Nitrogen gas pressure synthesis and photoluminescent properties of orange-red SrAlSi4N7:Eu2+ phosphors for white light-emitting diodes

Article abstract of DOI:10.1111/j.1551-2916.2010.04104.x

Nitride phosphor SrAlSi₄N₇:Eu²⁺ was synthesized by gas pressure sintering of powder mixtures of Sr₃N ₂, AlN, α-Si₃N₄, and EuN at 1750°C under 0.48 MPa N₂. The photolum

Full text of DOI:10.1111/j.1551-2916.2010.04104.x

J. Am. Ceram. Soc., 94 [2] 536–542 (2011)
DOI: 10.1111/j.1551-2916.2010.04104.x
r 2010 The American Ceramic Society
ournal
J
Nitrogen Gas Pressure Synthesis and Photoluminescent Properties of
Orange-Red SrAlSi4N7:Eu21 Phosphors for White Light-Emitting Diodes
Jian Ruan,w Rong-Jun Xie,* Naoto Hirosaki,* and Takashi Takeda
Nano Ceramics Center, National Institute for Materials Science, 1-1Namiki, Tsukuba, Ibaraki 305-0044, Japan
Nitride phosphor SrAlSi4N7:Eu21 was synthesized by gas pres-
sure sintering of powder mixtures of Sr3N2, AlN, a-Si3N4, and
EuN at 17501C under 0.48 MPa N2. The photoluminescent
properties of SrAlSi4N7:Eu21 were measured and analyzed.
Two-peak emission from Eu21 located at two different Sr sites
in the SrAlSi4N7 host structure was observed. When the phorp-
hors were excited at 410 nm, the highest emission intensity was
found to be B126% of that in YAG:Ce31 excited at 460 nm.
The highest relative emission intensity at 1501C was B84.6% of
that at 301C. The highest external quantum efficiency acheived
was 58.5%. SrAlSi4N7:Eu21-based phosphors are potential for
white light-emitting diodes.
the detailed luminescent behaviors is also necessary for the
estimation of its potential application in warm white LEDs.
In this work, well crystalline SrAlSi4N7:Eu21 nitride phos-
phors were synthesized by gas pressure sintering. The lumines-
cent properties of SrAlSi4N7:Eu21 were investigated. Their
thermal quenching properties and internal and external QE
were also estimated; the results show SrAlSi4N7:Eu21 nitride
phosphors are potential for white LED.
II. Experimental Procedure
(1) Preparation of SrAlSi4N7:Eu21
SrAlSi4N7:Eu21-based phosphors were prepared by a solid-state
reaction of high-purity reagents Sr3N2 (Kojundo Chem. Lab.
Co. Ltd., Tokyo, Japan), AlN (Tokuyama Chem. Co. Ltd.,
Tokyo, Japan), a-Si3N4 (SN-E10, Ube Industries, Tokyo,
Japan), and EuN. Starting materials (B2 g) with chemical
composition of Sr1ꢀxEuxAlSi4N7 (referred to as SASN—x,
where x 5 0.01, 0.02, 0.03, 0.05, 0.075, and 0.10, respectively)
were mixed properly in an agate mortar and then filled into a
BN crucible under a continuously purified N2 atmosphere in a
glove box. Then, the mixtures were fired at 17501C for 2 h under
0.48 MPa N2 atmosphere in a gas pressure sintering furnace
(FVPHR-R-10, FRET-40, Fujidempa Kogyo Co. Ltd., Osaka,
Japan) with a graphite heater. Subsequently, the samples were
mixed properly with additive AlN with contents equal to
the first time and refired for 6 h under the same conditions,
respectively. EuN was synthesized by the reaction of metallic Eu
with N2 at 8501C for 6 h in a tube furnace.
I. Introduction
NERGY-EFFICIENT solid-state lighting technology has been
highly promoted by the development of white light-
E
emitting diodes (white LEDs) with characteristics of high
efficiency, simple structure, and long life in the past decade.1
The most common commercial white LED is combined of a
blue-emitting InGaN chip and a yellow Ce31-doped yttrium
aluminum garnet (YAG:Ce31) phosphor.2 YAG:Ce31-based
white LED exhibits a high luminous efficiency, but its
color-rendering index is limited to relatively low level by the
color deficiency in the red- and blue-green of the phosphor.3 An
alterative way to overcome this weakness is incorporation with
additional phosphors, and extensive investigations have been
devoted to these materials. Among them, rare-earth-doped
(oxy)nitride phosphors have attracted increasing attention in
recent years.3–15 They exhibit broad excitation bands covering
the emission of the commercial InGaN near-UV and blue LED
chips because of the strong nephelauxetic effect and large
crystal-field splitting caused by the coordination of nitrogen
atoms. Furthermore, higher thermal and chemical stabilities of
nitride phosphors are expected because of their stiff frameworks
in the host lattices based on [SiN4] or [AlN4] tetrahedra.10
Recently, novel nitrioalumniosilicate SrAlSi4N7 was re-
ported.15 As shown in Fig. 1, SrAlSi4N7 has an orthorhombic
structure (Pna21 symmetry) with unit cell parameter of a 5 11.7
(2) Structure Characterization
The X-ray diffraction (XRD) pattern of the as-synthesized pow-
der was analyzed using Rigaku RINT2000 system (Rigaku Co.,
˚
Tokyo, Japan) using CuKa1 (l 5 1.5408 A), operating at 40 kV
and 40 mA. Data were collected over 2y range of 10–601, with a
step width of 0.021 and count time of 12 s/step. The powder mor-
phology was investigated by field-emission scanning electron
microscopy (FESEM, JEOL-840A, JEOL Co., Tokyo, Japan).
