Journal of Luminescence 158 (2015) 294300

Contents lists available at ScienceDirect

Journal of Luminescence
journal homepage: www.elsevier.com/locate/jlumin

Photoluminescence of rhodamine 6G in plasmonic eld

of Au nanoparticles: Temperature effects
Oleg A. Yeshchenko n, Illya S. Bondarchuk, Viktor V. Kozachenko, Mykhaylo Yu. Losytskyy
Physics Department, Taras Shevchenko National University of Kyiv, 64/13 Volodymyrs'ka Str., 01601 Kyiv, Ukraine

art ic l e i nf o

a b s t r a c t

Article history:
Received 23 June 2014
Received in revised form
26 September 2014
Accepted 8 October 2014
Available online 18 October 2014

Inuence of temperature on the photoluminescence of rhodamine 6G deposited on 2D array of the gold

nanoparticles was studied in the temperature range of 78278 K. The factor of surface plasmonic
enhancement of rhodamine luminescence was found to decrease monotonically with increasing
temperature. Electronphonon scattering and thermal expansion of the gold nanoparticles were
considered as two competing physical mechanisms of the temperature dependence of plasmonic
enhancement factor. The calculations showed the signicant prevalence of the electronphonon
scattering. The temperature induced increase of the scattering rate leads to higher plasmon damping
that causes the decrease of plasmonic enhancement of rhodamine 6G luminescence.
& 2014 Elsevier B.V. All rights reserved.

Keywords:
Gold nanoparticles
Surface plasmon resonance
Rhodamine 6G photoluminescence
enhancement
Temperature effects

1. Introduction
Noble metal nanoparticles (NPs) exhibit unique optical properties, such as resonant absorption and scattering of light, not found
in bulk counterparts [13]. Collective coherent excitations of the
free electrons in the conduction band are responsible for the
strong absorption and scattering of the light by the particles. These
coherent oscillations, also known as surface plasmon resonance
(SPR), lead to an enhanced local electric eld close to the surface of
the particles (both inside and outside) [15]. It is well known that
such strong local electric eld enhances the optical processes such
as Raman scattering (SERS), absorption (SEIRA) and photoluminescence (SEPL), etc. of molecules and clusters functionalized on
the surface of metal nanoparticles or close to a rough metal
surface. These methods received considerable experimental and
theoretical attention [614].
The theory of local enhanced electric eld was developed [15]
to describe the nonlinear second harmonic generation and Raman
scattering by rough metal surfaces. An extension of the theory was
developed for non-interacting hemispherical metal particles on a
surface [16]. Such local plasmonic eld can enhance the photoluminescence (PL) of molecules and clusters located in vicinity of
the metal NP as well as PL of the NP itself. The enhancement of the
PL was considered in Ref. [16] to be the result of the enhancement
of the electric elds of exciting laser photon and emitted photon of
luminescence occurring at the SPR excitation in the metal

Corresponding author. Tel.: 380 44 5264587; fax: 380 44 5264036.

E-mail address: [email protected] (O.A. Yeshchenko).

http://dx.doi.org/10.1016/j.jlumin.2014.10.018
0022-2313/& 2014 Elsevier B.V. All rights reserved.

nanoparticle. According to theory [16], the local electric eld in

vicinity of metal NP is enhanced by a factor known as the local
eld correction factor. The plasmonic enhancement factor depends
on the size, morphology, and optical constants of the NP.
The inuence of the NP's size on plasmonic enhancement of
SERS [6,7,17] and PL [8] from the molecules adsorbed on the rough
metal surfaces and on the surface of metal NPs as well as on PL from
the metal NPs themselves [18] was studied quite well. Meanwhile,
the inuence of temperature on the phenomenon of plasmonic
enhancement is still open for the study. In general, understanding
of the inuence of temperature on the SPR in metal NPs and related
phenomena, including the effect of SPR induced enhancement of
the various optical processes mentioned above, is crucial for pure
and applied science involving NPs [1,2,19]. The temperature dependence of the SPR must be further understood because of the recent
applications of noble metal NPs in thermally assisted magnetic
recording [20], thermal cancer treatment [2124], catalysis and
nanostructure growth [25], and computer chips [26]. The temperature effects on the SPR absorption band of metal NPs were studied
by Kreibig [2,27] and Doremus [28,29], and the origin of temperature effects upon the SPR was analyzed by Mulvaney [30]. Recently,
the inuence of temperature on SPR in Ag, Au and Cu NPs at high
temperatures was reported in our works [3134] and the results of
a similar study of Au-based plasmonic nanostructures at low temperatures were reported in Ref. [35], where appreciable red-shift
and broadening of the SPR with increasing temperature were
observed. Counterintuitive thermo-optical response of Au/SiO2
nanocomposites as a result of local eld enhancement was reported
in Refs. [36,37]. Recently, we reported the temperature dependence
of the quantum yield of the PL from silver [38] and copper [39] NPs.

