Near IR emitting BODIPY fluorophores with mega-stokes shifts

Article abstract of DOI:10.1039/c2cc31150j

A novel Near Infra-Red emitting BODIPY derivative is presented which exhibits the largest Stokes shift thus far reported for a BODIPY compound.

Full text of DOI:10.1039/c2cc31150j

View Online / Journal Homepage / Table of Contents for this issue
Dynamic Article Links
ChemComm
Cite this: Chem. Commun., 2012, 48, 5617–5619
www.rsc.org/chemcomm
COMMUNICATION
Near IR emitting BODIPY fluorophores with mega-stokes shiftsw
Aaron Martin,a Conor Long,b Robert J. Forsterab and Tia E. Keyes*ab
Received 15th February 2012, Accepted 29th March 2012
DOI: 10.1039/c2cc31150j
A novel Near Infra-Red emitting BODIPY derivative is presented
which exhibits the largest Stokes shift thus far reported for a
BODIPY compound.
acceptor and interesting because of its H-bond capability
which has been applied in DNA sensing.13 Compound 4
exhibits both a NIR emission and a mega-stokes shift of
approximately 185 nm. To our knowledge, this is the largest
Stokes shift ever reported for a BODIPY derivative.
The 4,4-difluoro-4-bora-3a,4a-diaza-indacene (BODIPY) deriva-
tives are a synthetically versatile class of fluorophore.1,2 They are
of particular interest in sensing and imaging because of their
notably high fluorescence quantum yields in the visible spectral
region and also their photostability.3,4 Indeed several BODIPY
derivatives have been commercialized for biological labelling.
Ideally, in bioimaging applications the dye should emit within
the biological window (650–900 nm) where light scattering,5
auto-fluorescence, and tissue absorption are minimised. The
simple BODIPY core emits ca. 500 nm, and strategies to
induce bathochromic shifts in these dyes have been developed
which have focussed on attachment of secondary units at the
pyrrole,6,7 meso,8 and N-ortho positions.9
Scheme 1 shows the synthetic steps leading to the novel
BODIPY derivatives 4 and 5. The p electron conjugation of the
central pyrrole rings is extended by attachment of dimethylamino
aryl groups to the 6 and 2 positions of the BODIPY core and a
naphthyridine substituent at the 8 position. The detailed
synthesis and structural characterisation is presented in the
ESI.w Briefly, bromonaphthyridine was obtained by treating
6-methoxypyridin-3-amine with diethyl ethoxymethylenemalonate
as reported previously.14 A Suzuki-Miyaura protocol15 was
Less widely explored for BODIPY derivatives is the imple-
mentation of a Stokes shift i.e. a separation between the
absorbance and emission maxima in the spectral properties of
the derivative.10 Indeed, although mega Stokes shifts (Stokes
shift exceeding 100 nm) are relatively common in luminescent
inorganic complexes such as ruthenium polypyridyl complexes,
which are formally phosphorescent,11 shifts of comparable magni-
tude are unprecedented in BODIPY derivatives. Mega-Stokes
shifted luminophores are nonetheless of significant value in
imaging and sensing applications. They eliminate the self-
quenching prevalent in high quantum yield fluorophores which
restricts their use at concentrations as low as 10ꢀ7 molar. Self-
quenching is particularly problematic in applications which
require multiple labels, for example conjugation of protein or
DNA or localisation of probes in sub-cellular structures such as
organelles or membranes. This is particularly true in quanti-
tative fluorescence sensing applications. Furthermore, as we
have recently demonstrated, mega Stokes shifted luminophores
have application in multimodal imaging.12
In this contribution, we describe two novel BODIPY push-
pull compounds incorporating a naphthyridine acceptor site
and dimethylamino aryl linked BODIPY donor (4 and 5,
shown in Scheme 1). The naphthyridine unit is a well known
Scheme 1 Synthetic route to NIR BODIPY dyes. (i) 2-fluoro-4-
formylphenylboronic acid, K2CO3, Pd(dppf)Cl2. DCM, dioxane/H2O
(3 : 1), 45 min, 79%; (ii) 2,4-dimethyl-1H-pyrrole, TFA, DCM, 4h;
(iii) tetrachlorobenzoquinone, Et3N. BF3ꢁOEt2, 12 h, 58%; (iv) 1,10-
azobis(cycloheaxanecarbonitrile), N-bromosuccinimide, CCl4, 4 h, 22%;
(v) 4-dimethylaminophenylboronic acid, K2CO3, Pd(dppf)Cl2ꢁDCM,
dioxane/H2O (3 : 1), 45 min, (30%, 4) and (24%, 5).
