Nanoscale

View Article Online

PAPER View Journal | View Issue
Published on 24 September 2013. Downloaded by University of St Andrews Library on 09/12/2013 07:49:12.

Optical trapping of NaYF4:Er3+,Yb3+ upconverting

fluorescent nanoparticles
Cite this: Nanoscale, 2013, 5, 12192

P. Haro-González,a B. del Rosal,a L. M. Maestro,a E. Martı́n Rodrı́guez,b R. Naccache,b

J. A. Capobianco,b K. Dholakia,c J. Garcı́a Soléa and D. Jaque*a

Received 16th July 2013
Accepted 21st September 2013
DOI: 10.1039/c3nr03644h
www.rsc.org/nanoscale
We report on the first experimental observation of stable optical trapping of dielectric NaYF4:Er3+,Yb3+
upconverting fluorescent nanoparticles (
26 nm in diameter) using a continuous wave 980 nm single-
beam laser. The laser serves both to optically trap and to excite visible luminescence from the
nanoparticles. Sequential loading of individual nanoparticles into the trap is observed from the analysis
of the emitted luminescence. We demonstrate that the trapping strength and the number of individual
nanoparticles trapped are dictated by both the laser power and nanoparticle density. The possible
contribution of thermal effects has been investigated by performing trapping experiment in both heavy
water and into distilled water. For the case of heavy water, thermal gradients are negligible and optical
forces dominate the trap loading behaviour. The results provide a promising path towards real three
dimensional manipulation of single NaYF4:Er3+,Yb3+ nanoparticles for precise fluorescence sensing in
biophotonics experiments.

A Introduction or Tm3+/Yb3+ co-doped hexagonal NaYF4 crystals exhibited the

Three-dimensional manipulation and control of single and
multiple nanoparticles (tens of nanometers in size) is key for a
variety of applications, including single molecule spectroscopy,
colloidal dynamics, tailored particle assembly, protein isola-
tion, high resolution surface studies, controlled investigation of
biological processes and surface enhanced spectroscopy.1 The
potential applications of nanoparticle manipulation may be
further extended if the nanoparticles (NPs) in question also act
as uorescent probes, so that their luminescence properties
may be used to provide physical and/or chemical cues. Such
single particle manipulation may be used for physical/chemical
imaging of a variety of systems with reduced sizes (ranging from
photonic devices to single cells).1a,d,2
Among the multitude of luminescent NPs that may act as
versatile optical sensors, lanthanide doped upconverting
nanoparticles (UCNPs) were introduced in the 90s and unlocked
a realm of new possibilities in the eld of biolabeling and
bioassays.3 Upconversion refers to the generation of higher
energy light from lower energy irradiation, usually near-infrared
(NIR), through the use of lanthanide ions doped into a solid
state host.4 Gudel et al. showed that micrometer sized Er3+/Yb3+
a
Fluorescence Imaging Group, Departamento de Fı́sica de Materiales, C-IV,
Universidad Autónoma de Madrid, C/Francisco Tomás y Valiente 7, Madrid, 28049,
Spain. E-mail: daniel.jaque@uam.es
b
Department of Chemistry and Biochemistry, Centre for NanoScience Research,
Concordia University, Montreal, Quebec, H4B 1R6, Canada
c
SUPA, School of Physics & Astronomy, University of St Andrews, Physical Science
Building, North Haugh, St. Andrews, KY16 9SS, UK
highest upconversion efficiencies.3c,5 Currently, upconverting
nanoparticles have numerous applications in the biological
eld.3a,6 In particular, Er3+/Yb3+ or Tm3+/Yb3+ co-doped hexag-
onal NaYF4 or NaGdF4 nanoparticles show a number of
advantages with respect to traditional uorescent biolabels,
including improved detection sensitivity due to the absence of
autouorescence, deeper penetration of near infrared (NIR)
light into biological tissue, and reduced photo-damage.
Furthermore, they offer the possibility of excitation via contin-
uous wave low-power NIR lasers that are compact and inex-
pensive when compared to their femtosecond counterparts
required for multiphoton excitation of other upconverting NPs
such as semiconductor quantum dots.3a,c,5a,6,7 Additional
advantages of UCNPs include their chemical, physical and
photochemical stabilities, size independent optical properties,
narrow emission peaks, long uorescence lifetimes, large anti-
Stokes shis (up to 500 nm) and low toxicity.
UCNPs have been successfully used for high resolution
intracellular dynamic studies, bioimaging and accurate intra-
cellular thermal measurements.6a,7,8 However, to the best of our
knowledge, optical trapping of single or multiple UCNPs has
not yet been demonstrated. Such a step would facilitate explo-
ration of new applications. As an example, intracellular uo-
rescence sensing and imaging are presently achieved by
intracellular incorporation of UCNPs, so that UCNPs are
distributed over the entire cell volume.9 Subsequently, the
uorescence of UCNPs at different intracellular locations is
obtained by laser scanning. This traditional approach has yiel-
ded good results but relies on the uncontrolled incorporation of

12192 | Nanoscale, 2013, 5, 12192–12199 This journal is ª The Royal Society of Chemistry 2013

