Raman spectroscopy of dipyrrins: Nonresonant, resonant and surface-enhanced cross-sections and enhancement factors

Article abstract of DOI:10.1002/jrs.3093

Detailed studies of the mechanism of surface-enhanced (resonance) Raman spectroscopy (SE(R)RS), and its applications, place a number of demands on the properties of SERS scatterers. With large Raman cross-sections, versatile synthetic chemistry and comple

Full text of DOI:10.1002/jrs.3093

Research Article
Received: 4 August 2011
Revised: 6 September 2011
Accepted: 6 September 2011
Published online in Wiley Online Library: 16 November 2011
(wileyonlinelibrary.com) DOI 10.1002/jrs.3093
Raman spectroscopy of dipyrrins: nonresonant,
resonant and surface-enhanced cross-sections
and enhancement factors
a
b
b
a
Tracey M. McLean, Deidre Cleland, Keith C. Gordon, Shane G. Telfer and
a
Mark R. Waterland *
Detailed studies of the mechanism of surface-enhanced (resonance) Raman spectroscopy (SE(R)RS), and its applications, place
a number of demands on the properties of SERS scatterers. With large Raman cross-sections, versatile synthetic chemistry and
complete lack of fluorescence, free dipyrrins meet these demands but the Raman and SE(R)RS spectroscopy of free dipyrrins is
largely unknown. The first study of the Raman spectroscopy of free dipyrrins is therefore presented in this work. The nonres-
onant Raman, resonant Raman and surface-enhanced Raman spectra of a typical meso aryl-substituted-dipyrrin are reported.
Absolute differential cross-sections are obtained for excitation wavelengths in the near infrared and visible region, in solution
phase and for dipyrrin adsorbed on the surface of silver nanoparticles. Raman enhancement factors for SERRS and resonance
Raman are calculated from the observed differential cross-sections. The magnitudes of the resonantly enhanced cross-sections
are similar to those recently reported for strong SERS dyes such as Rhodamine 6G and Crystal Violet. Free dipyrrins offer the
advantages of existing SERS dyes but without the drawback of strong fluorescence. Free dipyrrins should therefore find appli-
cations in all areas of Raman spectroscopy including fundamental studies of the mechanisms of SERS and bioanalytical and
environmental applications. Copyright © 2011 John Wiley & Sons, Ltd.
Supporting information may be found in the online version of this article.
Keywords: dipyrrin; resonance Raman spectroscopy; SERS; Raman cross-sections; Raman enhancement factors
For the development of reliable SERS applications, the Raman
Introduction
cross-sections and enhancement factors must be known. Since
[
10,11]
Dipyrrins are bispyrrolic compounds first reported by Fischer and
the first report of SERS from single molecules,
there has been
[1]
Orth in 1937. Dipyrrin research has attracted considerable
interest over many years and is currently experiencing a renais-
much debate on the magnitude of SERS enhancement factors. As
demonstrated by Le Ru et al. there are a large number of factors
that contribute to the SERS effect. Putting aside the issue of hot-
spot formation and number of molecules in the scattering volume,
in surface-enhanced resonance Raman scattering (SERRS) it is im-
portant to take into account both enhancement from resonance
with the electronic states of the molecule (the resonance Raman
effect) and enhancement from resonance with the plasmon
modes of the metallic nanostructure (the SERS effect). A further
complication is the presence of resonant enhancement via
electronic transitions between the molecule and metal states.
The vast majority of single-molecule SERS (SM-SERS) experiments
are performed under SERRS conditions. The molecular electronic
resonance is clearly a major factor in the extremely large Raman
cross-sections observed in SM-SERS experiments, but unfortunately,
[2,3]
sance.
The conjugated p system of the free dipyrrin unit, anal-
ogous to porphyrins, provides free dipyrrins and their metallo
complexes with interesting and useful optical properties. Free
dipyrrins and their complexes display promise as light harvest-
[4–7]
[4]
ers,
sensitizers for solar cells, and as components of
[
5,6]
chromophores in metal organic frameworks
and molecular
[7]
materials. Dipyrrins also display versatile structural properties
and various functional groups may be incorporated into the
[
2,8]
dipyrrin unit by substitution on the aryl and/or pyrrole rings.
Because of strongly allowed p–p* transitions, free dipyrrins have
large transition dipole strengths (absorption coefficients of 20
–1
–1
0
00 M cm and oscillator strengths of 0.5), which make them
ideal candidates for resonantly enhanced Raman spectroscopies
that rely on enhancement through coupling to dipole-allowed elec-
tronic transitions such as resonant Raman and surface-enhanced
Raman scattering (SERS). Surprisingly, there are no publications
involving the Raman spectroscopy of free dipyrrins. The SERS
spectra of dipyrrin–BF complexes have been reported and the only
reported Raman cross-sections for dipyrrin compounds are for the
recently reported ruthenium complex. Boron complexes of
dipyrrins are widely known for their excellent emission properties
*
Correspondence to: M. R. Waterland, MacDiarmid Institute for Advanced
Materials and Nanotechnology, Institute of Fundamental Sciences, Massey
University, Private Bag 11 222, Palmerston North, New Zealand 4410.
E-mail: M.Waterland@massey.ac.nz
2
[9]
a MacDiarmid Institute for Advanced Materials and Nanotechnology, Institute of
Fundamental Sciences, Massey University, Palmerston North, New Zealand
TM
and the success of BODIPY
complexes in fluorescence
applications in particular, may have obscured the promising
Raman properties of the free dipyrrin unit.
b Department of Chemistry and MacDiarmid Institute for Advanced Materials
and Nanotechnology, University of Otago, PO Box 56, Dunedin, New Zealand
J. Raman Spectrosc. 2011, 42, 2154–2164
Copyright © 2011 John Wiley & Sons, Ltd.

