonium ion
formation and hence blocks this nonradiative decay channel of
the LE state. This is accompanied by a large fluorescence
enhancement.
Compared to 1-4, there are still some discrepancies when
the tertiary amine function of 5 is protonated, while the
fluorescence excitation (511 nm) and emission (529 nm) maxima
of the protonated form of 5 are intermediate between those of
2 and 3 (see Table 1), indicating that 5-7 have a similar Stokes
shift to that of 1-4. However, the absorption spectrum of the
protonated form of 5 is much broader and its maximum (494
nm) is situated between 11 and 25 nm (440-975 cm-1) to

Photophysical Properties of BODIPY Analogues
J. Phys. Chem. A, Vol. 109, No. 33, 2005 7383
shorter wavelengths compared to the absorption maxima of 1-4.
This discrepancy suggests the presence of two different proto-
nated species at low pH. The observation of isosbestic points
in the absorption spectra indicates that the ratio between the
concentrations of these two species does not change as a function
of the pH. While the species which absorbs at 511 nm and emits
at 529 nm corresponds to protonation at the aniline nitrogen,
the nature of the species absorbing at 494 nm is not clear. The
differences between the excitation and absorption spectra (Figure
9e) suggest that both species interconvert only to a limited extent
in the excited state. A consequence of the presence of two
species is that the measured fluorescence quantum yield (φf) is
only an apparent value because a large part of the incident light
is absorbed by the nonfluorescent species with an absorption
maximum at 494 nm. Therefore, the apparent φf values are
expected to change as a function of excitation wavelength. It is
found experimentally (Figure 9f) that at wavelengths shorter
than the maximum of the excitation spectrum the quantum yield
values closely follow the ratio of the fluorescence excitation
signal over the absorbance. The photophysical behavior of 6 as
a function of pH is very similar to that of 5.
BODIPY moiety in the para position to the tertiary amine acts
as an electron withdrawing unit.
Conclusion
Our study reflects the effect of the para-substituent (R) of
the phenyl group on the fluorescence properties of BODIPY
deriviatives 1-4. The combination of the electron rich p-anisol
group and the electron deficient BODIPY fluorophore in
compound 1 has the highest fluorescence quantum effciency
among the four coumpounds. On the other hand, the electron
withdrawing p-cyano group in derivative 4 produces the smallest
φf value. However, the shape of the absorption and emission
spectra is barely affected by changing the para-substituent (R).
In polar solvents, the photophysical properties of free (i.e.,
uncharged) tertiary amines 5-7 are governed by a fast inter-
conversion to an ICT state, where the lone electrons of the amine
nitrogen are delocalized over the BODIPY, leading to strong
quenching of the fluorescence from the LE state. Upon
protonation of the amine nitrogen atom, the amine receptor
becomes an inductively electron withdrawing group and the ICT
process is suppressed, leading to a pH dependent fluorescence
enhancement of the LE emission.7,27 The variation of the para-
substituent group in compounds 5-7 has a large influence on
φf and pKa. The electron releasing tendency of piperidino >
NMe2 > morpholino is borne out in the pKa values: 6 (pKa
4.15) > 5 (pKa 2.37) > 7 (pKa 1.47).
The acidity constant (Ka) and the stoichiometry (n) of the
binding of H+ by the tertiary amine group of compound 5 were
determined by fitting eq 4 to the fluorescence excitation or
emission spectral data. The results of the fluorimetric titration
of 5 using excitation data (λem ) 560 nm, λex ) 511 nm) are
displayed in Figure 9d and yield a pKa value of 2.37 and a well-
defined 1:1 stoichiometry between 5 and H+, indicative of
ammonium ion formation. Similar curve fittings were performed
Acknowledgment. The authors are grateful to the University
Research Fund of the K.U.Leuven for IDO-grant IDO/00/001,
GOA 2001/2, and postdoctoral fellowships to W.Q. and M.B.