Paper
hundreds of UCNPs into a cell that could cause adverse effects
on cell dynamics or viability. The controlled optical trapping of
one or few UCNPs would provide the possibility of uorescence
imaging and sensing by simultaneous scanning of laser
spots and so the trapped UCNPs, thus obviating the need for
incorporating hundreds of UCNPs into the cell under study.
This would ensure the minimization of any adverse cell
Published on 24 September 2013. Downloaded by University of St Andrews Library on 09/12/2013 07:49:12.
perturbation.
Optical trapping (OT)10 is a non-contact sterile technique
exerting forces on mesoscopic particles using a tightly focused
laser beam. Particles of higher refractive indices than the
surrounding medium may be attracted into the high intensity
region created by the laser beam focus.10e,f Once in the focal
area, the optically trapped NPs may be manoeuvered at will by
controlling the focusing optics (the position of the laser focus).
Nevertheless, optical trapping of NPs is not an easy task. The
viability of optical trapping depends on the magnitude of
the optical potential well, Uopt. Assuming a Gaussian beam, the
potential depth Uopt, can be expressed as:10f–h
I0 jaNP j
Uopt ¼ (1)
cnm
where, I0 is the maximum beam intensity, c is the speed of light,
nm is the refractive index of the surrounding medium, and aNP
is the polarizability of the optically trapped particle.10f–h For
insulating isotropic crystals, aNP, this can be expressed as:
nNP 2  nm 2 nNP 2  nm 2
aNP ¼ 3V ¼ 4pr3 (2)
nNP þ 2nm
2 2 nNP 2 þ 2nm 2
where V is the particle volume, r is the particle radius, and nNP is
the refractive index of the trapped particle (at the wavelength of
the trapping laser). Combining eqn (1) and (2) the magnitude
of the optical potential is found to be proportional to the third
power of the NP radius, i.e. Uopt f r3. Therefore, the optical
potential for NPs (r < 100 nm) is weak, making optical trapping
difficult in view of the fact that the thermal energy may be large
and we cannot satisfy the criterion Uopt > kBT, where, kB and T
are the Boltzmann constant and the medium temperature (in
Kelvin), respectively.10f,11 We have estimated that for dielectric
NaYF4 nanoparticles (nNP z 1.6) with a particle diameter of 30
nm (that leads to a particle polarizability of 5.5  1024 m3) and
assuming our experimental beam spot size of 0.7 mm, optical
trapping may be possible for laser powers in the range of
hundreds of milliwatts. This simple calculation suggests that
optical trapping and manipulation of dielectric NPs such as
NaYF4 would be possible by using diode lasers of low to
moderate powers, which is the topic of this present study.
However, when dealing with dielectric nanoparticles, as in our
case, the trapping process is only one of the multiple hurdles
encountered. Detecting trapped UCNPs is a further challenge
given that the optical scattering is expected to be weak due to
the reduced particle size.
In this work we experimentally demonstrate the optical
trapping of water dispersible NaYF4:Er3+,Yb3+ UCNPs by using a
simple experimental set-up including a single diode laser with a
moderate output laser power (<400 mW). The wavelength of the
trapping beam was xed at 980 nm, matching that of the
This journal is ª The Royal Society of Chemistry 2013
View Article Online
Nanoscale
excitation spectrum peak for the two-photon excited visible
luminescence of NaYF4:Er3+,Yb3+ UCNPs, corresponding to the
absorption maximum of Yb3+ ions. Under these conditions, we
observe sequential “particle-by-particle” incorporation into the
optical trap by analysing the time evolution of the visible
luminescence generated at the laser focus. The trapping rate
and force, estimated by the hydrodynamic drag method, have
been measured as a function of trapping power and the density
of the nanoparticles. We consider the role played by thermal
effects by performing the experiments both in distilled and
heavy (D2O) water.
B Experimental
The NaYF4:Er3+,Yb3+ UCNPs (with dopant concentrations of 2%
Er3+ and 18% Yb3+), used in this work were synthesized via a
solvothermal synthetic route.9 The polyethyleneimine (PEI)-
capped nanoparticles crystallize in the cubic phase (a-NaYF4
phase) as conrmed by XRD analysis (a Scintag XDS-2000
diffractometer equipped with a Si(Li) Peltier-cooled solid state
detector, a CuKa source at a generator power of 45 kV and
40 mA). Transmission electron microscopy (TEM, a JEOL JEM-
2011 microscope operating at 200 kV) analysis showed that the
nanoparticles are nearly spherical in shape (Fig. 1(a), inset) with
an average diameter of 26 nm and with size dispersion close to
20% (Fig. 1(a)). In order to avoid agglomeration of the UCNPs,
the solution was sonicated for 30 minutes prior to conducting
the optical trapping experiments. Fig. 1(c) shows a digital
photograph of the NaYF4:Er3+,Yb3+ aqueous dispersion when
optically excited using a 980 nm focused laser beam, revealing
the presence of a bright two-photon excited visible (green)
emission at the laser focus. This luminescence is produced
because of an efficient energy transfer from Yb3+ to Er3+ (energy
transfer upconversion): NIR radiation (980 nm) is efficiently
absorbed by Yb3+ ions while the visible emission is generated by
the Er3+ ions.5a,7,9
Fig. 2 details the experimental set-up used for optical trap-
ping of the UCNPs. In these experiments, a colloidal dispersion
of UCNPs in water was placed in a 5 mm width, 100 mm height
microchannel (Ibidi Inc., m-Slide I, catalog number: 80106) and
excited at 980 nm using a single-mode ber-coupled laser diode.
The laser radiation coming out from the single-mode ber
(SMF) was collimated using a ber-port (Thorlabs PAF-X-7-B).
The collimated beam was expanded by using a 2X telescope
(Thorlabs BE02M) resulting in a 2 mm beam diameter. This
matches the back aperture of the 100 long working distance
microscope objective of 0.8 NA that was used to focus the
trapping beam into a microchannel (Olympus LMPLFLN100
with a working distance of 3.4 mm) and that constitutes an
upright microscope as can be observed in Fig. 1. The 980 nm
spot radius was estimated to be close to 0.7 mm. The spatial
distribution and the visible upconversion luminescence inten-
sity generated from UCNPs were recorded using a CCD camera.
A dichroic mirror was used to separate the visible luminescence
and the 980 nm (excitation and trapping) radiation.
For thermal imaging experiments, the microchannel used
for optical trapping was lled with a solution containing two
Nanoscale, 2013, 5, 12192–12199 | 12193