Raman spectroscopy of dipyrrins
the best-performing SM-SERS scatterers (e.g. Rhodamine 6G (R6G),
Crystal Violet, boron-dipyrromethene dyes) have excellent fluores-
cence quantum yields, which largely precludes the measurement
of the resonance Raman cross-sections in solution. Recently,
Computational and Experimental Procedures
Computational methods
[19]
All calculations were performed using GAUSSIAN09. DFT was used,
[12]
femtosecond stimulated Raman spectroscopy (FSRS),
and the
[20]
with the Becke’s three-parameter exchange functional (B3)
in
[13]
SERS effect itself have been used to estimate resonant Raman
cross-sections of R6G in solution. Although FSRS provides accurate
values, it is technically very demanding while using the SERS effect
carries a number of caveats. An advantage of using free dipyrrins for
SERS measurements is the complete absence of fluorescence, which
allows the resonance Raman cross-sections to be measured simply
and accurately, which in turn allows a more accurate assessment
of the enhancement mechanisms in SERS.
combination with the correlation functional of Lee, Yang and Parr
[21]
[22]
(
LYP); O3LYP, which has essentially the same form as B3LYP
[23]
but the OPTX density functional replaces the Becke parameter;
and a new hybrid exchange correlation functional M06. 6-311
[24]
(+)G(2d,2p) basis sets were used for all optimisation and Raman
wavenumber calculations. Self-consistent reaction field methods,
more specifically the polarisable continuum model (CPCM) of Cossi
[25]
et al.
were used to model solvent effects. All geometry opti-
Until recently the concept of enhancement factor was not well
defined and as a result SERS enhancement factors reported in
mizations, wavenumber, and time-dependent calculations
were carried out using the CPCM model. A comparison of the
B3LYP, O3LYP and M06 functionals was carried out using mean
average deviation between the experimental Raman wave-
numbers and the calculated Raman wavenumbers (Fig. S1,
Table S1) and a comparison of the experimental and calculated
absorption spectra (Fig. S2, Table S2). From these evaluations
the B3LYP functional gave the closest agreement between
4
14
.
the literature covered a wide range, from 10 –10
SERS
enhancement factors depend largely on the precise conditions
of the SERS experiment: the substrate (metal surface – gold/silver,
nanoparticles (size)/roughed surface), excitation wavelength
(
proximity to plasmon resonance or analyte resonance), detection
ꢀ
setup (scattering geometry – 135/180 degree backscattering),
properties of the analyte, SERS analyte adsorption properties
[26]
calculated and experimental parameters. GAUSSSUM 2.2.4
(
surface coverage, distance from surface, adsorption orientation),
was used to extract vibrational wavenumbers and Raman
activities from the calculations. The calculated Raman spectrum
is a convolution of a series of d-functions, scaled by the cal-
culated intensities and located at the calculated wavenumbers,
[14,15]
etc.
enhancement seen in SERS were subject to much debate.
Recently, Lombardi and Birke
Similarly, the physical mechanisms responsible for the
[16,17]
have provided a theoretical de-
[18]
scription of surface-enhanced Raman using Albrecht’s approach
and a Gaussian function with full width at half maximum
for resonantly enhanced Raman scattering. Lombardi and Birke
note that in addition to the familiar A-term scattering observed
in molecular resonant Raman spectra, the less common B-term
and C-term resonances can also make substantial contributions
to the enhancements observed in SERS (and SERRS) experiments
via charge-transfer transitions between the molecule and metal
nanoparticle.
–1
(
FWHM) of 5 cm . Raman wavenumbers were scaled using
[27–29]
literature values for each functional and basis set.
The
results of the above calculations were used to analyse and
identify the vibrational wavenumbers, normal modes, and
Raman activities. Calculations for the electronic excited states
were performed using the time-dependent DFT (TD-DFT) calcu-
lations with the B3LYP/6-311(+)G(2d,2p) model. The calculated
absorption spectrum is a convolution of a series of d-functions,
located at the transition wavenumber and scaled by the calcu-
lated oscillator strengths, and a Gaussian with a FWHM value
determined from the FWHM values of the corresponding peaks
of the experimental absorption spectrum.
In this work, we measure the nonresonant, resonant, and
surface-enhanced Raman cross-sections and calculate the en-
hancement factors of a simple dipyrrin system (Scheme 1) follow-
[
15]
ing the recent definition given by Le Ru et al.
We present
assignments of the vibrational modes of dipyrrins using density
functional theory (DFT) calculations of the nonresonant Raman
spectra. We use these assignments, along with time-dependent
DFT calculations to explain the enhancements observed in the
resonant Raman spectra and the surface-enhanced Raman Experimental Procedures
spectra of dipyrrins on silver nanoparticles within the unified
picture of SERS as presented by Lombardi and Birke.
Preparation of dipyrrin ligand: 1
[
30,31]
1
was prepared following literature procedures.
Briefly, the
appropriate aldehyde was condensed with pyrrole via an HCl
catalyzed condensation at room temperature, which was followed
by oxidation of the resulting dipyrromethane with DDQ. Purification
was achieved via column chromatography on deactivated alumina
using CH Cl as the eluent.
2
2
Preparation of silver nanoparticles
Silver nanoparticles were prepared via the trisodium citrate
[
32]
reduction of silver nitrate method of Lee and Meisel.
A 1-mM
stock solution of silver nitrate and 34 mM solution of trisodium
citrate were prepared in Milli-Q water. Five millilitres of the silver
ꢀ
nitrate stock solution was heated to approximately 90 C and
0.1 mL of trisodium citrate was added and the solutions were
Scheme 1. 5-(4-Methoxycarbonylphenyl)-4,6-dipyrrin, 1.
J. Raman Spectrosc. 2011, 42, 2154–2164
heated in a hot water bath for 1 h.
Copyright © 2011 John Wiley & Sons, Ltd.
wileyonlinelibrary.com/journal/jrs

T. M. McLean et al.
[
15]
Preparation of samples for SERS
accurately calculated using density functional techniques.
A
B3LYP/6-311(+)G(2d,2p) calculation (with the CPCM solvent
The conditions required for single molecule SERS of R6G on silver
–32
2
–1
model) of methanol gives a value of 6.9 ꢁ 10 cm sr for
[33]
nanoparticles were used, as reported by Le Ru et al.
0 mM solution of KCl was prepared in Milli-Q water. Of the KCl stock
solution, 0.25 mL was added to a 0.5 mL volume of freshly prepared
Briefly, a
–1
1
064 nm excitation for the 1035 cm band, in very close agreement
4
–32
2
–1
with the value of Moran et al., and we use 7.8 ꢁ 10 cm sr in this
work. The cross-sections obtained by Moran and Kelley and the data
reported here were obtained using similar back-scattering
geometries. Dichloromethane was used as an internal standard
for measurements of resonant Raman cross-sections. The absolute
–6
silver nanoparticles, followed by 0.25 mL of 2.0 ꢁ 10 M solution of
dipyrrin in water, giving a final concentration of 10 mM KCl and
–
7
5
.0 ꢁ 10 M dipyrrin. This procedure consistently produces large
SERS intensities for the cationic R6G (sufficient for single-molecule
SERS) but we have not attempted to adjust these conditions to
optimize the SERS enhancement of dipyrrins.
differential Raman cross-sections of dichloromethane are reported
[40]
in the literature.
The cross-section of sodium sulfate in water
(
concentration = 0.100 M) was used as an external standard for the
[41]
calculation of the SERS cross-sections.
