A first approach toward understanding the targeted design of molecular photoacoustic contrast agents (MPACs) is presented. Optical and photoacoustic Z-scan spectroscopy was used to identify how nonlinear (excited-state) absorption contributes to enhancing the photoacoustic emission of the curcuminBF2 and bis-styryl (MeOPh)2BODIPY dyes relative to Cy3.
A first approach toward understanding the targeted design of molecular photoacoustic contrast agents (MPACs) is presented. Optical and photoacoustic Z-scan spectroscopy was used to identify how nonlinear (excited-state) absorption contributes to enhancing the photoacoustic emission of the curcuminBF2 and bis-styryl (MeOPh)2BODIPY dyes relative to Cy3.
Combining
the advantages of
both ultrasound and optical imaging, photoacoustic tomography (PAT)[1−5] and photoacoustic microscopy (PAM)[6−9] are emerging as highly promising imaging
alternatives. Based upon the classical photoacoustic (PA) effect,
photoacoustic imaging relies upon a materials generation of acoustic
waves in response to absorption of electromagnetic radiation.[10] A major advantage of PAT and PAM over their
fluorescence counterparts is that the output acoustic waves are far
less susceptible to scatter than optical waves, allowing for deeper
penetration, which is particularly advantageous during photoacoustic
tomography (PAT) in vivo.[5] However, application
of both PAT and PAM is currently limited due to a lack of available
contrast agents.[11] A common assumption
is that any dye with a low fluorescence quantum yield will make for
a suitable PA contrast agent.[12,13] In this respect, cyanine
dyes have been highly studied as molecular photoacoustic contrast
agents (MPACs) with metallic and polymeric nanodimensional materials
also attracting much interest of late.[12,14−21] Contrary to this assumption, we aim to demonstrate how a strongly
fluorescent bis-styryl (MeOPh)2BODIPY dye (Φfl = 0.719; 1τ = 5.41 ns) can display an enhanced
PA signal, far exceeding that of the Cy3 (Φfl = 0.025; 1τ = 0.28 ns) cyanine dye. More importantly, with respect
to the future design of efficient MPACs, an excited-state sequential
absorption mechanism is identified as responsible for this PA enhancement,
highlighting the advantage of a long-lived S1 excited state
combined with a high quantum yield to facilitate a strong PA emission.To aid in this study the “naked” BODIPY dye, lacking
3,5-styryl substituents, and the curcuminBF2 dyes were
also investigated (Figure 1). BODIPY and curcuminoid
dyes are deemed excellent candidates for MPACs due to their ease of
synthetic functionalization and their strong, visible, tunable absorption.
Figure 1
Molecular
structures of the BODIPY, (MeOPh)2BODIPY,
curcuminBF2, and Cy3 chromophores.
Molecular
structures of the BODIPY, (MeOPh)2BODIPY,
curcuminBF2, and Cy3 chromophores.CurcuminBF2 in particular was chosen due to its
structural
similarity to that of the bis-styryl (MeOPh)2BODIPY system.
Their π-conjugated styryl arms introduce the potential for a
large excited-state structural volume change and PA response following
photoexcitation.[22,23] Indeed, curcuminBF2 has a more pronounced full width at half-maximum (fwhm = 3053 cm–1) indicating a greater access to its S1 vibrational states relative to BODIPY (fwhm = 807 cm–1), hinting on a greater potential for nonradiative decay and PA response
(Figure 2; Table S1, Supporting
Information).
Figure 2
Electronic absorption spectra (solid) and fluorescence
emission
spectra (dashed) recorded in acetonitrile (ε(MeOPh)2BODIPY was determined in 9:1 v/v acetonitrile/dichloromethane due to poor
solubility).
