Submitted to

DOI: 10.1002/adma.((please add manuscript number))

A luminescent molecular thermometer for long-term absolute temperature

measurements at the nanoscale

By Carlos D. S. Brites, Patricia P. Lima, Nuno J. O. Silva, Angel Millán, Vitor S.

Amaral, Fernando Palacio* and Luís D. Carlos*

Keywords: Luminescent molecular thermometer, Hybrid Materials, Magnetic

[*] Mr. C. D. S. Brites, Dr. P. P. Lima, Dr. N. J. O. Silva, Prof. V. S. Amaral, Prof. L. D. Carlos
Departamento de Física and CICECO, Universidade de Aveiro
3810–193 Aveiro (Portugal)
E-mail: lcarlos@ua.pt

Dr. A. Millán, Prof. F. Palacio

1
 Submitted to

Temperature is a fundamental thermodynamic variable, the measurement of which is

crucial in countless scientific investigations and technological developments,

accounting at present for 75%–80% of the sensor market throughout the world.[1] The

traditional liquid-filled and bimetallic thermometers, the thermocouples, the pyrometers

and the thermistors are generally not suitable for temperature measurements at scales

below 10 μm. This intrinsic limitation has encouraged the development of new non-

contact accurate thermometers with micrometric and nanometric precision, a

challenging research topic increasingly hankered for.[2-9]

Current developments[4] of thermometers operating below a few micrometers and

with high spatial resolution include: Ga-filled carbon/MnO nanotubes,[2] temperature

dependent conformation modifications of molecular spring superstructures,[3] Raman[10]

and infrared[11] spectroscopic probing, scanning thermal microscopy probes,[6] and

luminescence-based measurements.[5-9, 12-29]

Nanotube based systems require a transmission electron microscope for the read-out

and calibration, an evident limitation for practical applications. Infrared observation is

restricted by diffraction resolution and other spectroscopic methods (such as Raman),

and scanning microscopy systems have a read-out rate limited by material surface

conditions and probe motion (typically 30 ms per pixel,[29] making real-time

temperature mapping unfeasible). By contrast, the thermal dependence of phosphor

(organic dyes or trivalent lanthanides, Ln3+) luminescence is a noninvasive and accurate

alternative technique that works remotely by way of an optical detection system, even in

biological fluids, strong electromagnetic fields and fast-moving objects. Among the

luminescent thermometers proposed so far only a handful of them illustrate effectively

the temperature sensing/mapping at a submicrometer scale: a cellular thermometer with

an organic dye as the temperature probe;[8] a scanning thermal microscope with

2
 Submitted to
3+ 3+ [6]
Er /Yb co-doped fluoride glass small particles or PbF2 nanoparticles[29] glued at the
5
tip extremity; 2D distribution maps using D0 rise-time temporal response in

Y2O3:Eu;[26] and siloxane hybrid nanoparticles (size ranging from 20 to 30 nm)

incorporating a Eu3+ tris(β-diketonate) complex.[9]

Here, we report the development of a unique Eu3+/Tb3+ luminescent

nanothermometer[30] combining a number of properties, which make it a step forward

relatively to the single Ln3+-based devices proposed so far. This new thermometer is

self-referencing, allowing absolute measurements in the 10-350 K temperature range. Its

temperature sensitivity is up to 4.9%·K-1, 1.5 times larger than the highest value

reported previously [9], and it exhibits high photostability in long-term use. The variation

of the Eu3+/Tb3+ ratio affords tunability to the temperature working range. Alternatively,

tunability is also accomplished by changing the host matrix, thus modifying the

interaction between the Ln3+ and the host matrix energy levels. To obtain a high

resolution 2-dimensional temperature mapping, a judicious choice of the host matrix

enables processing the thermometer material as a film. Finally, the combination of the

molecular thermometer with a nanometric magnetic/luminescent host matrix provides

the device multifunctionality at the nanoscale.

The present luminescent molecular thermometer consists of

[Eu(btfa)3(MeOH)(bpeta)] and [Tb(btfa)3(MeOH)(bpeta)] β-diketonate chelates[30] (Fig.