Nanoscale
Published on 24 September 2013. Downloaded by University of St Andrews Library on 09/12/2013 07:49:12.
Fig. 1 (a) Size distribution of the UCNPs used in optical trapping experiments.
The inset shows a TEM image revealing the quasi spherical shape of the NPs. (b)
Digital image of the UCNP aqueous colloidal dispersion and (c) digital image of
the same solution as that of (b) under 980 nm laser excitation in which the two-
photon visible luminescence generated by erbium ions after optical excitation of
ytterbium ions can be observed.
different uorescent dyes: Rhodamine B (RhoB) and Rhoda-
mine 110 (Rho110). Both dyes were provided by SigmaTech Inc.
and the concentrations in the solution were set to 0.25 mM and
0.05 mM for RhoB and Rho110, respectively. In our experiments
we used RhoB as a temperature dependent dye and Rho110 as a
reference dye. Thermal quenching causes a monotonous
decrease in the intensity ratio of both dyes that can be used to
determine the solution temperature with an accuracy of 1  C
(under our experimental conditions). The spatial variation of
temperature can be then obtained from the spatial variation of
the intensity ratio that was measured by using a commercial
epi-uorescence microscope. A high-voltage mercury lamp was
used as an excitation source and an adequate combination of
uorescence lters was used to obtain the spatial variation of
the ratio between the uorescence intensities generated from
Rho110 and RhoB. In particular, these two images were
obtained by using two sets of lters. For acquisition of a uo-
rescence image at 530 nm we used a set composed of a 450–500
excitation bandpass lter, a 500 nm longpass lter and a 510–
550 emission bandpass lter. For acquisition of a uorescence
12194 | Nanoscale, 2013, 5, 12192–12199
View Article Online
Paper
Fig. 2 Left: schematic diagram of the experimental set-up used for optical trap-
ping of UCNPs in an aqueous colloidal dispersion. Note that the presence of UCNPs
at the optical trap is monitored by recording their two-photon excited visible
luminescence. Right fluorescence image of the optical trap at different times after
switching on the 980 nm trapping beam. Sequential incorporation into the optical
trap of UCNPs is evidenced by the discrete increments in the intensity of visible
luminescence. Scale bar is 5 mm. 980 nm laser power was 200 mW and NaY-
F4:Er3+,Yb3+ UCNP concentration in water was 1010 UCNPs per cm3.
image at 580 nm we used a set composed of a 525–560 nm
excitation bandpass lter, a 575 nm longpass lter and a 580–
650 emission bandpass lter.
C Results and discussion
Optical trapping experiments were rst conducted on a
colloidal dispersion of UCNPs in distilled water with a particle
concentration of 1010 particles per cm3. The presence of UCNPs
at the optical trap is detected by a camera that images the visible
(upconversion) luminescence generated from the UCNP solu-
tion in the surroundings of the optical trap. The column on the
right in Fig. 2 shows a series of images captured by the CCD
camera revealing the temporal evolution of the luminescence
aer switching on the trapping and excitation 980 nm laser (the
laser power in this case was 200 mW). As can be observed,
immediately aer the 980 nm laser is switched on (t ¼ 0 s), no
visible luminescence was detected, indicating the absence of
UCNPs at the laser focus. Aer 40 seconds, a weak visible
luminescence at the laser focus is detected, an indication of the
trapping of UCNPs. Longer times lead to a more intense visible
luminescence, indicating the subsequent capture of additional
UCNPs into the optical trap. Thus, the emitted intensity at the
focal spot is attributed to UCNPs that are being accumulated by
laser trapping. Although Fig. 2 demonstrates optical trapping of
UCNPs, our experimental set-up does not allow us to determine
This journal is ª The Royal Society of Chemistry 2013