O.A. Yeshchenko et al. / Journal of Luminescence 158 (2015) 294300

Another interesting SP related thermal effect is that noble metal

NPs, when illuminated at a wavelength close to their SP resonance, can serve as highly efcient localized heat sources at the
nanometer-length scale [4042]. This is a unique property that
promises numerous applications in solar energy, chemical catalysis,
protein imaging, and biomedicine. The nature of such heating is the
electronphonon scattering in metal NP which leads to the conversion of the energy of electronic excitation (SP) to the thermal one.
In the present work we studied the inuence of the temperature on the plasmonic enhancement of PL from the object external
to the metal NP. As such an external model object we have chosen
the rhodamine 6G (R6G), a well-known dye widely used in many
areas of fundamental and applied science. Optical properties of
dye molecule/metal NP systems are the object of high interest
now. Recently, energy transfer between dye molecules and metal
nanoparticles has gained interest in nding out the potential
application in biophotonics. Such transfer (surface energy transfer
SET) was considered in Persson and Lang's model [43] as a result
of the depopulation of excited state of the dye molecule with the
simultaneous scattering of an electron in the nearly metal to above
Fermi level. SET does not require a resonant electronic transition
like the Frster process and originates from the interaction of the
electromagnetic eld of the donor dipole with the free conduction
electrons of the accepting metal. Several theoretical and experimental studies have been published on energy transfer from a dye
to metal surface, e.g. see Refs. [4449]. Specically, the energy
transfer from R6G molecules to gold NPs was studied in Refs.
[44,45]. The signicant inuence of the SP eld of metal NPs on
the efciency of photogeneration and lifetime of excitons in
polymers coupled with the NPs was reported in Refs. [42,50]. On
the contrary to these cited works we report here the results of
study of the inverse process, namely an inuence of metal NP
electronic excitation (SP) on the luminescence of dye (R6G)
molecule. The temperature dependence of the quantum yield of
PL from R6G deposited on Au NPs array was studied. The
plasmonic factor of enhancement of the rhodamine PL was found
to decrease monotonically with increasing temperature that is due
to increase of the electronphonon scattering rate in the gold NPs.

2. Experimental: sample preparation and characterization

The samples were prepared as follows. Four glass plates with
area 1  1 cm2 were cut from the same plate. Gold lm with
thickness of 5 nm was fabricated by thermal vacuum deposition on
two glass plates at the same time. Then Au lm was annealed at
the temperature of 370 1C during half an hour. The annealing leads
to transformation of the gold lm to array of the gold NPs. The
existence of Au NPs on glass substrate was conrmed by atomicforce microscopy (AFM) (Fig. 1(a)) and absorption spectra (see
below, Fig. 2(a)). AFM measurements were performed with an
NT-MDT Ntegra microscope in semi-contact tapping mode using Si
cantilevers with a tip apex radius of  10 nm. The studied samples
contain the Au NPs with average height about 10 nm, while
average lateral size was about 22 nm. Since at absorption and PL
measurements the electric vector of incident light beam was
parallel to Au NPs array plane, only the lateral SP mode was
excited in the NPs. So, the gold NP lateral size distribution should
only be considered. The histogram of the lateral size distribution of
Au NPs is shown in Fig. 1(b). One can see that size distribution is
quite wide: o d 4 22 nm with standard deviation d 8 nm.
Then, one of obtained samples with gold NPs array and one of
remaining initial glass substrates were covered by rhodamine 6G/
shellac hybrid lm with a thickness of 10 nm at the same time. We
used hybrid lm because molecular aggregates of R6G, in the form
of dimers or higher aggregates, act to quench photoluminescence

295

Fig. 1. (a) AFM image of Au NPs array on glass substrate. (b) Lateral size
distribution of Au NPs.

[51]. Thus, we had four samples. The rst is initial glass plate, the
second one gold NPs on glass, the third R6G/shellac lm
covering gold NPs, and the fourth R6G/shellac lm on glass. The
lms were deposited as follows. The metal was vacuumevaporated from simple sources, consisting of tungsten crucibles
with shutters heated by an electric current. The crucibles were
lled with a 99.99% purity metal Au. R6G/shellac hybrid lm was
vacuum-evaporated from two simple sources at the same time.
The temperature in the crucible with a R6G powder was 400 1C,
and in the crucible with a shellac was 200 1C, while substrates
were kept at 20 1C. The pressure in the vacuum chamber was
about 10  5 Torr. The distance between substrates and laments
was about 15 cm. This large distance between substrates and
laments allowed producing quite uniform lms. The thickness
of lms was controlled by both in-situ via deposition monitors
using quartz microbalances and by multi-angle incident ellipsometry ex-situ.
A tungsten-halogen incandescent lamp was used as the light
source for absorption measurements. The single spectrometer
MDR-3 was used for the registration of absorption spectra. PL
spectra of the R6G and R6G/Au samples were measured using a
Cary Eclipse (Varian Inc.) uorescence spectrophotometer. The PL
was excited by a single monochromatic line with wavelength of
480 nm (spectral half-width of 5 nm) cut from the continuous
emission spectrum of a pulse Xe lamp. The repetition rate of the
pulses was 80 Hz, pulse duration was 2 s, and the pulse power
density was 1.3  103 W/cm2. During measurements, the samples
were placed in a nitrogen bath cryostat Optistat DN (Oxford
instruments). The PL spectra were measured at various temperatures ranging from 78 to 298 K. The temperature was maintained
and measured with the use of an Intelligent Temperature