S1a in the Supporting Information) embedded into organic-inorganic hybrid

nanoclusters (100-400 nm) formed by a maghemite (γ-Fe2O3) magnetic core

(hydrodynamic size of 21.0±4.0 nm, Fig. 1a) coated with a tetraethyl

orthosilicate/aminopropyltriethoxysilane (TEOS/APTES) organosilica shell (see the

experimental section for details about synthesis and labeling). The hybrid host was

rationally designed in order to present an excited triplet state with energy slightly above

3
 Submitted to
3+ 5
that of the Tb D4 emitting state, thus warranting the occurrence of thermally-driven
5
D4→host energy transfer and, consequently, the determination of the temperature

dependence of the 5D4 emission. As the energy difference between that triplet state and

the Eu3+ 5D0 emitting level is too large to permit thermally-driven depopulation, the

Tb3+/Eu3+ relative intensity guarantees the absolute measurement of temperature, with

spatial resolution adjustable by the size of the nanoclusters to which the luminescent

probes are anchored. We remark that similar host-to-Ln3+ energy resonance conditions

may be obtained using other β-diketonate chelates with different ligands. The self-

calibration and the temperature measurement by way of relative intensities (an

inexpensive, much easier and less time-consuming method than the measurements of

the emission quantum yield and the excited state lifetime) overcome the well-known

drawbacks of intensity-based measurements, such as variation of the sensor

concentration and drifts of the optoelectronic system.[9]

The multifunctional nanoparticles are dispersible in water and show a bimodal size

distribution centered at 120 and 390 nm as seen in DLS measurements (Fig. 1a).

Nanoparticles with these characteristic sizes are also observed in SEM and TEM images

(Figs. 1 b,c and Figure S2 in the Supporting Information). Aggregation of the

nanoparticles occur during evaporation of the dispersions on the sample holder, with

these particles showing regions with high and low contrast, corresponding to

maghemite/Eu3+/Tb3+ cores and to the siliceous matrix, respectively (Fig. 1c). EDS

mappings show Eu3+ and Tb3+ distributions with contours and shapes similar to those of

the nanoparticles (Fig. 1d), indicating that most of the nanoparticles contain both Eu3+

and Tb3+. The magnetic properties of the γ-Fe2O3@TEOS/APTES nanoparticles reveal

superparamagnetic behaviour above 200 K (Figure S3 in the Supporting Information).

4
 Submitted to
In optical microscopy observations, the Eu /Tb co-doped γ-Fe2O3@TEOS/APTES
3+ 3+

nanoparticles appear as red spots with sizes of the order of 5 μm, as illustrated in Fig 1e

for the NP3-1.3 nanoparticles (see experimental section for sample labeling), with each

spot corresponding to the emission either of a nanoparticle or a nanoparticle aggregate −

the red color being the result of the dominance of the Eu3+ emission at room-

temperature (Figs. S4 and S5 in the Supporting Information). Since the emission color is

a measure of the absolute temperature − as it is shown below − the use of this

luminescent molecular thermometer coupled with an optical microscope constitutes a

thermometer with micrometer spatial resolution. The larger spot size as compared to the

nanoparticles aggregates is direct consequence of the limited space resolution of the

instrument.

The temperature dependence of the emission spectra (14-300 K) of the co-doped

Eu,Tb γ-Fe2O3@TEOS/APTES NP3 nanoparticles is illustrated in Fig. 2a (and in Figs.

S6 and S7 in the Supporting Information), and whereas the intensity of the green Tb3+

luminescence strongly decreases as the temperature increases, the intensity of the red

Eu3+ lines starts to increase at precisely the same temperature at which the Tb3+

emission for NP3-1.1, NP3-1.2 and NP3-1.3 starts to decrease, remaining

approximately constant for NP3-1.10 (see experimental section for sample labeling). As

this temperature dependence − appropriately illustrated by the calculus of the

Commission Internacionale d'Éclairage (CIE) (x,y) color coordinates (Fig. 2b) −

strongly depends on the relative proportion of the 1 and 2 complexes anchored at the

surface of the nanoparticles (Figs. S6 and S7 in the Supporting Information), the ability

to fine-tune the nanocomposites emission from the green area towards the red region of

Fig. 2b as a function of the temperature is readily managed by changing the relative

proportion of embedded complexes. The temperature dependence of the Eu3+ emission

5
 Submitted to
is rationalized assuming a concerted two–step process involving only the Boltzmann

energy factor ΔE. First, the excited triplet state of the TEOS/APTES layer with energy

above that of the 5D4 emitting state is populated through thermally-driven 5D4-to-host

energy transfer (diminishing, therefore, the 5D4→7F5 intensity). Then, the triplet host

level transfers a fraction of the absorbed energy to Eu3+ ions (through the multipolar and

exchange mechanisms[31]). For diluted Eu3+ nanoparticles, the probability of the host-to-

Eu3+ energy transfer mechanism should be strongly reduced (due to a larger average

Tb3+-Eu3+ distance) resulting in that the 5D0→7F2 intensity should be practically

temperature independent up to 250-275 K. This is exactly what is observed for NP3-

1.10 (Fig. S7 in the Supporting Information), thus bearing out the above interpretation.