Raman Spectroscopy
Nonresonant Raman spectra (1064 nm) were collected on powder
samples and solutions in methanol (11.2 mM) using a Bruker IFS-55
FT-interferometer bench equipped with an FRA/106 Raman
accessory and using OPUS (version 4.0) software. A Nd:YAG laser
with 1064 nm excitation wavelength was used. An InGaAs diode
Results and Discussion
Absorption spectrum and TD-DFT calculations
The optimised structure of 1 with the atom labels used for mode
assignment is displayed in Fig. 1. The pyrrolic rings are planar
while the phenyl ring makes a torsion angle of 66 with the pyrro-
lic rings. As predicted by Brückner et al.
interactions between the hydrogen atoms (H , H ; H , H ).
(
D424) operating at room temperature was used to detect
ꢀ
Raman photons. Spectra were typically measured using 256
[
42]
–
1
this reduces the steric
scans at a power of 90 mW and a resolution of 4 cm .
Resonance Raman and SERS spectra were collected using
either a Crystallaser (444.2 nm excitation) or a ModuLaser
Stellar-Pro argon laser (457.9 and 488.0 nm excitation) and
25 11
6
18
The position and intensity of the electronic resonance of the
molecule plays a key role in determining the enhancement in
[
43]
[16,17,44–47]
[34–37]
both resonant Raman
and SERS.
The experimental
spectra were collected using a system previously described.
and calculated absorption spectra of 1 in methanol is given in
Fig. 2. The TD-DFT calculations show several configurations that
contribute to the electronic transitions of 1 (Table 1, Fig. S3).
The broad dominant transition at 432 nm is assigned as a p–p*
transition based on the results of the TD-DFT calculation, which
predicts the transition at 408 nm. There are two higher energy
transitions at 300 nm and 231 nm. The calculation predicts a
transition at 339 nm, two closely spaced transitions at 311 and
Excitation power ranged from 6–20 mW at the sample. Solutions
for resonance Raman were prepared in dichloromethane
(
0.461 mM). Raman and Rayleigh scattering light was collected
ꢀ
from the sample cell using either 180 backscattering geometry
or 135 back-scattering geometry. Rayleigh scattering was rejected
ꢀ
using Raman edge filters from Iridian Technologies or notch filters
from Kaiser Optical Systems. The scattered photons were focused
W
onto the entrance slit of an Acton Research SpectraPro 2550i,
313 nm, and 248 nm, which can all be assigned as p–p* type
0.500 m imaging single stage monochromator/spectrograph and
transitions.
detected with a Roper Scientific Spec-10:100B CCD detector
controlled by WINSPEC software. A combination of cyclohexane
and a 50/50 v/v mix of toluene and acetonitrile (ASTM E 1840
Raman Shift Standard) were used for the wavenumber calibration.
The electron density difference maps for the strongest transitions
are shown in Fig. 3. These maps plot regions of charge-density
depletion (light) and accumulation (dark) between the ground
and excited-states. They are a useful guide to interpreting the
resonant Raman spectra, as vibrational modes located in regions
of large charge-density changes should receive the greatest
resonant enhancement. Clearly they are all pꢂp* type transitions.
–1
Wavenumber calibration was accurate to 0.5–1 cm . Resolution
–1
ranged from 2–4 cm depending on spectral position. Peak
–1
positions were reproducible to within 1 cm .
Raman Cross-sections
For nonresonant Raman cross-section measurements in metha-
nol solution, solute absolute differential Raman cross-sections
were obtained relative to solvent cross-sections by
ꢀ
ꢁ
ꢀ
ꢁ
dsRS
dΩ
dsRS
dΩ
Isample creference
¼
:
:
(1)
Ireference csample
sample
reference
using the solvent (methanol) as an internal standard. The differ-
–1
ential cross-section of the 1035 cm band of methanol at
–
32
2
–1
1
064 nm (7.8 ꢁ 10 cm sr ) is obtained from the A-Term para-
[38] [39]
meters reported by Moran et al. Colles and Griffith reported
an absolute differential cross-section for the 1035 cm band of
–1
–
30
2
–1
methanol with 488 nm excitation of 2.5 ꢁ 10 cm sr (Moran
–
30
2
–1
and Kelley’s parameters give 2.3 ꢁ 10 cm sr with 488 nm
excitation). Le Ru et al. showed that the nonresonant Raman
cross-sections of 2-bromo-2-methylpropane (2B2MP) can be
Figure 1. Optimised structure of 1 (DFT/B3LYP 6-311 + G (2d, 2p)) with
atom labels adopted for mode assignments.
wileyonlinelibrary.com/journal/jrs
Copyright © 2011 John Wiley & Sons, Ltd.
J. Raman Spectrosc. 2011, 42, 2154–2164

Raman spectroscopy of dipyrrins
À
Á
2
2
Despite the nonplanarity of the phenyl ring there is also significant
movement of charge-density onto the phenyl ring during the
pꢂp* transition. Similar changes in charge-density over the
phenyl ring have been proposed to explain the enhanced
transannular torsional mode intensity in the resonant Raman
where C is a constant C ¼ p =45e and e ꢃ permittivity
o
o
Â
Ã
ꢂ 12
ꢂ 1
2
i
2
2
of free space = 8.8542ꢁ 10
F m , B ¼ ðh=8p c ꢀu iÞ kg m is
the square of the zero point amplitude for the normal mode in
reduced mass coordinates, ꢀu R is the wavenumber of the Stokes-
Â
Ã
2
o
4
ꢂ1
shifted Raman signal and Ri e m kg
is the Raman scattering
[
9,31]
spectra of the ruthenium complexes of dipyrrin ligands.
It is worth noting that strong exciton coupling is observed in
[
15]
ꢂ1
activity from Gaussian output.
KðTÞ ¼ ½1 ꢂ expðꢂhc ꢀu i=kTÞꢄ
[
48]
accounts for the thermal population of the vibrational state.
Equation (2) gives the Raman cross-section in the gas phase. To
obtain the predicted liquid phase cross-section a local field correc-
metallodipyrrin complexes with multiple dipyrrin ligands.
This may be relevant for SERS studies (see below) as the free
dipyrrins adsorbed on a nanoparticle may be in sufficiently
close proximity for strong electronic interactions and exciton
coupling. Similar effects have been observed with other dyes
that are known to exhibit exciton coupling, e.g. dimer formation
2
4
tion factor, L, is applied where L = [(n + 2)/3] and n is the refractive
index of the solvent.
The experimental nonresonant Raman intensities (in methanol
with 1064 nm excitation) are obtained using the solvent line at
[
46]
in R6G.
–1
1
035 cm as an intensity standard. Only the five strongest dipyrrin
modes (q66, q69, q71/ 72, q82/83, q84) are observed. The Raman cross-
q
sections for these modes are given in Table 2 and Fig. 5 shows the
Cartesian displacement vectors for these modes. The nonresonant
Raman spectrum obtained as a KBr pellet (Fig. 4(b)) displays several
Nonresonant cross-sections
The calculated and experimental nonresonant Raman spectra of
1
are shown in Fig. 4. The calculated wavenumbers have been
scaled by 0.9679. This scaling factor is recommended for the
[
49]
B3LYP/6–311 + G (d,p) basis sets and larger.