296

O.A. Yeshchenko et al. / Journal of Luminescence 158 (2015) 294300

spectrum of Au NPs array (red line) shows the clear SPR band
located at 543 nm that is common value for gold NPs on glass
substrate. The absorption spectrum of R6G/shellac hybrid lm
(black line) contains several bands characteristic for R6G. To
separate these bands from each other we decomposed the total
spectrum to elementary bands (Fig. 2(b)). One can see that the
spectrum consists of 5 bands denoted as Ma, D1a, D2a, C1a, and
C2a. Based on data of Refs. [5256], these bands can be rationalized as follows: Ma band with maximum at 535 nm originates
from the absorption transitions in single R6G molecule (monomer). D1a (494 nm) and D2a (458 nm) ones originate from the
transitions in R6G dimers of different types. 1a (627 nm) and 2a
(687 nm) bands belong to R6G aggregates (complexes). Fig. 2(a)
shows as well the absorption spectrum of R6G/shellac hybrid lm
deposited on Au NPs array (blue line). One can see that this spectrum
reects the peculiarities of the spectra of both R6G and Au NPs.
The main task of this work was to study the temperature effects
in the inuence of the SP eld on the PL spectrum of R6G
deposited on Au NPs array. Figs. 3 and 4(a) show the behavior of
the PL spectrum of hybrid R6G/shellac lm and one deposited on
the gold NPs array correspondingly with the monotonous increase
of temperature from 78 to 298 K. Note, all the spectra presented in
Figs. 3 and 4 belong to R6G species. Indeed, the spectra of Fig. 3
were obtained by subtraction of the PL spectrum of shellac lm on
glass from the total PL spectrum and the ones of Fig. 4 by
subtraction of the spectrum of shellac lm on Au NPs array. It is
seen that PL spectra of R6G both without the gold NPs and

Fig. 2. (a) Absorption spectra of R6G/shellac hybrid lm (black line), Au NPs array
(red line), and R6G/shellac hybrid lm deposited on Au NPs array (blue line).
(b) Decomposition of the spectrum of R6G/shellac hybrid lm to the elementary
bands. The spectra were measured at room temperature. (For interpretation of the
references to color in this gure legend, the reader is referred to the web version of
this article.)

Controller ITC503S (Oxford instruments). Each spectrum was

measured at its own stabilized temperature. The temperature of
the sample during the registration of the PL spectra was measured
with a temperature sensor in the cryostat. We assume in this work
that the temperature of the sample was equal to the ambient one
measured by the temperature sensor. However, as was noted in
Section 1, the illumination of metal NPs at wavelengths close to SPR
leads to their quite efcient plasmon-induced heating [4042].
So, one can suppose that due to plasmonic heating of Au NPs the
temperature of the samples in our experiments was somewhat
higher than the average temperature measured by the cryostat
thermal sensor. However, taking into account that the excitation
wavelength in our experiments (480 nm) was quite far from SPR in
Au NPs (543 nm), it is reasonable to assume that real sample
temperatures were not noticeably higher than ones measured by
the thermal sensor.

3. Results and discussion

3.1. Temperature dependence of PL from rhodamine 6G on Au NPs
array: experiment
Primarily, the absorption spectra of the samples were studied.
Fig. 2(a) shows the spectra of R6G, array of Au NPs, as well as R6G
on Au NPs array measured at room temperature. The absorption

Fig. 3. (a) Evolution of PL spectrum of R6G/shellac thin lm with monotonous

increase of temperature from 78 to 298 K. (b) Decomposition of PL spectrum of R6G
in shellac measured at 78 K to the elementary bands. Excitation wavelength was
480 nm.

O.A. Yeshchenko et al. / Journal of Luminescence 158 (2015) 294300

Fig. 4. (a) Evolution of PL spectrum of R6G/shellac thin lm deposited on Au NPs

array with monotonous increase of temperature from 78 to 298 K.
(b) Decomposition of PL spectrum of R6G/shellac on Au NPs array measured at
78 K to the elementary bands. Excitation wavelength was 480 nm.

deposited on Au NPs array consist of three bands denoted as Ml, Dl,

and Cl. This is clearly seen from Figs. 3 and 4(b) where the total PL
spectra were decomposed to the elementary bands. These bands
correspond to the respective absorption ones. Namely, following
the data of Refs. [5256], the Ml one with maximum at 570 nm
originates from the light emission from the single R6G molecules,
the Dl one (605 nm) is due to PL from R6G dimers, and Cl (635 nm)
is the PL band of R6G aggregates. Note that in PL spectrum of R6G
species in shellac on glass the Ml band of single rhodamine
molecule is quite intensive; its intensity is comparable with ones
of dimers and aggregates. Meanwhile, the deposition of the R6G/
shellac lm on the Au NPs array leads to appreciable decrease of
the molecule Ml band intensity with synchronous increase of the
intensity of Dl and Cl bands related to R6G aggregates. Probably,
the higher intensity of PL from rhodamine aggregates in sample
with Au NPs can be explained as follows. The increase of the
temperature leads to red-shift of SPR [3135], i.e. to the shift
towards the Dl and Cl bands. Such shift would lead to higher
overlap of the SPR in Au NPs and the PL bands of rhodamine
aggregates causing the higher plasmonic enhancement of intensity
of the respective Dl and Cl bands. Another cause of the higher
intensity of aggregate related PL bands for samples with R6G on
Au NPs could be that gold NPs facilitate the aggregation of
rhodamine molecules. So, in this case the relative number of
monomers could be lower in the samples with Au NPs that would
lead to lower relative intensity of monomer band to one of
aggregates.