˚
˚
˚
A, b 5 21.31 A, and c 5 4.95 A and infinite chains of edge-shar-
ing [AlN4] tetrahedra in crystal structure. It is obvious that there
are two different Sr lattice sites in SrAlSi4N7. Emission around
635 nm was observed in the SrAlSi4N7:Eu21 and high quantum
efficiency (QE) was also expected. However, the reported syn-
thesis of SrAlSi4N7:Eu21 based on radio-frequency furnace is
usually baffled by the byproducts of microcrystalline Sr2Si5N8.
For practical applications, both high purity and crystallinity are
required for the synthesized product. Therefore, applicable syn-
thesis methods for the novel SrAlSi4N7:Eu21 phosphor are still
urgently required for mass production. Besides this, the study of
(3) Luminescence Properties Measurement
The photoluminescence (PL) and photoluminescence excitation
(PLE) spectra were measured at room temperature using a
fluorescent spectrophotometer (F-4500, Hitachi Ltd., Tokyo,
Japan) with a 200 W Xe lamp as the excitation source. The QE
of the samples were measured using a multi-channel spectro-
photometer (MCPD-7000, Otsuka Electronics, Tokyo, Japan)
with a 200 W Xe lamp as the excitation source. A white BaSO4
powder was used as a standard reference for correction in the
measurements of QE. The temperature-dependent photolumines-
cence (TDPL) spectra (30–2001C) were also conducted by the
MPCD-7000 machine under excitation at 410 nm. The measure-
ments of room-temperature time-resolved photoluminescence
(TRPL) were carried out in a time-controlled single photon count-
ing (TCSPC) system (FluoroHub, Horiba Jobin Yvon, Edison,
NJ) with a TBX picoseconds photon detection module. The
J. Ballato—contributing editor
Manuscript No. 28090. Received May 26, 2010; approved July 30, 2010.
*Member, The American Ceramic Society.
wAuthor to whom correspondence should be addressed. e-mail: bowkeeper@163.com
536

February 2011
Orange-Red SrAlSi4N7:Eu21 Phosphors
537
Fig. 1. Crystal structure of Sr1ꢀxEuxAlSi4N7. The Sr/Eu, Al, Si, and N
atoms are marked in green, red, blue, and light gray, respectively.15
interchangeable nanoLED (370 nm, Horiba Jobin Yvon) was used
as an excitation source.
III. Results and Discussion
Fig. 2. (a) XRD patterns of samples after the first firing. Standard
Sr2Si5N8 and SrAlSi4N7 patterns were also shown for comparison. (b)
XRD patterns of samples after the second firing. Standard Sr2Si5N8 and
SrAlSi4N7 patterns were also shown for comparison.
(1) Synthesis and Structural Characteristics of SASN—x
Phosphors
It was found that the pure SrAlSi4N7 phase was difficult to be
acheived with starting raw materials mixed stoichiometrically.
Additional AlN may be helpful to solve this problem. Figures
2(a) and (b) shows the XRD patterns of samples after the first
firing (stoichiometrical compositon) and the second firing (ad-
ditional AlN introduced). For comparison, the standard XRD
patterns of Sr2Si5N8 and SrAlSi4N7 (downloaded from ICSD
website) were also shown in Figs. 2(a) and (b). The residual
Si3N4 and AlN and a small amount of as-synthesized SrAlSi4N7
have been marked in Fig. 2(a). Very little amount of residual
Sr2Si5N8 has also been marked in Fig. 2(b). For better illumi-
nation, the main phases in those patterns were not marked, but
they can be identified easily. It is clear that the main phase of
the first firing products is Sr2Si5N8 and only a little amount of
SrAlSi4N7 can be obtained. On the other hand, the main phase
of the second firing product is SrAlSi4N7 and almost no AlN or
Sr2Si5N8 phase can be recognized. The content of AlN seems to
be very important for the synthesis of the SrAlSi4N7 phase.