The differential cross-sections were determined from the cal-
[
50–54]
culated Raman scattering activities (R
i
)
using Eqn (2), where
the differential cross-section for a vibrational mode with wave-
[
55]
number ꢀu i can be expressed as
dsRS
dΩ
2
i
4
R
¼
CB Ri ꢀu KðTÞ
(2)
408 nm
339 nm
3
13 nm
311 nm
248 nm
Figure 2. Absorption spectra of 1 in methanol. (A) Experimental and (B)
simulated line shape using calculated oscillator strengths obtained from
DFT calculations.
Figure 3. Electron density difference plots for the strongest transitions
of 1. Light represents depletion of electron density and dark represents
accumulation of electron density.
Table 1. Calculated and experimental absorption parameters for 1
Experimental
Calculated
–1
–1
l / nm
e / L mol cm
l / nm
Configuration (% contribution)
Oscillator strength
a
b
4
32
00
31
18 900
408
HOMO ! LUMO (95)
HOMO ! LUMO + 1 (95)
HOMO-3 ! LUMO (97)
0.536
0.105
0.162
0.054
0.365
3
39
313
11
248
3
8 200
3
HOMO-5 ! LUMO (ꢂ23), HOMO-2 ! LUMO (71)
2
15 600
HOMO-7 ! LUMO (50), HOMO-3 ! LUMO + 1 (ꢂ32), HOMO-2 ! LUMO + 1 (ꢂ11)
a
HOMO, highest occupied molecular orbital.
b
LUMO, lowest unoccupied molecular orbital.
J. Raman Spectrosc. 2011, 42, 2154–2164
Copyright © 2011 John Wiley & Sons, Ltd.
wileyonlinelibrary.com/journal/jrs

T. M. McLean et al.
because the nonresonant cross-sections are not typically measured
and the most common SERS dyes have very strong fluorescence
cross-sections that preclude measurement of resonance Raman
spectra. There are several reports of resonant cross-sections for
strongly enhanced Raman scatterers with visible excitation such as
–25
2
–1 [56]
betaine-30 (maximum cross-section 1.6 ꢁ 10 cm sr ), the non-
linear chromophore julolidinyl-n-N,N′-diethylthiobarbituric acid (JTB)
–24
2
–1 [38]
with a maximum cross-section of 7 ꢁ 10 cm sr , the azo dye
–25
2 [57]
Disperse Red (a maximum total cross-section of 6.0 ꢁ 10 cm ),
2
+
–24
2
–1 [58]
and [Ru(bpy)
3
]
(1.0 ꢁ 10 cm sr ), which are similar in magni-
tude to the values reported here for the free dipyrrin systems. The
cross-sections reported here are also in good agreement with the
cross-sections for the ruthenium complex of a meso-aryl-substi-
–
24
2
–1
–1
[9]
tuted-dipyrrin (1.6 ꢁ 10 cm sr for the 409 cm band). Reso-
–1
nant Raman cross-sections for fluorescent dyes, R6G (1510 cm
–24
2
–1
band, 1.5 ꢁ 10 cm sr with 532 nm excitation), and Crystal Violet
–1
–25
2
–1
(
1650 cm band, 6.7 ꢁ 10 cm sr with 568 nm excitation) com-
monly used in SM-SERS have been estimated by using the SERS ef-
Figure 4. Calculated and experimental Raman spectrum of 1 with some
[
13]
normal modes highlighted. (A) Calculated spectrum using DFT/B3LYP
fect to quench spontaneous emission.
FSRS gives a value of
[[49]]
–24
2
–1
–1
6-311 + G (2d,2p). Wavenumbers are scaled by 0.9679.
(B) Solid-state
1
.5 ꢁ 10 cm sr for the 1510 cm band of R6G with 532 nm exci-
nonresonant Raman spectrum as a KBr pellet. (C) Solution phase nonres-
onant Raman spectrum in methanol (S = methanol peaks).
[12]
tation. This suggests that dipyrrins have the requisite electronic
enhancement to be candidates for single-molecule SERS investiga-
tions. The ability to obtain a complete characterization of nonreso-
nant Raman, resonant Raman and SERS is a very attractive feature
of dipyrrin dyes.
other Raman active modes. We obtain nonresonant cross-sections
for the observed bands in Fig. 4(b) by assuming the band at
The strongest modes in the resonance Raman spectra are mode
2 + 83 and mode 69. Electronic structure calculations (TD-DFT)
–1
1576 cm has the same cross-section in the solution and solid
8
phases (a reasonable approximation given the experimental error
associated with the cross-section measurements). Using a dipyrrin
mode as its own internal standard avoids issues with polarization
and concentration when dealing with solid samples. Table 2
displays the calculated and experimental nonresonant Raman
cross-sections for 1, mode assignments and mode descriptions as
the percentage contribution of redundant internal coordinates to
each of the normal mode. The calculated and experimental values
show excellent agreement which provides confidence in the verac-
ity of the electronic structure calculations. As expected, modes
associated with the dipyrrin core dominate the nonresonant
Raman spectrum with some contribution of modes from the
phenyl group.
show that during the pꢂp* transition substantial electron density
changes occur in the pyrrole rings and over the phenyl ring. Mode
83 is a phenyl ring deformation and mode 69 and 82 are pyrrole
ring deformations. Therefore, the strong enhancement of these
modes is consistent with the changes in electron density predicted
by the TD-DFT calculations. The low symmetry of the dipyrrin sys-
tems and the extensive delocalization of the electron density
changes provide an explanation for the enhancement of nearly
all the vibrational modes. Consequently, all enhanced modes are
enhanced via A-term scattering (therefore the ground and
excited pꢂp* state are Frank–Condon coupled). Under Franck–
Condon coupling, the oscillator strength of the electronic transition
and the mode displacements between the ground and excited
states determine the mode intensities. The large intensity of mode
8
3 implies substantial structural reorganization of the phenyl ring in
Resonance Raman, resonance cross-sections and enhance-
ment factors
[
9]
the pꢂp* excited state. As shown in dipyrrin complexes the struc-
tural reorganization of the phenyl ring is mostly associated with
rotation of the ring about the transannular bond. It is appropriate
to make some comparison between the electronic structures and
resonant Raman features of phenyl-substituted dipyrrins and tetra-
phenylporphyrin species. In contrast to the strong enhancement of
the phenyl modes in the dipyrrin spectra reported here, phenyl
modes in tetraphenylporphyrin are only weakly enhanced via reso-
nance with the strongly allowed Soret bands (A-term enhance-
ment), and the enhancement is typically ascribed to intensity
borrowing from nearby pyrrole modes. Weak electronic conju-
gation between the phenyl ring and the porphine p-system and
poor coupling between the phenyl and pyrrole vibrational modes
are the causes of the weak enhancement. However, in the dipyrrin
system the strong enhancement of the phenyl modes and the
@electronic structure calculations both indicate that the phenyl
and pyrrole p-systems interact strongly. This may be due to the
conformational flexibility of the free dipyrrin unit that allows orbital
overlap between the phenyl substituent and the dipyrrin core.