297

Fig. 5. (a) Experimental temperature dependences of intensity of the total PL

spectrum of rhodamine 6G in shellac lm without Au NPs (blue circles and solid
line), R6G in shellac lm deposited on Au NPs array (red squares and solid line), and
calculated one for R6G on Au NPs array (magenta dashed line). (b) Temperature
dependence of the ratio of PL intensity of R6G on Au NPs to one of R6G without Au
NPs: experimental blue points and solid line, calculated dashed green line.

Fig. 5(a) shows the obtained temperature dependences of

intensity of total PL spectrum of R6G without Au NPs (blue points
and line) and one deposited on Au NPs array (red points and line).
One can see that the dependence for R6G without gold NPs is
slightly decreasing in temperature range of 78198 K and becomes
sharply decreasing in the range of 198258 K that is in full
agreement with data reported in Ref. [57]. Meanwhile, the further
increase of the temperature from 258 to 298 K leads to abrupt
change of the intensity decrease to its sharp increase. The similar
effect was observed for R6G in Refs. [58,59]. It is seen that
qualitatively similar dependence is observed for R6G on gold NPs
array. Note that increase of R6G PL intensity at increase of the
temperature in the range of 258298 K is accompanied by sharp
increase of intensity of molecular (monomer) Ml band and
simultaneous sharp decrease of dimer and aggregates Dl and Cl
ones, Figs. 3 and 4(a). Based on this, one can suppose that such
increase of PL intensity of monomers with temperature is caused
by thermally induced dissociation of R6G aggregates that leads to
increase of the number of free rhodamine molecules. It is well
known that the PL quantum yield decreases at the aggregation of
rhodamine molecules. Thus, thermally induced increase of number
of R6G free molecules would lead to increase of the quantum yield
of all the rhodamine species and, correspondingly, to the increase
of their PL intensity.

298

O.A. Yeshchenko et al. / Journal of Luminescence 158 (2015) 294300

Let us consider the temperature dependence of PL from R6G on

gold NPs array (Figs. 4 and 5(a)). It is seen that qualitatively the
dependence is similar to one of R6G without Au NPs. However, the
temperature induced decrease of the PL intensity in range of 78
258 K is considerably stronger and the increase in range of 258
298 K is considerably smaller for R6G with Au NPs than without
them. Note the rather interesting temperature dependence of ratio
of the total PL intensity of R6G on Au to one of R6G without Au,
Fig. 5(b) blue points and line. One can see that this ratio
decreases monotonically with temperature. Let us analyze the
cause of such an effect. The R6G molecules in sample with Au NPs
are located in the local eld of surface plasmons excited in gold
nanoparticles. The plasmonic electric eld enhances the PL from
rhodamine. Thus, one can conclude that difference in the temperature dependences of PL intensity from R6G on Au NPs and one
from R6G without them is caused by the temperature dependence
of the plasmonic enhancement factor of PL from R6G. The
magnitude of enhancement factor is determined by the magnitude
of plasmonic eld in vicinity of R6G molecules located near Au
NPs. So, let us consider the physical mechanisms that can cause
the temperature dependence of the magnitude of the plasmonic eld.

It is well known that optical processes in molecules and

clusters located near the surface of metal NPs are strongly
enhanced. Such enhancement is caused by a strong local electric
eld near the metal NP where surface plasmons are excited. The
electric elds of incoming (exciting) and outgoing (emitted)
photons are enhanced via coupling to the SPR. Thus, tuning the
magnitude of plasmonic enhancement gives the opportunity to
tune the magnitude of the optical response of such molecules and
clusters. In this work, we analyze the inuence of the temperature
on the plasmonic enhancement factor in Au NPs and, correspondingly, on the quantum yield of PL from rhodamine 6G deposited on
Au NPs array. According to Boyd's theory [16], the local electric
eld outside a metal NP is enhanced by a factor known as the local
eld correction factor

D1
h
1=2 i;
4 2 V 1  m
1

1i
3
m  1 D

where D is the depolarization factor (D 1=3 for spherical particles), is the dielectric permittivity of the metal particle, m is the
dielectric permittivity of the surrounding medium, is the light
wavelength, V 4 r 3 =3 is the NP volume, and r is its radius. The
intensity of PL I l from a molecule located near the metal NP
excited with a photon of energy exc is given by




I l I 0 l L2 exc L2 l ;
2
where I 0 l is a function describing the intrinsic PL spectrum of
the molecule, i.e. in the absence of metal NP.
The plasmonic factor L depends on T due to temperature
dependence of dielectric permittivity of NP and NP's volume V.
The dielectric permittivity of the metal NP is

; r; T ib D ; r; T;

D ; r; T 1 

2p r; T
;
2 i r; T

where p r; T and r; T are the size and temperature dependent

plasma frequency and damping constant of the metal NP respectively. Lets us consider the physical mechanisms of such
dependences.
Let us consider the SPR damping constant. It can be expressed as

