Crystal structure characterization, optical and photoluminescent properties of
tunable yellow- to orange-emitting Y
2
(Ca,Sr)F
4
S
2
:Ce
3+phosphors for
solid-state lighting†
Yun-Chen Wu,
aYi-Chin Chen,
aTeng-Ming Chen,*
aChi-Shen Lee,
bKuo-Ju Chen
cand Hao-Chung Kuo
cReceived 28th December 2011, Accepted 8th February 2012 DOI: 10.1039/c2jm16882k
In this study, a new efficient Ce3+-doped fluorosulfide phosphor, Y
2(Ca,Sr)F4S2:Ce3+, was obtained by
using solid-state methods in a sealed silica ampoule. The synthesized Y2(Ca,Sr)F4S2:Ce3+was
characterized by powder X-ray diffraction and refined with Rietveld methods. Y2(Ca,Sr)F4S2:Ce3+can
be excited by blue light (440–470 nm) and shows yellow-to-orange broadband emission peaking at 553– 590 nm with a quantum efficiency of 16–31%. Non-radiative transitions between Ce3+ions in
Y2CaF4S2:Ce3+and Y2SrF4S2:Ce3+hosts have also been demonstrated to be attributable to dipole–
dipole interactions, and the critical distances were calculated to be 18.9 and 19.3 A. The possible mechanism of the tunable luminescence properties was described on the basis of band structure calculations. In addition, a white LED device was fabricated by using Y2(Ca,Sr)F4S2:Ce3+phosphor
pumped with a 460 nm blue chip. The CRI value and CCT were measured to be 74–85 and 3500–8700 K, respectively, showing promising potential for solid-state lighting.
Introduction
In recent years, remarkable advancements have been made in the development of commercially realized efficient white light-emitting diodes (LEDs) for their merits of energy-efficient, life-durable and environmentally friendly characteristics.1–3
The most common approach for white light relies on a blue LED chip and a blue-light excitable yellow phosphor, (Y,Gd)3(Al,Ga)5O12:Ce3+ (YAG:Ce).4,5 Nevertheless, it is also
characterized as cool white light because of its high correlated color temperature (CCT) of >8000 K, and poor color rendering index (CRI, Ra) of 70–75.4,6–8To improve the mentioned
prob-lems, Ce3+or Eu2+-doped sulfide phosphors were reported for
their applicability of generating white light, because they show strong blue absorption and green- to red-emitting color.9Jia and Wang summarized a series of rare-earth doped alkali-earth sulfide, AES:Ce3+or Eu2+(AE¼ Ca, Sr), which is suitable for
blue LED pumping.10A similar type of white light was achieved
by Guo et al. using (Ca,Sr)S:Eu2+, Ln3+ (Ln ¼ La, Pr–Nd,
Sm–Yb). A relatively high CRI value of 82–92 and low CCT of 3600–4800 K were reported.11–13However, this class of binary materials shows a relatively strong thermal quenching and a limited stability with moisture. Recently, several ternary sulfide phosphors were reported in order to ameliorate the aforemen-tioned drawbacks, e.g., Ce3+ or Eu2+-doped thiosilicates and
thiogallates.14,15 Accordingly, it is essential to develop a new
sulfide phosphor which can be effectively excited by the blue light and also possess better chemical property.16The Ce3+emission
usually consists of an asymmetric broad band due to the parity allowed characteristics of the transition between the lowest crystal field components’ 5d excited state and the 4f ground state (2F
7/2and2F5/2), and can be varied from ultraviolet to yellow in
its emitting color, depending on the different host lattices.17–19 In this study, we present a new Ce3+-doped mixed-anion
fluorosulfide Y2(Ca,Sr)F4S2:Ce3+ phosphor. The quaternary
Ln2AEF4S2(Ln¼ Ce, Sm, Eu, Yb; AE ¼ Ca, Sr) halogensulfide
material has been originally recognized as a ceramic/glass pigment, however, its luminescence property has yet to be investigated.20–22In Y2(Ca,Sr)F4S2host lattice, the Ce3+is
coor-dinated by both fluoride and sulfide anion. With low phonon energy in fluoride anion and strong covalency in sulfide anion, the efficient emission due to the minor quenching process from multi-phonon relaxation and longer wavelength emission for Ce3+ is expected in Y
2(Ca,Sr)F4S2:Ce3+.23,24 In this work, we
report on the luminescence properties of Y2CaF4S2:Ce3+and it
exhibits an absorption band in the blue region with a broad orange emission peaking at 590 nm, making it an appropriate candidate for our pursuit of a new phosphor component for
aPhosphors Research Laboratory, Department of Applied Chemistry and
Institute of Molecular Science, National Chiao Tung University, Hsinchu 30010, Taiwan. E-mail: tmchen@mail.nctu.edu.tw
bDepartment of Applied Chemistry and Institute of Molecular Science,
National Chiao Tung University, Hsinchu 30010, Taiwan
cSemiconductor Laser Technology Laboratory, Department of Photonics,
National Chiao Tung University, Hsinchu 30010, Taiwan
† Electronic supplementary information (ESI) available: EDX spectra, SEM image, XRD patterns, structural parameters, and decay curves. See DOI: 10.1039/c2jm16882k
Materials Chemistry
Cite this: J. Mater. Chem., 2012, 22, 8048
www.rsc.org/materials
PAPER
white LED. We also thoroughly studied the Sr2+substitution in
Y2CaF4S2:Ce3+in an attempt to realize the effect of
composi-tional modification on spectroscopic properties. Furthermore, we utilized the structural parameters of Y2CaF4S2, Y2SrF4S2,
Ce2CaF4S2, and Ce2SrF4S2 to calculate their densities of state
(DOS), which in turn allowed us to investigate distribution of energy levels in fluorosulfide systems and to propose their working luminescence mechanisms. Finally, temperature dependence and the LED device using Y2(Ca,Sr)F4S2:Ce3+
phosphors with blue chip were investigated to demonstrate the applicability of the Y2(Ca,Sr)F4S2:Ce3+ phosphor as a color
conversion material.