Paper
the actual size of the UCNPs being trapped. Indeed, based on
the results shown in Fig. 1(a), the solution contains UCNPs
which may be up to 40 nm in diameter.
In order to demonstrate the sequential loading of UCNPs
into the optical trap we have plotted the emission intensity
produced at the laser focus (i.e. inside the optical trap) as a
function of time (Fig. 3, le frame). From the intensity plot we
Published on 24 September 2013. Downloaded by University of St Andrews Library on 09/12/2013 07:49:12.
observe discrete intensity jumps and a variation in step heights.
In Fig. 3 (right frame) we have also included the intensity
derivative (dI/dt) curves which provide further evidence of the
presence of the intensity jumps (Fig. 3, le frame). According to
previous studies, dealing with optical trapping of single metallic
nanoparticles and submicrometre spheres, and the results
obtained in this work we can state that the presence of the
intensity jumps clearly reveals one-by-one incorporation of
UCNPs into the 980 nm optical trap.10b,f,12 During the initial
period (20–50 s, depending on the trapping power) no particles
are incorporated into the trap and only the detector readout
noise was recorded. Shortly aer this time, a single UCNP enters
and remains in the trap. This behaviour is analogous to the
previous results reported by Oddershede et al.12a describing
optical trapping of quantum dots. Our loading process is
repeated, leading to successive discrete increments in the
visible luminescence generated at the focus, so that each
intensity increment can be attributed to the incorporation of an
additional UCNP into the trap. The height of the observed
intensity steps has not been found to be homogeneous reecting
the polydispersity of the UCNPs (close to 35%) although other
Fig. 3 Left: time evolution of the two-photon excited luminescence generated by
the optically trapped UCNPs as obtained for three different trapping powers. Dots
are experimental data and solid lines act as guides for the eye. Right: time evolution
of the time derivative of the two-photon luminescence generated at the optical trap.
The sequential incorporation of single UCNPs is evidenced by the derivate peaks.
Dots represent experimentally observed data and solid lines act as guide for the eye.
All the data correspond to optical trapping experiments conducted in the aqueous
dispersion with NaYF4:Er3+,Yb3+ UCNPs at a concentration of 1010 UCNPs per cm3.
This journal is ª The Royal Society of Chemistry 2013
View Article Online
Nanoscale
effects (such as spatial re-location of the UCNPs into the optical
trap due to the presence of several UCNPs) might also be
contributing factors.12b,c
Although sequential trapping of individual UCNPs is
observed for laser powers above 50 mW, the UNCP loading rate
has been found to be strongly dependent on the trapping power.
If we dene the trapping period (Ttrap) as the elapsed time
between two consecutive UCNP trapping events (shown in
Fig. 3, right column) it is clear that Ttrap decreases with
increasing laser power (see Fig. 4). Indeed, this is not unex-
pected, since an increase in trapping power results in an
increase in laser intensity at focus, thus increasing the magni-
tude of the optical potential well (see expression (1)).11,12 This, in
turn, raises the trapping probability and reduces the time
interval between two consecutive trapping events. As a matter
of fact Ttrap is, indeed, proportional to the inverse of the trap-
ping probability. According to Askin's formalism, this proba-
bility depends on the magnitude of the optical potential
Uopt in such a way that the probability for a given particle to be
trapped is proportional to exp(Utrap/kBT). Thus, in a rst
order approximation, Ttrap is expected to be proportional to
exp(Utrap/kBT). According to expression (1), the magnitude of
the optical potential is proportional to the laser trapping power
so that if we write Uopt ¼ gPtrap we conclude that the trapping
period should decrease with the trapping power with an expo-
nential trend (i.e. Ttrap f exp(gPtrap/kBT)). The guide for the
eye included in Fig. 4 corresponds, indeed, to an exponential
decay curve, revealing the correctness of this rst order
approximation.
The trapping rate does not only depend on the laser power
employed but it is also expected to depend on the concentration
of UCNPs. Higher concentrations of UCNPs would reduce the
average distance between UCNPs and the optical trap, thus
increasing the trapping probability and, consequently, reducing
the time interval between two consecutive events. Therefore,
solutions with larger concentrations of UCNPs should show
shorter Ttrap values. This has been experimentally corroborated
in our experiments by measuring the trapping rates for two
different concentrations of UCNPs. Results, not shown for the
Fig. 4 Trapping period (Ttrap) as a function of the trapping power. Dots are
experimental data and the dashed line acts as a guide that corresponds to a
simple exponential decay.
Nanoscale, 2013, 5, 12192–12199 | 12195