; r; T e  ph T e  e ; T S r; T;

where e  ph and e  e are the rates of electronphonon and

electronelectron scattering respectively and S is the rate of the
scattering of free electrons on surface of the NP. Let us analyze the
impact of each term in Eq. (5) on the temperature behavior of the
damping constant. The electronphonon scattering rate e  ph
increases with increasing temperature due to increase of the
phonon population in the metal. The temperature dependence of
electronphonon scattering rate is given by [35,61]
"
#
 5 Z =T 4
2
T
x dx
e  ph T S 4
;
6
5
ex  1

0
where is the Debye temperature for bulk metal ( 165 K for bulk
gold [62]) and S is the constant. S can be calculated as

3.2. Temperature dependence of plasmonic enhancement factor

Lout

theory as [60]

where ib ib1 iib2 is the contribution of the bound

electrons (interband transitions) to the permittivity of metal. Here
we suppose that ib is independent of the temperature.
D ; r; T is the size and temperature dependent contribution of
the free electrons which can be expressed according to Drude

e  ph T 0
 5 R
=T 0
T0
2
0
54

x4 dx
ex  1

where T 0 298 K. The electronphonon scattering rate at room

temperature can be estimated as e  ph T 0 vF =, where vF is the
Fermi velocity (vF 1.29  108 cm/s for bulk gold [62]), and is the
free electron mean free path ( 36 nm for bulk gold at 298 K). So,
we obtain e  ph T 0 3:6  1013 s  1. Second term e  e in Eq. (5),
the rate of electronelectron scattering is expressed as the sum of the
temperature dependent and the frequency dependent terms [63]

e  e ; T Te T fe ;

and, correspondingly, is neglected in our calculawhere

tions, while frequency dependent term is expressed as [64]
T
e T e  ph

fe

2
:
4 2 p

Thus, electronelectron scattering rate in fact is temperature

independent. Finally, the rate of surface scattering of the free
electrons is given as [1,2,65]
vF
;
rT

S A

10

where r T is the NP radius, and A is the parameter whose value is

determined by both surface and interface scattering effects
[2,65,66] (A 1 was taken for calculation). Thus, the plasmon
damping constant would depend on the temperature both directly
through the rate of electronphonon scattering and indirectly
through the temperature dependence of the NP size. The physical
mechanism of this size dependence is the thermal expansion of
NP occurring with increase of the temperature. The NP radius
depends on temperature as
rT r 0 1 T1=3 ;

11

where r0 is the NP radius at room temperature, T T  T 0 is the

temperature variation from room one (T 0 298 K), and is the
volume thermal expansion coefcient (4.17  10  5 K  1 for bulk
gold [62]).

O.A. Yeshchenko et al. / Journal of Luminescence 158 (2015) 294300

Note that plasma frequency is size and temperature dependent

as well
r
p0
4 nT e2
p r; T
12
p;
mn
1 T
where the concentration of free electrons is temperature dependent as nT N=V T and the NP volume depends on temperature
due top
the
thermal expansion of the NP, VT V 0 1 T. Here,

p0 4 n0 e2 =mn is the plasma frequency (p0 9.1 eV for bulk

Au, Ref. [60]) and n0 is the concentration of free electrons for bulk
metal at T 0 298 K, mn is the effective electron mass, and e is the
electron charge.
Based on the above expressions one can calculate the PL
spectra of R6G on Au NPs array in dependence on the temperature.
Note that for calculation of NP permittivity by Eq. (3) the
contribution of the bound electrons ib is needed. ib was calculated as the result of subtraction of Drude term calculated for
T 0 298 K from tabular dielectric permittivity of bulk gold at
T 0 298 K taken from Ref. [67]. It has been noted above that PL
spectrum of R6G is strongly temperature dependent, Fig. 3(a). So,
as I 0 l we used in Eq. (2) the PL spectrum of R6G without Au NPs
at a certain temperature presented in Fig. 3(a). We have noted in
Section 2 that size distribution of the Au NPs is quite wide. Since
the plasmonic local eld correction factor is size dependent
through the dielectric permittivity of NP and its volume (see Eq.
(2)), the size distribution can affect the L factor, and, correspondingly, the PL spectrum of R6G on Au NPs array. Therefore, the PL
spectrum was calculated using the value of L averaged by NP size
E
D


I l I 0 l L2 exc ; r L2 l ; r 


R1  2

2
0 L exc ; r L l ; r f r dr
R1
;
13
I 0 l
0 f r dr
where f r is a lateral size distribution function presented in Fig. 1(b).
3.3. Discussion
Based on above theoretical model we calculated the PL spectra
of R6G on Au NPs inuenced by SP eld at various temperatures in
range of 78298 K. The obtained behavior of the spectra is shown
in Fig. 6. Fig. 5(a) presents the calculated temperature dependence
of PL intensity of R6G on Au NPs. One can see that calculated
dependence (dashed magenta line) is in good agreement with
experimental one (red squares and solid line) that proves correctness of the used theoretical model. The good agreement is
obtained as well for the experimental and calculated temperature

Fig. 6. Calculated evolution of PL spectrum of rhodamine 6G on Au NPs array with

monotonous increase of temperature from 78 to 298 K.

