Photoluminescence of nanocrystalline SrMgF4 prepared by a solution chemical route

Article abstract of DOI:10.1016/j.materresbull.2007.02.010

SrMgF₄ was prepared by precipitation in aqueous solution. Alkaline earth metal acetates and ammonium fluoride were used as precursors. After drying and annealing the samples at different temperatures and times, single phase SrMgF₄ was obtained. By varying the annealing conditions, the mean crystallite size could be adjusted. Furthermore, the thermally treated samples displayed UV-excited intensive broad band luminescence in the visible region. The emissions colour and intensity can be adjusted by the tempering conditions. X-Ray diffraction, TEM-microscopy, fluorescence and IR-spectroscopy were used for analysis.

Full text of DOI:10.1016/j.materresbull.2007.02.010

Materials Research Bulletin 43 (2008) 168–175
www.elsevier.com/locate/matresbu
Photoluminescence of nanocrystalline SrMgF4 prepared
by a solution chemical route
a
b
C.Veitsch a, , F. Kubel , H. Hagemann
*
a
¨
Institut fur Chemische Technologien und Analytik, TU Wien, Getreidemarkt 9/164, A 1060 Vienna, Austria
´
b
´ `
Departement de Chimie Physique, Universite de Geneve, 30, quai E. Ansermet, 1211 Geneva 4, Switzerland
Received 8 November 2006; received in revised form 18 December 2006; accepted 6 February 2007
Available online 11 February 2007
Abstract
SrMgF4 was prepared by precipitation in aqueous solution. Alkaline earth metal acetates and ammonium fluoride were used as
precursors. After drying and annealing the samples at different temperatures and times, single phase SrMgF4 was obtained. By
varying the annealing conditions, the mean crystallite size could be adjusted. Furthermore, the thermally treated samples displayed
UV-excited intensive broad band luminescence in the visible region. The emissions colour and intensity can be adjusted by the
tempering conditions. X-Ray diffraction, TEM-microscopy, fluorescence and IR-spectroscopy were used for analysis.
# 2007 Elsevier Ltd. All rights reserved.
Keywords: A. Fluorides; B. Chemical synthesis; C. Electron microscopy; C. X-ray diffraction D. Luminescence
1. Introduction
Fluorides have been studied for a very long time. In particular, their transparency in the UV region of the
electromagnetic spectrum makes them an interesting substance class for optical applications. In many cases it is
possible to dope fluorides with rare earth ions to achieve UV-fluorescence. Mixed alkaline earth fluorides have been
shown to be interesting host crystals for divalent rare earth ions:
One example for such a host matrix is the mixed alkaline earth fluoride SrMgF4 [1–3]. This compound is the only
ternary compound reported for the SrF2–MgF2 phase diagram. SrMgF4 melts incongruently at about 900 8C [4]. Its
crystal structure has been described by Banks et al. [4] as orthorhombic. In 2001, Ishizawa et al. [5] described a
monoclinic superstructure of SrMgF4 for a crystal containing also Ce3+ ions. The aforementioned authors used melting
or powder synthesis methods to prepare the fluoride compound, which involve high temperatures (875 8C [4]) and
protective gas atmosphere.
In the present work we have established a solution based synthesis approach for the preparation of SrMgF4. It has
the advantage of lower reaction temperatures, shorter reaction times and it results reproducibly in a well defined single
phase product.
* Corresponding author. Tel.: +43 15880117112; fax: +43 15880117199.
E-mail address: cveitsch@mail.zserv.tuwien.ac.at (C.Veitsch).
0025-5408/$ – see front matter # 2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.materresbull.2007.02.010

