Upconversion luminescence in sub-10 nm β-NaGdF4:Yb3+,Er3+ nanoparticles: An improved synthesis in anhydrous ionic liquids

Article abstract of DOI:10.1039/c9ra05950d

Sub-10 nm β-NaGdF₄:18% Yb³⁺,2% Er³⁺ nanoparticles were synthesized in ethylene glycol and various ionic liquids under microwave heating. The products were characterized by powder X-ray diffraction, electron microscopy, and upconversion (UC) luminescence spectroscopy. After Yb³⁺ excitation at 970 nm, Er³⁺ ions are excited by energy transfer upconversion and show the typical green and red emission bands. The UC luminescence intensity was optimized with respect to reactant concentrations, solvents, and reaction temperature and time. The strongest UC emission was achieved for sub-20 nm core-shell nanoparticles which were obtained in the ionic liquid diallyldimethylammonium bis(trifluoromethanesulfonyl)amide from a two-step synthesis without intermediate separation. Strictly anhydrous reaction conditions, a high fluoride/rare earth ion ratio, and a core-shell structure are important parameters to obtain highly luminescent nanoparticles. These conditions reduce non-radiative losses due to defects and high energy acceptor modes of surface ligands. A low power excitation of the core-shell particles by 70 mW at 970 nm results in an impressive UC emission intensity of 0.12% compared to the bulk sample.

Full text of DOI:10.1039/c9ra05950d

RSC Advances
View Article Online
View Journal | View Issue
PAPER
Upconversion luminescence in sub-10 nm b-
NaGdF4:Yb3+,Er3+ nanoparticles: an improved
Cite this: RSC Adv., 2019, 9, 34784
synthesis in anhydrous ionic liquids†
ab
c
a
* *
and Karl W. Kramer
¨
Gabriella Tessitore,
Anja-Verena Mudring
Sub-10 nm b-NaGdF4:18% Yb3+,2% Er3+ nanoparticles were synthesized in ethylene glycol and various
ionic liquids under microwave heating. The products were characterized by powder X-ray diffraction,
electron microscopy, and upconversion (UC) luminescence spectroscopy. After Yb3+ excitation at
970 nm, Er3+ ions are excited by energy transfer upconversion and show the typical green and red
emission bands. The UC luminescence intensity was optimized with respect to reactant concentrations,
solvents, and reaction temperature and time. The strongest UC emission was achieved for sub-20 nm
core–shell nanoparticles which were obtained in the ionic liquid diallyldimethylammonium
bis(trifluoromethanesulfonyl)amide from a two-step synthesis without intermediate separation. Strictly
anhydrous reaction conditions, a high fluoride/rare earth ion ratio, and a core–shell structure are
important parameters to obtain highly luminescent nanoparticles. These conditions reduce non-
radiative losses due to defects and high energy acceptor modes of surface ligands. A low power
excitation of the core–shell particles by 70 mW at 970 nm results in an impressive UC emission
intensity of 0.12% compared to the bulk sample.
Received 31st July 2019
Accepted 25th September 2019
DOI: 10.1039/c9ra05950d
rsc.li/rsc-advances
nine-fold coordination, which implies non-centrosymmetric RE
sites. Microscopic disorder gives rise to line broadening and
1. Introduction
a higher probability for spectroscopic overlap of the UC steps.
Due to longer RE–F distances and a higher coordination
number, hexagonal b-NaREF4 has lower effective phonon
energy than the cubic a-phase. Accordingly, the UC emission
intensity of b-NaREF4 is about an order of magnitude stronger
compared to the respective a-NaREF4 host lattice.8 UC mecha-
nisms were investigated in detail,4 and recently it was shown
¨
About a decade ago, the work of Haase and Gudel et al. initiated
the investigation of upconversion (UC) luminescence from
transparent solutions of nanoparticles.1,2 They reported
a synthetic route to cubic a-NaYF4 : Yb3+, Er3+ and hexagonal b-
NaGdF4:Yb3+,Er3+ nanoparticles dispersed in dimethyl sulf-
oxide. Aer 980 nm laser excitation of Yb3+, blue, green, and red
UC emission from Er3+ was observed from clear, colorless
solutions. Since then the number of publications on upcon-
version grew exponentially, about half of them investigating
nanoparticles, most prominent with Yb3+ and Er3+ cooping.3
Auzel gave a comprehensive review on UC ions and processes.4
Hexagonal sodium rare earth uorides, b-NaREF4 (RE ¼ rare
earth), are among the most efficient UC host lattices.5–7 Two
crystallographically distinct RE sites were structurally and
spectroscopically characterized in b-NaGdF4:Er3+.8 Both have
that the Er3+ 4 9/2 / 4I15/2 emission in b-NaYF4 originates from
F
cross-relaxation but not from multi-phonon relaxation.9,10
Syntheses for sodium rare earth uoride nanoparticles were
reported by many groups.11–14
UC nanoparticles were used in various applications, e.g. bio-
logical imaging, photodynamic therapy, and nano-thermom-
etry.15–20 Continuous improvements of the syntheses yielded
nanoparticles of smaller size and stronger luminescence;21–26
water-dispersible particles are more prone for biological imaging
applications.27–30 UC luminescence nanoparticles provide several
advantages as biological uorophores. They show low light
scattering, no auto bleaching, and near infrared excitation light
has a deep tissue penetration depth.15 Both in vitro and in vivo
experiments led to encouraging results without signicant hints
for any cytotoxicity of sodium rare earth uoride nanoparticles.15
Syntheses of UC nanoparticles in high boiling solvents or
under hydrothermal conditions are meanwhile well estab-
lished.1,2,11–28 Oen, oleic acid is used as coordinating ligand.