Nanoscale
sake of brevity, revealed how Ttrap increases by almost 4 times
when the concentration was reduced from 1011 and 1010 UCNPs
per cm3. Further investigation is required to gain a full under-
standing on the relationship between the trapping rate and
UCNP and the role played by different processes such as clus-
tering and trapping.11 In any case, data included in Fig. 3 and 4
reveal that an accurate control over the trapping rate is essential
Published on 24 September 2013. Downloaded by University of St Andrews Library on 09/12/2013 07:49:12.
for achieving stable single-particle optical trapping.11 This can
be achieved by an appropriate choice of both trapping power
and nanoparticle concentration.
The trapping force acting on the UCNPs was determined by
using the so-called hydrodynamic-drag method.10g This consists
of the creation of a controlled uid ow around the trapped
nanoparticle and then measuring the uid speed needed to
remove the nanoparticle from the optical trap. This occurs
when the drag force (Fdrag) created by the ow rate exceeds the
optical trapping force (Ftrap) at a specic laser power. This
condition allows us to estimate the optical trapping force from
the experimentally determined critical velocity using the
following equation:10g,h
Ftrap ¼ Fdrag(nmax) ¼ 6prhnmax (3)
where r is the radius of the nanoparticle and h is the uid (water)
viscosity. The required ow was achieved by mounting the sample
chamber on a translating linear stage. For each laser trapping
power we measured the translating speed (nmax) required to
remove the UCNPs from the trap (monitored by the disappear-
ance of the visible luminescence generated from trap).13
At this point, it is important to note that water absorption at
980 nm (absorption coefficient
0.5 cm1) is not negligible. As a
consequence, local heating in the surroundings of the trap is
expected to occur due to absorption of laser light by the
medium.15 The magnitude of laser-induced trap heating has
been previously determined (under experimental conditions
identical to ours) using quantum dot uorescence thermom-
etry, as described elsewhere.15b We have experimentally
obtained that, due to water absorption, the temperature of the
trap increases as a function of the laser trapping power at a rate
close to 6  C per 100 mW.15b Thus for a trapping power of 100
mW, the temperature increment at the trap position was
measured to be close to 6  C. This laser-induced thermal
loading is of relevance when calculating the optical force using
eqn (3) since the kinematic viscosity of water decreases with
temperature. Therefore, to estimate the trapping force exerted
for a laser power of 100 mW the viscosity of water at 26  C
(0.7 mPa s) must be used instead of that at room temperature
(20  C, 0.8 mPa s). This temperature-induced modication of
water viscosity is not negligible and must be considered in order
to estimate a reliable trapping force value for each laser trap-
ping power. Thus, in the calculation of the trapping force we
have taken into account this varying water viscosity as a func-
tion of laser trapping power. The obtained trapping forces are
shown in Fig. 5. For low laser powers <50 mW trapping was not
observed. We conclude that for such low trapping powers, the
thermal energy is larger than the optical potential well and
stable optical trapping is not possible. At laser power higher
12196 | Nanoscale, 2013, 5, 12192–12199
View Article Online
Paper
Fig. 5 Power dependence of the trapping force acting on UCNPs (dispersed in
both distilled and heavy water) as determined by the so-called hydrodynamic-
drag method. Dots are experimental data and dashed lines are the best fit to force
data obtained for trapping laser powers above 50 mW. The inset shows the
numerical simulation of the force field acting on 20 nm radius NaYF4:Er3+,Yb3+
UCNPs in an aqueous medium for a laser trapping power of 100 mW as obtained
by the simulation software developed by Mas et al.14
than 50 mW, optical trapping was observed with a high degree
of stability. For such trapping powers, the trapping force was
found to increase almost linearly with the laser power at a rate
of 200 fN W1 (see dashed lines in Fig. 5). The linear trend
suggests the absence of dominant heating effects. When heat-
ing effects are present, a superlinear behaviour between the
trapping force and laser power appears as reported previously
during the optical trapping of gold nanoparticles.16
At this point, it should also be noted that the 980 nm laser
trapping radiation is absorbed by both the medium and nano-
particles. This could give rise to an unequal temperature
distribution around the nanoparticle and will affect the overall
trapping force measured. There are two different photothermal
effects that could lead to particle trapping: photophoresis and
thermophoresis, associated with the absorption of light by
either the suspended particles or the medium.17 Photophoresis
is related to the creation of a non-uniform temperature distri-
bution within the suspended particles due to light absorptio-
n.17a,d In our case, photophoresis is expected to be a second
order effect due to the nanometric size of UCNPs in comparison
to that of the trapping wavelength. This, in conjunction with the
characteristic low absorption cross section of the lanthanide
ions, leads to negligible single particle absorption. On the other
hand, thermophoresis does not depend on the particle but on
the absorption of the surrounding medium (water). Thermo-
phoresis affects the motion of the particle due to the presence of
a thermal gradient.17b,c The presence of a thermal gradient in
the surroundings of the laser spot can not only affect particle
motion by either photophoresis or thermophoresis but it can
also contribute to particle accumulation through the creation of
convection currents.18 In the present study, thermal gradients
around the optical trap are known to exist. Therefore, the
contribution of the above mentioned thermal effects cannot be
discarded a priori. In order to elucidate the possible
This journal is ª The Royal Society of Chemistry 2013