The Fonds Voor Wetenschappelijk Onderzoek-Vlaanderen is
thanked for grant G.0320.00. The continuing support through
IAP-V-03 is gratefully acknowledged. Ms. A. Stefan is thanked
for technical help with the single-photon timing measurements.
to estimate pKa and n values, using emission spectra at λex
)
496 nm (λem ) 529, 539, and 549 nm) and excitation spectra at
λem ) 560 nm (λex ) 501 and 521 nm). The pKa values obtained
from the emission spectra are 2.46, 2.42, and 2.37, respectively.
The corresponding pKa values from the excitation spectra are
2.37 and 2.24, respectively. These results show that the pKa
values are independent of the wavelengths used. The average
estimated pKa value of the ammonium form of 5 is summarized
in Table 3. All estimated n values are indicative of a 1:1 binding
between 5 and H+.
References and Notes
(1) (a) Desvergne, J.-P., Czarnik, A. W., Eds. Chemosensors of Ion
and Molecule Recognition; Kluwer: Dordrecht, The Netherlands, 1997. (b)
Haugland, R. P. Handbook of Fluorescent Probes and Research Chemicals,
9th ed.; Molecular Probes, Inc.: Eugene, OR, 2002.
The fluorescence excitation and emission spectra of 6 as a
function of pH (between pH 1.12 and 4.32) are analogous to
those of 5 shown in parts b and c of Figure 9, respectively. The
intensities of the emission (with λmax ) 528 nm) and excitation
spectra (with λmax ) 511 nm) increase at lower pH without any
appreciable spectral shift. BODIPY analogue 6 has a lower φf
value (0.008 at low pH between 1.12 and 2.93, Table 3) than
5. The pKa values obtained from the fluorimetric titrations of 6
indicate that the average estimated pKa value of the ammonium
form of 6 (4.15) is about 1.8 units higher than that measured
for 5 (2.37).
(2) Treibs, A.; Kreuzer, F.-H. Liebigs Ann. Chem. 1968, 718, 208-
223.
(3) BODIPY is a registered trademark of Molecular Probes, Inc.,
Eugene, OR.
(4) (a) Kollmannsberger, M.; Gareis, T.; Heinl, S.; Breu, J.; Daub, J.
Angew. Chem., Int. Ed. Engl. 1997, 36, 1333-1335. (b) Karolin, J.;
Johansson, L. B.-Å.; Strandberg, L.; Ny. T. J. Am. Chem. Soc. 1994, 116,
7801-7806. (c) Pavlopoulos, T. G.; Shah, M.; Boyer, J. H. Appl. Opt. 1988,
27, 4998-4999. (d) Johnson, I. D.; Kang, H. C.; Haugland, R. P. Anal.
Biochem. 1991, 198, 228-237.
(5) (a) Chang, T. C.; Kuo, C.-T.; Chiang, C. C.; Cheng, J.-Y.; Yan,
C.-S.; Peck, K. Phys. Chem. Chem. Phys. 1999, 1, 3783-3787. (b) Yang,
Y.-P.; Kuo, C.-T.; Yan, C.-S.; Lin, K.-C.; Huang, W.-C.; Chang, T.-C. Phys.
Chem. Chem. Phys. 2000, 2, 5271-5274.
Finally, from Figure 10b and c, it is clear that compound 7
does not show any spectral shift upon protonation; there is only
an enhancement of the fluorescence emission and excitation
intensities with increasing H+ concentration, as shown for 5.