Oxide ions from its carboxyl groups can substitute for uoride
aUniversity of Bern, Department of Chemistry and Biochemistry, Freiestrasse 3, 3012
Bern, Switzerland. E-mail: karl.kraemer@dcb.unibe.ch
bConcordia University, Department of Chemistry and Biochemistry, 7141
Sherbrooke W., H4B1R6, Montreal, QC, Canada
cStockholm University, Department of Materials and Environmental Chemistry, Svante
¨
Arrhenius vag 16 C, 106 91 Stockholm, Sweden. E-mail: anja-verena.mudring@mmk.
su.se
† Electronic supplementary information (ESI) available: Thermogravimetric
analyses and XRD patterns of rare earth acetates, XRD patterns and UC
luminescence spectra, power density of the laser diode. See DOI:
10.1039/c9ra05950d
34784 | RSC Adv., 2019, 9, 34784–34792
This journal is © The Royal Society of Chemistry 2019

View Article Online
Paper
RSC Advances
in the rare earth coordination sphere on the particle surface due DADMA Cl results in the precipitation of AgCl, which was
to the similar size of O and F. High energy vibration modes of ltered off. DADMA NTf2 (bis(triuoromethanesulfonyl)amide)
the ligands increase losses by multi-phonon relaxation.31 The was synthesized from LiNTf2 (Alfa Aesar, 98+%), as reported in
high surface-to-volume ratio results in a signicant number of the literature.45 The products were decolorized over an active
(surface) defects in nanoparticles which decrease the lumines- carbon column,49 and nally dried at 80 ꢁC under dynamic
cence compared to the bulk material. Accordingly, the UC vacuum overnight. DADMA ILs were obtained as clear, trans-
emission from nanoparticles is still about three orders of parent, colorless liquids, and stored under nitrogen
magnitude lower than for the respective mm-size phosphor atmosphere.
powder.21,32 A shell of undoped host material efficiently protects
2.1.2. Ethylene glycol purication. EG (Dr Grogg Chemie
the active core and reduces quenching from surface defects and AG, Switzerland) was puried and dried as described for the ILs
ligands.31,33 Finally, surface functionalization is required to above. This equivalent treatment avoids any bias in the
achieve water-dispersible nanoparticles.31
comparison between the different solvents due to the purica-
Microwave heating was successfully applied for syntheses in tion process.
polar or ionic solvents.34 The homogeneous heating results in
2.1.3. Core nanoparticle synthesis. b-NaGdF4:18% Yb3+,2%
lower temperature and shorter time for the reaction.35,36 Er3+ nanoparticles were synthesized from rare earth acetates,
Ethylene glycol (EG) and ionic liquids (ILs) were prociently NaCl, and a uoride source in various solvents. The rare earth
used in microwave syntheses of luminescent materials.37–43 acetates were prepared from the oxides by dissolving Gd2O3,
Uniform nucleation of particles was achieved which results in Er2O3, and Yb2O3 (Metal Rare Earth Limited, 5 N) in a mixture of
a narrow particle siꢀze distribution. ILs with complex uoride glacial acetic acid (Merck, 99.9%) and bidistilled water at
anions such as BF4 have a dual use as solvent and uoride 120 ꢁC. Aer a transparent solution was obtained, the liquid was
source. Cubic a-NaYF4 : Yb3+, Er3+ nanoparticles were obtained dried up on a heating plate. The dry powder was converted into
from a microwave synthesis in BF4 ILs.39,44 A synthesis in the the anhydrous acetates at 180 C under vacuum. The reaction
ꢁ
imidazolium IL 1-butyl-3-methylimidazolium tetrauoroborate, was monitored by thermogravimetric analysis and powder X-ray
(C4mim) (BF4), yields phase pure hexagonal nanoparticles.44 diffraction, see Fig. S1 and S2 (ESI†).
The ions determine the heating efficiency36 and the properties
In a typical reaction, 30 mg (0.085 mmol) anhydrous rare
of the IL, e.g. thermal stability, dielectric constant, ionic earth acetate RE(AcO)3 (RE ¼ Gd, Er, Yb) was ground with 10 mg
conductivity, and water content.45–47 The dielectric heating (0.17 mmol) NaCl (Reagent grade, Merck) under inert atmo-
depends on the dipole relaxation and ionic motion of the IL and sphere and then mixed with 0.5 ml solvent (EG or IL). The molar
increases with the microwave absorption.
between tetrauoroborate (BF4)
A
and
comparison ratios were xed to 80/2/18 for Gd/Er/Yb and 2/1 for Na/RE in all
bis(tri- samples. These ratios were optimized for bulk samples before.6,7
uoromethanesulfonyl)amide (NTf2) ILs revealed that asym- The acetate solution was heated to 60 ꢁC for 5 minutes with
metric NTf2 heat more efficiently than symmetric BF4 anions,36 a Biotage Initiator+ Robot Eight microwave synthesis system
because the dipole relaxation takes place only in asymmetric (Biotage AB, Uppsala, Sweden). Fluoride-containing ILs or NH4F
anions. Hydrophobic anions such as NTf2 result in lower (Alfa Aesar, 96%) ꢀwere used as uoride sources; the effects of
moisture content of the IL than for hydrophilic anions48 and the reactant (BF4 or NH4F) and the Fꢀ concentration (60–
such reduce the number of OHꢀ defects in uoride products. In 180 mg, 1.62–4.86 mmol) were evaluated. In a microwave glass
the present paper, we investigate the inuence of ILs with dia- vial, the uoride source was added to 1.5 ml of an EG/IL solu-
llyldimethylammonium cations and several different counter tion or the IL alone and stirred up to homogeneity. The acetate
anions on the synthesis of b-NaGdF4:Yb3+,Er3+ nanoparticles suspension was added dropwise to the microwave vial and
with respect to phase purity and particle size. The synthetic stirred overnight before the microwave heating. The tempera-
parameters, as well as the core, core/shell, and ligand design, ture in the reactor was controlled by dry compressed air (4 bar),
were optimized for UC luminescence intensity.
and microwave powers of 400 W or 50 W were used for EG/IL
solutions or pure IL, respectively. For separation, the nano-
particles were sequentially washed several times with methanol
and ethanol. They were collected by centrifugation at 3000 rpm
for 5 min and dried at 80 ꢁC overnight. Alternatively, the
2. Experimental
2.1. Sample preparation
2.1.1. Ionic liquid synthesis and purication. The ILs for nanoparticles were kept in solution for subsequent growth of an
our investigations were prepared by a metathesis reaction from undoped shell. Experimental conditions for all samples are
diallyldimethylammonium chloride (DADMA Cl, Sigma Aldrich, summarized in Table 1.