299

dependences of the ratio of PL intensity of R6G on Au to one of

R6G without Au (Fig. 5(b)). One can see that the dependence of the
ratio is monotonically decreasing with temperature. This indicates
the fact of monotonic decrease of plasmonic enhancement factor
with temperature. Let us discuss it.
As one has been discussed above, the temperature affects both
the SPR width and frequency. The temperature dependences of
both width and frequency are caused by two physical mechanisms.
Those are the electronphonon scattering and the thermal expansion of metal NP. The SPR width is caused by the plasma oscillation
damping constant. The increase of temperature leads to increase of
electronphonon scattering rate and, correspondingly, to increase
of SPR damping, Eq. (6). Meanwhile, the temperature increase
leads to the thermal expansion of the NP which leads to decrease
of the rate of electron surface scattering and, correspondingly, to
decrease of damping constant, Eq. (10). However, as we have
shown in our previous works [3234] the contribution of the
electronphonon scattering in the temperature dependence of SP
damping constant exceeds greatly the one of the surface
scattering.
As well, the thermal expansion of NP leads to red-shift of SPR,
Eq. (12). The red-shift of SPR with increasing temperature would
increase the overlap of SPR (maximum at 543 nm at 298 K) and PL
bands of R6G with maxima at 570 nm, 605 nm and 635 nm. This
effect would lead to increase of plasmonic enhancement factor
and, correspondingly, would lead to the increase of R6G PL
quantum yield. Note, however, that the spectral separation of the
SPR in Au NPs and PL spectrum of R6G is quite large. So, it is
reasonable to expect that this effect is not so strong to affect
signicantly the plasmonic enhancement factor. Besides this, an
additional mechanism of the inuence of temperature induced
red-shift of SPR on plasmonic enhancement factor exists.
Hubenthal showed recently [68] that interband transitions are
signicant additional mechanism of SPR damping when SPR
frequency is close to the onset of interband transitions. At photon
energies higher than onset of interband transitions the energy of
the plasmon can be transferred to a single electron, which makes
an interband transition, as was previously suggested by Dalacu
and Martinu [69]. It is clear that for photon energies well below
the onset of the interband transition, the plasmon energy is too
low to permit an interband transition and, correspondingly, such a
mechanism of plasmon damping is not important. For bulk gold
the onset of interband transitions is at about 660 nm [67]. At
wavelength shorter than 660 nm the increase of imaginary part of
dielectric permittivity of gold 2 occurs that reects the increase of
contribution of the interband transitions. Note, however, that
strong increase of 2 starts at r 520 nm while in wavelength
range from 660 to 520 nm 2 increases quite slightly. It is obvious
that the red-shift of SPR with increasing temperature would lead
to decrease of the contribution of interband transitions to plasmon
damping constant at the frequency of SPR. Such a decrease of SPR
damping would lead to the increase of plasmon enhancement
factor and, correspondingly, to the increase of PL intensity. However, taking into account that SPR wavelength (about 540 nm) is in
the range of slight spectral dependence of 2 (520660 nm) it is
reasonable to expect that this effect is not so strong to affect
signicantly the plasmonic enhancement factor. Therefore, proceeding from above considerations one can conclude that the
electronphonon scattering is prevailing mechanism of the temperature dependence of plasmonic enhancement factor. Correspondingly, the temperature induced increase of electronphonon
scattering rate leads to decrease of quantum yield of PL from
rhodamine 6G on gold NPs array. Note that this effect observed for
Au NPs is opposite to one observed for the PL from copper NPs
where the increase of temperature leads to increase of PL quantum
yield [39]. An origin of such surprising behavior of PL from copper