However, it is difficult to distinguish those residual AlN from
XRD patterns after the second firing (Fig. 2(b)). Further inves-
tigation may be necessary for better understanding of the func-
tion and behavior of AlN in the growth of SrAlSi4N7 crystalline.
The designed compositions are Sr1ꢀxEuxAlSi4N7 (x 5 0.01,
0.02, 0.03, 0.05, 0.075, and 0.10) (SASN—x). The theoretical
compositions of the final products should be Sr1ꢀxEuxAl2Si4N7
(x 5 0.01, 0.02, 0.03, 0.05, 0.075, and 0.10) in mol% because of
the additional AlN. The chemical composition of the sample
was measured by an induction coupled plasma atomic emission
spectroscopy method and an oxygen–nitrogen analyzer (TC-
436, LECO, St. Joseph, MI). The results show that the actual
compositions are almost the same as the theoretical ones. For
example, the actual and the theoretical compositions of the
SASN—0.05 sample are [Sr: 22.070.1%, Eu: 2.1170.01%, Al:
14.670.1%, Si: 30.070.1%, N: 28.770.02%, O: 09770.06%,
rest are unknown] and [Sr: 22.55%, Eu: 2.06%, Al: 14.61%, Si:
30.43%, N: 30.35%] in wt%, respectively. Little oxygen impu-
rity was found because it is usually unavoidable from the start-
ing materials and from the second contamination in the
processing for all the nitride compounds.
The intensity distribution of diffraction peaks shown in
Fig. 2(b) seems to be a little different from the standard pattern,
especially for those samples with higher Eu concentrations. This
may be ascribed to the preferred growth of Sr1ꢀxEuxAlSi4N7
crystallines along the c-direction [001] because of the infinite
chains of edge-sharing [AlN4] tetrahedra in the crystal structure
as shown in Fig. 1.
The assumption can be supported by FESEM observations.
Figure 3 shows the particle morphology images of SASN—0.01,
0.05, and 0.10 samples. It can be found that crystallines with a
larger particle size and a more cylindrical shape were obtained with
an increase in the Eu concentration. This is because lower tem-
perature of the solid-state reaction can be estimated when higher
content of Eu was used to replace the Sr in the Sr1ꢀxEuxAlSi4N7
crystalline. Figures 3(d), (e), and (f) show the crystalline particles
with different shapes. As shown in Fig. 3(d), the crystallines with
larger size usually show smooth surface on their head faces and
adhered dozens of smaller particles on their side faces. Figure 3(e)
indicates that the crystalline can epitaxially grow starting from
the same face. Figure 3(f) shows an FESEM image with zoom

538
Journal of the American Ceramic Society—Ruan et al.
Vol. 94, No. 2
Fig. 3. (a)–(c) Morphology images of SASN—0.01, SASN—0.05, and SASN—0.10. (d)–(f) crystalline particles with different shapes ((a)–(c)ꢁ 1000, (d)–
(e) ꢁ 3000, and (f) ꢁ 20 000).
multiple of 20 000. Two types of particles can be observed: cyl-
inder particle A with a diameter of B1.0 mm and a length of
B4.0 mm, and amorphous particles B adhered in the two head
sides of particle A. It can be observed that the side face of
particle A is very smooth. This indicates that the axial growth
of the cylindrical crystalline particles may play a dominant role
in forming such a morphology. Although the particle mor-
phology was found to be cylindrical in SASN:Eu21 phosphors,
we believed that the influence of phosphor sharp on their
practical application will become weaker once those phosphors
can be well milled and classified. A similar case can be found in
Eu21-doped b-SiAlON phosphors, which have been applied
successfully.
the dependence of PL intensity and emission peak position of
SASN—x samples as a function of Eu concentration. As seen,
the emission peak shifted from B610 to B650 nm with an in-
crease in the Eu concentration, while the PL intensity increased
at first till x 5 5% and then decreased gradually. The PL max-
imum value was observed in the SASN—0.05 sample. When it
(2) Photoluminescent Properties of SASN—x Phosphors
The PL and PLE spectra of SASN—x samples are shown in
Fig. 4(a). The PL spectra were measured with irradiations at 410
nm, while the PLE spectra were monitored at 620, 623, 625, 640,
645, and 650 nm for SASN—x (x 5 0.01, 0.02, 0.03, 0.05, 0.075,
and 0.10) samples, respectively. The intensity has been normal-
ized according the emission intensity of a standard YAG:Ce31
(P46Y3) with excitation at 460 nm for comparison. As shown in
Fig. 4(a), all samples exhibit broadband excitation and emission
spectra. The excitation spectra consist of four bands at 320, 355,
410, and 460 nm. With an increase in the Eu concentration, only
difference in excitation intensity was observed in the first two
bands. On the other hand, both peak shifts and excitation in-
tensity differences can be found in the last two bands. It is in-
teresting that a quite plate excitation band from 350 to 400 nm
was observed in the SASN—0.1 sample.