65% w/w in water) and the respective sodium, silver, or lithium
2.1.4. Core-shell nanoparticle synthesis. Aer the core
salt in saturated aqueous solution with a 2% stoichiometric synthesis, another 1/3 of the NH4F mass used in the rst step
excess of the salt.45 DADMA OTf (triuoromethanesulfonate) was added to the nanoparticle solution and stirred. A solution
and DADMA BF4 (tetrauoroborate) were obtained from the of 30 mg gadolinium acetate and 10 mg NaCl in 0.5 ml of the
respective sodium salt (NaOTf, Sigma Aldrich, 98%; NaBF4, respective solvent was prepared and heated to 60 ꢁC for 5 min by
Sigma Aldrich, 98%), following an established synthesis route microwave irradiation. The solution was added dropwise to the
for DADMA BF4.45 DADMA TFA (triuoroacetate) was prepared core nanoparticle solution and stirred for 10 minutes. Aer-
ꢁ
from a solution of AgTFA (Merck, 98+%). The reaction with ward, the solution was heated to 200 C for 5 min with 50 W
This journal is © The Royal Society of Chemistry 2019
RSC Adv., 2019, 9, 34784–34792 | 34785

View Article Online
RSC Advances
Paper
Table 1 Reaction conditions and compositions of the samples
RE(AcO)3
(mg)
NH4F
(mg)
Sample
EG (ml)
IL (ml)
IL anion
NaCl (mg)
T (ꢁC)
Time (min)
P (W)
Phase
EG1
EG2
EG3
EG4
EG5
EG6
EG7
EG8
EG9
EG10
EG11
EG12
EG13
IL1
IL2
IL3
IL4
IL5
IL6
IL7
IL8
IL9
1
1
1
1
BF4
OTf
OTf
OTf
OTf
OTf
OTf
OTf
OTf
OTf
OTf
NTf2
NTf2
OTf
NTf2
NTf2
NTf2
NTf2
NTf2
NTf2
NTf2
NTf2
NTf2
60
60
60
60
60
60
60
60
30
30
30
30
30
30
30
30
30
30
30
30
30
30
30
20
20
20
20
20
20
20
20
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
0
80
60
80
200
200
120
120
120
120
160
200
200
200
200
200
200
200
200
225
225
225
225
200
250
250
200
30
30
30
30
30
30
30
30
30
30
30
30
30
30
30
5
5
5
5
5
5
10
5
400
400
400
400
400
400
400
400
400
400
400
400
400
50
50
50
50
50
50
50
50
50
a
b
0.5
0.5
0.5
0.5
0.5
0.5
0.5
1.5
2
0.5
1.5
0
0
0
0
0
0
0
0
0
0
1.5
1.5
1.5
1.5
1.5
1.5
1.5
0.5
0
1.5
0.5
2
2
2
2
2
2
2
2
2
0.5
b
b
100
120
120
120
120
120
120
120
120
120
120
60
100
140
180
140
140
140
33
b
b
b
b
b
b
b
b
b
b
b
a/b
b
b
b
b
a/b
a/b
b
CS_IL4
50
microwave power. The core–shell nanoparticles were separated Veleta side-mounted CCD camera (Olympus) with 4 megapixels
as described above.
at 30 000-fold magnication (pictures with 500 nm scale),
100 000-fold magnication (picture with 200 nm scale), and
300 000-fold magnication (pictures with 50 nm scale). High-
resolution TEM micrographs were recorded on a G2 F20 FE-
TEM (Tecnai) using a ZrO2/W (100) Schottky eld emitter gun.
A Falcon II direct electron detector (FEI) with 4096 ꢂ 4096 pixels
and an Orius SC200 CCD camera (Gatan Inc.) with 2048 ꢂ 2084
pixels were used. The specimens were prepared from diluted
ethanol solutions on carbon-coated copper grids (300 mesh).
The pictures were analyzed by the ImageJ 1.49v soware
(National Institute of Health, USA).
2.2. Experimental methods
X-ray diffraction (XRD) patterns were measured on a StadiP
powder diffractometer (STOE, Darmstadt, Germany) in trans-
mission geometry at room temperature. Cu Ka1 radiation (l ¼
ꢀ
1.540 598 A) was used from a 40 kV 40 mA X-ray source and
a focusing Ge (111) monochromator. Layers of nanoparticles
were prepared on thin plastic foil. Data were recorded by
a Mythen 1K linear position sensitive detector (Dectris, Baden,
Switzerland) with 0.01ꢁ resolution in 2-theta.
Upconversion luminescence spectra were measured in
a standardized setup. Dry powder samples were lled in glass
tubes of 1 mm inner diameter. The nanoparticles were excited
at 970 nm by an SDL-6380-L2 ber-coupled cw NIR laser diode
(maximum power 2.5 W) and an SDL-820 highly stabilized laser
diode driver (Spectra Diode Labs). The unfocused excitation
light was transmitted to and the luminescence collected from
the sample by a QP 400-2-VIS-BX Y-type glass ber (Ocean
The thermal decomposition of rare earth acetates was
investigated under a nitrogen ow of 20 ml minꢀ1 and a heating
rate of 5 K minꢀ1 by thermogravimetric analysis using a TGA/
SDTA 851e (Mettler-Toledo, Schwerzenbach, Switzerland).
3. Results and discussion
3.1. b-NaGdF4:Yb3+,Er3+ nanoparticles from EG/IL mixtures
Optics) with 400 mm core diameter. The luminescence was X-ray diffraction patterns of cubic a- and hexagonal b-
recorded by an HR-2000-CG-UV-NIR ber spectrometer (Ocean NaGdF4:18% Yb3+,2% Er3+ nanoparticles are shown in Fig. 1.
Optics). For comparative measurements of the luminescence Samples were obtained from EG/DADMA BF4 (EG1) and EG/
ꢀ
intensity, only the sample tubes were exchanged. The reliability DADMA NTf2 (EG2) solutions with BF4 and NH4F as uoride
and repeatability of the measurements were better than 5%. All sources, respectively, see Table 1. Both samples were heated for
spectra were measured at room temperature and corrected for 30 min at 200 ꢁC. Tetrauoroborate anions decompose at 160 ꢁC
the spectral response of the detection system by calibration with and release Fꢀ ions into the solution. Sample EG1 crystallizes in
a tungsten lamp.50
the cubic a-NaGdF4 phase. The calculated peak positions are
Transmission electron microscopy (TEM) analyses were shown as red ticks below the top trace in Fig. 1 a-NaGdF4 crys-
accomplished on a G2 Spirit BioTWIN TEM (Tecnai) using tallizes in space group Fm3m in a structure related to CaF2.51,52 Its
ꢁ
a LaB6 120 kV electron beam. Images were recorded by an SIS strongest diffraction peak (111) is located at 28ꢁ 2-theta.