300

O.A. Yeshchenko et al. / Journal of Luminescence 158 (2015) 294300

NPs is that for Cu NPs the interband transitions and spectral

overlap of SPR with PL band give considerably larger contribution
in the temperature dependence of plasmonic enhancement factor
than for gold NPs.
4. Conclusions
In conclusion, we made the comparative study of the temperature dependence of the PL from rhodamine 6G deposited on 2D
array of gold NPs and PL from R6G without Au NPs in the
temperature range of 78298 K. The PL from R6G on Au NPs array
is enhanced by the coupling of exciting and emitted photons to the
surface plasmons excited in the NPs. There is a signicant
difference in the temperature behavior of the PL spectra from
R6GAu and R6G samples. The ratio of PL intensity of R6G on Au to
one of R6G without Au was found to be monotonically decreasing
with temperature. This indicates the fact of monotonic decrease of
the factor of plasmonic enhancement of PL from R6G on Au NPs
with temperature. The theoretical model was proposed to explain
the observed temperature dependence of plasmonic enhancement
factor. Two possible physical mechanisms of the observed temperature dependence were considered. First one is the electron
phonon scattering which causes the increase of plasmonic damping constant with increase of the temperature leading to the
weakening of the plasmonic enhancement. Second one is the
thermal expansion of the gold NPs. The thermal expansion
increases the NP size leading to decrease of the plasmonic
damping caused by surface scattering of electrons and to corresponding strengthening of plasmonic enhancement. As well, the
thermal expansion of Au NPs leads to red-shift of SPR in Au NPs
with increasing temperature. The red-shift causes the increase of
the spectral overlap of SPR with PL band of R6G as well as decrease
of the contribution of interband transitions in gold NPs in
plasmonic damping constant. Thus, the temperature induced
red-shift of SPR would lead to increase of the plasmonic enhancement of R6G PL. The calculations of the inuence of temperature
on the PL spectrum of R6G in plasmonic eld of Au NPs were
performed. The results of calculations agree with the results of the
experimental observations proving the model used. The calculations showed that the electronphonon scattering is prevailing
mechanism in the temperature dependence of the factor of
plasmonic enhancement of PL from R6G on Au NPs as compared
to effects related to the thermal expansion of Au NPs. Correspondingly, such prevalence of the electronphonon scattering causes
the temperature induced decrease of plasmonic enhancement of
R6G PL observed.
Acknowledgments
O. A. Yeshchenko and I. S. Bondarchuk acknowledge the partial
support by the State Agency on Science, Innovations and Informatization of Ukraine (Grant # M/325-2013).
References
[1] V.V. Klimov, Nanoplasmonics, Pan Stanford Publishing, Singapore, 2013.
[2] U. Kreibig, M. Vollmer, Optical Properties of Metal Clusters, Springer-Verlag, Berlin
Heidelberg, 1995.
[3] S.A. Maier, Plasmonics: Fundamentals and Applications, Springer, New York, 2007.
[4] B.G. Ershov, E. Janata, A. Henglein, A. Fojtik, J. Phys. Chem. 97 (1993) 4589.
[5] A. Henglein, J. Phys. Chem. 97 (1993) 5457.
[6] M. Fleischmann, P.J. Hendra, A.J. McQuillan, Chem. Phys. Lett. 26 (1974) 163.
[7] Y. Fang, N.H. Seong, D.D. Dlott, Science 321 (2008) 388.
[8] E. Ozbay, Science 311 (2006) 189.
[9] O.L. Muskens, V. Giannini, J.A. Sanchez-Gil, J.G. Rivas, Nano Lett. 7 (2007) 2871.
[10] Y.X. Zhang, K. Aslan, M.J.R. Previte, C.D. Geddes, Appl. Phys. Lett. 90 (2007)
173116.