a Biomedical Diagnostics Institute, Dublin City University,
Glasnevin, Dublin 9, Ireland
b School of Chemical Sciences, Dublin City University, Glasnevin,
Dublin 9, Ireland. E-mail: Tia.Keyes@dcu.ie
w Electronic supplementary information (ESI) available: Experimental,
and DFT calculations. See DOI: 10.1039/c2cc31150j
c
This journal is The Royal Society of Chemistry 2012
Chem. Commun., 2012, 48, 5617–5619 5617

View Online
then followed combining 4-formylarylboronic acid (1.5 equiv)
with Pd(dppf)Cl2ꢁDCM (4 mol%) as catalyst and K2CO3
(2 equiv) as a base in aqueous dioxane/H2O (3 : 1) which
was heated to reflux for ca. 45 min (Scheme 1). This gave
8-(4-formylaryl)-2-methoxynaphthyridine 1 in high yield.16
Typically reactions came to completion with no unwanted
by-products (by TLC). The desired product was isolated by
flash column chromatography (hexane : EtOAc 4 : 1) as a
yellow oil and recrystallised from pentane to yield 1 as a white
the complex insensitive to O2. The quantum yield of emission
is 0.06 (ꢃ0.003), which is considerably lower than conven-
tional BODIPY compounds. However, it is significantly higher
than many luminescent metal complexes, such as the Ru(II)
polypyridyls which exhibit comparable Stokes shifts, but
whose emissions are rarely so red-shifted. The radiative, kf
and non-radiative decay rates, knr for 4 were estimated from
this data to be 1.2 ꢂ 108 and 1.9 ꢂ 109 sꢀ1 respectively.18 The
asymmetric bromo compound, 5, exhibits comparable absorp-
tion to 4, albeit with a blue shift to 523 nm [e, 7 ꢂ 104 Mꢀ1 cmꢀ1
(ꢃ4800)]. However, it exhibits a more intense emission and
longer emission lifetime [f = 0.14 (ꢃ0.008), t = 1.9 ns]. This
data yields kf and knr values of 7.4 ꢂ 107 and 4.5 ꢂ 108 sꢀ1
respectively for 5. Interestingly, unlike 4, where strong correl-
ation exists between the absorption and excitation spectra, for
5, the maximum of the absorption spectra is red shifted by
roughly 20 nm compared with the excitation spectrum. The
emission and excitation spectra for 4 and 5 were studied over
the concentration range 10ꢀ7 to 10ꢀ5 M and found not to
change, indicating this behaviour is not due to aggregation or
excimer formation. This suggests that the red end of the
absorbance spectrum contains contributions from an optical
transition that does not yield an emitting state in 5.
solid.
A Knoevenagel reaction with 2,4-dimethylpyrrole
produced 2 which was then purified on silica gel (20%
EtOAc:hexane). The insertion of bromine17 at the 2,6 positions
of the BODIPY was achieved by reacting 2 with 1,10-azobis-
(cycloheaxanecarbonitrile) (4 equiv) and N-bromosuccinimide
(4 equiv.) to yield 3. The bromo derivative was then reacted
with p-dimethylaminophenyl boronic acid using the same
Suzuki-Miyuara conditions described above to afford
compounds 4 and 5. The symmetric form 4 was the dominant
product (25 mg, 30%) which was isolated as a dark blue solid,
following purification via flash column chromatography using
hexane/EtOAc, 4 : 1.
Fig. 1(a) and (b) show the absorption (green), excitation and
emission spectra of 4 and 5 respectively in CH2Cl2. BODIPY is
typically characterised by intense and narrow absorption and
fluorescence bands, often with mirror image vibronic structure
and by notably small Stokes shifts. Whereas both 4 and 5
exhibit markedly broad absorption and emission features,
which lack mirror symmetry. Furthermore, both exhibit sub-
stantial Stokes shifts, which in light of the broad FWHM of
the emission maxima are more consistent with charge transfer
transitions. Indeed compound 4 exhibits an absorption at
554 nm [e, 3 ꢂ 103 Mꢀ1 cmꢀ1, (ꢃ 180)] and an emission with
a maximum in dichloromethane at 740 nm. The Stokes shift is
therefore 186 nm. This is, to our knowledge, the largest
reported for a BODIPY derivative and is indeed remarkable,
more generally, among organic fluorophores. 4 exhibits a
fluorescence lifetime of 0.5 ns, such a short lifetime renders
Both compounds were modelled using Density Functional
Theory (DFT), full details of these calculations are provided in
the ESI.w The hydrid functional B3LYP was used along
with the 6-311G(d) basis set.19–22 Structural parameters were
optimised without symmetry or other constraints. Each molecule
was divided into two fragments. The first (Frag 1) consisted of
the boron heterocycle and associated dimethylamino aryl
groups, and the second (Frag 2) consisted of the meso-
substituent i.e. the fluorophenyl link and the naphthyridyl
substituent. The nature of the lowest energy excited states was
classified according to the contribution of the fragment orbitals
to the excited state. The results of these calculations are outlined
in Fig. 2 where the orbitals which are mainly centred on Frag 1
are indicated in red while those centred on Frag 2 are drawn in
blue. For both compounds, the HOMO is centred on Frag 1.