The temperature dependence of the 5D0 and 5D4 lifetimes is shown in Fig. S8 in the

Supporting Information and in Fig. 2c, respectively (with the corresponding decay

curves being well modeled by single exponential functions; Fig. S9 in the Supporting

Information). Whereas the 5D0 lifetime remains approximately constant (0.53-0.75 ×10-3

s, Fig. S8 in the Supporting Information, excluding, therefore, Tb3+-to-Eu3+ energy

transfer[32]), the 5D4 lifetime decreases approximately one order of magnitude, (Fig. 2c),

a behavior that can be understood considering the presence of a thermally activated non-

radiative mechanism associated to the emitting centers of the TEOS/APTES layer,

whose energy is centered at ∼22500±1500 cm-1 (Fig. S10 in the Supporting

Information), in accord with the one previously reported for aminopropylsilica

nanospheres annealed at 673 K.[33, 34] The non-radiative de-excitation probability of the

Tb3+ 5D4 level (20590±5 cm-1) may be approximately described by the Mott-Seitz

model, which expresses the temperature dependence of the experimental lifetime as: [32]

6
 Submitted to
(1)

where τr is the radiative lifetime, k is the migration energy rate, ΔE is the activation

energy between the triplet host level and the 5D4 state, kB is the Boltzmann constant and

T is the absolute temperature. The activation energies obtained for NP3-1.1, NP3-1.2

and NP3-1.3 are 615±19, 521±18 and 687±14 cm-1, respectively, in agreement with the

above suggestion that, for these nanoparticles, the thermally activated non-radiative

mechanism involves the emitting levels of the TEOS/APTES layer.

The absolute temperature measurement is inferred defining an experimental

parameter Δ ( ), related to the integrated areas ITb and IEu of the 5D4→7F5

(ITb at 545 nm) and the 5D0→7F2 (IEu at 612 nm) transitions. Figure 3a shows the

temperature dependence of Δ for the NP3 nanoparticles, while Fig. 3b presents the

temperature dependence of the sensitivity (defined in the experimental section) as a

function of the relative Tb3+-to-Eu3+ ratio. The best value (4.9%·K-1) is the largest one

measured so far for molecular thermometers, being 1.5 times larger than the highest

value previously reported using the same intensity ratio technique.[9] Setting, as quality

limit, a minimum sensitivity value of 1%·K-1, leads to an unprecedent optimal operation

range of ∼60-80 K around the temperature of maximum sensitivity (120-190 K).

Furthermore, the absolute emission quantum yields of the NP3 nanoparticles − ranging

from 0.20±0.02 (NP3-1.3) to 0.37±0.04 (NP3-1.2) (Table S1 in the Supporting

Information) − are sufficiently high so as to permit the use of commercial fiber optics

for excitation and detection.

To make absolute measurements at higher temperatures feasible, the activation

energy ΔE must be increased, which can accomplished either by increasing the

7
 Submitted to
[33]
TEOS/APTES ratio or by designing a different host matrix presenting an excited

triplet state located at higher energy, relatively to that of the TEOS/APTES layer. Here,

we report this latter approach using Eu3+/Tb3+ co-doped di-ureasils[35, 36] (allowing the

processability of the thermometer as films), whose excitation and emission spectra are

shown in Fig. S11 in the Supporting Information. Increasing ΔE to ≈2000 cm-1 (as

calculated using Eq. 1), the thermometer range of operation is now shifted to room

temperature (Figs. 3b,c). The magnetic properties of the γ-Fe2O3 nanoparticles

embedded into the di-ureasil host (U1.3) also display superparamagnetism above 200 K

(Figure S3 in the Supporting Information). The maximum absolute emission quantum

yield and the maximum sensitivity of U1.3 are 0.16±0.02 (Table S1 in the Supporting

Information) and 1.9%·K-1 (at 201 K, Fig. 3b), respectively. The Eu3+/Tb3+ co-doped di-

ureasils are, therefore, sufficiently sensitive to be used in the physiological temperature

range, as illustrated by the calibration curve of U1.3 between 298-328 K (Fig. S12b in

the Supporting Information).