Therefore, substituents on the phenyl ring should have a strong
The resonant Raman excitation wavelengths (444.2 nm, 457.9 nm,
4
88.0 nm) are strongly resonant with the p–p* transition and also
on the red edge of the electronic transition of interest, which
avoids any complication of preresonance enhancement from
the higher energy transitions. As can be seen from Fig. 6 most
features that are in the nonresonant Raman spectrum are also
common to all the resonant Raman spectra.
Table 3 gives the resonant Raman cross-sections and resonant
Raman enhancement factors for 1, for 444.2 nm, 457.9 and
[
59]
4
88.0 nm excitation. At 444.2 nm the resonant Raman cross-sections
3
5
are on the order of 10 –10 times larger than the nonresonant
Raman cross-sections, consistent with resonance with a strongly
allowed pꢂp* transition. As the laser excitation moves away from
the centre of the electronic transition (at 432 nm), the resonant
Raman cross-sections decrease by at least an order of magnitude.
Despite their importance in studies of SERS, with only two notable
[12,13]
exceptions,
there are no reports of resonant Raman enhance-
ment factors for commonly used dyes in SERS experiments, first
wileyonlinelibrary.com/journal/jrs
Copyright © 2011 John Wiley & Sons, Ltd.
J. Raman Spectrosc. 2011, 42, 2154–2164

Raman spectroscopy of dipyrrins
J. Raman Spectrosc. 2011, 42, 2154–2164
Copyright © 2011 John Wiley & Sons, Ltd.
wileyonlinelibrary.com/journal/jrs

T. M. McLean et al.
influence on the electronic properties of dipyrrin compounds and
[
60]
this is indeed the case.
Surface-enhanced Raman spectra
There are a number of factors that affect the magnitude of the
SERS enhancement. A complicating factor is the definition of the
enhancement factor itself. Most studies of SERS enhancement
[
61]
factors define the average SERS enhancement factor as
ISERS=NSurf
EF ¼
IRSNVol
where NVol =cRSV is the average number of molecules in the scattering
volume (V) for the non-SERS measurement and NSurf is the average
number of molecules adsorbed in the scattering volume for the SERS
measurements. Le Ru et al. have presented the problems associated
[61]
with this definition, with the consequence being that there could
be large variations in reported enhancement factors.
Silver nanoparticles prepared via the method reported by Lee
and Meisel typically have a heterogeneous size distribution with
diameters between 10–200 nm with various shapes such as rods
[
62]
and triangles. To confirm the size and shape of our nanoparticles,
we obtained TEM images of the silver nanoparticles used in the
SERS measurements (Fig. S4).
Knowledge of the size and shape distribution of the nanoparticles
enables an estimation of the concentration of the adsorbate required
for monolayer coverage of the silver nanoparticles and therefore the
SERS enhancement factors can be accurately calculated. TEM images
such as shown in Fig. S4 were used to determine the number and
types of nanoparticles (i.e. spherical nanoparticles vs nanorods), in a
typical sample, along with an estimate of the variation in particle size
(
we thank one of the reviewers for this suggestion). In the samples that
were analysed, approximately 70% were nanoparticles (diameter =
8 ꢅ 9 nm, most with well-defined crystal faces) and 30% were nano-
4
rods (diameter = 41 ꢅ 7 nm, length = 84 ꢅ 28 nm). The total available
surface area can be expressed by assuming all nanoparticles are
spherical with a diameter of 52 ꢅ 10 nm. These uncertainties, along
with an estimate in the uncertainty of the Raman intensities, allow an
estimate of the uncertainty in the enhancement factors reported
below, which we estimate to be approximately 20%. The footprint of
the dipyrrin molecule can be estimated using the DFT calculation.
The ratio of the surface areas gives the number of dipyrrin molecules
required per nanoparticle and finally the concentration of dipyrrin
required for monolayer coverage.
For monolayer coverage of nanoparticles with a 52 nm effec-
4
tive diameter we estimate for NSurf a value of 3.02 ꢁ 10 dipyrrin
molecules per silver nanoparticle. With the nanoparticle concen-
tration estimated from the total silver concentration and average
particle size, a dipyrrin concentration of 7.9 mM is required for
monolayer coverage.
Table 4 displays the Raman cross-sections for SERS and elec-
tronic and SERS enhancement factors for 1. The observed SERS
cross-sections are approximately 1–2 orders of magnitude larger
than the resonant Raman cross-sections. Also, the corresponding
electronic and SERS enhancement factors have similar relative
values. This illustrates that for free dipyrrins systems, the majority
of the enhancement has electronic origins at least for the condi-
tions used in this study. From detailed studies of the SERS cross-
sections and by considering the effects of preresonance
enhancement, the maximum average SERS enhancement factors
10
12 [61]
have been estimated to be in the range of 10 –10 .
In this
work no attempt was made to maximise the position of the
wileyonlinelibrary.com/journal/jrs
Copyright © 2011 John Wiley & Sons, Ltd.
J. Raman Spectrosc. 2011, 42, 2154–2164

Raman spectroscopy of dipyrrins
5
4
66
69
-1
-1
-1
1
054 cm
1279 cm
1331 cm
7
1
72
81
-1
-1
-1
1
385 cm
1385 cm
1560 cm
8
2
83
84
-1
-1
-1
1
576 cm
1576 cm
1609 cm
Figure 5. Selected normal modes of 1. Arrows represent the displacement of atoms.
plasmon resonance with respect to the excitation wavelength or
to match the plasmon resonance with the molecular resonance
frequency. The maximum values (approx. 10 ) reported here
suggest that there is ample scope for improving the intensity of
the SERS intensities of dipyrrin compounds.
8
The SERS spectra of 1 with mode assignments is given in Fig. 7.
Modes were assigned based on the DFT calculation; however,
there were discrepancies where there were interactions of some
modes with the silver nanoparticle surface. For instance the
modes 81 and 82 + 83 have significantly shifted wavenumbers
but they can be assigned by a consistent downshift from the
nonresonant spectrum.