The HOMO is mostly localised on the dimethylaminoaryl
unit on 5 whereas for 4 it is extended over both of dimethyl-
aminoaryl fragments. Interestingly, the LUMO switches from
the naphthyridine fragment, Frag 2, in 4 to Frag 1 in 5.
Calculations show the lowest energy optical transitions for
Fig. 2 The orbital energy levels for the two highest occupied orbitals
and four lowest energy virtual orbitals for 4 and 5. The green arrow
indicates the transition to the lowest energy singlet excited state which
in the case of 4 involves a transition from Frag 1 (red) to Frag 2 (blue),
the equivalent transition in 5 is centred on Frag 2 (red).
Fig. 1 Absorption (green), emission (red) and excitation (red) spectra
of (a) 4 (lex = 553 nm) and (lem = 720 nm) (b) 5 (lex = 550 nm) and
(lem = 585 nm) in CH2Cl2 solution. Excitation and emission spectra
were collected with a slit width of 10 nm for 4 and 5 nm for 5.
c
5618 Chem. Commun., 2012, 48, 5617–5619
This journal is The Royal Society of Chemistry 2012

View Online
In contrast, both DFT and resonance Raman indicate that the
state excited under 473 nm irradiation is likely to be the
HOMO to LUMO+1 transition and from the excitation
spectrum it appears this is the origin of the luminescence.
Overall, we conclude, that the emissive state in both cases
arises from a charge transfer transition between the dimethyl-
aminoaryl substituted boron heterocycle moiety and the
naphthyridyl units in both compounds. However, for 5 this
transition is not the lowest energy optical absorbance.
A Stokes shift for both complexes is observed, but particularly
for 4 which at 185 nm is the largest reported for a 4-Difluoro-
4-bora-3a,4a-diaza-indacene derivative.
This material is based upon work supported by the Science
Foundation Ireland Grant No. 10/CE/B1821 and by the Irish
Programme for Research in Third Level Institutions, Cycle 4,
Ireland’s EU Structural Funds Programmes 2007–2013.
Fig. 3 Resonance Raman spectroscopy of (a) 5 excited at 473 nm,
(b) 5 excited at 532 nm, (c) 4 excited at 532 nm, (d) 4 excited at 473 nm.
Samples were approx. 1% w/w of compound dispersed in KBr.
both compounds have significant charge transfer character. For
5 this transition involves charge transfer from the dimethyl-
aminoaryl group to the BODIPY whereas for 4 charge transfer
occurs between fragment 1 and 2. The energy of these optical
transitions are reasonably well correlated with the DFT
calculations, although the red shift of the optical transition
for 4 with respect to 5 is not reflected in the calculations.
To further elucidate the optical transitions and the disparity
between excitation and absorbance spectra in 5, resonance
Raman spectroscopy of the BODIPY derivatives were investi-
gated, exciting at 473 and 532 nm, Fig. 3, black and blue
respectively. Resonance Raman spectroscopy is uniquely
enabled in these fluorophores by their large Stokes shifts.
473 and 532 nm coincide reasonably well with the maxima
of the absorbance and excitation spectra respectively of 5.
532 nm correlates well with both the excitation and absor-
bance frequencies of 4 as these match. However, 473 nm is
pre-resonant with the absorbance/excitation maxima of 4 and
post-resonant with a higher frequency transition centred
around 438 nm. Correspondingly, the similarities in the spectra
shown in Fig. 3(c) and (d) indicate that both 473 and 532 nm are
resonant with the same transition of 4. Although there are
some changes to relative intensity, all spectral features are
present in both spectra. Vibrational bands characteristic of
both Frag 1 and Frag 2 are resonantly enhanced indicating
charge transfer between these fragments. In contrast, there are
significant differences in the resonance Raman spectra excited
at 473 compared to 532 nm for 5. Under 473 nm excitation,
which is resonant with the excitation spectrum maximum, the
Raman signature is strongly reminiscent of the resonance
Raman spectra shown for 4 consistent with charge transfer
between Frag 1 and Frag 2. However, under 532 nm excita-
tion, i.e. resonance with the absorbance maximum of 5, a
number of key modes observed under 473 nm excitation are
much weaker or missing. Most notably features at 1400 and
1520 cmꢀ1, which are associated, on the basis of DFT calcula-
tions, with the naphthyridyl unit, the features are marked with
arrows. The remaining bands are associated predominantly
with Frag 1 suggesting a charge transfer between the dimethyl-
aminoaryl moiety and the substituted boron heterocycle,
consistent with the HOMO–LUMO transition predicted for 5.