34786 | RSC Adv., 2019, 9, 34784–34792
This journal is © The Royal Society of Chemistry 2019

View Article Online
Paper
RSC Advances
compared to the symmetric BF4ꢀ. The luminescence from the
DADMA TFA sample was too weak to be compared with the
others. Also, the hygroscopicity of the IL is relevant. Samples
from hydrophobic DADMA NTf2 show a stronger luminescence
than those from hygroscopic ILs. Lower water contamination of
the IL decreases the number of OHꢀ impurities in the nano-
particles and non-radiative losses from these defects. The
crucial importance of strictly dry reaction conditions is further
supported by samples obtained from anhydrous rare earth
acetates versus those from acetate hydrates, see Fig. S5 (ESI†).
The UC luminescence from nanoparticles synthesized from dry
precursors is much stronger than from hydrates. Fig. 2 shows
the inuence of uoride concentration, temperature, and rare
earth concentration on the UC luminescence. The NH4F content
was varied from 60 mg to 120 mg for samples EG3 to EG6, see
Table 1. The stability of the hexagonal phase increases with the
Fꢀ concentration, as shown in Fig. S6† (ESI). The weak (111)
diffraction peak of the cubic phase around 28ꢁ 2-theta disap-
pears by raising the NH4F content from 60 to 100 mg. For higher
Fꢀ concentrations, only the hexagonal phase is obtained. As an
Fig.
1 Powder XRD patterns of nanocrystalline a-NaGdF4:18%
Yb3+,2% Er3+ (EG1, top trace) and b-NaGdF4:18% Yb3+,2% Er3+ (EG2,
bottom trace). Samples were synthesized in EG/IL solutions with
DADMA BF4 and NH4F, respectively, see Table 1. The calculated peak
positions for cubic a-NaGdF4 (Fm3ꢁm) and hexagonal b-NaGdF4 (P6ꢁ)
are shown as red and blue ticks below the traces, respectively.
The pure hexagonal b-NaGdF4 phase was obtained from an
NH4F solution for sample EG2, see the bottom trace in Fig. 1.
The calculated peak positions are shown as blue ticks. b-
NaGdF4 crystallizes in the b-NaNdF4 structure with space group
ꢀ
ꢀ
ꢁ
P6 and lattice parameters of a ¼ 6.0304 (9) A and c ¼ 3.6111 (7) A
at room temperature.8,53 A Fꢀ excess favors the formation of the
hexagonal b-NaGdF4 phase.53 BF4 as uoride source yields
ꢀ
preferentially the cubic a-phase in a microwave synthesis, since
ꢀ
the slow BF4 decomposition during heating provides low
uoride concentration, only. The UC luminescence of cubic a-
NaGdF4:2% Er3+,18% Yb3+ bulk material is about an order of
magnitude less intense compared to the hexagonal b-phase.9
Accordingly, NH4F was chosen as uoride source in order to
obtain hexagonal nanoparticles with strong UC emission.
The inuence of DADMA ILs on the microwave synthesis was
investigated. The thermal stability of the IL depends on the
anion and increases along the series TFA < BF4 < OTf < NTf2. The
stability determines the maximum reaction temperature,
ranging from 120 ꢁC (TFA) to 200 ꢁC (BF4, OTf, and NTf2) for EG/
IL solutions and 400 W microwave power. Higher temperatures
result in partial decomposition of the IL, and the colorless
solutions turn brown. Samples were synthesized under identical
conditions in different ILs. Their phase purity and UC lumi-
nescence are compared in Fig. S3–S4 (ESI†). All products show
a pure hexagonal phase in powder XRD, see Fig. S3 (ESI†). The
anions strongly inuence the b-NaGdF4:18% Yb3+,2% Er3+ UC
luminescence which increases with the thermal stability of the
ILs along the series BF4 < OTf < NTf2, see Fig. S4 (ESI†). Asym-
metric anions such as TFA, OTf, and NTf2 have a higher
microwave absorption, which results in more effective heating
Fig. 2 Upconversion luminescence of b-NaGdF4:18% Yb3+,2% Er3+
nanoparticles from EG/DADMA OTf syntheses in dependence on
fluoride concentration (bottom), reaction temperature (middle), and
rare earth acetate concentration (top). The intensity scale increases
from bottom to top. Sample names refer to Table 1. The UC lumi-
nescence was excited at 970 nm with 580 mW (unfocused) laser
power.
This journal is © The Royal Society of Chemistry 2019
RSC Adv., 2019, 9, 34784–34792 | 34787

View Article Online
RSC Advances
Paper
impurity, NaF was identied, which does not inuence the UC
luminescence. The NaF diffraction peaks have a narrow line
width and are marked with asterisks for the top trace of Fig. S6.†
The crystallite size was evaluated from the XRD patterns in
Fig. S6,† ranging from 6 to 14 nm.
For the UC luminescence spectra in Fig. 2, Yb3+ ions are
excited at 970 nm. Excited Yb3+ ions can transfer energy reso-
nantly to neighboring Er3+ ions in several steps. The energy
4
transfer upconversion (ETU) rst populates the I11/2 state and
4
in a second step the F7/2 state of Er3+, which decays via multi-
2
4
phonon relaxation to the H11/2 and S3/2 states. These states
Fig. 3 Upconversion luminescence of b-NaGdF4:18% Yb3+,2% Er3+
nanoparticles from EG/IL syntheses with DADMA OTf (A) and DADMA
NTf2 (B) for various EG/IL ratios. The sample names refer to Table 1.
The UC luminescence was excited at 970 nm with 580 mW (unfo-
cused) laser power. Microwave absorption; compared to EG/IL
mixtures, the microwave power could be reduced from 400 W to
50 W.
4
are in thermal equilibrium, and their emission to the I15/2
ground state yields the typical green Er3+ UC luminescence.4,7
The 4F9/2 state is populated via cross relaxation.9,10 Its emission
to the ground state adds a red component to the luminescence.