Experimental section
SynthesisThe polycrystalline samples of Y2xCexCaF4S2, Y2xCexSrF4S2,
and Y1.98Ce0.02Ca1ySryF4S2were prepared by solid-state
reac-tions using YF3 (Alfa), CaF2 (Aldrich), SrF2 (Aldrich), CaS
(Alfa), SrS (Alfa), Y2S3 (Alfa), and CeF3 (Aldrich) as raw
materials. The stoichiometric amounts of the starting materials were thoroughly mixed and loaded into Al2O3tubing, which was
transferred into a vertically positioned quartz ampoule, fully evacuated to 103torr and sealed off. The quartz ampoule was heated to 900–1000C for 8–12 h and then cooled down slowly to room temperature. The obtained product was then annealed at 500–600C for 2–8 h under 1–5% H2/Ar atmosphere.
Characterizations
The phase purity of the reaction product was analyzed by powder X-ray diffraction (XRD) using a Bruker AXS D8 advanced automatic diffractometer with Cu Ka radiation (l¼ 1.5418 A, 40 kV 40 mA). The powder diffraction data were subjected to analysis by a computer software General Structure Analysis System (GSAS) package.25 Refined structure parameters
comprised overall scale factors, lattice parameters, and fractional coordinates. The morphology and energy-dispersive X-ray spectroscopy (EDX) spectrum were measured by a JEOL JSM-7401F conventional thermal field-emission scanning electron microscope. The diffuse reflection spectra were measured with a Hitachi 3010 double-beam ultraviolet-visible (UV-Vis) spec-trometer (Hitachi Co., Tokyo, Japan) equipped with a ø60 mm integrating sphere. The photoluminescence (PL) and photo-luminescence excitation (PLE) spectra were recorded with a Spex Fluorolog-3 spectrofluorometer (Jobin Yvon Inc/specx) equip-ped with a 450 W Xe lamp and analyzed by a Jobin-Yvon spectrometer HR460 with a multichannel charge-coupled device detector. The Commission Internationale de l’Eclairage (CIE) chromaticity coordinates were determined by a Laiko DT-100 color analyzer equipped with a CCD detector (Laiko Co., Tokyo, Japan). The program Linear Muffin Tin Orbital (LMTO)26,27 was used to calculate the electronic structures of Y2CaF4S2, Y2SrF4S2, Ce2CaF4S2, and Ce2SrF4S2. The density
functional theory is used with the local density approximation (LDA).28The integration in k point is performed by an improved
tetrahedron method on a grid of 40 40 40 k points for Y2CaF4S2 and Ce2CaF4S2, and 16 16 16 k points for
Y2SrF4S2 and Ce2SrF4S2 of the first Brillouin zone. The
crystallographic parameters used for the calculation were derived from the GSAS refinement data. The final results from the calculations are examined to obtain the electronic properties, such as the band structure, total and partial densities of states. The phosphor-converted white LED devices were fabricated using commercial blue InGaN-based LED (lmax¼ 460 nm) with
an intimate mixture of as-synthesized phosphors and silicone resin. Electroluminescence (EL) spectra were recorded at a forward-current and measured by using a SphereOptics integrating sphere with LED measurement starter packages (Onset, Inc.).
Results and discussion
Structural characterizations and crystallographic parameters of the Y1.98Ce0.02AEF4S2(AE ¼ Ca, Sr) phosphors
The Rietveld refinements were accomplished in order to obtain the detailed crystal information about Y1.98Ce0.02AEF4S2
(AE¼ Ca, Sr) and ensure the purity of the samples phase. The initial structural model was first built with crystallographic data of isotypic single-crystal Eu2(III)Eu(II)F4S2.22 According to the
similarity of effective cationic radii in the same coordination environment,29the Eu3+can be substituted by the Y3+ion; also,
the Eu2+can be substituted by Ca2+and Sr2+ions. Therefore, the
title hosts Y2CaF4S2 and Y2SrF4S2 (hereafter referred to as
YCFS and YSFS) were demonstrated. The fractional occupan-cies of each component atom were adjusted to the nominal stoichiometry. Fig. 1 presents the Rietveld refinement results for Y1.98Ce0.02CaF4S2 (YCFS:Ce3+) and Y1.98Ce0.02SrF4S2
(YSFS:Ce3+) from the observed XRD patterns, indicating the
final converged weighted-profile of Rwp ¼ 6.8% and 4.96%,
respectively. Both structures were found to crystallize tetrago-nally in the space group I4/mmm (no. 139) with Z¼ 2. In the crystal structure of YCFS and YSFS, both Ce and Y atoms occupy only the 4e position site, as presumed, due to the inability of Ce3+ions achieving the divalent oxidation state. Furthermore,
the 2b site is fully occupied by Ca or Sr atoms, the 4e site is fully occupied by S atoms and the 16n site is estimated to be half-filled by F atoms. The final refined structural parameters, selected atomic distances, and bond angles of YCFS:Ce3+and YSFS:Ce3+
are summarized in Tables 1 and 2. The grain size and morphology of YCFS:Ce3+particles were characterized by SEM,
which indicates that the as-synthesized phosphor was composed of many irregular granular microcrystals with an average size of 5 to 10 mm. The nominal stoichiometry was also verified accurately by EDX measurement (see Fig. S1 in the ESI†).