C.Veitsch et al. / Materials Research Bulletin 43 (2008) 168–175
169
2. Experimental
The title compound SrMgF4 was prepared by mixing an aqueous solution containing 0.2 M Sr2+ and Mg2+ ions with
an aqueous 0.3 M NH4F solution in an ultrasonic bath at room temperature. The Sr/Mg solution was either prepared by
dissolving the corresponding acetates Sr(CH3COO)2 (Aldrich 99%) and Mg(CH3COO)2Á4H2O (Aldrich 99%) in water
or by dissolving the (basic) carbonates SrCO3 (Merck, rein) and Mg5(CO3)4Á(OH)2Á5H2O (Aldrich 99%) in diluted
acetic acid (Loba Chemie, p.a.). The different precursors had no impact on the final product. NH4F (Aldrich, 98+%) was
used with 10 wt.% excess relative to its stoichiometric quantity. Usually the resulting precipitate was very fine and had a
gel-like appearance, so it did not settle in the beaker. After precipitation, the product was therefore put into a drying
oven over night at 50 8C to evaporate the water. At these conditions the byproduct, ammonium acetate, decomposes and
sublimates. After this step, the samples were annealed at different temperatures between 200 and 500 8C for different
durations varying between 1.5 min and 18 h. The samples were left to cool down to room temperature in the oven,
which took approximately 4 h. The products phase composition and crystal size was analyzed using a Phillips X’Pert
powder diffractometer (Cu Ka) and the software package Topas. TEM-micrographs were taken with a JEOL100CX
electron microscope. A phosphorescence-decay-curve was recorded in a dark box with a standard photomultiplier tube.
Fluorescence spectra of the samples were recorded using a Jobin Yvon Fluorolog-3-22-spectrofluorometer equipped
with two Czerny Turner double monochromators and a 450 W Xe-lamp. IR-spectroscopy was carried out with a Perkin-
Elmer Spectrum One FT-IR spectrometer using a Specac ‘‘Golden Gate’’ single reflection ATR cell.
3. Results and discussion
3.1. X-ray diffraction
X-ray diffraction patterns of the over-night-dried samples had an amorphous shape with two very broad signals at
29 and 4482Q, which indicates a very small / nanocrystalline grain size distribution. In a next step the samples were
annealed to achieve a higher degree of crystallisation and to induce grain growth. The annealing was carried out at
relatively low temperatures under atmospheric conditions. Fig. 1 shows the resulting X-ray patterns after 30 min
annealing at different temperatures between 200 and 500 8C. At 350 8C the diffraction pattern still has its amorphous
Fig. 1. X-ray powder diffraction data for SrMgF4 annealed for 30 min at different temperatures.

170
C.Veitsch et al. / Materials Research Bulletin 43 (2008) 168–175
Table 1
Crystallite size at different temperatures and times (res = amorphous residue; values in nm; esds ꢀ1%)
1.5 min
7 min
15 min
30 min
60 min
3 h
6 h
18 h
350 8C
400 8C
450 8C
500 8C
amorphous
res/30.9
64.4
amorphous
res/50.6
71.1
amorphous
56.5
res/28.9
66.6
99
res/30.2
73
amorphous
68.6
63.5
82.8
92
104
amorphous
90.4
95.2
108
137
140
shape, while at 400 8C part of the sample already crystallized. At 450 8C the sample is completely crystalline and can
be identified as single phase SrMgF4. Refinement of the resulting structure yielded a mean crystal size of 64.6(8) nm.
Further experiments showed, that it is possible to achieve the same degree of crystallisation at lower temperatures and
longer annealing times. Crystal size refinements for different durations and temperatures are given in Table 1 and
Fig. 2. For temperatures higher than 500 8C and annealing times longer than 30 min the samples start to decompose
and SrF2 is formed (see last graph at 26,682Q in Fig. 1). A sample treated at 800 8C for 4 h at atmospheric conditions
consisted solely out of SrF2, MgF2 and MgO. When this treatment is conducted under Argon-atmosphere, no MgO is
formed during the decomposition.
3.2. TEM
A TEM-micrograph of a precipitated and dried sample is depicted in Fig. 3 (50 k magnification). Apparently the
powder is composed of roughly spherical grains which are agglomerated to form larger units. The subunits are
approximately in the range between 10 and 50 nm. Due to magnification limits it is not clear, if the grains are separated
or if they form a continuous unit of matter in the agglomerates. TEM-photographs of samples annealed for 30 min at
different temperatures are depicted in Fig. 4 at a magnification of 33k. It is obvious that the mean particle size rises
with higher temperatures. The diffraction mode gives amorphous rings when focused on the unannealed samples.
Using the diffraction mode on samples, which were annealed at temperatures higher than 400 8C, delivers clearly
visible diffraction spots (see Fig. 5). This corresponds with the X-ray data (Fig. 1).
3.3. Luminescence
As an interesting side-effect, the SrMgF4 samples show intensive room temperature UV-fluorescence after the heat
treatment described above. Additionally a weak afterglow of up to 40 s can be observed (a typical room temperature
phosphorescence-decay-curve is depicted in Fig. 6). The wavelength of the fluorescence emission, excited at 254 nm,
changes gradually with annealing time and temperature. The general trend is depicted in Fig. 7. While at low annealing
Fig. 2. Correlation between annealing time and crystallite size (the experimental error is within the size of the symbols).