A demonstration of the spatial resolution of the Eu3+/Tb3+ thermometer was

performed mapping the temperature of an integrated circuit covered with a U1.3 layer

(Figs. 4a,b). An optical fiber is used to excite and collect the hybrid emission which is

converted in absolute temperature using the U-1.3 calibration curve. A map of the

temperature of the circuit was also recorded with a state-of-the-art commercial IR

camera (see the supporting information for details). The comparison between the two

mappings is illustrated in Figs. 4c,d showing a better spatial resolution for the Eu3+/Tb3+

thermometer, despite the 450 µm of the optical fiber’s inner diameter being ∼2.5 times

larger than the camera pixel size. In our experiment, the spatial resolution achieved, 35

µm (Fig. 4d), is estimated by the minimal distance resolved by the thermometer, when

the temperature changes above the temperature uncertainty (0.5 K in this case;

8
 Submitted to
Supporting Information). This procedure shows, therefore, that the thermometer

proposed here permits measuring absolute temperatures in regions with high

temperature gradients (line B in Fig. 4b), as well as in regions presenting complex

temperature distributions (line A in Fig. 4b). We draw attention to the fact that although

the spatial resolution of the thermometer obviously depends upon the spatial resolution

of the optical detector (≈1 µm in the case of the optical microscope in Fig. 1e, for

example), the ultimate resolution limit of the thermometer is not reached yet, since the

thermographic technique used in this experiment can − with appropriate

sensing/mapping devices − detect temperature variations in nanosized regions.

In summary, the luminescent molecular thermometer introduced here combines: i)

ability to fine-tune emission color as a function of temperature and Eu3+/Tb3+

proportion; ii) self-referencing that allows absolute measurements; iii) 4.9%·K-1

maximum temperature sensitivity (better than 0.5%·K-1 in the physiological temperature

range); iv) high photostability for long-term use; v) flexibility to be processed as thin

films for sensing/mapping large areas with a spatial resolution limited by the size of the

optical detectors (~1-10 µm for commercial optical fibers and CCD cameras); vi) a

temperature uncertainty of 0.5 degree; and vii) multifunctionality as it can be hosted in

silica-coated magnetic nanoparticles. When compared this thermometer with the Ln3+-

based ones proposed so far it clearly represents a step forward in thermometry at the

nanoscale. We anticipate that the synergetic outcomes arising by combining temperature

sensing/mapping and superparamagnetism opens the way for new exciting applications,

especially in the biomedical field. In particular, such association will provide a unique

instrument to map, in a non-invasive way, temperature distributions in biological tissues

(e.g., in tumors) during heat release, due to the application of an ac field to magnetic

9
 Submitted to
[37]
nanoparticles (magnetic hyperthermia ), this being, with no doubt, a powerful tool for

the study of biochemical micro-processes occurring within a cell.

10
 Submitted to

Experimental

Synthesis of Ln3+-containing γ-Fe2O3@TEOS/APTES nanoparticles. The Ln3+-

containing γ-Fe2O3@TEOS/APTES nanoparticles (Ln=Eu, Tb or Eu/Tb), named as

NP1, NP2 and NP3, respectively, were prepared following the procedure described in

detail in the Supporting Information. Eu,Tb co-doped nanoparticles were produced

using Eu:Tb ratios of 2:1, 1:1, 1:2, 1:3 and 1:10, named as NP3-2.1, NP3-1.1, NP3-1.2,

NP3-1.3 and NP3-1.10, respectively. The corresponding ferrofluid was prepared by

dispersing nanoparticle powders in water by sonication.

Synthesis of the Di-ureasil hybrid sensing film. The first step of the synthesis

(detailed in the Supporting Information) involves the formation of a urea cross-linked

organic-inorganic hybrid precursor[35] (Fig. S1b in the Supporting Information). In the

second step, the [Eu(btfa)3(MetOH)(bpeta)] (1) and [Tb(btfa)3(MetOH)(bpeta)] (2)

complexes were incorporated as ethanolic solutions together with water and HCl for

promoting the hydrolysis of the alkoxysilane precursor. The sample, named U-1.3, was

prepared as a thick (~10 µm) film, aged for seven days at 318 K.