As with the resonant Raman spectra of the free dipyrrin mole-
cules, the SERS spectra show that most of the vibrational modes
are enhanced. The pattern of enhancement is somewhat differ-
ent and the strong enhancement of mode 54 is particularly
evident along with modes 46, 51 and to a lesser extent, modes
63 and 64. We do not attempt a detailed analysis of the underly-
ing enhancement mechanisms in this work (a wider range of
excitation wavelengths would be required at least) but we can
suggest possible explanations for the strong enhancement of
mode 54. This arises from the interaction of 1 with silver nanopar-
ticles. The different intensity patterns under the nonresonant,
resonant and surface-enhanced conditions reflect the mechan-
isms responsible for generating the Raman scattering in each
case. Under nonresonant conditions, the polarisability in the
ground state controls the intensity of the observed modes, and
therefore modes with large polarisabilities will have large intensi-
ties. Under resonant conditions, the origin of the electronic
transition controls the mode intensities, and for strongly allowed
electronic transitions, such as pꢂp*, the scattering will be
Figure 6. Nonresonant Raman and resonant Raman spectra of 1 with
normal modes assigned. Resonant Raman spectra were collected with
(
[
A) 444 nm excitation, (B) 458 nm excitation, and (C) 488 nm excitation.
S denotes solvent peaks (CH Cl ), # denotes laser lines]. (D) Nonresonant
2
2
Raman spectrum was recorded as a KBr pellet.
J. Raman Spectrosc. 2011, 42, 2154–2164
Copyright © 2011 John Wiley & Sons, Ltd.
wileyonlinelibrary.com/journal/jrs

T. M. McLean et al.
Table 3. Experimental Raman cross-sections for nonresonant Raman and resonant Raman data, and resonant Raman enhancement factors for 1
4
44 nm
458 nm
488 nm
a
b
4
a
b
4
a
b
4
Normal
modes
Wavenumber
dsRRS/dΩ
EF [10 ]
dsRRS/dΩ
EF [10 ]
dsRRS/dΩ
EF [10 ]
–
1
–28
2
–1
–28
2
–1
–28
2
–1
[cm
]
[10 cm sr
]
[10 cm sr
]
[10 cm sr ]
3
4
4
4
5
5
5
5
5
6
6
6
6
6
6
7
7
7
7
8
8
8
8
6
819
870
1 015
2 093
1 607
1 172
1 681
2 398
1 793
1 876
1 757
1 691
900
5.4
22.1
49.1
19.7
32.0
13.3
15.8
35.6
24.0
—
539
750
102
—
2.9
7.9
3.1
—
119
219
302
—
0.6
2.3
9.2
—
4
6
895
9
938
1
989
573
720
504
693
239
475
508
521
590
1 731
5 236
—
10.9
4.0
4.4
13.1
3.3
—
123
241
177
243
36
2.3
1.3
1.6
4.6
0.5
—
3
1 007
1 047
1 074
1 099
1 123
1 178
1 222
1 255
1 275
1 329
1 355
1 390
1 410
1 518
1 538
1 561
1 580
1 612
4
5
7
0
80
1
6.6
3.7
1.1
—
91
0.7
—
3
255
0.5
—
4
166
—
58
—
6
3 234
8 882
1 863
10 126
629
3.8
2.0
5.2
—
208
484
—
0.2
0.5
—
9
8.8
0
—
1 + 72
9.3
2 007
—
1.8
—
253
—
0.2
—
4
1.5
9
936
18.5
—
380
—
7.5
—
9
0.2
—
0
841
62
1
3 889
13 608
544
10.5
12.1
2.1
1 573
3 636
42
4.2
3.2
0.2
355
605
10
1.0
0.5
0.04
2 + 83
4
a
[38]
Raman cross-sections were calculated using Eqn (1) and the solvent band of dichloromethane at 701/740 as in
.
b
EF = Enhancement Factor = (dsRRS/dΩ)/(dsnRRS/dΩ).
c
EFs are relative to the nonresonant (solid) Raman cross-sections.
[
18]
dominated by Albrecht A-term scattering.
The selection rules
for A-term dictate that only modes that are totally symmetric
within the common group of the ground and excited states will
be resonantly enhanced. Although the totally symmetric modes
are only a subset of the modes that could appear in the nonres-
onant spectra, the large electronic enhancement factors mean
that many modes that were below the level of noise in the non-
resonant spectrum appear with significant intensity in the reso-
nant Raman spectrum. A similar number of modes are observed
between the resonant and surface-enhanced spectra. However,
the pattern of enhancements is different and this is a result of addi-
tional enhancements from resonance with the plasmon modes of
the silver nanoparticles. Under plasmon enhancement conditions,
Herzberg–Teller coupling between the ground and other electronic
excited-states of the system leads to enhancement of nontotally
symmetric and totally symmetric modes and this explains the
changes in the appearance of the resonant and surface-enhanced
spectra. The same effect can be observed in comparing the differ-
ences in intensities obtained with different excitation wavelengths
in the SERS measurements.
Figure 7. Surfaced-enhanced Raman spectra of 1 on silver nanoparticles
with normal modes highlighted. (A) 458 nm and (B) 488 nm.
Table 4 shows that the SERS enhancement factors are greater
with 488 nm excitation than with 458 nm excitation. This seems
to be at odds with the peak of the silver plasmon resonance for
the Lee and Meisel Ag nanoparticles around 420 nm. However,
scattering from hot-spots dominates the total Raman scatter-
at 488 nm is that the enhancement is dominated by the red-
shifted plasmon resonance. On the other hand at 488 nm
excitation, electronic resonance with the pꢂp* transition
decreases, but this decrease is compensated for by an increase
in the plasmon enhancement as the excitation moves to lower
frequencies. Those modes that show noticeable changes in
intensity between 458 and 488 nm excitations are modes 46,
50, 51, 79 and 81.
[63]
[64]
ing,
and plasmon coupling
between the small fraction of
nanoparticles that form the hot-spots red-shifts the plasmon res-
onance. A possible explanation for the large SERS enhancements
wileyonlinelibrary.com/journal/jrs
Copyright © 2011 John Wiley & Sons, Ltd.