[11] J. Kundu, F. Le, P. Nordlander, N. Halas, Chem. Phys. Lett. 452 (2008) 115.
[12] A. Whitney, R. Van Duyne, F. Casadio, J. Raman Spectrosc. 37 (2006) 993.
[13] D. Stuart, C. Yonzon, X. Zhang, O. Lyandres, N. Shah, M. Glucksberg, J. Walsh,
R. Van Duyne, Anal. Chem. 77 (2005) 4013.
[14] J. Hulteen, M. Young, R. Van Duyne, Langmuir 22 (2006) 10354.
[15] G.T. Boyd, T. Rasing, J.R.R. Leite, Y.R. Shen, Phys. Rev. B 30 (1984) 519.
[16] G.T. Boyd, Z.H. Yu, Y.R. Shen, Phys. Rev. B 33 (1986) 7923.
[17] S.M. Nie, S.R. Emery, Science 275 (1997) 1102.
[18] O.A. Yeshchenko, I.M. Dmitruk, A.A. Alexeenko, M.Yu. Losytskyy, A.V. Kotko,
A.O. Pinchuk, Phys. Rev. B 79 (2009) 235438.
[19] U. Kreibig, Appl. Phys. B 93 (2008) 79.
[20] W.A. Challener, C. Peng, A.V. Itagi, D. Karns, W. Peng, Y. Peng, X.M. Yang,
X. Zhu, N.J. Gokemeijer, Y.T. Hsia, G. Ju, R.E. Rottmayer, M.A. Seigler, E.C. Gage,
Nat. Photon 3 (2009) 303.
[21] D. ONeal, L. Hirsch, N. Halas, J. Payne, J. West, Cancer Lett. 209 (2004) 171.
[22] L.R. Hirsch, R.J. Stafford, J.A. Bankson, S.R. Sershen, B. Rivera, R.E. Price,
J.D. Hazle, N.J. Halas, J.L. West, Proc. Natl. Acad. Sci. USA 100 (2003) 13549.
[23] A. Lowery, A. Gobin, E. Day, N. Halas, J. West, Breast Cancer Res. Treat. 100
(2006) S289.
[24] A. Lowery, A. Gobin, E. Day, N. Halas, J. West, Int. J. Nanomed. 1 (2006) 149.
[25] L. Cao, D.N. Barsic, A.R. Guichard, M.L. Brongersma, Nano Lett. 7 (2007) 3523.
[26] W. Cai, J.S. White, M.L. Brongersma, Nano Lett. 9 (2009) 4403.
[27] U. Kreibig, J. Phys. F 4 (1974) 999.
[28] R.H. Doremus, J. Chem. Phys. 40 (1964) 2389.
[29] R.H. Doremus, J. Chem. Phys. 42 (1965) 414.
[30] P. Mulvaney, in: K.J. Klabunde (Ed.), Nanoscale Materials in Chemistry, Wiley &
Sons, New York, 2001, p. 121.
[31] O.A. Yeshchenko, I.M. Dmitruk, A.A. Alexeenko, A.V. Kotko, J. Verdal,
A.O. Pinchuk, Plasmonics 7 (2012) 685.
[32] O.A. Yeshchenko, I.S. Bondarchuk, A.A. Alexeenko, A.V. Kotko, Funct. Mater. 20
(2013) 357.
[33] O.A. Yeshchenko, I.S. Bondarchuk, V.S. Gurin, I.M. Dmitruk, A.V. Kotko, Surf.
Sci. 608 (2013) 275.
[34] O.A. Yeshchenko, Ukr. J. Phys. 58 (2013) 249.
[35] J.S.G. Bouillard, W. Dickson, D.P. OConnor, G.A. Wurtz, A.V. Zayats, Nano Lett.
12 (2012) 1561.
[36] M. Rashidi-Huyeh, B. Palpant, Phys. Rev. B 74 (2006) 075405.
[37] B. Palpant, M. Rashidi-Huyeh, B. Gallas, S. Chenot, S. Fisson, Appl. Phys. Lett. 90
(2007) 223105.
[38] O.A. Yeshchenko, I.S. Bondarchuk, M.Yu. Losytskyy, A.A. Alexeenko, Plasmonics
9 (2014) 93.
[39] O.A. Yeshchenko, I.S. Bondarchuk, M.Yu. Losytskyy, J. Appl. Phys. 116 (2014)
054309.
[40] A.O. Govorov, H.H. Richardson, Nano Today 2 (2007) 30.
[41] Z. Fang, Y.R. Zhen, O. Neumann, A. Polman, F.J. Garca de Abajo, P. Nordlander,
N.J. Halas, Nano Lett. 13 (2013) 1736.
[42] B. Choudhuri, A. Mondal, J.C. Dhar, N.K. Singh, T. Goswami, K.K. Chattopadhyay,
Appl. Phys. Lett. 102 (2013) 233108.
[43] B.N. Persson, N.D. Lang, Phys. Rev. B 26 (1982) 5409.
[44] T. Sen, A. Patra, J. Phys. Chem. C 116 (2012) 17307.
[45] T. Sen, S. Sadhu, A. Patra, Appl. Phys. Lett. 91 (2007) 043104.
[46] C.S. Yun, A. Javier, T. Jennings, M. Fisher, S. Hira, S. Peterson, B. Hopkins,
N.O. Reich, G.F. Strouse, J. Am. Chem. Soc. 127 (2005) 3115.
[47] T.L. Jennings, M.P. Singh, G.F. Strouse, J. Am. Chem. Soc. 128 (2006) 5462.
[48] E. Dulkeith, A.C. Morteani, T. Niedereichholz, T.A. Khar, J. Feldmann, S.A. Levi,
F.C.J.M. Veggel, D.N. Reinhoudt, M. Moller, D.I. Gittins, Phys. Rev. Lett. 89
(2002) 203002.
[49] A.P. Alivisatos, D.H. Waldeek, C.B. Harris, J. Chem. Phys. 82 (1985) 541.
[50] J.L. Wu, F.C. Chen, Y.u.S. Hsiao, F.C. Chien, P. Chen, C.H. Kuo, M.H. Huang,
C.S. Hsu, ACS Nano 5 (2011) 959.
[51] R. Vogel, P. Meredith, M.D. Harvey, H. Rubinsztein-Dunlop, Spectrochim. Acta
A 60 (2004) 245.
[52] V.V. Bryukhanov, S.A. Elistratova, V.C. Laurinas, J. Appl. Spectrosc. 70 (2003)
450.
[53] G.S.S. Saini, S. Kaur, S.K. Tripathi, C.G. Mahajan, H.H. Thanga, A.L. Verma,
Spectrochim. Acta A 61 (2005) 653.
[54] R. Gvishi, G. Ruland, P.N. Prasad, Opt. Commun. 126 (1996) 66.
[55] W. Holzer, H. Gratz, T. Schmitt, A. Penzkofer, A. Costela, I. Garcia-Moreno,
R. Sastre, F.J. Duarte, Chem. Phys. 256 (2000) 125.
[56] D. Toptygin, B.Z. Packard, L. Brand, Chem. Phys. Lett. 277 (1997) 430.
[57] H.T. Oh, H.-S. Kam, T.Y. Kwon, B.K. Moon, S.I. Yun, Mater. Lett. 13 (1992) 139.
[58] M.A. Ali, J. Moghaddasi, S.A. Ahmed, Appl. Opt. 29 (1990) 3945.
[59] R.B. Alaverdyan, T.K. Dadalyan, A.S. Karapetyan, N.S. Torosyan, J. Phys.: Conf.
Ser. 350 (2012) 012033.
[60] N.W. Ashcroft, N.D. Mermin, Solid State Physics, Brooks/Cole, Belmont, 1976.
[61] M.D. Tillin, J.R. Sambles, J. Phys.: Condens. Matter 2 (1990) 7055.
[62] C. Kittel, Introduction to Solid State Physics, Willey, New York, 2005.
[63] G.R. Parkins, W.E. Lawrence, R.W. Christy, Phys. Rev. B 23 (1981) 6408.
[64] D.E. Aspnes, E. Kinsbron, D.D. Bacon, Phys. Rev. B 21 (1980) 3290.
[65] A. Pinchuk, U. Kreibig, A. Hilger, Surf. Sci. 557 (2004) 269.
[66] A. Pinchuk, U. Kreibig, New J. Phys. 5 (2003) 151.
[67] P.B. Johnson, R.W. Christy, Phys. Rev. B 6 (1972) 4370.
[68] F. Hubenthal, Plasmonics 8 (2013) 1341.
[69] D. Dalacu, M. Martinu, J. Opt. Soc. Am. B 18 (2001) 85.