Characterization. The morphology, chemical composition and size distribution of the

particles were characterized by scanning and transmission electron microscopy and

dynamic light scattering (see the Supporting Information for details). Optical

microscopy was carried out on a Carl Zeiss Axiovert 200 M fluorescence microscope

equipped with Apotome, using a 365 nm excitation wavelength. The samples were

prepared by dispersing the NP3 powders in water by sonication, dropping this

dispersion in a microscope slide and drying it in open air. Emission spectra were

recorded on a Fluorolog–3® 2-Triax, Horiba Scientific, with a modular double grating

excitation spectrometer and a TRIAX 320 single-emission monochromator coupled to a

R928 Hamamatsu photomultiplier, using the front face acquisition mode. The excitation

11
 Submitted to
source was a 450W Xe arc lamp. The emission spectra were corrected for detection and

optical spectral response of the spectrofluorimeter, and the excitation spectra were

corrected for the spectral distribution of the lamp intensity using a photodiode reference

detector. The time-resolved experiments were performed with the setup described for

the luminescence spectra using a pulsed Xe-Hg lamp (6 µs pulse at half width and 20-30

µs tail). The CIE (x,y) color coordinates were calculated from the emission spectra

using the 2nd observer data. A layer of U-1.3 was deposited onto an integrated circuit

with tracks of different widths (down to 200 µm), allowing temperature gradients up to

0.03 degree·µm-1. A translation stage makes possible moving the circuit with a

precision of 100 µm. An optical fiber (spot diameter ≈900 µm) is used to illuminate the

hybrid layer at 355 nm and collect the emission spectra, which was converted in

absolute temperature.

Acknowledgements

We acknowledge Fundação para a Ciência e a Tecnologia (FCT, Portugal), COMPETE

and FEDER programs (PTDC/CTM/101324/2008) and Integrated Spanish-Portuguese

Action PT2009-0131 for financial support. The work in Zaragoza has been supported by

the grants MAT2007-61621 and CONSOLIDER CSD2007-00010 from the Ministry of

Education. CDSB (SFRH/BD/38472/2007) and PPL (SFRH/BPD/34365/2006) thank

FCT for grants. NJOS acknowledges CSIC for a I3P contract. We thank A.L.L. Videira

for a critical reading of the manuscript; Carlos Vicente and André Martins (IT-Aveiro)

for the help with the tests in the integrated circuit; and Victor Sorribas (Zaragoza) for

the collaboration with the optical microscopy observations.

Received: ((will be filled in by the editorial staff))

12
 Submitted to

Figure Captions

Figure 1. (a) DLS data showing the diameter distribution of the γ-Fe2O3 and Ln3+-
doped γ-Fe2O3@TEOS/APTES nanoparticles, respectively. (b, c) SEM and TEM
images, respectively, of NP3-1.3. (d) EDS mappings of the relative Eu and Tb content
in the SEM image of NP3-1.3 shown bellow. (e) Image of the optical microscope (UV
excitation at 365 nm) of NP3-1.10. The red color corresponds to the real color.

Figure 2. (a) Emission spectra of NP3-1.3 excited at 357 nm and recorded between 14
and 300 K. The sharp lines assigned to 1, 2, 3, 4 and 5 correspond to the 5D4→7F6,5
(Tb3+) and 5D0→7F2-4 (Eu3+) transitions, respectively. In the area marked with an
asterisk there is an overlap of the Eu3+ (5D0→7F0,1) and Tb3+ (5D4→7F4) emissions. (b)
CIE chromaticity diagram showing the temperature dependence of the (x,y) color
coordinates of NP3-1.3 and U1.3. (c) Temperature dependence of the 5D4 lifetime of
NP3-1.1, NP3-1.2, NP3-1.3 and U-1.3, monitored at 547 nm and excited at 357 nm.
The lines correspond to the best fit of the experimental data using Eq. (1) (r2>0.983).

Figure 3. (a) Normalized Δ parameter (divided by its value at room temperature Δ0) of
NP3-1.1, NP3-1.2, NP3-1.3, NP3-1.10 (excitation wavelength of 357 nm) and U-1.3
(excitation wavelength of 365 nm). (b) Sensitivity of NP3-1.1, NP3-1.2, NP3-1.3, NP3-
1.10 and U-1.3. (c) Integrated areas of the 5D4→7F5 (green squares) and 5D0→7F2 (red
squares) transitions as a function of temperature for U-1.3 (excitation wavelength of 365
nm). The green line corresponds to the best fit of the experimental data using Eq. (1)
(correlation coefficient r2>=0.978) and the red line is a guide for the eyes.