The UC luminescence intensity increases 2-fold from sample
EG3 to EG6 for higher Fꢀ concentration, see Fig. 2 bottom and
Table 1. Another 2-fold increase of the UC intensity is achieved
by rising the reaction temperature from 120 ꢁC to 200 ꢁC for
samples EG6 to EG8, see Fig. 2 middle. The emission intensity
rises 16-fold by halving the rare earth acetate content from
60 mg to 30 mg for samples EG8 and EG9, see Fig. 2 top. As for
a higher Fꢀ content, a smaller rare earth ion concentration
increases the F/RE ratio. A higher reaction temperature results
in larger and better crystalline nanoparticles.24 In summary,
a high F/RE ratio and a high reaction temperature strongly
increase the UC luminescence intensity.
3.2. b-NaGdF4:Yb3+,Er3+ nanoparticles from IL solutions
Reactions in pure ILs yield material with a better UC lumines-
cence compared to EG or EG/IL syntheses, cf. Fig. 2 and 3, but
From the literature, it is known that higher NH4F concen-
trations increase the aggregate size of b-NaGdF4:Eu3+ and Er3+
nanoparticles synthesized at room temperature in EG.54,55
Aggregation reduces the exposed particle surface; concomi-
tantly, the non-radiative relaxation decreases and UC lumines-
cence is not quenched. The same was reported for nanoparticles
from octadecene/oleic acid solutions.11 Since nanoparticles
from EG solutions are oen affected by agglomeration,54,55
reactions in EG/IL mixtures and pure ILs, in particular the
hydrophobic DADMA NTf2, were investigated in order to reduce
the aggregate size and increase the UC luminescence.
UC luminescence spectra of b-NaGdF4:18% Yb3+,2% Er3+
nanoparticles from EG/IL mixtures with DADMA OTf and
DADMA NTf2 are shown in Fig. 3. The UC emission intensity
increases strongly when the EG content is reduced. The by far
best results were obtained from syntheses in pure ILs, see the
green traces in Fig. 3. Among the pure ILs, the product from
hydrophobic DADMA NTf2 (sample IL2) is signicantly better
than from DADMA OTf (sample IL1). The UC luminescence
intensity of sample IL2 is 70-fold higher than for EG11. As
discussed above, strictly anhydrous reaction conditions reduce
the number of OHꢀ defects in uorides and restrict non-
radiative losses of the luminescence intensity. Avoiding
solvents with OH groups, such as EG, is equally important.
These solvents may introduce traces of water into the reaction
mixture or act as surface ligands with high energy vibration
acceptor modes. Another bonus of the synthesis in pure ILs is
an increased microwave absorption; compared to EG/IL
mixtures, the microwave power could be reduced from 400 W to
50 W.
Fig. 4 Upconversion luminescence of b-NaGdF4:18% Yb3+,2% Er3+
nanoparticles from IL syntheses in dependence on fluoride concen-
tration (bottom), reaction temperature (middle), and time (top). The
sample names refer to Table 1. All samples were excited at 970 nm with
330 mW (unfocused) laser power.
34788 | RSC Adv., 2019, 9, 34784–34792
This journal is © The Royal Society of Chemistry 2019

View Article Online
Paper
RSC Advances
Fig. 5 TEM and HR-TEM micrographs of representative samples: (A) IL3 and (B) IL6 from syntheses with 60 mg and 180 mg NH4F at 225 ꢁC; (C)
IL8 and (D) IL9 from syntheses with 140 mg NH4F at 250 ꢁC for 5 min and 10 min; (E–G) IL4 from a synthesis with 100 mg NH4F at 225 ꢁC for
5 min; (H) CS_IL4 core–shell nanoparticles. The sample names refer to Table 1.
a different solvent requires the adjustment of reaction param- has a weaker UC emission. The higher temperature favors the
eters. Accordingly, the uoride concentration, temperature, and formation of the metastable a-phase.
reaction time were optimized for syntheses in DADMA NTf2
Doubling the reaction time from 5 to 10 min (IL8 vs. IL9)
solutions, see Fig. 4. The ionic liquid has a vapor pressure close reduces the amount of a-phase, see Fig. S7 (ESI†). A longer
to zero and superb thermal stability, which allows higher reaction time supports completion of the a / b phase transi-
reaction temperatures. No hints for any decomposition of the IL tion, which enhances the UC luminescence, see Fig. 4, top.
were detected up to 250 ꢁC, the highest accessible temperature Overall, sample IL6 shows the best UC emission. Optimized
ꢁ
for the microwave system. Due to the stronger microwave reaction parameters are a high uoride concentration, 225 C,
absorption of the IL, the reaction time was reduced to 5 min and 10 min reaction time.
compared to 30 min for EG/IL mixtures.
TEM micrographs of representative samples are shown in
Increasing the uoride concentration gives rise to a stronger Fig. 5. Particle size and agglomeration increase with uoride
UC luminescence, as demonstrated for samples IL3 to IL6 in concentration, see Fig. 5A and B for samples IL3 and IL6. The
Fig. 4, bottom, and in good agreement with Fig. 2. However, crystallite size increases from 3–4 nm to 15–18 nm, and
a comparison of Fig. S7 and S6 (ESI†) reveals the presence of a- agglomerates form in the mm size range. Longer reaction times
phase in the XRD diagrams of several IL samples. The a-phase result in larger aggregates, see Fig. 5C and D. From higher
content is reduced from IL3 to IL5 and IL6 shows a pure b- magnication in Fig. 5E–G, an individual crystallite size in the
phase. A higher uoride concentration is required in the IL in order of 4.6 ꢃ 1.1 nm becomes visible for sample IL4. It does not
order to obtain the pure b-phase. For identical uoride show signicant aggregation and provides a good UC lumi-
concentration, e.g. EG3 vs. IL3, an IL sample shows a much nescence. This sample was further used for a core–shell
stronger UC luminescence because of a reduced number of synthesis, where an undoped b-NaGdF4 shell was grown on top
defects due to the anhydrous reaction conditions, despite an a/ of the b-NaGdF4:2% Er3+,18% Yb3+ particles. The core–shell
b phase mixture. Higher uoride concentrations along the IL3 particles (CS_IL4) have a crystallite size <20 nm and show rather
to IL6 series increase the nanoparticle size as evidenced by the small aggregates, see Fig. 5H. Sample IL4 contains a minor
narrower diffraction peaks.
fraction of the a-phase, as evident from its XRD pattern, see
The reaction temperature has a stronger effect on the UC Fig. S7 (ESI†). In the subsequent reaction step towards the core–
luminescence than the NH4F concentration, see Fig. 5A middle. shell sample CS_IL4, a complete conversion to the b-phase was
At 225 ꢁC (IL5) the UC luminescence intensity is 34-fold higher achieved, see Fig. S8 (ESI†).