Fig. 2 shows the exact 1 1 1 unit cell crystal structure of the YCFS lattice viewed from the [010] and the Y atomic sites along with their corresponding neighboring atoms from the refined result. The structure of YCFS can be depicted as an ordered intergrowth YF4S5/CaF8/YF4S5polyhedra layer
struc-ture similar to the Ln3F4S2-family (Ln¼ Eu, Yb),20,22which was
composed of one fluoride-analogous (AE¼ Ca, Sr, Ba) sheet and two PbFCl-type sheets along the c-axis repeat. The coordination polyhedron YF4S5is comprised of five S, four F atoms and one Y
atom situated in the center of the monocapped square antiprism (CN ¼ 9). According to the point charge theory, the five degenerate 5d orbitals of Ce3+are regarded as having split into
five distinct d levels in the increasing order of energy levels dx2y2, dz2, dyz, dxz, and dxyin C4v symmetry.19,30The crystal
splitting of 5d levels can be further changed by the geometrical translation, that is, the so-called nephelauxetic effect. The isotropic volume contraction strengthens crystal field splitting, accordingly lowering the lowest-lying 5d level and altering the excitation and emission wavelength, whereas isotropic volume expansion produces the opposite effect. In our case, the varia-tions in PLE/PL spectra coupled with changing Ca/Sr substitu-tional ratio were observed; this part is to be discussed later. Spectroscopic study of Y2AEF4S2:Ce3+(AE ¼ Ca, Sr)
Fig. 3 shows the diffuse reflection spectrum of as-synthesized polycrystalline YCFS and the PLE/PL spectra of YCFS:Ce3+.
For the YCFS host, the diffuse reflection spectrum shows a status of high reflection in the wavelength ranging from 400 to 800 nm and decreasing intensity from 250 to 400 nm. The Kubelka–Munk absorption coefficient (K/S) relation was used to calculate the absorption edge from the measured reflectance (R) of the YCFS host:31
K S¼
ð1 RÞ2
2R (1)
where K represents the absorption coefficient, S represents the scattering coefficient, and R represents the reflectivity. The fundamental band gap energy of the YCFS was calculated to be approximately 3.39 eV by extrapolation. The observed color of YCFS is light yellow, which then turns yellow with doping Ce3+
ions. For YCFS:Ce3+, a typical PLE and PL spectra were
observed. In order to study the Ce 5d and 4f energy levels in YCFS:Ce3+, we first discriminated each preferential site for Ce3+
dopants in YCFS:Ce3+. Comparing the emission bands in
YCFS:Ce3+and CaF
2:Ce3+,19,32–34 both have a nearly identical
chemical environment around the Ca2+ion, viz., the same
coor-dination number of eight, same atom symmetry (D4h) and
a similar Ca–F bond distance, but occur with different emission wavelengths, widths and Stoke shifts. This suggests that the
Fig. 1 XRD profiles for Rietveld refinement results of (a) Y1.98Ce0.02CaF4S2 and (b) Y1.98Ce0.02SrF4S2. Observed intensities
(cross), calculated patterns (red line), Bragg positions (tick mark), and difference plot (blue line) are presented.
Table 1 Crystal structural data and isotropic displacement parameters of Y1.98Ce0.02CaF4S2and Y1.98Ce0.02SrF4S2crystal systems from
Riet-veld refinementa Formula Y1.98Ce0.02CaF4S2 Lattice parameters a¼ 384.97(5), c ¼ 1891.53(4), V ¼ 280.33(6) Rwp 6.8% Rp 4.92% c2 2.02
Atom Wyck. x/a y/b z/c S.O.F. U Y 4e 0 0 0.1563(2) 0.982 0.0071(2) Ca 2b 0 0 0.5 1.005 0.0081(2) S 4e 0.5 0.5 0.1933(8) 0.997 0.0093(6) F 16n 0 0.4997(2) 0.0674(2) 0.478 0.0204(2) Ce 4e 0 0 0.1563(2) 0.018 0.0017(2) Formula Y1.98Ce0.02SrF4S2 Lattice parameters a¼ 393.46(5), c ¼ 1907.23(4), V ¼ 295.25(4) Rwp 4.96% Rp 3.57% c2 0.79
Atom Wyck. x/a y/b z/c S.O.F. U Y 4e 0 0 0.1569(2) 0.981 0.0066(2)
Sr 2b 0 0 0.5 0.987 0.0112(2)
S 4e 0.5 0.5 0.1938(8) 0.998 0.0109(8) F 16n 0 0.5 0.0659(2) 0.49 0.0106(2) Ce 4e 0 0 0.1565(2) 0.019 0.015(2)
aTetragonal; space group: I4/mmm (no. 139); lattice parameters: a and c
in pm, V in 106pm3; a¼ b ¼ g ¼ 90; T¼ 298 K; Z ¼ 2; Cu Ka, l ¼
1.5418; total reflections ¼ 5375; Site Occupancy Fraction (S.O.F.); U in A2.
Table 2 Selected interatomic bond distancesaof Y
1.98Ce0.02CaF4S2and Y1.98Ce0.02SrF4S2 Y1.98Ce0.02CaF4S2 Y1.98Ce0.02SrF4S2 (Y/Ce)–S 281.08(3) (4) (Y/Ce)–S 286.94(2) (4) (Y/Ce)–S 284.63(7) (1) (Y/Ce)–S 284.71(3) (1) (Y/Ce)–F 255.38(3) (4) (Y/Ce)–F 262.41(1) (4) Ca–F 255.58(3) (4) Sr–F 233.45(1) (4) Ca–F 255.58(3) (4) Sr–F 233.45(1) (4) a Bond distances in pm.