Figure 4. (a, b) Scheme of the temperature measurement using an optical fiber sensing
(spot diameter 0.9 mm) on an integrated circuit with tracks of different widths (down to
200 µm) covered with a layer of U1.3. (c, d) Temperature profiles obtained with U1.3
(red circles, the size corresponds to the temperature uncertainly of 0.5 degree) along the
lines A and B compared with the measurements performed using an IR camera (blue
squares).

13
 Submitted to

Figure 1

30
γ−Fe2O3 ferrofluid Tb3+
25 NP3-1.10
20
(a) (d)
number /%

Eu3+
15

10

0
10 100 1000 5 μm 5 μm
size /nm

(b) (c) (e)

1 μm 40 μm

14
 Submitted to
Figure 2
0.9
1.0
2 (a) U-1.3 (b)

Normalised Intensity
0.8
0.8
NP3-1 .3
3 0.6
0.7

0.6
0.4 10K
1
0.5 10K
4 5 0.2
y
0.4 300K
0.0
0.3 300K
100

/K
re
200 0.2

atu
er
*

mp
300 0.1
480 520 560 600 640 680 720
Wavelength /nm Te 0.0
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8
0.75
(c) x
τ ( D4 ) /10 s
-3

0.50
NP3-1.1
NP3-1.2
0.25
NP3-1.3
5

U-1.3
0.00
0 50 100 150 200 250 300
Temperature /K

15
 Submitted to
Figure 3
1 .0 5. 0
N P3- 1 .1 0
N P3 -1 .1
N P3- 1 .3
N P3 -1 .2
0 .8 4. 0 N P3- 1 .2
N P3 -1 .3
N P3- 1 .1

Sensitivity (% per K)
N P3 -1 .1 0 ´
U ET-1 .3
U -1 .3
Normalised Δ

0 .6 3. 0

0 .4 2. 0

0 .2 1. 0

0 .0
(a) 0. 0
(b)
0 50 100 1 50 2 00 25 0 30 0 0 50 100 150 2 00 2 50 30 0

Temperature /K Temperature /K
1. 0 2 .0

(c)
0. 8 1 .6
Intens ity / a.u.