J. Raman Spectrosc. 2011, 42, 2154–2164

Raman spectroscopy of dipyrrins
Table 4. Normal modes and Raman cross-sections for experimental SERS data for 1
4
58 nm
488 nm
a
b
d
a
b
d
Normal
modes
Wavenumber
dsSERS/dΩ
Total EF
[10 ]
Electronic EF
dsSERS/dΩ
Total EF
[10 ]
Electronic EF
–
1
–28
2
–1
4
4
–28
2
–1
4
4
[cm
]
[10 cm sr
]
[10 ]
[10 cm sr
]
[10 ]
1
1
2
2
2
2
2
2
3
3
3
4
4
5
5
5
6
6
6
6
7
7
8
8
8
8
402
410
454
492
587
603
655
697
726
772
822
872
903
990
6 499
10 618
1 055
2 851
398
9
1
2
3 + 24
11 985
5
1 503
5 916
4 946
1 415
4 244
4 863
7 504
11 464
13 528
9 326
30 036
4 187
5 738
15 177
4 500
7 236
1 758
7 822
19 782
5 707
7
8 126
9 734
9
1 + 32
2 411
5
6 098
6
25.8
79.2
2.9
7.9
3.1
10.9
4.0
4.4
3.7
1.1
1 688
9.0
195.2
557.8
416.9
62.3
0.6
2.3
9.2
2.3
1.3
1.6
0.7
4
18 483
18 270
21 903
11 255
97 646
9 864
6
350.0
257.5
51.6
1
3
1 021
1 054
1 173
1 207
1 238
1 338
1 389
1 485
1 525
1 555
1 609
4
264.8
30.7
861.0
72.4
1
3
12.0
9 672
20.2
4
22 375
17 404
9 607
e
9
Undefined
6.6
5.2
1.8
7.5
4.2
3.2
0.2
17.3
8.8
0.5
0.2
0.2
1.0
0.5
0.04
1
9
34.7
2 093
41.3
39.5
60.1
49.4
1
21.0
14 689
67 653
12 914
2 + 83
4
17.6
21.8
a
[41]
Raman cross-sections were calculated using Eqn (2) and Na
2
4
SO as an external reference.
b
Total Enhancement Factor = (dsSERS/dΩ)/(dsnRRS/dΩ).
c
EFs are relative to the nonresonant (solid) Raman cross-sections.
d
e
Electronic Enhancement Factor = (dsRRS/dΩ)/(dsnRRS/dΩ)
.
Obscured by laser line.
[
3] J. K. Laha, S. Dhanalekshmi, M. Taniguchi, A. Ambroise, J. S. Lindsey,
Org. Process Res. Dev. 2003, 7, 799.
Conclusion
[
4] S. Hattori, K. Ohkubo, Y. Urano, H. Sunahara, T. Nagano, Y. Wada, N. V.
With strong resonant enhancement, negligible fluorescence,
facile synthetic chemistry and an established track record in
biological applications, dipyrrins display all the desired properties
for not only fundamental studies of SERS but a range of bioana-
lytical applications.
Tkachenko, H. Lemmetyinen, S. Fukuzumi, J. Phys. Chem. B 2005, 109,
1
5368.
[5] S. J. Garibay, J. R. Stork, Z. Q. Wang, S. M. Cohen, S. G. Telfer, Chem.
Commun. 2007, 4881.
[
6] S. G. Telfer, J. D. Wuest, Chem. Commun. 2007, 3166.
7] J. D. Hall, T. M. McLean, S. J. Smalley, M. R. Waterland, S. G. Telfer,
Dalton Trans. 2010, 39, 437
[
[
8] G. Ulrich, R. Ziessel, A. Harriman, Angew. Chem., Int. Ed. 2008,
Acknowledgements
4
7, 1184.
[
9] T. M. McLean, D. M. Cleland, S. J. Lind, K. C. Gordon, S. G. Telfer, M. R.
The authors wish to acknowledge The MacDiarmid Institute for
Advanced Materials and Nanotechnology and Massey University
for financial support. We also thank Mr Doug Hopcroft from the
Manawatu Microscopy and Imaging Centre (MMIC).
Waterland, Chem. Asian J. 2010, 5, 2036.
[
10] S. Nie, S. R. Emory, Science 1997, 275, 1102.
[11] K. Kneipp, Y. Wang, H. Kneipp, L. T. Perelman, I. Itzkan, R. R. Dasari,
M. S. Feld, Phys. Rev. Lett. 1997, 78, 1667.
[
[
12] S. Shim, C. M. Stuart, R. A. Mathies, Chemphyschem 2008, 9, 697.
13] S. A. Meyer, E. C. L. Ru, P. G. Etchegoin, J. Phys. Chem. A 2010,
Supporting information
1
14, 5515.
[
14] E. C. Le Ru, P. G. Etchegoin, Principles of Surface-Enhanced Raman Spec-
troscopy and related plasmonic effects, Elsevier, Great Britain, 2009.
15] E. C. Le Ru, E. Blackie, M. Meyer, P. G. Etchegoin, J. Chem. Phys. C
Supporting information may be found in the online version of
this article.
[
2
007, 111, 13794.
[16] J. R. Lombardi, R. L. Birke, Acc. Chem. Res. 2009, 42, 734.
[17] J. R. Lombardi, R. L. Birke, J. Phys. Chem. C 2008, 112, 5605.
[18] A. C. Albrecht, M. C. Hutley, J. Chem. Phys. 1971, 55, 4438.
[19] M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuseria, M. A. Robb, J. R.
Cheeseman, G. Scalmani, V. Barone, B. Mennucci, G. A. Petersson,
References
[1] H. Fischer, H. Orth, Die Chemie des Pyrrols, Akademische Verlagsge-
sellschaft, Leipzig, 1937.
[2] T. E. Wood, A. Thompson, Chem. Rev. 2007, 107, 1831.
J. Raman Spectrosc. 2011, 42, 2154–2164
Copyright © 2011 John Wiley & Sons, Ltd.
wileyonlinelibrary.com/journal/jrs

T. M. McLean et al.
H. Nakatsuji, M. Caricato, X. Li, H. P. Hratchian, A. F. Izmaylov, J.
Bloino, G. Zheng, J. L. Sonnenberg, M. Hada, M. Ehara, K. Toyota,
R. Fukuda, J. Hasegawa, M. Ishida, T. Nakajima, Y. Honda, O. Kitao,
H. Nakai, T. Vreven, J. Montgomery Jr., J. E. Peralta, F. Ogliaro, M.
Bearpark, J. J. Heyd, E. Brothers, K. N. Kudin, V. N. Starverov, R.
Kobayashi, J. Normand, K. Raghavachari, A. Rendell, J. C. Burant,
S. S. Iyengar, J. Tomasi, M. Cossi, N. Rega, J. M. Millam, M. Klene,
J. E. Knox, J. B. Cross, V. Bakken, C. Adamo, J. Jaramillo, R. Gomperts, R.
E. Stratmann, O. Yazyev, A. J. Austin, R. Cammi, C. Pomelli, J. W. Ochterski,
R. L. Martin, K. Morokuma, V. G. Zakrzewski, G. A. Voth, P. Salvador, J. J.
Dannenberg, S. Dapprich, A. D. Daniels, O. Farkas, J. B. Foresman, J. V.
Ortiz, J. Cioslowski, D. J. Fox, Gaussian, Inc., Wallingford CT, 2009.
20] A. D. Becke, J. Chem. Phys. 1992, 97, 9173.