Paper
contribution/presence of thermally assisted particle accumula-
tion effects in this study the same trapping experiments were
performed but using heavy water (D2O) as the solvent instead of
water. The reason for this is that its absorption coefficient of 980
nm for heavy water is <0.02 cm1 (see Fig. 6(a) in which we have
included the absorption coefficient of both D2O and H2O).
Therefore, medium heating, while relevant to H2O, is expected
Published on 24 September 2013. Downloaded by University of St Andrews Library on 09/12/2013 07:49:12.
to be negligible in D2O. This has been indeed corroborated by
uorescence thermal imaging of a 980 nm optical trap when
created in H2O and D2O. Fig. 6(c) and (d) show the temperature
spatial distribution in the surroundings of the 980 nm laser
spots in H2O and D2O (laser power was in both cases equal to
200 mW). Images were obtained by using the thermosensitive
uorescence dye described in the Experimental section and
analyzing the spatial dependence of dye emission in a double
beam epi-uorescence microscope as also described else-
where.19 Fig. 6(d) clearly demonstrates the absence of any
thermal loading when optical trapping is performed in D2O at
variance with thermal loading produced in water. Nevertheless,
we have been able to observe particle-by-particle optical trapping
of NaYF4:Er3+,Yb3+ NPs in D2O. Fig. 6(b) shows the time evolu-
tion of the trap luminescence when the 980 nm laser beam
(370 mW) was focused into a solution of UCNPs in D2O. The
observation of optical trapping in D2O reinforces our view that
optical forces are the main mechanism leading to the observed
trapping of the nanoparticles.
We have also measured the trapping force as a function of
the laser trapping power for UCNPs dispersed in heavy water
Fig. 6 (a) Absorption coefficient of distilled and heavy water. The wavelength of
the laser used for trapping in this work (980 nm) is indicated by a dashed line.
Note that the absorption coefficient of heavy water vanishes at this wavelength.
(b) Time evolution of the two-photon excited luminescence generated by
the optically trapped UCNPs dispersed in heavy water as obtained for a
trapping power of 370 mW. Dots are experimental data and dashed lines act
as guide to the eye. (c) Spatial temperature distribution in the surroundings of a
980 nm laser beam when it is focused (with the same optics as those used for
optical trapping) into a micro-channel filled with distilled water. (d) Spatial
temperature distribution in the surroundings of a 980 nm laser beam when it is
focused (with the same optics as those used for optical trapping) into a micro-
channel filled with heavy water. Dashed lines in thermal images indicate the
position of the 980 nm laser beam.
This journal is ª The Royal Society of Chemistry 2013
View Article Online
Nanoscale
(Fig. 5). As for UCNPs dispersed in H2O, the trapping force
increases linearly with the laser power. The maximum trapping
force measured in the absence of laser induced medium heat-
ing (in D2O) was found to be close to 10 fN for a trapping laser
power of 125 mW (maximum laser power achievable during
force measurements). The absence of any thermal gradient
allows us to unequivocally relate the measured trapping force to
a pure optical force. Assuming the trap radius to be equal to the
980 nm beam radius at focus (i.e. 0.7 mm) we have estimated an
optical trapping force constant (ktrap) close to 0.1 fN nm1 W1.
This trapping force constant is lower than the values previously
reported for other dielectric nanoparticles such as latex beads
(1 pN nm1 W1),10c silicon nanoparticles (0.1 pN nm1 W1),13
polystyrene nanobeads (2 fN nm1 W1),20 and semiconductor
quantum dots (2.4 pN nm1 W1).12a The experimentally
obtained optical trapping force constant was compared to that
predicted from theory. Based on the formalism described by
Svoboda et al.10f ktrap can be written as:
pffiffiffi
2jaNP j e
ktrap ¼ (4)
pnm 2 wl 4 c
where |aNP| is the nanoparticle polarizability that depends on
the NP size (see eqn (2)), nm is the refractive index of the
medium (water in our case) and wl is the beam waist of the
trapping laser. Using eqn (4) a ktrap value of 1 fN nm1 W1 was
obtained using a maximum particle size of 40 nm in diameter
(see Fig. 1(a)). This value can be compared with the experi-
mental value of 0.1 fN nm1 W1. To further establish an
agreement between the measured trapping force and the theo-
retical prediction we have also performed numerical simula-
tions of the force eld acting on the UCNPs by using the
computer program developed by Mas et al.14 In the present
study, the diameter of the trapped particle is signicantly
smaller than the wavelength of light; therefore the simulations
treat the nanoparticle as a point dipole in an inhomogeneous
electromagnetic eld. The inset in Fig. 5 shows the calculated
force eld acting on a nanoparticle with a 20 nm radius placed
at the focus of a 100 mW, 980 nm laser beam focused using a
0.8 NA microscope objective. The results obtained from the
simulation show that the maximum force acting on the UCNPs
is approximately 20 fN, which is in good agreement with the
experimentally determined trapping force of 15 fN for a laser
power of 100 mW (Fig. 5). Based on this good agreement
between the simulation and the experimental result we
conclude that the main mechanism leading to optical trapping
of the UCNPs is the interaction between the laser electric eld
and the induced dipole in the nanoparticles. This discrepancy
between the experimental and calculated trapping forces was
also found in previous studies dealing with optical trapping on
NPs.12a It may be attributed to either the uncertainty due to
particle size (due to dispersion) or the non-accurate determi-
nation of the nanoparticle polarizability (due to the presence of
different surface coatings and/or the additional contribution of
resonantly excited ytterbium ions to the overall NP
polarizability).12a
Finally, it should be noted that in those applications in
which optical trapping of UCNPs should be achieved with
Nanoscale, 2013, 5, 12192–12199 | 12197