0. 6 1 .2

0. 4 0 .8

0. 2 0 .4

0. 0 0 .0
0 50 1 00 1 50 200 25 0 30 0

Temperature /K

16
 Submitted to

Figure 4

Fib re Bund le
(a) (b)
M C
F In te grated Circuit

D Optical Layer

A
B

M – Monocromator C – Colimating Lens

33 0
(c) 340
(d)
32 0 330
Temper ature /K

Temperatur e /K

320
31 0

310

30 0
300

29 0 290
0 5 10 15 20 25 3.0 4.5 6.0 7.5 9.0
Position /mm Position /m m

17
 Submitted to

[1] P. R. N. Childs, J. R. Greenwood, C. A. Long, Rev. Sci. Instrum. 2000, 71, 2959.
[2] Y. H. Gao, Y. Bando, Nature 2002, 415, 599.
[3] J. Lee, A. O. Govorov, N. A. Kotov, Angew. Chem. Int. Edit. 2005, 44, 7439.
[4] J. Lee, N. A. Kotov, Nano Today 2007, 2, 48.
[5] P. Löw, B. Kim, N. Takama, C. Bergaud, Small 2008, 4, 908.
[6] E. Saïdi, B. Samson, L. Aigouy, S. Volz, P. Löw, C. Bergaud, M. Mortier,
Nanotechnology 2009, 20, 115703.
[7] S. Singh, K. Kumar, S. Rai, Sensor Actuat. A-Phys. 2009, 149, 16.
[8] C. Gota, K. Okabe, T. Funatsu, Y. Harada, S. Uchiyama, J. Am. Chem. Soc. 2009,
131, 2766.
[9] H. S. Peng, M. I. J. Stich, J. B. Yu, L. N. Sun, L. H. Fischer, O. S. Wolfbeis, Adv.
Mater. 2010, 22, 716.
[10] R. Schorer, E. Friess, K. Eberl, G. Abstreiter, Phys. Rev. B 1991, 44, 1772.
[11] W. Ryba-Romanowski, J. Phys. Chem. Solids 1993, 54, 153.
[12] S. Allison, G. Gillies, Rev. Sci. Instrum. 1997, 68, 2615.
[13] N. Chandrasekharan, L. Kelly, J. Am. Chem. Soc. 2001, 123, 9898.
[14] G. Walker, V. Sundar, C. Rudzinski, A. Wun, M. Bawendi, D. Nocera, Appl.
Phys. Lett. 2003, 83, 3555.
[15] S. Allison, G. Gillies, A. Rondinone, M. Cates, Nanotechnology 2003, 14, 859.
[16] B. Zelelow, G. Khalil, G. Phelan, B. Carlson, M. Gouterman, J. Callis, L. Dalton,
Sensor Actuat. B-Chem. 2003, 96, 304.
[17] M. Mitsuishi, S. Kikuchi, T. Miyashita, Y. Amao, J. Mater. Chem. 2003, 13,
2875.
[18] G. A. Baker, S. N. Baker, T. M. McCleskey, Chem. Commun. 2003, 2932.
[19] G. E. Khalil, K. Lau, G. D. Phelan, B. Carlson, M. Gouterman, J. B. Callis, L. R.
Dalton, Rev. Sci. Instrum. 2004, 75, 192.
[20] A. Heyes, S. Seefeldt, J. Feist, Opt. Laser Technol. 2006, 38, 257.
[21] H. Aizawa, T. Katsumata, S. Komuro, T. Morikawa, H. Ishizawa, E. Toba, Sensor
Actuat. A-Phys. 2006, 126, 78.
[22] S. Uchiyama, A. P. de Silva, K. Iwai, J. Chem. Educ. 2006, 83, 720.
[23] X. Wang, X. G. Kong, Y. Yu, Y. J. Sun, H. Zhang, J. Phys. Chem. C 2007, 111,
15119.
[24] S. Katagiri, K. Manseki, Y. Tsukahara, K. Mitsuo, Y. Wada, J. Alloy. Compd.
2008, 453, 1.
[25] C. Gota, S. Uchiyama, T. Yoshihara, S. Tobita, T. Ohwada, J Phys Chem B 2008,
112, 2829.
[26] A. Khalid, K. Kontis, Meas. Sci. Technol. 2009, 20, 025305.
[27] V. Tikhomirov, K. Driesen, V. Rodriguez, P. Gredin, M. Mortier, V.
Moshchalkov, Opt. Express 2009, 17, 11794.
[28] S. Singh, K. Kumar, S. Rai, Appl. Phys. B: Lasers Opt. 2009, 94, 165.
[29] L. Aigouy, E. Saïdi, L. Lalouat, J. Labéguerie-Egéa, M. Mortier, P. Löw, C.
Bergaud, J. Appl. Phys. 2009, 106, 4301.
[30] F. Palacio, A. Millán, N. J. Silva, L. D. Carlos, V. Amaral, P. P. Lima, C. D. S.
Brites, Spain Patent P200930367, 2009.
[31] P. P. Lima, S. S. Nobre, R. O. Freire, S. A. Junior, R. A. Sa Ferreira, U. Pischel,
O. L. Malta, L. D. Carlos, J Phys Chem C 2007, 111, 17627.
[32] L. D. Carlos, R. A. Ferreira, J. P. Rainho, V. de Zea Bermudez, Adv. Funct.
Mater. 2002, 12, 819.
[33] A. M. Jakob, T. A. Schmedake, Chem. Mater. 2006, 18, 3173.

18
 Submitted to
[34] L. Wang, M. C. Estevez, M. O'Donoghue, W. H. Tan, Langmuir 2008, 24, 1635.
[35] V. D. Bermudez, L. D. Carlos, L. Alcacer, Chem. Mater. 1999, 11, 569.
[36] L. D. Carlos, R. A. S. Ferreira, V. D. Bermudez, S. J. L. Ribeiro, Adv. Mater.
2009, 21, 509.
[37] A. Jordan, R. Scholz, K. Maier-Hauff, M. Johannsen, P. Wust, J. Nadobny, H.
Schirra, H. Schmidt, S. Deger, S. Loening, W. Lanksch, R. Felix, J. Magn. Magn.
Mater. 2001, 225, 118.

19