[41] R. L. McCreery, Raman Spectroscopy for Chemical Analysis,
vol. 157, Wiley Interscience, New York, 2000.
[42] C. Brückner, V. Karunaratne, S. J. Rettig, D. Dolphin, Can. J. Chem.-Rev.
Can. Chim. 1996, 74, 2182.
[43] A. B. Myers, in Laser Techniques in Chemistry, vol. 23 (Eds.: A. B. Myers,
T. R. Rizzo), John Wiley & Sons, Inc., New York, 1995, pp. 325.
[44] J. R. Lombardi, R. L. Birke, J. Chem. Phys. 2007, 126, 4709.
[45] C. L. Haynes, R. P. Van Duyne, J. Phys. Chem. B 2003, 107, 7426.
[
46] J. Zhao, L. Jensen, J. H. Sung, S. L. Zou, G. C. Schatz, R. P. Van Duyne,
J. Am. Chem. Soc. 2007, 129, 7647.
[
47] A. J. Haes, S. L. Zou, J. Zhao, G. C. Schatz, R. P. Van Duyne, J. Am.
Chem. Soc. 2006, 128, 10905.
[
[
[
[
[48] S. G. Telfer, T. M. McLean, M. R. Waterland, Dalton Trans. 2011, 40, 3097.
[49] M. P. Andersson, P. Uvdal, J. Phys. Chem. A 2005, 109, 2937.
[50] G. A. Guirgis, P. Klaboe, S. Y. Shen, D. L. Powell, A. Gruodis, V. Aleksa,
C. J. Nielsen, J. Tao, C. Zheng, J. R. Durig, J. Raman Spec. 2003,
21] C. Lee, W. Yang, R. G. Parr, Phys. Rev. B 1988, 37, 785.
22] W.-M. Hoe, A. J. Cohen, N. C. Handy, Chem. Phys. Lett. 2001, 341, 319.
23] N. C. Handy, A. J. Cohen, in Molecular Physics, vol. 99, Taylor &
Francis Ltd, 2001, pp. 403.
3
4, 322.
[
[
[
[
24] Y. Zhao, D. G. Truhlar, Theor. Chem. Acc. 2008, 120, 215.
25] M. Cossi, N. Rega, G. Scalmani, V. Barone, J. Comput. Chem. 2003, 24, 669.
26] N. M. O’Boyle, A. L. Tenderholt, K. M. Langner, J. Comput. Chem. 2007, 29, 839.
27] Y. Tantirungrotechai, K. Phanasant, S. Roddecha, P. Surawatanawong,
V. Sutthikhum, J. Limtrakul, Theochem-J. Mol. Struct. 2006, 760, 189.
28] M. P. Andersson, P. Uvdal, J. Phys. Chem. A 2005, 109, 2937.
29] I. M. Alecu, J. Zheng, Y. Zhao, D. G. Truhlar, J. Chem. Theory Comput.
[
[
[
[
51] G. A. Guirgis, Y. E. Nashed, J. R. Durig, Spectroc. Acta Pt. A-Molec.
Biomolec. Spectr. 2000, 56, 1065.
52] G. A. Guirgis, X. D. Zhu, Z. H. Yu, J. R. Durig, J. Phys. Chem. A 2000,
1
04, 4383.
53] T. K. Gounev, G. A. Guirgis, T. A. Mohamed, P. Q. Zhen, J. R. Durig,
J. Raman Spec. 1999, 30, 399.
54] J. R. Durig, J. P. Xiao, J. B. Robb, F. F. D. Daeyaert, J. Raman Spec. 1998,
[
[
2
010, 6, 2872.
2
9, 463.
[30] T. Rohand, E. Dolusic, T. H. Ngo, W. Maes, W. Dehaen, Arkivoc 2007, 307.
[31] S. J. Smalley, M. R. Waterland, S. G. Telfer, Inorg. Chem. 2009, 48, 13.
[32] P. C. Lee, D. Meisel, J. Phys. Chem. 1982, 86, 3391.
[33] E. C. Le Ru, M. Meyer, P. G. Etchegoin, J. Chem. Phys. B 2006, 110, 1944.
[34] D. M. Cleland, G. Irwin, P. Wagner, D. L. Officer, K. C. Gordon, Chem.
Eur. J. 2009, 15, 3682.
[55] R. Horvath, K. C. Gordon, Coord. Chem. Rev. 2010, 254, 2505.
[56] X. Zhao, J. A. Burt, J. L. McHale, J. Chem. Phys. 2004, 121, 11195.
[57] N. Biswas, S. Umapathy, J. Chem. Phys. 2003, 118, 5526.
[58] L. C. T. Shoute, G. R. Loppnow, J. Am. Chem. Soc. 2003, 125, 15636.
[59] R. S. Czernuszewicz, K. A. Macor, X. Y. Li, J. R. Kincaid, T. G. Spiro,
J. Am. Chem. Soc. 1989, 111, 3860.
[35] N. J. Lundin, P. J. Walsh, S. L. Howell, A. G. Blackman, K. C. Gordon,
Chem. Eur. J. 2008, 14, 11573.
[
[
[
60] I. V. Sazanovich, C. Kirmaier, E. Hindin, L. Yu, D. F. Bocian, J. S. Lindsey,
D. Holten, J. Am. Chem. Soc. 2004, 126, 2664.
61] E. C. Le Ru, E. Blackie, M. Meyer, P. G. Etchegoin, J. Phys. Chem.
C 2007, 111, 13794.
[36] S. J. Lind, K. C. Gordon, S. Gambhir, D. L. Officer, Phys. Chem. Chem.
Phys. 2009, 11, 5598.
[
[
37] S. J. Lind, K. C. Gordon, M. R. Waterland, J. Raman Spectrosc. 2008, 39, 1556.
38] A. M. Moran, C. Delbecque, A. M. Kelley, J. Phys. Chem. A 2001,
62] D. Pristinski, S. Tan, M. Erol, H. Du, S. Sukhishvili, J. Raman Spectrosc.
2
006, 37, 762.
1
05, 10208.
[
[
63] A. M. Michaels, M. Nirmal, L. E. Brus, J. Am. Chem. Soc. 1999, 121, 9932.
64] N. Vogel, J. Fischer, R. Mohammadi, M. Retsch, H.-J. R. Butt,
K. Landfester, C. K. Weiss, M. Kreiter, Nano Lett. 2011, 11, 446.
[39] M. J. Colles, J. E. Griffith, J. Chem. Phys. 1972, 56, 3384.
[40] D. S. Egolf, M. R. Waterland, A. M. Kelley, J. Phys. Chem. B 2000, 104,
10727.
wileyonlinelibrary.com/journal/jrs
Copyright © 2011 John Wiley & Sons, Ltd.
J. Raman Spectrosc. 2011, 42, 2154–2164