Notes and references
1 R. Ziessel, G. Ulrich and A. Harriman, New J. Chem., 2007,
31, 496.
2 F. Nastasi, F. Puntoriero, S. Campagna, J.-H. Olivier and
R. Ziessel, Phys. Chem. Chem. Phys., 2010, 12, 7392.
3 T. Komatsu, D. Oushiki, A. Takeda, M. Miyamura, T. Ueno,
T. Terai, K. Hanaoka, Y. Urano, T. Minenoc and T. Nagano,
Chem. Commun., 2011, 47, 10055.
4 A. B. Nepomnyashchii, M. Broring, J. Ahrens and A. J. Bard,
¨
J. Am. Chem. Soc., 2011, 133, 8633.
5 (a) V. Ntziachristos, J. Ripoll and R. Weissleder, Opt. Lett., 2002,
27, 333–335; (b) J. Sowell, L. Strekowski and G. Patonay,
J. Biomed. Opt., 2002, 7, 571–575.
6 R. P. Haughland and H. C. Kang, US Patent US4774339, 1988;
F. J. Monsma, A. C. Barton, H. C. Kang, D. L. Brassard,
R. P. Haughland and D. R. Sibley, J. Neurochem., 1989, 52,
1641–1644.
7 L. H. Thoresen, H. Kim, M. B. Welch, A. Burghart and
K. Burgess, Synlett, 1998, 11, 1276–1278.
8 (a) G. Ulrich and R. Ziessel, J. Org. Chem., 2004, 69, 2070;
(b) W. Qin, M. Baruah, M. van der Auweraer, F. C. de Schryver
and N. Boens, J. Phys. Chem. A, 2005, 109, 7371;
(c) T. Venkateshwar Goud, A. Tutar and J. F. Biellmann,
Tetrahedron, 2006, 62, 5084.
9 (a) M. Baruah, W. Qin, R. A. L. Vallee, D. Beljonne, T. Rohand,
´
W. Dehaen and N. Boens, Org. Lett., 2005, 7, 4377–4380;
(b) T. Rohand, W. Qin, N. Boens and W. Dehaen, Eur. J. Org.
Chem., 2006, 4658.
10 M. Brc¸ ring, R. Kruger, S. Link, C. Kleeberg, S. Kohler, X. Xie,
¨
¨
B. Ventura and L. Flamigni, Chem.–Eur. J., 2008, 14, 2976.
11 U. Neugeburger, Y. Pellegrin, N. Moran, W. Signac, R. J. Forster
and T. E. Keyes, Chem. Commun., 2008, 5307.
12 L. Cosgrave, U. Neugebauer, R. J. Forster and T. E. Keyes,
Chem. Commun., 2010, 46, 103.
13 K. Ghosh and T. Sen, J. Inclusion Phenom. Macrocyclic Chem.,
2009, 67, 271.
14 W. H. Miller, M. B. Rouse and M. A. Seefeld, WO Pat.
2006081178, 2006.
15 Z. Dost, S. Atilgan and E. U. Akkaya, Tetrahedron, 2006, 62, 8484.
16 H. Ioannidou, A. Martin, A. Gollner and P. A. koutentis, J. Org.
Chem., 2011, 76, 5113.
17 T. Yogo, Y. Urano, Y. Ishitsuka, F. Maniwa and T. Nagano,
J. Am. Chem. Soc., 2005, 127, 12162–12163.
18 Estimated from; kf = ff/t and knr = (1 ꢀ ff)/t, where ff and t are
the fluorescent quantum yield and lifetime respectively.
19 A. D. Becke, J. Chem. Phys., 1993, 98, 5648–5652.
20 C. T. Lee, W. T. Yang and R. G. Parr, Phys. Rev. B, 1988, 37, 785.
21 A. D. McLean and G. S. Chandler, J. Chem. Phys., 1980, 72,
5639.
22 R. Krishnan, J. S. Binkley, R. Seeger and J. A. Pople, J. Chem.
Phys., 1980, 72, 650.
c
This journal is The Royal Society of Chemistry 2012
Chem. Commun., 2012, 48, 5617–5619 5619