A single and broad emission band peaking at 620–650 nm
was observed in all samples when the samples were excited at
355, 410, or 460 nm. For instance, the PL spectra of SASN—
0.05 with excitation at the above-mentioned wavelengths are
given in Fig. 4(b). The full-width at half-maximum (FWHM) is
115, 119, and 114 nm, respectively. With excitation at 410 nm,
the FWHM is 125, 122, 121, 119, 114, and 113 nm for SASN—x
(x 5 0.01, 0.02, 0.03, 0.05, 0.075, and 0.10), respectively, as seen
in Fig. 4(a). The emission should be ascribed to 4f65d1–4f tran-
sition of Eu21 ions because no characteristic emission of Eu31
(sharp lines between 580 and 650 nm) was observed. It is sug-
gested that Eu31 were reduced to Eu21 in a reducing carbon
monoxide gas atmosphere using a graphite heater and in a ni-
trogen atmosphere.
Fig. 4. (a) Photoluminescence (PL) and photoluminescence excitation
(PLE) spectra of SASN—x (x 5 0.01, 0.02, 0.03, 0.05, 0.075, and 0.1)
samples. (b) PL spectra of SASN—0.05 with excitation at 355, 410, and
460 nm.
In addition, the emission peak also exhibits an obvious red-
shift with an increase in the Eu concentration. Figure 5 shows

February 2011
Orange-Red SrAlSi4N7:Eu21 Phosphors
539
relaxation and energy transfer among the active ions.16 The
similar phenomenon observed in Figs. 4 and 5 could also be
explained as the re-absorption processes.
In order to further understand the re-absorption process, a
series of lifetime decay curves were recorded by monitoring the
SASN—0.01 and SASN—0.10 samples in different wavelengths
from 450 to 750 nm at 3 nm interval, and the TRPL spectra
shown in Figs. 6(a) and (b) were obtained by slicing them, re-
spectively. When the time interval was 1.4 ns, both the emission
peaks A and B were apparently observed in the spectra. The
occurrence of the two emission peaks may be assigned to the
Eu21 emission from the two different Sr lattice sites. The Eu21
can enter the structure and occupy two distinct Sr21 sites, in
which one (referred to as Sr1) is coordinated by eight nitrogen
˚
ions with an average distance of 2.87 A, and the other (referred
to as Sr2) is coordinated by six nitrogen ions with an average
˚
distance of 2.71 A. Because lower ligand field strength is ex-
Fig. 5. Eu concentration-dependent photoluminescence (PL) intensity
and peak position of SASN—x (x 5 0.01, 0.02, 0.03, 0.05, 0.075, and 0.1)
samples with excitation at 355, 410, and 460 nm.
pected in the former position based on its larger coordination
number, Eu21 occupying the Sr1 position should exhibit a
higher-energy (shorter wavelength) emission peak. Therefore,
the observed emission peaks A and B should be originated from
the Eu21 ions located in Sr1 and Sr2 positions, respectively.
As shown in Fig. 6, with the time interval prolonging, it was
interesting to observe that the relative emission intensity of peak
A rapidly descended, whereas that of peak B increased till the
time interval was 2.9 ns and then descended gradually. An ap-
parent red-shift of emission peak B was also observed in the
both SASN—0.01 and SASN—0.1, while no apparent red-shift
of emission peak A was found. In comparison with SASN—
0.01, a higher relative intensity of emission peak B versus peak A
was observed in the SASN—0.1 sample. This indicates that
Eu21 ions will occupy Sr2 sites taking precedence over Sr1 sites,
when the Eu concentration increased from 1% to 10%.
The thermal quenching properties of SASN—x samples were
also investigated. Figure 7(a) presents the TDPL spectrum of
SASN—0.01 with 410 nm excitation. A slight blue-shift from
600 to 594 nm was observed in the spectra when the temperature
increased from 301 to 2001C. This phenomenon can be ascribed
to the thermally active phonon-assisted tunneling from the ex-
cited states of lower-energy emission band to those of higher-
energy emission band in the configuration coordinate diagram.17
The temperature-dependent emission intensities of all
SASN—x samples under 410 nm excitation are shown in
Fig. 7(b). It can be found that the emission intensity of
SASN—x (x 5 0.01, 0.02, 0.03, 0.05, 0.075, and 0.1) samples
at 1501C remains 84.6, 83.6, 83.1, 77.2, 72.9, and 68.1% of
that measured at 301C, respectively. The activation energy for
the thermal quenching was also estimated by the equation as
follows18:
was excited at 410 and 460 nm, the PL intensity was B126%
and B111% of that observed in YAG:Ce31 (P46Y3) with 460
nm excitation, respectively.