orange emission is originated from the substitution of the Ce3+
ion for the Y3+ ion rather than the Ca2+ ion. Since the Ce3+
emission can be only observed on the 4e site, the asymmetric emission band in the PL spectrum was fitted into two Gaussian curves centering respectively at 567 and 632 nm, ascribed to the transitions from the lowest 5dx2y2level to the two2F7/2and2F5/2
ground states of the Ce3+ions.35,36Then, the energy gap between 2F
7/2and2F5/2associated with spin-orbit splitting was calculated
to be 1814 cm1(usually2000 cm1).19On the other hand, the
PLE spectrum was characterized by two main excitation bands in the UV-vis range: one band in UV is centered at340 nm (range from 250 to 400 nm) with two shoulders at 275 and 370 nm and another band in visible is centered at470 nm (range from 420 to 530 nm). The excitation band in the UV range overlapped partially with the host absorbance; however, different Ce3+
concentrations can still affect the intensity and thereby confirm the transition from 4f to 5d states of Ce3+. The PLE spectrum can
be further decomposed into several Gaussian functions at270, 308, 339, and 368 nm, which are found to be correlated with the 4f1/ 4f05d1(b
2, e and a1) transitions and the excitation band at
472 nm is owing to the 4f1/ 4f05d1(b
1) transitions. The center
of gravity (COG) of the excitation bands is calculated to be approximately 29.47 103cm1(3.66 eV). Also, we can estimate
the crystal field splitting (CFS) to be about 15.85 103 cm1
(1.97 eV) based upon the highest and lowest absorption bands in the PLE spectrum. Depending upon the lowest excitation and the higher emission energy, the Stoke shift is obtained to be 3550 cm1(440 meV).34
Fig. 4 shows the PLE and PL spectra of Y2xCexCaF4S2and
Y2xCexSrF4S2 (x ¼ 0.01, 0.02, 0.04, and 0.06). The relative
intensity of both the PLE and PL spectra varies in accordance with different doped Ce3+concentrations, and an optimal value,
also called critical concentration (xc), was obtained for x¼ 0.02
(ca. 1 mol%). Because each activator ion is introduced solely into one site, there is on average one activator per V/xcN when
considering the concentration quenching caused by energy transfer mechanisms, such as exchange interaction, radiation reabsorption, or multipole–multipole interaction.37The critical
transfer distance (Rc) is equal to approximately twice the radius
of a sphere with the volume:37,38
Rcz2 3V 4pxcN 1 3 (2) where V represents the volume of unit cell, xc represents the
critical concentration, and N represents the number of total Ce3+
sites in the unit cell. According to the crystal structure of the YCFS:Ce3+and YSFS:Ce3+compounds, R
cvalues are reckoned
to be 18.8 and 19.2 A. If the rapid migration of Ce3+ions occurs,
quenching tends to be proportional to the Ce3+concentration;
this is not observed in PL spectra since the PLE and PL spectra do not overlap very well and the exchange interaction generally takes place in forbidden transition (the Rcis typically 5 A).
Therefore, based on the Dexter theory, we can infer that the non-radiative concentration quenching between the two nearest Ce3+
centers occurs via electric multipolar interactions.39 For the emission intensity per activator concentration, the following equation can be described:40–42
I x¼
k
1þ bðxÞq=3 (3)
where I represents the quenching intensity; x represents the Ce3+concentration; k and b represent constants for individual
electric multipolar interactions; and q ¼ 6, 8, 10 correlate correspondingly to the dipole–dipole, dipole–quadrupole, and quadrupole–quadrupole interactions. Assuming that b(x)q/3[
1, the correlation between log(I/x) with log(x) can be fitted linearly within the PL spectra of Y2xCexCaF4S2 and
Y2xCexSrF4S2(x¼ 0.02, 0.04, and 0.06) and the values of q are
determined to be 6.12 and 5.67 from the slopes (q/3). In partic-ular, the calculated value for both Y2xCexCaF4S2 and
Y2xCexSrF4S2is close to 6, which implies that the concentration
quenching mechanisms in Ce3+emission are strongly accounted
for in the dipole–dipole interaction. For the electric dipole– dipole mechanism, the transfer probability can be defined:37
PDD Ce--Ce¼ 0:63 10 28 QA sSOR6Ce--CeE4 ð FSðEÞFAðEÞdE (4)
where QA¼ 4.8 1016fdis the absorption cross-section of Ce3+;
fd z 0.01 is the electric dipole oscillator strength for Ce3+;
E (in eV) is the maximum energy of spectral overlap; integrated area represents the spectral overlap between the normalized
Fig. 2 (a) Schematic unit cell crystal structure of Y1.98Ce0.02CaF4S2and
(b) coordination environment around YF4S5and CaF8. Pink, blue, green,
and yellow spherical balls describe Y/Ce, Ca, F, and S atoms.
Fig. 3 Diffuse reflection spectrum of YCaF4S2host and PLE/PL spectra
of Y1.98Ce0.02CaF4S2. The deconvolutions of excitation and emission
bands are illustrated as dashed lines.
shapes of Ce3+emission F
S(E) and Ce3+excitation FA(E), and it is
estimated to be about 0.0481 and 0.0739 eV1. The critical distance Rc of energy transfer between Ce3+ is defined as
the distance for which the probability of transfer equals the probability of radiative emission of Ce3+, i.e., the distance for
which PCe–CesSO¼ 1. Therefore, Rccan be calculated using the
following equation:40 R6 c¼ 0:63 10 28QA E4 ð FSðEÞFAðEÞdE (5)
The Rcs of energy transfer between Ce3+ in YCFS:Ce3+ and
YSFS:Ce3+were calculated to be about 18.98 and 19.34 A, which
is close to that of the value obtained from concentration spectra.