Nanoscale
minimum thermal contribution, such as those involving bio-
systems, the trapping wavelength could be tuned from 980 nm
to wavelengths such as 920 nm which show lower absorption by
water but are still absorbed by ytterbium ions (so that two-
photon emission would still be generated for particle tracking).
D Conclusions
Published on 24 September 2013. Downloaded by University of St Andrews Library on 09/12/2013 07:49:12.
We have demonstrated the optical trapping of upconverting
NaYF4:Er3+,Yb3+ nanoparticles by using a single 980 nm laser
beam. This laser beam has two main functions: creating an
optical trap and exciting dopant lanthanide ions resulting in the
characteristic green upconversion emission of Er3+ ions. As a
result of the dual role of the 980 nm laser, it is possible to
observe particle-by-particle incorporation into the optical trap by
simply monitoring the visible upconversion luminescence
generated at the trap volume. It has also been found that the
trapping rate (rate at which the NaYF4:Er3+,Yb3+ UCNPs are
incorporated into the trap) can be controlled by adjusting
the trapping power and/or the initial concentration of
NaYF4:Er3+,Yb3+ nanoparticles in the solution. The comparison
between the experimentally measured trapping forces (20 fN per
100 mW of laser trapping power) and numerical simulation in
the dipolar approximation suggests that photothermal effects
are not the main mechanisms causing optical trapping. This
fact has been corroborated by the demonstration of optical
trapping of NaYF4:Er3,Yb3+ UCNPs dispersed in heavy water
(D2O), which do not absorb 980 nm laser radiation and hence
avoid the creation of any thermal gradient.
The results reported here provide the possibility of achieving
single particle trapping, manipulation and direct observation of
the trapped particle thanks to the luminescence generated by
the upconverting nanoparticle.
Acknowledgements
This work was supported by the Universidad Autónoma de
Madrid and Comunidad Autónoma de Madrid (Project S2009/
MAT-1756), by the Spanish Ministerio de Educacion y Ciencia
(MAT2010-16161) and by Fundación Dr Manuel Morales.
L.M.M. thanks the Spanish Ministerio de Economia y com-
petitividad (MINECO) for a FPI grant and B.R. acknowledges the
Universidad Autónoma de Madrid for a FPI grant. K.D. is a
Royal Society-Wolfson Merit Award Holder. JAC is a Concordia
University Research Chair in Nanoscience and is grateful to
Concordia University for nancially supporting his research.
JAC is also grateful for nancial support from the Natural
Science and Engineering Research Council (NSERC) of
Canada. EMR acknowledges nancial support from the Marie
Curie Fellowship Program (FP7-PEOPLE-2010- IOF-274404
“LUNAMED”).
Notes and references
1 (a) D. J. Stevenson, F. Gunn-Moore and K. Dholakia, Light
forces the pace: optical manipulation for biophotonics,
J. Biomed. Opt., 2010, 15(4), 041503; (b) P. J. Reece,
12198 | Nanoscale, 2013, 5, 12192–12199
View Article Online
Paper
S. Paiman, O. Abdul-Nabi, Q. Gao, M. Gal, H. H. Tan and
C. Jagadish, Combined optical trapping and
microphotoluminescence of single InP nanowires, Appl.
Phys. Lett., 2009, 95(10), 101109; (c) A. Koniger and
W. Kohler, Optical funneling and trapping of gold colloids
in convergent laser beams, ACS Nano, 2012, 6(5), 4400–
4409; (d) T. T. Perkins, Optical traps for single molecule
biophysics: a primer, Laser Photonics Rev., 2009, 3(1–2),
203–220.
2 (a) P. C. Ashok and K. Dholakia, Optical trapping for
analytical biotechnology, Curr. Opin. Biotechnol., 2012,
23(1), 16–21; (b) K. Dholakia, P. Reece and M. Gu, Optical
micromanipulation, Chem. Soc. Rev., 2008, 37(1), 42–55.
3 (a) F. Wang, D. Banerjee, Y. S. Liu, X. Y. Chen and X. G. Liu,
Upconversion nanoparticles in biological labeling, imaging,
and therapy, Analyst, 2010, 135(8), 1839–1854; (b) A. Gnach
and A. Bednarkiewicz, Lanthanide-doped up-converting
nanoparticles: merits and challenges, Nano Today, 2012,
7(6), 532–563; (c) J. F. Suyver, J. Grimm, M. K. van Veen,
D. Biner, K. W. Kramer and H. U. Gudel, Upconversion
spectroscopy and properties of NaYF4 doped with
Er3+,Tm3+ and/or Yb3+, J. Lumin., 2006, 117(1), 1–12; (d)
M. Yu, F. Li, Z. Chen, H. Hu, C. Zhan, H. Yang and
C. Huang, Laser scanning up-conversion luminescence
microscopy for imaging cells labeled with rare-earth
nanophosphors, Anal. Chem., 2009, 81(3), 930–935.
4 F. Auzel, Upconversion and anti-Stokes processes with f and
d ions in solids, Chem. Rev., 2004, 104(1), 139–173.
5 (a) J. F. Suyver, J. Grimm, K. W. Kramer and H. U. Gudel,
Highly efficient near-infrared to visible up-conversion
process in NaYF4:Er3+,Yb3+, J. Lumin., 2005, 114(1), 53–59;
(b) K. W. Kramer, D. Biner, G. Frei, H. U. Gudel,
M. P. Hehlen and S. R. Luthi, Hexagonal sodium yttrium
uoride based green and blue emitting upconversion
phosphors, Chem. Mater., 2004, 16(7), 1244–1251.
6 (a) Y. M. Bae, Y. I. Park, S. H. Nam, J. H. Kim, K. Lee,
H. M. Kim, B. Yoo, J. S. Choi, K. T. Lee, T. Hyeon and
Y. D. Suh, Endocytosis, intracellular transport, and
exocytosis of lanthanide-doped upconverting nano-
particles in single living cells, Biomaterials, 2012, 33(35),
9080–9086; (b) F. Wang, D. K. Chatterjee, Z. Q. Li,
Y. Zhang, X. P. Fan and M. Q. Wang, Synthesis of
polyethylenimine/NaYF4 nanoparticles with upcon-
version uorescence, Nanotechnology, 2006, 17(23), 5786–
5791.
7 L. M. Maestro, E. M. Rodriguez, F. Vetrone, R. Naccache,
H. L. Ramirez, D. Jaque, J. A. Capobianco and J. G. Sole,
Nanoparticles for highly efficient multiphoton uorescence
bioimaging, Opt. Express, 2010, 18(23), 23544–23553.
8 Y. Zhang, W. Li, in Upconverting Fluorescent Nanoparticles
for Bioimaging and Therapy, Optical Society of America,
2012, p. AS3E.1.
9 F. Vetrone, R. Naccache, A. J. de la Fuente, F. Sanz-Rodriguez,
A. Blazquez-Castro, E. M. Rodriguez, D. Jaque, J. G. Sole and
J. A. Capobianco, Intracellular imaging of HeLa cells by non-
functionalized NaYF4:Er3+,Yb3+ upconverting nanoparticles,
Nanoscale, 2010, 2(4), 495–498.
This journal is ª The Royal Society of Chemistry 2013