Table I shows the Eu21 5d excitation bands, emission band
maximum, crystal-field splitting, and Stokes shifts of SASN—x
(x 5 0.01, 0.02, 0.03, 0.05, 0.075, and 0.10) samples. The crystal-
field splitting was estimated from the energy difference between
the highest and the lowest observed 5d excitation levels of Eu21
.
Stokes shifts were calculated from the energy difference between
the lowest 5d excitation level (labs) and the emission band of
Eu21 (lem). Limited to the resolutions of excitation spectra at
room temperature, it is difficult to avoid the extra errors if the
labs was determined directly from the spectra. Therefore, the
mirror-image relationship may be available to estimate the labs
for this work based on the Gaussian sharp of the observed
emission peaks. The labs was calculated by the formula as fol-
lows:
labs ¼ 2E0 ꢀ lem
where E0 is the zero-phonon line that can be determined at the
intersection of the emission and excitation spectra. Then, the
Stokes shifts were calculated as DS 5 labsꢀlem. As shown in
Table I, the crystal-field splitting and Stokes shifts increased and
decreased with an increase in the Eu concentration, respectively.
The energy levels of Eu21 ions have been extensively studied.
The broadband emission from the 4f65d1–4f transition of Eu21
ions has been observed in the color range from ultraviolet to red,
which depends on the ligand-field splitting of the 5d levels
caused by the host lattice. In this work, orange-red emission
was observed in SASN—x samples because of the relatively
large ligand-field splitting of the 5d levels and the low-energy
center of gravity of the 5d states due to the high formal charge of
N3ꢀ and the nephelauxetic effect. The red-shift of emission
peaks is commonly observed in rare-earth-doped phosphors,
as the doping concentration increases. This phenomenon has
been assigned to the re-absorption processes caused by the cross-
Io
ꢀ
ꢁ
IðTÞ ¼
ꢀDE
1 þ A exp
kB
T
where Io is the initial intensity, I(T) is the intensity at a given
temperature T, A is a constant, DE is the activation energy for
thermal quenching, and kB is Boltzmann’s constant. As shown
in the inset of Fig. 6(b), it can be concluded that the DE for
Table 1. The Eu21 5d Excitation Bands, Zero-Phonon Line (E0), Emission Band Maximum (lem), Crystal-Field Splitting and Stokes
Shifts (DS) of SASN—x (x 5 0.01, 0.02, 0.03, 0.05, 0.075, and 0.10, Respectively) Samples
Eu21 5d excitation
Crystal-field
splitting (cmꢀ1
FWHM (nm)
at 410 nm
SASN—x
bands (nm)
E0 (nm)
lem (nm)
)
DS (cmꢀ1
)
x 5 0.01
x 5 0.02
x 5 0.03
x 5 0.05
x 5 0.075
x 5 0.10
31472, 35372, 40672, 47272 (labs
31372, 35372, 40672, 47972 (labs
31472, 35572,40772, 48172 (labs
31472, 35672,41072, 49672 (labs
31772, 35672, 40872, 50572 (labs
31572, 35872, 41272, 51272 (labs
)
)
)
)
)
)
54672
55172
55372
56872
57572
58172
62072
62372
62572
64072
64572
65072
10 1557100
10 5467100
10 3887100
10 4807100
10 7127100
10 9567100
50577100
48257100
47907100
45367100
42987100
41477100
125
122
121
119
114
113

540
Journal of the American Ceramic Society—Ruan et al.
Vol. 94, No. 2
Fig. 7. (a) The temperature-dependent photoluminescence (TDPL) of
SASN—0.01 measured with increasing temperature from 301 to 2001C.
(b) TDEL emission intensity of SASN—x (x 5 0.01, 0.02, 0.03, 0.05,
0.075, and 0.1) samples. The inset shows an Arrhenius fitting of the
emission intensity and calculated DE. All the spectra were measured with
410 nm excitation.
Fig. 6. (a) Time-resolved photoluminescence (TRPL) emission spectra
of SASN—0.01. (b) TRPL emission spectra of SASN—0.1.
SASN—0.01 is 0.235 eV, which is a little higher than the value
(0.20 eV) reported in Sr2Si5N8:Eu21 phosphors.