C.Veitsch et al. / Materials Research Bulletin 43 (2008) 168–175
171
Fig. 3. TEM-micrograph of untreated SrMgF4 at 50k magnification.
temperatures and short times a bluish colour dominates, it changes to a white and afterwards a yellow tone at higher
annealing temperatures and longer times. At 500 8C/18 h there is nearly no luminescence left.
Fig. 8 shows emission spectra of a sample annealed at 350 8C for 15 min. The wavelength of maximum emission
intensity is dependent on the excitation wavelength: while at lower excitation wavelengths (260 nm) the emission
maximum is located at 440 nm, it shifts towards 390 at 320 nm excitation wavelength. Generally one observes that the
emission band is quite broad, and that the emission intensity decreases with longer excitation wavelengths. The
Fig. 4. TEM-micrographs of samples annealed for 30 min at different temperatures—33k magnification.

172
C.Veitsch et al. / Materials Research Bulletin 43 (2008) 168–175
Fig. 5. TEM-micrographs of samples annealed for 30 min at different temperatures—diffraction mode.
corresponding excitation spectra (Fig. 9) have a complicated structure as well. It is obvious, that there are several
different effects contributing to the overall spectrum, resulting in different excitation maxima (e.g. at 270 and 308 nm)
which change their intensities at different emission wavelengths.
In Fig. 10 the emission spectra at 300 nm from several samples, which were tempered at 350 8C for varying periods
of time, are compared. As expected from visual examination under the UV-lamp, the emission maximum shifts to
longer wavelengths with increased annealing times (The emission intensity is normalised for better comparability). A
fixed annealing time of 30 min with variable temperatures between 250 and 500 8C resulted in very similar emission
spectra. Since there is no self-luminescence reported for SrMgF4 and contamination with luminescence inducing ions
was prevented, the emission has to be caused by some kind of defect. Angelov et al. [6] describe a defect luminescence
Fig. 6. Room temperature luminescence spectra of SrMgF4, excited at 254 nm.The sample was annealed for 3 min at 350 8C.

C.Veitsch et al. / Materials Research Bulletin 43 (2008) 168–175
173
Fig. 7. UV-luminescence-color scheme of samples, annealed for different times and at different temperatures.
Fig. 8. Emission spectra of SrMgF4, annealed for 30 min at 350 8C, observed at different excitation wavelengths ranging from 260 to 420 nm.
caused by CO2À radicals in SrCO3 obtained by thermal decomposition of strontium oxalate. Since strontium acetate
shows a similar behaviour, and IR spectra (see Fig. 11) show, that there are acetate traces present in the sample
(probably adhering to the nanosized SrMgF4-particles), these radicals could be one source of the luminescence
observed. This theory is supported by the fact, that samples, which were precipitated from alkaline earth chloride or
nitrate solutions showed much weaker luminescence intensity. There are also reports of oxygen/vacancy defect pairs in
earth alkali fluoride crystals [7,8] like BaF2 and SrF2 which cause photoluminescence in similar spectral regions as
observed here. Such defects could also be a reason for the visible luminescence observed, since samples which were
annealed in Ar showed less luminescence intensity than those annealed in air. In an unpublished thesis [9], oxygen was
Fig. 9. Excitation spectra of SrMgF4, annealed for 30 min at 350 8C, observed at different emission wavelengths ranging from 356 to 440 nm.

174
C.Veitsch et al. / Materials Research Bulletin 43 (2008) 168–175
Fig. 10. Emission spectra (excited at 300 nm) of different samples, annealed at 350 8C.
Fig. 11. IR-spectrum of a sample annealed for 30 min at 250 8C, in comparison with spectra of several different acetates.
Fig. 12. Comparison: SrMgF4 [5] and SrMgF4 prepared in this work (super structure reflexes are magnified).