The fluorescence quantum yield is the lowest of the three studied
pH sensitive compounds (0.001 in the pH range from 1.05 to
1.39). The average estimated pKa value of the ammonium form
of 7 is the lowest (1.47). The pKa values of 6 (pKa 4.15) > 5
(pKa 2.37) > 7 (pKa 1.47) are in accord with the electron
releasing effect of the alkyl substituents on the nitrogen of
tertiary amines 5-7 (piperidino > NMe2 > morpholino).26 This
clarifies also why the LE transition in the deprotonated form
of 7 will be at the longest wavelength of the series 5-7 while
its CT transition will be at the shortest wavelengths. The
(6) (a) Lo´pez Arbeloa, T.; Lo´pez Arbeloa, F.; Lo´pez Arbeloa, I.; Garc´ıa-
Moreno, I.; Costela, A.; Sastre, R.; Amat-Guerri, F. Chem. Phys. Lett. 1999,
299, 315-321. (b) Costela, A.; Garc´ıa-Moreno, I.; Gomez, C.; Sastre, R.;
Amat-Guerri, F.; Liras, M.; Lo´pez Arbeloa, F.; Ban˜uelos Prieto, J.; Lo´pez
Arbeloa, I. J. Phys. Chem. A 2002, 106, 7736-7742. (c) Lo´pez Arbeloa,
F.; Ban˜uelos Prieto, J.; Mart´ınez Mart´ınez, V.; Arbeloa Lo´pez, T.; Lo´pez
Arbeloa, I. ChemPhysChem 2004, 5, 1762-1771.
(7) Kollmannsberger, M.; Rurack, K.; Resch-Genger, U.; Daub, J. J.
Phys. Chem. A 1998, 102, 10211-10220.
(8) Ban˜uelos Prieto, J.; Lo´pez Arbeloa, F.; Mart´ınez Mart´ınez, V.;
Arbeloa Lo´pez, T.; Amat-Guerri, F.; Liras, M.; Lo´pez Arbeloa, I. Chem.
Phys. Lett. 2004, 385, 29-35.
(9) Wagner, R. W.; Lindsey, J. S. Pure Appl. Chem. 1996, 68, 1373-
1380.
(10) Olmsted, J. J. Phys. Chem. 1979, 83, 2581-2584.
(11) (a) O’Connor, D. V.; Phillips, D. Time-correlated Single Photon
Counting; Academic Press: New York, 1984. (b) Boens N. In Luminescence

7384 J. Phys. Chem. A, Vol. 109, No. 33, 2005
Qin et al.
Techniques in Chemical and Biochemical Analysis; Baeyens, W. R. G., De
Keukeleire, D., Korkidis, K., Eds.; Marcel Dekker: New York, 1991; pp
21-45.
(22) (a) Grabowski, Z. R.; Rotkiewicz, K.; Rettig, W. Chem. ReV. 2003,
103, 3899. (b) Jurczok, M.; Plaza, P.; Rettig, W.; Martin, M. M. Chem.
Phys. 2000, 256, 137. (c) Mac, M.; Kwiatkowski, P.; Pischel, U. Chem.
Phys. Lett. 2002, 357, 440. (d) Onkelinx, A.; De Schryver, F. C.; Viaene,
L.; Van der Auweraer, M.; Iwai, K.; Yamamoto, M.; Ichikawa, M.;
Masuhara, H.; Maus, M.; Rettig, W. J. Am. Chem. Soc. 1996, 118, 2892.
(e) Herbich, J.; Kapturkiewicz, A. Chem. Phys. 1991, 158, 143.
(12) Maus, M.; Rousseau, E.; Cotlet, M.; Schweitzer, G.; Hofkens, J.;
Van der Auweraer, M.; De Schryver, F. C.; Krueger, A. ReV. Sci. Instrum.
2001, 72, 36-40.
(13) Crovetto, L.; Orte, A.; Talavera, E. M.; Alvarez-Pez, J. M.; Cotlet,
M.; Thielemans, J.; De Schryver, F. C.; Boens, N. J. Phys. Chem. B 2004,
108, 6082-6092.