Paper
10 (a) M. L. Juan, R. Gordon, Y. J. Pang, F. Eekhari and
R. Quidant, Self-induced back-action optical trapping of
dielectric nanoparticles, Nat. Phys., 2009, 5(12), 915–919;
(b) A. Usman, W. Y. Chiang and H. Masuhara, Optical
trapping and polarization-controlled scattering of dielectric
spherical nanoparticles by femtosecond laser pulses,
J. Photochem. Photobiol., A, 2012, 234, 83–90; (c) Y. J. Pang
Published on 24 September 2013. Downloaded by University of St Andrews Library on 09/12/2013 07:49:12.
and R. Gordon, Optical trapping of 12 nm dielectric
spheres using double-nanoholes in a gold lm, Nano Lett.,
2011, 11(9), 3763–3767; (d) A. H. J. Yang, S. D. Moore,
B. S. Schmidt, M. Klug, M. Lipson and D. Erickson, Optical
manipulation of nanoparticles and biomolecules in sub-
wavelength slot waveguides, Nature, 2009, 457(7225), 71–
75; (e) A. Ashkin, J. M. Dziedzic, J. E. Bjorkholm and
S. Chu, Observation of a single-beam gradient force optical
trap for dielectric particles, Opt. Lett., 1986, 11(5), 288–290;
(f) K. Svoboda and S. M. Block, Optical trapping of metallic
Rayleigh particles, Opt. Lett., 1994, 19(13), 930–932; (g)
K. C. Neuman and S. M. Block, Optical trapping, Rev. Sci.
Instrum., 2004, 75(9), 2787–2809; (h) A. Jonas and
P. Zemanek, Light at work: the use of optical forces for
particle manipulation, sorting, and analysis,
Electrophoresis, 2008, 29(24), 4813–4851; (i) M. Dienerowitz,
M. Mazilu and K. Dholakia, Optical manipulation of
nanoparticles: a review, J. Nanophotonics, 2008, 2, 021875.
11 C. Hosokawa, H. Yoshikawa and H. Masuhara, Cluster
formation of nanoparticles in an optical trap studied by
uorescence correlation spectroscopy, Phys. Rev. E: Stat.,
Nonlinear, So Matter Phys., 2005, 72(2), 021408.
12 (a) L. Jauffred, A. C. Richardson and L. B. Oddershede,
Three-dimensional optical control of individual quantum
dots, Nano Lett., 2008, 8(10), 3376–3380; (b) F. Hajizadeh
and S. N. S. Reihani, Optimized optical trapping of gold
nanoparticles, Opt. Express, 2010, 18(2), 551–559; (c)
K. C. Toussaint, M. Liu, M. Pelton, J. Pesic, M. J. Guffey,
P. Guyot-Sionnest and N. F. Scherer, Plasmon resonance-
based optical trapping of single and multiple Au
nanoparticles, Opt. Express, 2007, 15(19), 12017–12029.
13 M. G. Donato, M. A. Monaca, G. Faggio, L. D. Stefano,
P. H. Jones, P. G. Gucciardi and O. M. Marago, Optical
This journal is ª The Royal Society of Chemistry 2013
View Article Online
Nanoscale
trapping of porous silicon nanoparticles, Nanotechnology,
2011, 22(50), 505704.
14 J. Mas, A. Farre, J. Cuadros, I. Juvells and A. Carnicer,
Understanding optical trapping phenomena: a simulation
for undergraduates, IEEE Trans. Educ., 2011, 54(1), 133–140.
15 (a) Y. Liu, D. K. Cheng, G. J. Sonek, M. W. Berns,
C. F. Chapman and B. J. Tromberg, Evidence for localized
cell heating induced by infrared optical tweezers, Biophys.
J., 1995, 68(5), 2137–2144; (b) P. Haro-González,
W. T. Ramsay, L. M. Maestro, B. del Rosal, K. Santacruz-
Gomez, M. del Carmen Iglesias-de la Cruz, F. Sanz-
Rodrı́guez, J. Y. Choo, P. R. Sevilla, M. Bettinelli,
D. Choudhury, A. K. Kar, J. G. Solé, D. Jaque and
L. Paterson, Quantum dot-based thermal spectroscopy and
imaging of optically trapped microspheres and single cells,
Small, 2013, 9(12), 2162–2170.
16 Y. Seol, A. E. Carpenter and T. T. Perkins, Gold
nanoparticles: enhanced optical trapping and sensitivity
coupled with signicant heating, Opt. Lett., 2006, 31(16),
2429–2431.
17 (a) S. Tehranian, F. Giovane, J. Blum, Y.-L. Xu and
B. Å. S. Gustafson, Photophoresis of micrometer-sized
particles in the free-molecular regime, Int. J. Heat Mass
Transfer, 2001, 44(9), 1649–1657; (b) R. T. Schermer,
C. C. Olson, J. P. Coleman and F. Bucholtz, Laser-induced
thermophoresis of individual particles in a viscous liquid,
Opt. Express, 2011, 19(11), 10571–10586; (c) Y. Lamhot,
A. Barak, O. Peleg and M. Segev, Self-trapping of optical
beams through thermophoresis, Phys. Rev. Lett., 2010,
105(16), 163906; (d) A. Ashkin, Acceleration and trapping
of particles by radiation pressure, Phys. Rev. Lett., 1970,
24(4), 156–159.
18 S. Duhr and D. Braun, Two-dimensional colloidal crystals
formed by thermophoresis and convection, Appl. Phys.
Lett., 2005, 86(13), 131921.
19 D. Jaque and F. Vetrone, Luminescence nanothermometry,
Nanoscale, 2012, 4(15), 4301–4326.
20 E. J. Peterman, F. Gittes and C. F. Schmidt, Laser-induced
heating in optical traps, Biophys. J., 2003, 84(2 Pt 1), 1308–
1316.
Nanoscale, 2013, 5, 12192–12199 | 12199