With increasing Ce3+ concentration, the wavelength of
excitation and emission bands remains practically unchanged for both materials. In contrast, the excitation and emission bands in the visible region denote an obvious blueshift in YSFS:Ce3+
compared to YCFS:Ce3+, which is related to the changed crystal
field strength. The model that describes the crystal field splitting on account of the shape and size of the polyhedron can be determined in the following relationship:42,43
Dq¼
3Ze2r4
5R5 (6)
where Dq represents crystal field strength, Z represents the
valence of the anion ligand, e represents the charge, r represents radius of frontier d wave function, and R represents the bond length between a center ion and ligands. The substitution of Ca2+
ion is employed by isovalent substituent Sr2+ion, with both ions
known to be compatible with the phosphor host. Here, we note
that the above substitution leads to size expansion of the lattice volume, which is induced by larger Sr2+ionic size, thus changing
four Ce–S and four Ce–F bonds to 286.94 and 262.41 pm within the internal YF4S5polyhedra. The bond lengths of apical Ce–S,
284.63 and 284.71 pm are almost the same in YCFS:Ce3+and
YSFS:Ce3+ (Table 2). In such cases, the Ce3+ ion experiences
a weaker crystal field splitting due to the expansion of YF4S5
polyhedra in YSFS:Ce3+; therefore, it is reasonable that the
blueshift PLE and PL spectra are observed. The full width at half
Fig. 4 The PLE and PL spectra of Y2xCexCaF4S2(a and b) and Y2xCexSrF4S2(c and d) with different Ce3+concentrations, x. The insets show the
correlation between log(I/x) with log(x).
Fig. 5 Temperature-dependent PL intensity of commercial CaS:Ce3+,
Y1.98Ce0.02CaF4S2, and Y1.98Ce0.02SrF4S2. The inset shows the fitted PL
intensity and the calculated thermal activation energy (DE) as a function of temperature.
maximum (FWHM) of the emission bands are both found to be larger than 100 nm. The results depict that both Ce3+-doped
Y2CaF4S2and Y2SrF4S2fluorosulfides can achieve good color
rendering when incorporated in phosphor-converted white LED.
Fig. 5 shows the temperature-dependent PL intensity of CaS:Ce3+, YCFS:Ce3+, and YSFS:Ce3+in the range of 25C to
225C. The PL intensity of all samples is found to be diminished as compared to that observed at room temperature. This is due to the increasing thermal energy which ionized the electrons from the lowest state of the conduction band. Depending on the PL results, the intensity of fluorosulfide is comparable to (or even more stable than) that of binary sulfides. The introduction of F atoms into the sulfide host lattice may lower the extent of thermal quenching as a result of the softer phonon modes.30To
investi-gate the origin of temperature dependent emission intensity, the activation energy (Ea) of the electrons being excited from the 4f
level to the lowest 5d level of Ce3+ can be described in the
following equation:44,45 IðTÞ ¼ I0 1þ Aexp Ea kT (7)
where I0and I(T) represent the PL intensity at room temperature
and any temperature, respectively; k represents the Boltzmann constant. The values of Eafor YCFS:Ce3+and YSFS:Ce3+are
estimated to be 0.3741 and 0.3829 eV, respectively. The YSFS:Ce3+ shows higher activation energy characteristics
compared to those of YCFS:Ce3+. The results indicated that the
fluorosulfide could be a potential phosphor for solid-state lighting.
Tunable optical properties of Y1.98Ce0.02(Ca1ySry)F4S2
phosphors
To design a potentially compatible phosphor for a blue LED chip, the appropriate excitation band should be as close to 460 nm as possible. On the basis of the PLE results, the Ca/Sr ratio can be varied in order to obtain the desired excitation and emission wavelength of fluorosulfide phosphors. In all compo-sitions, viz., YCFS:Ce3+, YCSFS-y:Ce3+(y¼ 0.1, 0.25, 0.5, and
0.75), and YSFS:Ce3+, the single-phase samples were obtained in
the same synthetic condition. The Sr2+ion substitution for Ca2+
causes the shift of diffraction peaks to a lower angle position, which is in agreement with the refined data. The lattice
Table 3 Spectral and structural parameters for Y1.98Ce0.02CaF4S2,
Y1.98Ce0.02(Ca1ySry)F4S2and Y1.98Ce0.02SrF4S2phosphors; lex, lemand
average interatomic distances of Ce–S and Ce–Fa
Sample Composition lex lem dCe–S dCe–F
YCFS:Ce3+ Y 1.98Ce0.02CaF4S2 471 590 281.1 255.4 YCSFS-0.1:Ce3+ Y 1.98Ce0.02Ca0.9Sr0.1F4S2 470 588 281.9 255.9 YCSFS-0.25:Ce3+ Y 1.98Ce0.02Ca0.75Sr0.25F4S2 467 578 283.2 258 YCSFS-0.5:Ce3+ Y 1.98Ce0.02Ca0.5Sr0.5F4S2 461 575 285 260 YCSFS-0.75:Ce3+ Y 1.98Ce0.02Ca0.25Sr0.75F4S2 451 563 285.9 261 YSFS:Ce3+ Y 1.98Ce0.02SrF4S2 440 553 286.9 262.4 CaS:Ce3+ 456 502–560 a
Wavelength in nm; average interatomic distances in pm. Fig. 7 The total (DOS) and partial (PDOS) density of states and the projected DOS curves for (a) YCFS, (b) YSFS, (c) CeCFS, and (d) CeSFS.