ꢁ
compared to 200 C (IL7), but it drops again by a factor of 12
The UC luminescence of samples from EG (EG11) vs. core
towards 250 ꢁC (IL8). The increase from 200 ꢁC to 225 ꢁC is due (IL4) and core–shell particles (CS_IL4) from IL syntheses is
to a higher crystallinity and bigger particle size of the b-phase. compared in Fig. 6 and 7. Samples from IL syntheses show
At 250 ꢁC, a signicant amount of the a-phase is observed which strongly enhanced UC intensities over the EG sample, see Fig. 6.
This journal is © The Royal Society of Chemistry 2019
RSC Adv., 2019, 9, 34784–34792 | 34789

View Article Online
RSC Advances
Paper
The green-to-red ratio is representative for the level of multi-
4
phonon relaxation losses in the sample. The F9/2 state can be
populated by cross-relaxations or multi-phonon relaxation from
the 4S3/2 state.4,9 In macroscopic samples, e.g. single crystals, the
multi-phonon relaxation is negligible at room temperature for
the b-NaGdF4 host lattice due to its low effective phonon energy.
Therefore, the red emission is indicative of impurities or lattice
defects, which introduce acceptor modes of high-energy,
4
depopulate the S3/2 state, and strongly reduce the UC emis-
sion in total. The major enhancement of the UC luminescence
from sample EG11 to IL4 is the result of a signicant reduction
of defects and acceptor modes due to an anhydrous synthesis
and the absence of ligands with OH groups.
A further signicant increase of the UC emission is achieved
by growing an undoped b-NaGdF4 shell around the b-
NaGdF4:18% Yb3+,2% Er3+ core, see the blue trace in Fig. 6 for
sample CS_IL4. The shell saturates any partial coordination at
the core surface and separates the particle surface from the
optically active core. This approach reduces losses by multi-
phonon relaxation via high energy vibration modes of ligand
molecules and surface defects.
Fig. 6 Upconversion luminescence of b-NaGdF4:18% Yb3+,2% Er3+
nanoparticles from an EG synthesis (EG11, black trace) compared to
core (IL4, red trace) and core–shell (CS_IL4, blue trace) nanoparticles
The UC luminescence from IL4 and CS_IL4 nanoparticles is
from IL syntheses. The sample names refer to Table 1. The UC lumi- further compared to mm-size b-NaYF4: 20% Yb3+, 2% Er3+ bulk
material,9 see Fig. 8. The samples were excited at 970 nm by only
nescence was excited at 970 nm with 580 mW (unfocused) laser
power.
70 mW (unfocused) laser power. Despite a low excitation density
of 2.6 W cmꢀ2, see Fig. S9 and Table S1 (ESI†) for excitation
density calculation, the nanoparticle samples (red and blue
traces) show a well detectable UC emission with a spectral
distribution close to the bulk material (black trace). The spectra
The growth of an undoped b-NaGdF4 shell (CS_IL4, blue trace)
around the Er3+, Yb3+ doped core (IL4, red trace) results in
doubling of the UC intensity. The impressive differences
of the nanoparticle samples are scaled by a factor ꢂ500 in
become evident from the photographs in Fig. 7. Note that the
Fig. 8. The UC luminescence from the core (IL4) and core–shell
excitation power is reduced almost by half from 580 mW to 330
nanoparticles (CS_IL4) are 0.067 (4)% and 0.122 (5)% of the
mW for EG/IL and IL samples, respectively. Dry powder samples
bulk material. The green-to-red intensity ratio is 1.80 for the
of the nanoparticles were lled in glass tubes and are illumi-
bulk vs. 1.66 and 1.52 for the core–shell and core nanomaterials,
respectively. The slight decrease of the green-to-red intensity
nated from the le-hand side by unfocused laser light from
a glass ber. As discussed above, infrared excitation of Yb3+ and
ratio along this series is due to increasing multi-phonon relax-
energy transfer upconversion between Yb3+ and Er3+ ions result
ation losses. These results are very good for sub 20 nm particles
in Er3+ emission in the visible spectral range. Dominant is the
and compare well with literature.
green emission from the 4S3/2 and 2H11/2 states, weaker the red
emission from the 4F9/2 state. Higher order processes give rise to
a small population of the 2H9/2 state and result in a further weak
violet UC emission. The high green-to-red UC intensity ratio
demonstrates the excellent quality of the nanomaterial with
respect to chemical purity, phase purity, and crystallinity.
A synthesis of <6 nm b-NaGdF4:30% Yb3+,2% Er3+ core–shell
particles in oleic acid and octadecene at 240 ꢁC for 20 min was
reported by Liu et al.26 The UC luminescence of these particles
was excited at 10 W cmꢀ2 5.5 nm b-NaYF4 : Yb3+, Er3+ particles
from an oleic acid, oleylamine, and octadecene solution at
310 C for 30 min were reported by Ostrowski et al.24 Their UC
ꢁ
Fig. 7 Photos of the UC luminescence from b-NaGdF4:18% Yb3+,2% Er3+ nanoparticles excited at 970 nm with 580 mW or 330 mW (unfocused)
laser power, respectively. (A) EG11 from EG/IL, (B) IL4 core, and (C) CS_IL4 core–shell particles from IL syntheses. The sample names refer to
Table 1.
34790 | RSC Adv., 2019, 9, 34784–34792
This journal is © The Royal Society of Chemistry 2019

View Article Online
Paper
RSC Advances
non-radiative losses due to defects and high energy acceptor
modes of ligand vibrations. Another critical factor is the archi-
tecture of the nanoparticles. Core–shell particles show a more
intense luminescence compared to the bare core particles. The
growth of an isostructural undoped shell saturates coordination
environments at the core surface and separates the core from
the particle surface. Sub 20 nm core–shell nanoparticles were
obtained in two reaction steps without intermediate separation.