C.Veitsch et al. / Materials Research Bulletin 43 (2008) 168–175
175
added either as BaO to the parent compound BaMgF4 prepared from the melt or by heating melt grown pure BaMgF4
under 50 Torr of saturated water vapour pressure to 850 8C for 3 days. The resulting samples show two broad
luminescence bands centered at 22000 cmÀ1 (454 nm) and 17300 cmÀ1 (580 nm) with a steadily increasing excitation
band with a maximum at 49000 cmÀ1 (205 nm). Similar emission bands can be seen in the SrMgF4 samples annealed
at 400 8C for 30 min at around 490 nm and 580 nm.
4. Conclusions
It is possible to prepare strontium magnesium fluoride via a solution chemical route. Since the precipitated powder
is very fine and/or amorphous it has to be annealed at a minimum temperature of 350 8C to achieve X-Ray powder
patterns with discrete reflections. Annealing at 450 8C yields single phase crystalline SrMgF4.
The superstructure-reflexes reported by Ishizawa [5] are visible as well. Fig. 12 shows, that the measured data
corresponds to a powder pattern derived from the single crystal data published in [5] (the superstructure-reflexes are
especially magnified). The annealing time and temperature can be used to adjust the mean particle size in this system.
TEM-micrographs confirm the mean particle size calculated with X-ray data refinement. One can also observe a
growth of the mean particle size along with higher annealing temperatures. This corresponds quite well with the
crystallite growth calculated from X-ray data. Further HR-TEM analysis is needed to determine the exact constitution
of the grains and to gain information about the size of the crystalline regions which are responsible for X-ray
diffraction.
When heating the precipitated powder to higher temperatures it starts to decompose at 500 8C. Compared to
previous results [4] this temperature is rather low. Therefore we conclude, that the heating speed and the grainsize
distribution have an impact on the compounds melting and decomposition properties. An intensive broadband
luminescence with afterglow, was observed after annealing the samples. This luminescence is probably caused by
defects, since the pure parent compound BaMgF4 is fully transparent from 180 nm (7 eV) to 8000 nm [10], and
probably even further in the VUV region (pure SrF2 and MgF2 are transparent even above 8 eV). It does not appear
likely that the small grain size can shift the absorption edge by as much as at least 2 eV to generate an absorption close
to 250 nm (ca. 5 eV) associated with the observed emission. The emissions maxima shift with different annealing
temperatures and times. Therefore it is concluded, that these factors must have some kind of impact on the defects
which are causing the fluorescence and phosphorescence. The most intense luminescence can be observed, if the
samples are treated for 30 min at 350 8C.
Acknowledgements
This work was supported by the Swiss National Science Foundation; TEM photographs were taken at USTEM—
University of Technology Vienna.
References
[1] H. Hagemann, F. Kubel, H. Bill, F. Gingl, J. Alloys Compd. 374 (2004) 194–196.
[2] Q. Bingyi, E. Banks, Mater. Res. Bull. 17 (9) (1982) 1185–1189.
[3] Y. Wu, C. Shi, Solid State Comm. 95 (5) (1995) 319–322.
[4] E. Banks, S. Nakajima, M. Shone, J. Electochem. Soc. 127 (10) (1980) 2234–2239.
[5] N. Ishizawa, K. Suda, B. Etschmann, T. Oya, N. Kodama, Acta Crystallogr. C57 (2001) 784–786.
[6] S. Angelov, R. Stoyanova, R. Dafinova, K. Kabasanov, J. Phys. Chem. Solids 47 (4) (1986) 409–412.
[7] E. Radzhabov, P. Figura, Phys. Status Solidi (B) 186 (1994) K37–K40.
[8] U. Rogulis, S. Schweizer, J.-M. Spaeth, J. Phys.: Condens. Matter 14 (2002) 6949–6956.
[9] J. Rey, Ph.D. Thesis, University of Geneva, Switzerland, 2000.
[10] K. Recker, F. Wallrafen, J. Cryst. Growth 26 (1) (1974) 97–100.