Fig. 6 Excitation (monitored at 550–590 nm) and emission (lex¼ 440–
470 nm) spectra of Y1.98Ce0.02CaF4S2, Y1.98Ce0.02(Ca1ySry)F4S2 (y¼
0.1, 0.25, 0.5, and 0.75), and Y1.98Ce0.02SrF4S2. The inset shows
Y1.98Ce0.02CaF4S2, Y1.98Ce0.02(Ca1ySry)F4S2 (y ¼ 0.1, 0.25, 0.5, and
0.75), and Y1.98Ce0.02SrF4S2photos taken under 365 nm excitation.
parameters for YCFS:Ce3+, YCSFS-y:Ce3+(y ¼ 0.1, 0.25, 0.5,
and 0.75), and YSFS:Ce3+as refined from the cell refinement are
presented (see Fig. S2 in the ESI†). With the same introduction of Ce3+, the a-axis, c-axis and the average interatomic distances of
Ce–S and Ce–F increase due to the Ca replacement of Sr in Y1.98Ce0.02Ca1ySryF4S2. Fig. 6 shows the excitation and
emis-sion bands of fluorosulfide phosphors. The excitation spectrum of the YCSFS-0.5:Ce3+sample can be well excited by the blue
light (461 nm) and the emission of the phosphor shifts to 575 nm. On the other hand, the emission intensity of the phosphors was enhanced as the Sr2+ion concentration increased. The spectral
and structural parameters were summarized in Table 3. The quantum efficiency (QE) of the synthesized phosphor is obtained at room temperature. The QE of Y1.98Ce0.02Ca1ySryF4S2
increases as expected from y¼ 0 to 1 and reaches 15.9% to the maximum of 31.1% under excitation at 470 to 440 nm, respec-tively. Furthermore, the decay curve of YCFS:Ce3+,
YCSFS-y:Ce3+(y¼ 0.1, 0.25, 0.5, and 0.75) and YSFS:Ce3+phosphors
excited at 440–470 nm and monitored at 553–590 nm are shown (see Fig. S3 in the ESI†). The corresponding luminescence decay can be calculated to be 25.7, 32.1, 39.1, 45.5, and 51.82 ns using the first-order exponential equation. These results illustrate that the Ce3+ions occupy only the Y site in YCFS, YCSFS-y (y¼ 0.1,
0.25, 0.5, and 0.75), and YSFS host. As a result, the absorption band of YCSFS-y:Ce3+matches well with the emission of the
blue LED chip; thus, fluorosulfide phosphors show the potential as a promising candidate for solid-state lighting compared to commercial CaS:Ce3+.
Band structure and density of states
The use of the first principle density functional theory (DFT) in understanding the electronic structure of the host lattice has been receiving much recent attention.46,47 It is well-known that to
avoid the transformation of the crystal structure, most of the studies focused on undoped host and minor substitution system. In the present work, we have attempted to examine the energy
levels of the Ce3+ion by means of totally substituting Y3+with
a Ce3+ion. Here, we consider two different systems including
YCFS, YSFS, Ce2CaF4S2(CeCFS), and Ce2SrF4S2(CeSFS) in
an attempt to understand the variation of crystal field strength and interatomic interactions in our host compounds. Fig. 7 shows the total (DOS) and partial (PDOS) density of states and the projected DOS curves for (a) YCFS, (b) YSFS, (c) CeCFS, and (d) CeSFS, respectively. As indicated in the DOS curve of the YCFS model, the contribution of electronic states near the highest valence bands (VBs) is dominated by the S(3p) orbital, with a minor F(2p) orbital. Meanwhile, the lowest conduction band (CB) is mainly dominated by the Y(4d) orbital. The calculated electronic structure indicates that a direct charge transfer may occur not only from the S(3p) to the Y(4d) orbital but also from the F(2p) to the Y(4d) orbital. The orbital contributions observed in the DOS plot of the YSFS model are almost the same with the only difference noted in the distributed location of the Sr atom. As both the S(3p) and F(2p) orbitals are located similarly to Y(4d) in the VB (about5 to 3 eV and CB about 1 to 3 eV), interactions within these orbitals give significant contribution to the Y band. These results demonstrate that these overlaps correspond to Y–S and Y–F bonding interactions. Hence, the anionic ligands play a very important role in advancing the ligand-to-metal charge transfer (LMCT) effec-tively to the Y3+ion, which is in agreement with the strong Ce3+
photoluminescence properties as observed in our PL results. The DOS curve of the CeCFS model shows another different case in studying the charge transfer mechanism; here, the Y atoms are all replaced by dopant Ce atoms. The contributions of the electronic state around the VB close to 0 eV are dominated by a sharp localized Ce(4f) orbital and a Ce(5d) orbital around the CB. The charge transfer may occur mostly from the Ce(4f) to the Ce(5d) orbital. In addition, the DOS plot of CeSFS is essentially similar to that of CeCFS and the charge transfer in CeSFS is also made of Ce(4f) and Ce(5d) orbitals. The results suggest that the contribution of band gap is consistent with the lowest absorption
Fig. 8 A plausible mechanism of electronic transition in Y2CaF4S2:Ce3+
and Y2SrF4S2:Ce3+system. The arrows represent the electronic
transi-tions from Ce(4f) to Ce(5d) and photoemission, respectively.
Fig. 9 (a) EL spectra of a phosphor-converted white LED using Y1.98Ce0.02Ca0.45Sr0.55F4S2 as the conversion phosphor with blue
chip (460 nm). (b) Variation in CIE chromaticity coordinates as a function of the fraction of phosphor/resin used. The inset shows the photos of Y1.98Ce0.02Ca0.35Sr0.65F4S2/blue chip (left) and
Y1.98Ce0.02Ca0.45Sr0.55F4S2/blue chip (right) taken under forward bias
current. The Planckian locus line and the points corresponding to color temperatures of 6000 and 3500 K are indicated.
energy. In particular, the tendency in calculated value of band gap for CeCFS and CeSFS is close to the tendency obtained for Ca1xSrxS:Eu2+.48In CaS and SrS host lattices, the energy gaps
are 4.3 and 4.41 eV due to the dissimilarity in ionic radius between Ca2+and Sr2+.49In the CeCFS and CeSFS system, the
larger ionic radius of the Sr2+ ion introduces the increment of
the volume, thus changing the bond length between the Ce3+and
S2ions and the shrinkage of Ce(5d) was observed in the DOS plot. With the decrement of crystal field strength in CeSFS, the Ce 5d levels split smaller and thus generate a higher band gap. These changes are in agreement with the experimental results, where we find that the excited wavelength of the YSFS:Ce3+is
found to have a blueshift. A brief scheme for the luminescent mechanism (see Fig. 8), in which electrons are excited from VB via Ce(4f) to CB via Ce(5d) in YCFS:Ce3+and YSFS:Ce3+, and
then through the nonradiative Stoke shift relaxation to the lower stage is shown. In the last step, the electron goes back to the VB; such a process may result in luminescence or it may be lost thermally.