Very low excitation power was required to excite the upconver-
sion luminescence, which renders these nanoparticles attrac-
tive for biological imaging. We proved that the proposed
synthesis provides an effective route to minimize the presence
of OH-defects, which have been demonstrated to represent the
main source of luminescence quenching in uorides. We
believe that further optimization of this microwave-assisted
synthesis will yield even more efficient and monodisperse
nanomaterials, providing a fast and easily scalable synthesis
route to brighter uoride upconversion nanoparticles.
Fig. 8 Upconversion luminescence spectra of mm-size b-NaYF4: 20%
Yb3+, 2% Er3+ bulk material7 (black trace) compared to core (IL4, red
trace) and core–shell (CS_IL4, blue trace) nanoparticles. The sample
names refer to Table 1. The UC luminescence was excited at 970 nm
with 70 mW (unfocused) laser power.
Conflicts of interest
There are no conicts to declare.
Acknowledgements
spectra were excited by 1 W. 5–20 nm b-NaYF4: Yb3+, Ln3+ (Ln ¼
Er, Tm or Ho) particles were obtained from a synthesis in tri-
octylphosphine oxide (TOPO),25 which allows higher reaction
temperatures of 330 ꢁC to 370 ꢁC. Samples were synthesized
within 30 to 90 min without impurities of the a-phase. 11 nm
particles were excited by a focused laser with 1 W power. 8–
10 nm b-NaYF4: Yb3+, Er3+ particles were synthesized in oleyl-
amine at 330 ꢁC for 1 hour by Boyer et al.21 They reported
quantum yields of 3% and 0.005% for bulk (>>100 nm) and 8–
10 nm particles for 20 W cmꢀ2 and 150 W cmꢀ2 excitation
density, respectively. The UC luminescence intensity of our
nanoparticles, see Fig. 8, was achieved by 70 mW unfocused
excitation power, corresponding to a power density of 2.6 W
cmꢀ2. This value is more than an order of magnitude lower than
most literature values cited above. Incomplete information on
the excitation conditions oen hampers a direct comparison of
UC emission intensities from the literature. Due to the non-
linearity of UC processes the measurement of absolute
quantum yields and a comparison between different excitation
densities is inherently difficult.56,57 We restrict our study to
a relative comparison of nanoparticles with bulk material,
whose high quality was demonstrated in previous studies.7,56
The nancial support from the European Community's Seventh
Framework Program (FP7/2007-2013) under Grant No. 316906
(ITN LUMINET) is gratefully acknowledged by all authors. The
authors thank the Microscopy Imaging Centre (MIC) at the
University of Bern for access to TEM instruments.
Notes and references
¨
¨
1 S. Heer, K. Kompe, H. U. Gudel and M. Haase, Adv. Mater.,
2004, 16, 2102–2105.
2 A. Aebischer, S. Heer, D. Biner, K. W. Kramer, M. Haase and
¨
¨
H. U. Gudel, Chem. Phys. Lett., 2005, 407, 124–128.
¨
3 K. W. Kramer, J. Lumin., 2017, 189, 78–83.
4 F. Auzel, Chem. Rev., 2004, 104(1), 139–174.
5 T. Kano, H. Yamamoto and Y. Otomo, J. Electrochem. Soc.,
1972, 119, 1561.
6 N. Menyuk, K. Dwight and J. W. Pierce, Appl. Phys. Lett., 1972,
21, 159.
¨
¨
7 K. W. Kramer, D. Biner, G. Frei, H. U. Gudel, M. P. Hehlen
¨
and S. R. Luthi, Chem. Mater., 2004, 16, 1244–1251.
8 A. Aebischer, M. Hostettler, J. Hauser, K. W. Kramer,
¨
¨
¨
T. Weber, H. U. Gudel and H. B. Burgi, Angew. Chem., Int.
Ed., 2006, 45, 2802–2806.
9 R. B. Anderson, S. J. Smith, P. S. May and M. T. Berry, J. Phys.
Chem. Lett., 2014, 5, 36–42.
4. Conclusions
The synthesis under anhydrous conditions in the ionic liquid 10 M. T. Berry and P. S. May, J. Phys. Chem. A, 2015, 119, 9805–
DADMA NTf2 is a promising path to b-NaGdF4:18% Yb3+,2%
9811.
Er3+ nanoparticles. A microwave reactor allows for efficient 11 J. C. Boyer, F. Vetrone, L. A. Cuccia and J. A. Capobianco, J.
heating and short reaction times. Anhydrous conditions and Am. Chem. Soc., 2006, 128(23), 7444–7445.
appropriate solvents are essential for the synthesis of highly 12 J. Cichos, L. Marciniak, D. Hreniak, W. Strek and
luminescent upconversion material. These conditions limit
M. Karbowiak, J. Mater. Chem. C, 2014, 2, 8244–8251.
This journal is © The Royal Society of Chemistry 2019
RSC Adv., 2019, 9, 34784–34792 | 34791

View Article Online
RSC Advances
Paper
¨
13 M. Haase and H. Schafer, Angew. Chem., Int. Ed., 2011, 50, 34 C. Gabriel, S. Gabriel, E. H. Grant, E. H. Grant,
5808–5829.
B. S. J. Halstead and D. M. P. Mingos, Chem. Soc. Rev.,
1998, 27, 213–224.
14 H.-X. Mai, Y.-W. Zhang, R. Si, Z. G. Yan, L. Sun, L.-P. You and
C.-H. Yan, J. Am. Chem. Soc., 2006, 128(19), 6426–6436.
´
35 R. Martınez-Palou, Mol. Diversity, 2010, 14, 3–25.
¨
15 J. Zhou, Z. Liu and F. Li, Chem. Soc. Rev., 2012, 41, 1323– 36 J. Hoffmann, M. Nuchter, B. Ondruschka and
1349. P. Wasserscheid, Green Chem., 2003, 5, 296–299.
16 C. Li, Z. Hou, Y. Dai, D. Yang, Z. Cheng, P. Ma and J. Lin, 37 D. Wang, L. Ren, X. Zhou, X. Wang, J. Zhou, Y. Han and
Biomater. Sci., 2013, 1, 213–223. N. Kang, Nanotechnology, 2012, 23, 225705.