Besides the emission spectra, the external and internal QE of
SASN—x samples were also measured and calculated using the
method described in Hirosaki et al.3 With excitation at 410 nm,
the internal QE of SASN—x (x 5 0.01, 0.02, 0.03, 0.05, 0.075,
and 0.10) are about 84.0, 83.1, 80.4, 71.2, 62.3, and 56.3%, re-
spectively. The external QE in the wavelength range from 400
nm to 500 nm decreased gradually but the absorption increased
with an increase in the Eu concentration. This finally leads to an
initial increase in the external QE of SASN—x samples till
x 5 0.03 and a subsequent decrease. When excited at 410 nm,
the external QE of SASN—x (x 5 0.01, 0.02, 0.03, 0.05, 0.075,
and 0.10) are about 54.0, 58.0, 58.5, 57.5, 49.0, and 44.2%, re-
spectively. Figure 8 shows the internal and external QE and ab-
sorption of SASN—0.05 as a function of excitation wavelengths.
We believed that the external QE value can be improved by
further purification and particle classification.
tetrahedras in Sr2Si5N8 and CaSiAlN3 were connected by cor-
ner-sharing. And lower crystal-field splitting was achieved in
SrAlSi4N7 in comparison with Sr2Si5N8 and CaSiAlN3. Such a
difference in the host lattice may result in higher vibronic inten-
sity and stronger vibronic transition of Eu21 ions in SrAlSi4N7
compared with Sr2Si5N8 and CaSiAlN3. This will lead to a
Orange-red-emitting nitride phosphors have been extensively
studied and high external QE and thermal stability have been
achieved. For comparison, the crystal structures and lumines-
cent properties of Sr2Si5N8, CaSiAlN3, and SrAlSi4N7 are given
in Table II. It can be found that the FWHM of SrAlSi4N7 is
more than 25 nm broader than others, which indicates that a
higher color-rendering index may be achieved after the incor-
poration of phosphor. This may be ascribed to the difference in
their crystal structures. All of them are orthorhombic in the
crystal system, but the [AlN4] tetrahedras in SrAlSi4N7 con-
nected with each other by edge-sharing and formed infinite
chains in the crystal structure, while the [AlN4] or [SiN4]
Fig. 8. Internal and external quantum efficiency (QE) and absorption of
the SASN—0.05 sample.

February 2011
Orange-Red SrAlSi4N7:Eu21 Phosphors
541
Table II. Crystal Structures and Photoluminescent Properties of Sr2Si5N8, CaSiAlN3, and SrAlSi4N7
Formula
Sr1.98Eu0.02Si5N8
Ca0.992Eu0.008AlSiN3
Sr0.97Eu0.03AlSi4N7
Crystal system
Space group
Lattice parameters (A)
Orthorhombic
Pmn21 (No. 31)
a 5 5.710 (1)
b 5 6.822 (1)
c 5 9.341 (2)
Sr1 5 10,
Sr2 5 8
Sr1 5 2.928,
Sr2 5 2.865
Orthorhombic
Cmc21 (No. 36)
a 5 9.8007 (4)
b 5 5.6497 (2)
c 5 5.0627 (2)
Ca5 5
Orthorhombic
Pna21 (No. 33)
a 5 11.742 (2)
b 5 21.391 (4)
c 5 4.966 (1)w
Sr1 5 8,
˚
Coordination number (CN)
Avenge bond length of M-N
˚
(M5 Sr or Ca) (A)
Shortest band length of
˚
M-N (M5 Sr or Ca) (A)
Sr2 5 6w
Ca5 2.485
Ca5 2.405
Sr15 2.87,
Sr2 5 2.71w
Sr15 2.653
Sr2 5 2.504w
625 nmz
Sr1 5 2.542
Sr2 5 2.570
625 nm
Emission peaks
Excitation peaks
650 nm
335, 450, 500, 564 nm
14600
334, 395, 465, 505 nm
14200
3700
314, 355, 407, 481 nmz
103887100z
Crystal-field splitting (cmꢀ1
)
Stokes shifts (cmꢀ1
FWHM (nm)
Absorption
)
2500
47907100z
86 nm at 450 nm
80% at 450 nm
64% at 450 nm
86% at 450 nm
Mueller-Mach and colleagues9–11,19,20
89 nm at 460 nm
80% at 460 nm
76% at 460 nm
87% at 450 nm
Uheda and colleagues8,19
116 nm at 460 nmz
70% at 460 nmz
49.0% at 460 nmz
83.1% at 410 nmz
wHecht et al.15
external QE
Thermal quenching at 1501C
References
zThis work
broader emission observed in SrAlSi4N7. Similarly, it can be
found that the observed FWHM of Eu21-doped SrAlSi4N7
decreases with increasing crystal-field splitting when Eu concen-
tration increases, as shown in Table I.
Stokes shift is the reflection of energy consumption during the
electron relaxation from the lowest 5d–4f excitation band to
emission band. Based on the Stokes shift model, higher thermal
stability and internal QE of the PL should be achieved in the
phosphors with smaller Stokes shift. This is in agreement with
our results, that is, the Stokes shift occurs in the following se-
quence: CaSiAlN3oSr2Si5N8oSrAlSi4N7. The thermal stability
and internal QE will also be influenced by the impurity in the
crystalline. This is the reason why stronger thermal quenching
and lower internal QE were observed when the Eu concentration
increased.