Performance of LED devices based on Y1.98Ce0.02(Ca1ySry)
F4S2phosphors
Currently, using a binary complementary color system has the benefits regarding light quality, i.e., high luminous efficacy, simple packaging fabrication, and controllable uniform
phos-phor property. To demonstrate the potential of
Y1.98Ce0.02Ca1ySryF4S2 for a phosphor-converted white LED
application, the YCSFS-0.55 and YCSFS-0.65 phosphors were then utilized to fabricate LED devices with 460 nm LED chips, as illustrated in Fig. 9. When excited by a blue chip, the whole visible spectral region can be obtained from a blue emission from the LED chip and a broad yellow emission from the YCSFS-0.55 phosphor. With the increasing ratio of the encapsulant phosphor powder, the correlating color tempera-tures (CCT) of this dichromatic white LED were determined from 6962 to 4201 K. The Commission International de l’Eclairage (CIE) chromaticity coordinates were also obtained from (0.31, 0.21) to (0.37, 0.37). The detailed CCT and CRI of the LED devices using YCSFS-0.55 and YCSFS-0.65 phos-phors and the corresponding luminous efficiencies are shown in Table 4. Compared with the white LED using a conventional YAG:Ce3+phosphor having CRI values in the range from 70 to
75 and a color temperature of 6900 K, the generated dichro-matic white light in this work possesses rather improved properties, higher CRI and lower color temperature. Given that the overall device performance depends on numerous factors
starting with the phosphor manufacturing issues, the efficiency of the LED chip, the fabrication processes and so forth, we believe that the white LED performance can be still further enhanced by the optimization of the device structure.
Conclusion
In summary, a new viable yellow-emitting fluorosulfide phos-phor with chemical composition of Y2xCexCa1ySryF4S2(x¼
0, 0.01, 0.02, 0.04 and 0.06; y¼ 0, 0.1, 0.25, 0.5, 0.75 and 1) were synthesized and studied. The overall luminescence performances (i.e., PL intensity, quantum efficiency, thermal-quenching behavior, and its application in LED fabrication) were investi-gated. The detailed crystal structure and density of states calculation were also presented. The preliminary studies show that this novel yellow phosphor is excitable over a broad range from UV to blue light, and its emission can be adjusted from yellow to orange by changing Sr2+ ions ratio. Applying
Y1.98Ce0.02Ca0.45Sr0.55F4S2phosphor on blue chip, we can obtain
a warm white LED device with a high CRI value of 85 and a CCT value of 5320 K. With the interesting tunable emission property, Y2CaF4S2:Ce3+ phosphor has great application potential as
a potential candidate for white light solid-state lighting, espe-cially for generation of warm white-light.
Acknowledgements
This research was supported by National Science Council of Taiwan (ROC) under contract no. NSC98-2113-M-009-005-MY3. We thank Dr Ming-Yang Chung for assistance in the helpful suggestion.
References
1 P. Walterett, O. Brandt, A. Trampert, H. T. Grahn, J. Menniger, M. Ramsteiner, M. Reiche and K. H. Ploog, Nature, 2000, 406, 865. 2 J. M. Phillips, M. E. Coltrin, M. H. Crawford, A. J. Fischer, M. R. Krames, R. Mueller-Mach, G. O. Mueller, Y. Ohno, L. E. S. Rohwer, J. A. Simmons and J. Y. Tsao, Laser Photonics Rev., 2007, 1, 307.
3 S. Pimputkar, J. S. Speck, S. P. Denbaars and S. Nakamura, Nat. Photonics, 2009, 3, 180.
4 V. Bachmann, C. Ronda and A. Meijerink, Chem. Mater., 2009, 21, 2077.
5 A. B. Munoz-Garcia and L. Seijo, J. Phys. Chem. A, 2011, 115, 815. 6 R. Mueller-Mach, G. Mueller, M. R. Krames, H. A. H€oppe, F. Stadler, W. Schnick, T. Juestel and P. Schmidt, Phys. Status Solidi A, 2005, 202, 1727.
7 M. R. Krames, O. B. Shchekin, R. Mueller-Mach, G. O. Mueller, L. Zhou, G. Harbers and M. Craford, J. Disp. Technol., 2007, 3, 160. Table 4 Chromaticity and optical parameters for phosphor-converted white LED with Y1.98Ce0.02Ca0.45Sr0.55F4S2and Y1.98Ce0.02Ca0.35Sr0.65F4S2as
the conversion phosphor layer
Sample
Phosphor/resin
weight ratio (%) CIE x CIE y CCT/K CRI/Ra
Luminous efficacy/lm W1 YCSFS-0.55:Ce3+ 5 0.31 0.21 6962 83.5 19.8 10 0.33 0.33 5320 84.9 16.7 15 0.37 0.37 4201 81.6 14.2 YCSFS-0.65:Ce3+ 5 0.3 0.26 8772 83.9 29.18 10 0.32 0.31 5812 85 27 15 0.41 0.41 3511 74 17.43
8 A. A. Setlur, W. J. Heward, M. E. Hannah and U. Happek, Chem. Mater., 2008, 20, 6277.
9 P. F. Smet, A. B. Parmentier and D. Poelman, J. Electrochem. Soc., 2011, 158, R37.
10 D. Jia and X. Wang, Opt. Mater., 2007, 30, 375.
11 C. Guo, D. Huang and Q. Su, Mater. Sci. Eng., B, 2006, 130, 189. 12 R. Mueller-Mach, O. Mueller-Gerd, M. R. Krames and T. Trottier,
IEEE J. Sel. Top. Quantum Electron., 2002, 8, 339.