Electronic absorption spectra (solid) and fluorescence
emission
spectra (dashed) recorded in acetonitrile (ε(MeOPh)2BODIPY was determined in 9:1 v/v acetonitrile/dichloromethane due to poor
solubility).Similarly, its broad
and weak fluorescence (Φfl = 0.048; 1τ
= 0.72 ns) combined with the order
of magnitude Stokes shift enhancement for curcuminBF2 (3322
cm–1) also suggests this is a more viable S1 → S0 PA emitter relative to the rigid BODIPY
system (Φfl = 0.306; 1τ = 1.98 ns;
Stokes shift = 392 cm–1). Introduction of the p-methoxy styrylbenzene substituents in the (MeOPh)2BODIPY dye realizes an extended π-conjugation in both
its HOMO and LUMO levels responsible for the lowest energy S0 → S1 electronic transition (λmax 640 nm; ε = 4.34 × 104 M–1 cm–1), similar to the curcuminBF2 and
Cy3 dyes (viz. DFT analysis; Figure S4, Supporting
Information). However, comparable S0 → S1 oscillator strengths and full-width at half-maxima of the
BODIPY and (MeOPh)2BODIPY dyes (fwhm = 807 and 847 cm–1, respectively) indicate little divergence between
their ground and excited state geometries. This observation suggests
that analysis of the S1 and S0 electronic states
alone, viz UV/vis absorption and fluorescence emission spectroscopy,
is insufficient to completely explain the enhanced PA emission of
(MeOPh)2BODIPY.To gain a deeper understanding of
the photophysical pathway responsible
for an enhanced PA emission, beyond qualitative analysis of UV/vis
electronic absorption and fluorescence emission spectra, we used the
optical and photoacoustic Z-scan techniques while concurrently monitoring
the fluorescence signal (Scheme S1, Supporting
Information).[24,25] This allows correlation of both
the fluorescence and acoustic response to the linear/nonlinear optical
absorption properties of each dye while incrementing the laser fluence
in a controlled manner. Optical Z-scan experiments confirm a strong
nonlinear (reverse saturable) absorption behavior of (MeOPh)2BODIPY consistent with sequential absorption from its longer lived
S1 excited state (Figures 3 and
S6, Supporting Information). Decisively,
the strong similarity of Z-scan optical absorption and PA emission
profiles for each dye unambiguously demonstrates the correlation of
an enhanced PA emission with the nonlinear absorption of incident
photons (Figures 3 and S7, Supporting Information).
Figure 3
Relative nonlinear absorption (left) and
PA emission (right) as
a function of laser fluence for the BODIPY, (MeOPh)2BODIPY,
and curcuminBF2 dyes recorded at λexc =
532 nm. For clarity, Cy3 and standard crystal violet plots are provided
in the Supporting Information (Figure S7).
Relative nonlinear absorption (left) and
PA emission (right) as
a function of laser fluence for the BODIPY, (MeOPh)2BODIPY,
and curcuminBF2 dyes recorded at λexc =
532 nm. For clarity, Cy3 and standard crystal violet plots are provided
in the Supporting Information (Figure S7).Assuming Kashas rule[26] is still obeyed,
this would imply that an Sn → S1 nonradiative
decay is responsible for the observed enhancement in PA signal, followed
by typical radiative versus nonradiative competition for the S1 → S0 transition (Figures S8 and S9, Supporting Information). CurcuminBF2 displays a comparable but slightly weaker nonlinear absorption and
PA response, possibly due to its shorter lived excited state lifetime;
however, a weaker excited state absorption coefficient (yet to be
determined) may also play a role here. Contribution of sequential
ground + excited state absorption (as opposed to concerted two-photon
absorption) to the nonlinear optical Z-scan behavior has been confirmed
by dependence of the nonlinear absorption coefficient (β) on
the on-axis laser pulse intensity.[27] The
excited state absorption mechanism of (MeOPh)2BODIPY and
curcuminBF2 was here further confirmed by the addition
of 1 M 2,4-dinitrochlorobenzene (DNCB) acting as an excited-state
quencher. Steady state and time-resolved fluorescence show quantitative
quenching of their excited states in the presence of DNCB precluding
sequential absorption of a second photon (Figure 4a) and reducing their PA emission to that of a linear absorbing
chromophore (Figure 4b) akin to the PA standard
crystal violet dye. Crystal violet was identified here as an ideal
PA reference due to its strong absorbance (Figure S2, Supporting Information) at the operating laser
wavelength (λexc = 532 nm), rapid nonradiative relaxation
(1τ ≈ 6 ps),[28] linear
absorption, and PA response over a wide laser fluence range (Figures
S6 and S7, Supporting Information).
Figure 4
Quenching
studies of (MeOPh)2BODIPY in 9:1 v/v acetonitrile/dichloromethane
with 1 M DNCB monitored by (a) steady-state and time-resolved fluorescence,
(b) PAZ-scan, and (c) photoacoustic tomography (PAT).
Quenching
studies of (MeOPh)2BODIPY in 9:1 v/v acetonitrile/dichloromethane
with 1 M DNCB monitored by (a) steady-state and time-resolved fluorescence,
(b) PAZ-scan, and (c) photoacoustic tomography (PAT).In comparison, the Cy3 and BODIPY dyes display
saturable absorption
behavior and weak PA emission (Figures S6 and S7, Supporting Information) consistent with ground state bleaching
and negligible excited state absorption at 532 nm. Ultimately a true
comparison of the PA response of each dye is to compare their parallel
performance by photoacoustic tomography (PAT). PAT imaging at relatively
high laser fluence (366 mJ cm–2) shows a similar
trend to PAZ-scan experiments with (MeOPh)2BODIPY and curcuminBF2 showing the strongest contrast due to excited state absorption,
BODIPY and Cy3 showing the poorest contrast due to ground state bleaching,
and crystal violet showing an intermediate contrast due to its linear
optical/photoacoustic response (Figure 5).