It should also be noted that Sr0.9Eu0.1AlSi4N7 shows a rela-
tively plant excitation band in the wavelength range between 300
and 500 nm. This means that the sample is efficient for the ex-
citation of blue GaN and near-UV InGaN chips and compatible
with various blue, green, or yellow phosphors.21
As shown in Fig. 9, the emission peak position usually
depends on the crystal-field splitting and the nephelauxetic effect
in the Eu21-doped (oxy)nitride compounds. Both of them will
reduce the energy difference between the ground and the
excited state of Eu21 ions; hence, the excitation and emission
bands at lower energies (i.e. longer wavelength) were observed in
various phosphors. As shown in Table II, Eu21 ions can occupy
two different Sr lattice sites in SrAlSi4N7 with less CN and a
little shorter average band length than those in Sr2Si5N8. But
both of them are much more and longer than those in CaSiAlN3,
respectively. These indicate that stronger nephelauxetic effect
will be caused in SrAlSi4N7 than in Sr2Si5N8, but it will be
weaker than that in CaSiAlN3. And the weaker crystal-field
splitting in the SrAlSi4N7 will lead to higher energy of the lowest
5d–4f level and thus caused the emission observed at a wave-
length shorter than that in CaAlSiN3 and similar to that in
Sr2Si5N8.
IV. Conclusion
Novel orange-red SrAlSi4N7:Eu21 nitride phosphors were syn-
thesized by gas pressure sintering. The luminescent properties of
SrAlSi4N7:Eu21 were studied as a function of Eu concentration.
The TRPL and TDPL spectra of SrAlSi4N7:Eu21 were also
measured. The SrAlSi4N7:Eu21 phosphors were estimated to
show good thermal quenching properties and high external QEs,
which are potential orange-red down-conversion luminescent
materials for white LEDs. The crystal structure and photolumi-
nescent properties of SrAlSi4N7 were also compared with
Sr2Si5N8 and CaSiAlN3. The results may be useful to anticipate
the photoluminescent properties of other rare-earth ion-doped
SrAlSi4N7 compared with Sr2Si5N8 and CaSiAlN3.
Among the three above-mentioned nitride phosphors, SrAl-
Si4N7 shows the weakest crystal-field splitting and comparable
strong nephelauxetic effect with Sr2Si5N8 but weaker than
CaAlSiN3. This is the reason why excitation bands at higher
energy were observed in SrAlSi4N7.
References
1E. Schubert and J. Kim, ‘‘Solid-State Lighting Sources Getting Smart,’’ Sci-
ence, 308 [5726] 1274–8 (2005).
2S. Nakamura and G. Fasol, The Blue Laser Diode: GaN Based Light Emitters
and Lasers. Springer, Berlin, 1997.
3N. Hirosaki, R.-J. Xie, K. Kimoto, T. Sekiguchi, Y. Yamamoto, T. Suehiro,
and M. Mitomo, ‘‘Characterization and Properties of Green-Emitting b-Si-
AlON:Eu21 Powder Phosphors for White Light-Emitting Diodes,’’ Appl. Phys.
Lett., 86 [21] 211905, 3pp (2005).
4H. A. Hoppe, H. Lutz, P. Morys, W. Schnick, and A. Seilmeier, ‘‘Luminescence
in Eu21-doped Ba2Si5N8: Fluorescence, Thermoluminescence, and Upconver-
sion,’’ J. Phys. Chem. Solids, 61 [12] 2001–6 (2000).
Fig. 9. Schematic representation of the energy level of the Eu21 ion in
(oxy)nitride phosphors.

542
Journal of the American Ceramic Society—Ruan et al.
Vol. 94, No. 2
5R.-J. Xie, M. Mitomo, K. Uheda, F.-F. Xu, and Y. Akimune, ‘‘Preparation
(M5 Ca, Sr, Ba): A Promising Class of Novel LED Conversion Phosphors,’’
Chem. Mater., 17 [12] 3242–8 (2005).
and Luminescence Spectra of Calcium- and Rare-Earth (R5 Eu, Tb, and Pr)-
Codoped Alpha-SiAlON Ceramics,’’ J. Am. Ceram. Soc., 85 [5] 1229–34 (2002).
6R.-J. Xie, N. Hirosaki, K. Sakuma, Y. Yamamoto, and M. Mitomo,
‘‘Eu21-Doped Ca-Alpha-SiAlON: A Yellow Phosphor for White Light-Emitting
Diodes,’’ Appl. Phys. Lett., 84 [26] 5404–6 (2004).