13 H. Wu, X. Zhang, C. Guo, J. Xu, M. Wu and Q. Su, IEEE Photonics Technol. Lett., 2005, 17, 1160.
14 P. F. Smet, K. Korthout, J. E. Van Haecke and D. Poelman, Mater. Sci. Eng., B, 2008, 264, 146.
15 Y.-R. Do, K.-Y. Ko, S.-H. Na and Y.-D. Huh, J. Electrochem. Soc., 2006, 153, H142.
16 V. Petrykin, M. Okube, H. Yamane, S. Sasaki and M. Kakihana, Chem. Mater., 2010, 22, 5800.
17 P. Dorenbos, J. Lumin., 2000, 91, 155.
18 S. Gai, P. Yang, X. Li, C. Li, D. Wang, Y. Dai and J. Lin, J. Mater. Chem., 2011, 21, 14610.
19 Y. Q. Li, N. Hirosaki, R.-J. Xie, T. Takeda and M. Mitomo, Chem. Mater., 2008, 20, 6704.
20 T. Schleid, Z. Anorg. Allg. Chem., 2000, 626, 2429.
21 A. Demourgues, A. Tressauda, H. Laronzea and P. Macaudiere, J. Alloys Compd., 2001, 323, 223.
22 H. Grossholz, I. Hartenbach, G. Kotzyba, R. Pottgen, H. Trill, B. D. Mosel and T. Schleid, J. Solid State Chem., 2009, 182, 3071. 23 Y.-C. Wu, D.-Y. Wang, T.-M. Chen, C.-S. Lee, K.-J. Chen and
H.-C. Kuo, ACS Appl. Mater. Interfaces, 2011, 3, 3195.
24 Y.-C. Wu, Y.-C. Chen, D.-Y. Wang, C.-S. Lee, C.-C. Sun and T.-M. Chen, J. Mater. Chem., 2011, 21, 15163.
25 A. C. Larson and R. B. Von Dreele, General Structure Analysis System (GSAS), Los Alamos National Laboratory Report LAUR 86-748, Los Alamos National Laboratory, Los Alamos, NM, 1994. 26 O. K. Andersen, Phys. Rev. B: Solid State, 1975, 12, 3060.
27 O. Jepsen and O. K. Andersen, Z. Phys. B: Condens. Matter, 1995, 97, 35. 28 U. Von Barth and L. Hedin, J. Phys. C: Solid State Phys., 1972, 5,
1629.
29 R. D. Shannon, Acta Crystallogr., Sect. A: Cryst. Phys., Diffr., Theor. Gen. Crystallogr., 1976, 32, 751.
30 N. Kodama, M. Yamaga and B. Henderson, J. Appl. Phys., 1998, 84, 5820.
31 C. I. Aguirre, E. Reguera and A. Stein, ACS Appl. Mater. Interfaces, 2010, 2, 3257.
32 P. Dorenbos, J. Lumin., 2000, 91, 91.
33 G. Denis, X. Rocquefelte, P. Deniard, M. H. Whangbob and S. Jobic, J. Mater. Chem., 2009, 19, 9170.
34 P. Dorenbos, J. Lumin., 2002, 99, 283.
35 A. A. Setlur, E. V. Radkov, C. S. Henderson, J. H. Her, A. M. Srivastava, N. Karkada, M. S. Kishore, N. P. Kumar, D. Aesram, A. Deshpande, B. Kolodin, L. S. Grigorov and U. Happek, Chem. Mater., 2010, 22, 4076.
36 W. B. Im, S. Brinkley, J. Hu, A. Mikhailovsky, S. P. DenBaars and R. Seshadri, Chem. Mater., 2010, 22, 2842.
37 G. Blasse, Philips Res. Rep., 1969, 24, 131. 38 D. L. Dexter, J. Chem. Phys., 1953, 21, 836. 39 L. G. Van Uitert, J. Electrochem. Soc., 1963, 110, 46.
40 L. Ozawa, H. Forest, P. M. Jaffe and G. Ban, J. Electrochem. Soc., 1971, 118, 482.
41 G. Li, D. Geng, M. Shang, C. Peng, Z. Cheng and J. Lin, J. Mater. Chem., 2011, 21, 13334.
42 B. Henderson and G. F. Imbush, Optical Spectroscopy of Inorganic Solids, Clarendon, Oxford, 1989.
43 G. Blasse, Luminescence of Inorganic Solids, Plenum Press, New York, 1978.
44 R.-J. Xie, N. Hirosaki, N. Kimura, K. Sakuma and M. Mitomo, Appl. Phys. Lett., 2007, 90, 191101.
45 C.-H. Huang and T.-M. Chen, Inorg. Chem., 2011, 50, 5125. 46 C.-C. Lin, Z.-R. Xiao, G.-Y. Guo, T.-S. Chan and R.-S. Liu, J. Am.
Chem. Soc., 2010, 132, 3010.
47 C. Braun, M. Seibald, S. L. Borger, O. Oeckler, T. D. Boyko, A. Moewes, G. Miehe, A. Tucks and W. Schnick, Chem.–Eur. J., 2010, 16, 9646.
48 M. Nazarov and C. Yoon, J. Solid State Chem., 2006, 179, 2529. 49 T. Shosaku, J. Lumin., 1988, 40–41, 20.