(23) (a) Van Haver, P.; Helsen, N.; Depaemelaere, S.; Van der Auweraer,
M.; De Schryver, F. C. J. Am. Chem. Soc. 1991, 113, 6849. (b) Verhoeven,
J. W.; Scherer, T.; Wegewijs, B.; Hermant, R. M.; Jortner, J.; Bixon, M.;
De Paemelaere, S.; De Schryver, F. C. Recl. TraV. Chim. Pays-Bas 1995,
114, 443. (c) Onkelinx, A.; De Schryver, F. C.; Schweitzer, G.; Miyasaka,
H.; Van der Auweraer, M.; Asahi, T.; Masuhara, H.; Fukumura, H.;
Yashima, A.; Iwai, K. J. Phys. Chem. A 1997, 101, 5054. (d) Tominaga,
K.; Walker, G. C.; Kang, T. J.; Barbara, P. F.; Fonseca, T. J. Phys. Chem.
1991, 95, 10485. (e) Herbich, J.; Kapturkiewicz, A. Chem. Phys. 1991,
158, 143.
(14) Van den Zegel, M.; Boens, N.; Daems, D.; De Schryver, F. C.
Chem. Phys. 1986, 101, 311-335.
(15) Kowalczyk, A.; Boens, N.; Van den Bergh, V.; De Schryver, F. C.
J. Phys. Chem. 1994, 98, 8585-8590.
(16) Cielen, E.; Tahri, A.; Ver Heyen, A.; Hoornaert, G. J.; De Schryver,
F. C.; Boens, N. J. Chem. Soc., Perkin Trans. 2 1998, 1573-1580.
(17) (a) Pevenage, D.; Corens, D.; Van der Auweraer, M.; De Schryver,
F. C. Bull. Chem. Soc. Belg. 1997, 106, 56. (b) Noukakis, D.; Van der
Auweraer, M.; Toppet, S.; De Schryver, F. C. J. Phys. Chem. 1995, 99,
11860.
(24) (a) Boens, N.; Andriessen, R.; Ameloot, M.; Van Dommelen, L.;
De Schryver, F. C. J. Phys. Chem. 1992, 96, 6331. (b) Boens, N.; Van
Dommelen, L.; Ameloot, M. Biophys. Chem. 1993, 48, 301.
(18) Verbeek, G.; Depaemelaere, S.; Van der Auweraer, M.; De
Schryver, F. C.; Vaes, A.; Terrell, D.; De Meutter, S. Chem. Phys. 1993,
176, 195.
(25) (a) Boens, N.; Ameloot, M.; Hermans, B.; De Schryver, F. C.;
Andriessen, R. J. Phys. Chem. 1993, 97, 799. (b) Van Dommelen, L.; Boens,
N.; Ameloot, M.; De Schryver, F. C.; Kowalczyk, A. J. Phys. Chem. 1993,
97, 11738.
(19) (a) Lippert, E. Z. Naturforsch., A: Phys. Sci. 1955, 10, 541. (b)
Mataga, N.; Kaifu, Y.; Koizumi, M. Bull. Chem. Soc. Jpn. 1955, 28, 690.
(20) (a) Shen, Z.; Ro¨hr, H.; Rurack, K.; Uno, H.; Spieles, M.; Schulz,
B.; Reck, G.; Ono, N. Chem.sEur. J. 2004, 10, 4853. (b) Picou, C. L.;
Stevens, E. D.; Shah, M.; Boyer, J. H. Acta Crystallogr. 1990, C46, 1148.
(21) Vranken, N.; Jordens, S.; De Belder, G.; Lor, M.; Rousseau, E.;
Schweitzer, G.; Toppet, S.; Van der Auweraer, M.; De Schryver F. C. J.
Phys. Chem. A 2001, 105, 10196.
(26) Callan, J. F.; de Silva, A. P.; Ferguson, J.; Huxley, A. J. M.;
O’Brien, A. M. Tetrahedron 2004, 60, 11125-11131.
(27) Werner, T.; Huber, C.; Heinl, S.; Kollmannsberger, M.; Daub, J.;
Wolfbeis, O. S. Fresenius J. Anal. Chem. 1997, 359, 150-154.