14Y. Q. Li, G. de With, and H. T. Hentzen, ‘‘Luminescence Properties of
Ce31-Activated Alkaline Earth Silicon Nitride M2Si5N8 (M5 Ca, Sr, Ba)
Materials,’’ J. Lumin., 116 [1–2] 107–16 (2006).
15C. Hecht, F. Stadler, P. J. Schmidt, J. S. Auf der Gunne, V. Baumann, and
¨
7R.-J. Xie, N. Hirosaki, M. Mitomo, K. Takahashi, and K. Sakuma, ‘‘Highly
Efficient White-Light-Emitting Diodes Fabricated with Short-Wavelength Yellow
Oxynitride Phosphors,’’ Appl. Phys. Lett., 88 [10] 101104, 3pp (2006).
8K. Uheda, N. Hirosaki, Y. Yamamoto, A. Naoto, T. Nakajima, and
W. Schnick, ‘‘SrAlSi4N7:Eu21—A Nitridoalumniosilicate Phosphor for Warm
White Light (pc)LEDs with Edge-Sharing Tetrahedra,’’ Chem. Mater., 21 [8]
1595–601 (2009).
16K.-S. Sohn, S. Lee, R.-J. Xie, and N. Hirosaki, ‘‘Time-Resolved Photolumi-
nescence Analysis of Two-Peak Emission Behavior in Sr2Si5N8:Eu2,’’ Appl. Phys.
Lett., 95 [12] 121903, 3pp (2009).
H. Yamamoto, ‘‘Luminescence Properties of a Red Phosphor, CaAlSiN3:Eu21
,
for White Light-Emitting Diodes,’’ Electrochem. Solid-State Lett., 9 [4] H22–5
(2006).
17J. S. Kim, Y. H. Park, S. M. Kim, J. C. Choi, and H. L. Park, ‘‘Temperature-
Dependent Emission Spectra of M2SiO4:Eu21 (M 5 Ca, Sr, Ba) Phosphors
for Green and Greenish White-LEDs,’’ Solid State Commun., 133 [7] 445–8 (2005).
18S. Bhushan and M. V. Chukichev, ‘‘Temperature-Dependent Studies of
Cathodoluminescence of Green Band of ZnO Crystals,’’ J. Mater. Sci. Lett., 9
[4] 319–21 (1988).
9R. Mueller-Mach, G. Mueller, M. R. Crames, H. A. Hoppe, F. Stadler,
W. Schnick, T. Juestel, and P. Schmidt, ‘‘Highly Efficient All-Nitride Phosphor-
Converted White Light Emitting Diode,’’ Phys. Status A, 202 [9] 1727–32
(2005).
10W. Schnick, ‘‘Nitridosilicates, Oxynitridosilicates (sions), and Oxynitrido-
aluminosilicates (sialons)—New Materials with Promising Properties,’’ Int. J.
Inorg. Mater., 3 [8] 1267–72 (2001).
19R.-J. Xie, N. Hirosaki, N. Kimura, and K. Sakuma, ‘‘White Light-Emitting
Diodes (LEDs) Using (Oxy)nitride Phosphors,’’ J. Phys. D: Appl. Phys., 41 [14]
144013, 5pp (2008).
11X. Piao, T. Horikawa, H. Hanzawa, and K. Machida, ‘‘Characterization and
Luminescence Properties of Sr2Si5N8:Eu21 Phosphor for White Light-Emitting-
Diode Illumination,’’ Appl. Phys. Lett., 88 [16] 161908, 3pp (2006).
12Y. Q. Li, G. de With, and H. T. Hentzen, ‘‘Luminescence of a New Class of
UV–Blue-Emitting phosphors MSi2O2ꢀdN212d/3:Ce31(M5 Ca, Sr, Ba),’’ J. Ma-
ter. Chem., 15 [42] 4492–6 (2005).
20Y. Q. Li, J. E. J. van Steen, J. W. van Krevel, G. Botty, A. C. A. Delsing, F. J.
DiSalvo, G. de With, and H. T. Hintzen, ‘‘Luminescence Properties of Red-
Emitting M2Si5N8:Eu21 (M5 Ca, Sr, Ba) LED Conversion Phosphors,’’ J. Alloys
Compd., 417 [1–2] 273–9 (2006).
21R.-J. Xie, N. Hirosaki, N. Kimura, K. Sakuma, and M. Mitomo,
‘‘2-Phosphor-Converted White Light-Emitting Diodes Using Oxynitride/Nitride
13Y. Q. Li, A. C. A. Delsing, G. de With, and H. T. Hentzen, ‘‘Luminescence
Properties of Eu21-Activated Alkaline-Earth Silicon-Oxynitride MSi2O2–dN212d/3
Phosphors,’’ Appl. Phys. Lett., 90 [19] 191101, 3pp (2007).
&