Figure 5
PAT image
recorded at a laser fluence of 366 mJ cm–2 (λexc = 532 nm; dimension = 26.40 mm × 6.65
mm). The color scale represents the normalized acoustic intensity.
PAT image
recorded at a laser fluence of 366 mJ cm–2 (λexc = 532 nm; dimension = 26.40 mm × 6.65
mm). The color scale represents the normalized acoustic intensity.Remarkably, with a 3-fold increase
in laser fluence (100 vs 300
J cm–2), the excited state absorption capability
of (MeOPh)2BODIPY at 532 nm results in a 13-fold enhancement
in PA emission. In identical conditions, curcuminBF2 shows
a reasonable 5-fold PA enhancement, whereas crystal violet, BODIPY,
and Cy3 show a linear correlation to laser fluence. Considering the
recommended American National Standards Institute (ANSI) maximum permissible
exposure (MPE) limits of 20 mJ cm–2 at 532 nm, these
systems are currently not applicable for nonlinear PAT in vivo imaging.
However, as a proof-of-concept, they do merit future investigations
in the NIR region where higher MPE limits are accessible. At low laser
fluence (20 mJ cm–2) excited-state absorption is
minor, resulting in negligible difference between the PA contrast
of each dye (Figure S16, Supporting Information). This is consistent with the identical linear absorption coefficient
(α) of solutions employed for PAT experiments and the assumption
of identical Gruneisan coefficient of the medium for all samples.
Importantly, the use of laser fluences in the range of >100 mJ
cm–2 is common practice for in vitro PAM applications.[29,30] The future development of efficient and biocompatible MPACs therefore
holds great promise toward high resolution multiphoton PAM imaging
applications.[31,32]In conclusion, optical
and photoacoustic characterization has been
performed for a series of BODIPY, curcumin, cyanine, and crystal violet
dyes using optical and photoacoustic Z-scan experiments. (MeOPh)2BODIPY and curcuminBF2 show promise as efficient
PA emitters. BODIPY and Cy3 show weak PA emission due to ground-state
bleaching, whereas crystal violet displays a linear response due to
its short-lived excited state. A combination of quenching studies
with standard fluorescence, optical, and PAZ-scan techniques has pointed
to the role of sequential excited-state absorption followed by rapid
Sn → S1 nonradiative decay as being responsible
for this PA emission enhancement. On the basis of these observations,
the criteria for identification and design of efficient MPACs can
be redefined as requiring (i) a strong vis–NIR absorption,
(ii) a long-lived S1 excited state facilitating excited-state
absorption, (iii) a large excited-state absorption extinction coefficient,
and (iv) rapid Sn → S1 nonradiative decay.
In essence these criteria suggest that efficient optical-limiting
materials should make for desirable PA emitters. Considering the synthetic
versatility and electronic tunability of both the curcumin and BODIPY
class of molecules, these findings should inspire a new approach toward
the future design of MPACs, particularly for PAM applications where
relatively high laser fluences are accessible.
Authors: Elizabeth Huynh; Jonathan F Lovell; Brandon L Helfield; Mansik Jeon; Chulhong Kim; David E Goertz; Brian C Wilson; Gang Zheng Journal: J Am Chem Soc Date: 2012-08-01 Impact factor: 15.419
Authors: Samit Guha; Gillian Karen Shaw; Trevor M Mitcham; Richard R Bouchard; Bradley D Smith Journal: Chem Commun (Camb) Date: 2016-01-04 Impact factor: 6.222
Authors: Sheryl Roberts; Chrysafis Andreou; Crystal Choi; Patrick Donabedian; Madhumitha Jayaraman; Edwin C Pratt; Jun Tang; Carlos Pérez-Medina; M Jason de la Cruz; Willem J M Mulder; Jan Grimm; Moritz Kircher; Thomas Reiner Journal: Chem Sci Date: 2018-05-18 Impact factor: 9.825
Authors: Mina Nazari; Min Xi; Sarah Lerch; M H Alizadeh; Chelsea Ettinger; Hisashi Akiyama; Christopher Gillespie; Suryaram Gummuluru; Shyamsunder Erramilli; Björn M Reinhard Journal: Sci Rep Date: 2017-09-20 Impact factor: 4.996