17 J. Lee, T. S. Lee, J. Ryu, S. Hong, M. Kang, K. Im, J. H. Kang, 38 C. Chen, L.-D. Sun, Z.-X. Li, L.-L. Li, J. Zhang, Y.-W. Zhang
S. M. Lim, S. Park and R. Song, J. Nucl. Med., 2013, 54, 96–
103.
18 S. Cui, H. Chen, H. Zhu, J. Tian, X. Chi, Z. Qian, S. Achilefu
and Y. Gu, J. Mater. Chem., 2012, 22, 4861–4873.
and C.-H. Yan, Langmuir, 2010, 26, 8797–8803.
39 N. Niu, F. He, S. Gai, C. Li, X. Zhang, S. Huang, P. Yang,
F. P. Wang, J. Lin, D. Zhao and Y. Zhao, J. Mater. Chem.,
2012, 22, 21613–21623.
19 Y. Hu, B. Wu, Q. Jin, X. Wang, Y. Li, Y. Sun, J. Huo and 40 C. Mi, Z. Tian, C. Cao, Z. Wang, C. Mao and S. Xu, Langmuir,
X. Zhao, Talanta, 2016, 152, 504–512. 2011, 27, 14632–14637.
20 F. Wang, D. Banerjee, Y. Liu, X. Chen and X. Liu, Analyst, 41 C. Lorbeer, J. Cybinska and A.-V. Mudring, Chem. Commun.,
2010, 135, 1839–1854. 2010, 46, 571–573.
21 J.-C. Boyer and F. C. J. M. van Veggel, Nanoscale, 2010, 2, 42 C. Lorbeer, J. Cybinska and A.-V. Mudring, J. Mater. Chem. C,
1417–1419. 2014, 2, 1862.
22 N. J. J. Johnson, W. Oakden, G. J. Stanisz, R. S. Prosser and 43 C. Lorbeer and A.-V. Mudring, Chem. Commun., 2014, 50,
F. C. J. M. van Veggel, Chem. Mater., 2011, 23, 3714–3722. 13282–13284.
23 Q. Liu, Y. Sun, T. Yang, W. Feng, C. Li and F. Li, J. Am. Chem. 44 Q. Ju and A.-V. Mudring, RSC Adv., 2013, 3, 8172.
Soc., 2011, 133, 17122–17125.
24 A. D. Ostrowski, E. M. Chan, D. J. Gargas, E. M. Katz, G. Han,
45 V. Jovanovski, R. Marcilla and D. Mecerreyes, Macromol.
Rapid Commun., 2010, 31, 1646–1651.
P. J. Schuck, D. J. Milliron and B. E. Cohen, ACS Nano, 2012, 46 E. I. Izgorodina, M. Forsyth and D. R. MacFarlane, Phys.
6(3), 2686–2692. Chem. Chem. Phys., 2009, 11, 2452.
25 J. Shan, X. Qin, N. Yao and Y. Ju, Nanotechnology, 2007, 18, 47 T. Singh and A. Kumar, J. Phys. Chem. B, 2008, 112, 12968–
445607. 12972.
26 J. Liu, G. Chen, S. Hao and C. Yang, Nanoscale, 2017, 9, 91– 48 J. G. Huddleston, A. E. Visser, W. M. Reichert, H. D. Willauer,
98.
27 S. Wilhelm, M. Kaiser, C. Wurth, J. Heiland, C. Carrillo-
G. A. Broker, R. D. Rogers, R. L. Vaughn, J. S. Wilkes and
J. L. Williams, Green Chem., 2001, 3, 156–164.
¨
Carrion, V. Muhr, O. S. Woleis, W. J. Parak, U. Resch- 49 M. J. Earle, C. M. Gordon, N. V. Plechkova, K. R. Seddon and
Genger and T. Hirsch, Nanoscale, 2014, 7, 1403–1410. T. Welton, Anal. Chem., 2007, 79, 758–764.
28 N. J. Johnson, N. M. Sangeetha, J. C. Boyer and F. C. J. M. van 50 E. Ejder, J. Opt. Soc. Am., 1969, 59, 223–224.
Veggel, Nanoscale, 2010, 2, 771–777.
51 D. M. Roy and R. Roy, J. Electrochem. Soc., 1964, 111(4), 421–
29 Q. Ju, P. S. Campbell and A.-V. Mudring, J. Mater. Chem. B,
2013, 1, 179–185.
429.
52 J. H. Burns, Inorg. Chem., 1965, 4, 881–886.
30 Q. Ju and A.-V. Mudring, RSC Adv., 2013, 3, 8172–8175.
53 R. Thoma, H. Insley and G. Hebert, Inorg. Chem., 1966, 1005,
¨
¨ ¨
31 R. Arppe, I. Hyppanen, N. Perala, R. Peltomaa, M. Kaiser,
1222–1229.
¨
¨
¨
C. Wurth, S. Christ, U. Resch-Genger, M. Schaferling and 54 G. Tessitore, A.-V. Mudring and K. W. Kramer, J. Lumin.,
T. Soukka, Nanoscale, 2015, 7, 11746–11757.
32 D. J. Gargas, E. M. Chan, A. D. Ostrowski, S. Aloni, 55 G. Tessitore, A.-V. Mudring and K. W. Kramer, New J. Chem.,
M. Virginia, P. Altoe, E. S. Barnard, B. Sanii, J. J. Urban, 2018, 42(1), 237–245.
D. J. Milliron, B. E. Cohen and P. J. Schuck, Nat. 56 S. Fischer, B. Frohlich, H. Steinkemper, K. W. Kramer and
2017, 189, 91–98.
¨
¨
¨
Nanotechnol., 2014, 9, 300–305.
33 G. Yi and G. Chow, Chem. Mater., 2007, 19, 341–343.
J. C. Goldschmidt, Sol. Energy Mater. Sol. Cells, 2014, 122,
197–207.
¨
¨
57 S. Fischer, B. Frohlich, K. W. Kramer and J. C. Goldschmidt,
J. Phys. Chem. C, 2014, 118, 30106–30114.
34792 | RSC Adv., 2019, 9, 34784–34792
This journal is © The Royal Society of Chemistry 2019